Thank you so much Dave! I've argued this ad nauseum on a couple Facebook groups. Networking is my dayjob and I basically live with a DSX-5000 attached to my hip. I've tried to explain this and been told I'm wrong. Has to be CAT5e and you can't use better. The one thing I will say is you're increasing your loss with every connector, but really who keeps all those different lengths on hand to test. I would say you can fail the length but if is one actual cable it would probably work.... but as I typed this you talked about increasing the insertion loss because it was multiple extensions. Once again Dave, thanks for going total geek with a tester that most audio people just wouldn't buy because of the cost and need.
Thank you so much and it is so refreshing to meet people that understand and get it! I really tried to triangulate this from multiple angles and not leave much room for doubts
So glad this is pinned at the top. It matters, as does the crimping of the cable if you make them yourself. First thing I questioned myself: "how will you prove actual usable distance of a single run of any cable when you are increasing the loss with EACH separate connection?" I solder the shield drain to the J-45 to ensure continuity, knowing the shell of an ethercon may not do the job itself through friction alone.
I didn't set out to show the best possible results. This test is about showing worst case scenarios when people extend cables out and what to expect and what they can trust. Knowing what a specific length of cable can do and not having that length available to you is not of significant value when you're setting up a show
I'm not surprised you've had to argue this ad nauseum. My day job is in networking too, and I've stated the same points as you probably have. Very few people actually understand the properties of cables, yet will stand toe to toe with someone who does and argue with them. Thank you for taking the time to do this Dave. :) It's proven a point I've been trying to make to people for years.
Great video! My background is in RF and audio. One thing you always want to avoid is cable extensions. If at all possible a single run of the appropriate length will work best. Cable extensions add connectors and connectors effect insertion loss far more than the cable itself. In other words two 50 meter cables connected together will always be worse than one 100 meter cable (assuming they both are made from the same type of cable). Cable extensions should only be used when there is no other option and they should never be used in an install situation.
I was interested in that and actually tested the couplers. Adding a coupler and a short 6-in piece of cat cable showed less than a 0.1 dB differential. I also compared a 100 m super Cat XM plus two 10 m extensions and two couplers against 2x 60 m super Cat XM cables and a single coupler. Well the coupler was visible in graphs that showed cable lengths and reflections, the impact on insertion loss was less than .1 dB. All that said the goal of these tests is not to get the absolute best results possible. That's the opposite of my intent. The goal of these tests is to give real world information that is usable and reliable. In the real world people are going to extend cables to get their show up and running. That's where the valuable information is rather than the ideal situation that we often don't have the luxury to experience.
I came to the comments to find this same observation. It is interesting that Behringer specifies a max length of 100M, but they only sell reels of 50M. It might lead the consumer to believe that it's OK to couple 2 reels to stay within Behringer's spec. I agree that a single run of 100M should give much better performance sans couplers.
@@DaveRat Consider as well that at installation scenarios they may run the cable in the wall and terminate to wall jacks... So it might be interesting to see if your supercat is plenum rated, and any impact of neutrik wall jacks? Assume one wall jack at FOH and one at the stage... e.g. 75meters in the wall and 3 meter patch at each end
@huktek The Super Cat XM cable is not designed or meant for installation purposes. Cheaper stiffer plenum rated cable should be used for that. As that less expensive install grade cable is typically solid conductor as well so it's less expensive to make higher bandwidth cables. It's all about selecting the right tool for the job and using a pro touring cable for installs is not optimal just like using a solid conductor rigid install cable is a bummer for pro touring
Dave! Thank you for this video! I went down the rabbit hole on AES50 and cat cables after a failure on a recent tour. Turns out it was a bad AES50 card in the console. I did hack a hundred feet off my Supercat reel just to be safe. I have major jealousy over your Fluke tester! Thanks again for all you do!
@LucasWasson I will ponder that. The issue is that The fluke is focused on cable certification to specific specifications. The sound tools cables are designed for optimizing analog over cat while maintaining Cat5e or better specifications. That said just yesterday I put together a product grid for the sound tools to cables with the various specifications and product design focuses. That should be out soon on the sound tools site
Love this nerd-out so much! I've been saying for the past decade for so that if I knew I had to be a network engineer to do live sound, I might have just gone into networking. Well, touring is mostly more fun I think!
When I installed the M32 at my church 6 years ago I made sure to do my research. I spoke with Midas/Behringer directly and was told to NOT use Cat6. They told me at the time almost every problem they see is with people using Cat6 with their consoles. So we ordered high quality Cat5e and never had a single problem. The length was about 200 ft.
Interesting. Pretty much every cable works at 200ft so you are safe. At around 250 feet many or most cables have issues. If you watch the video I show how to test your cable and determine how far it will go. It's important not to just make assumptions that all Cat5e cables have similar performance because they do not, they are vastly different varying from manufacturer to manufacture and even within brands. Same with all the cat cables
@@DaveRat Exactly right Dave. If I had been closer to the max recommendation range id be testing cables. I did pull 2 spares in as well. Always a good idea to have spares. This comparison you did was very eye opening.
I think the test that I do prove that concept incorrect if you actually watch the video. Except for maybe cat5e unshielded, which I also show in the video. But this video tests exactly what you're talking about on shows the results
Thanks for doing this video Dave. There is a huge discussion on Behringer World about cabling and distance, ethercons, etc.. Very useful video. Also, thank you for dropping the Cat5e only bombshell at 2:56. Very useful information/video!
So cool and thank you and much appreciated! My long-form videos are not for everybody in this world of TikTok and such, but it makes me very happy with the imposed appreciated. Cheers to maximum nerds!
Thanks, for this great video, Dave. It's ALMOST the definitive video on this subject, which has been argued to death on at least two forums that I post on (Behringer World and Live Audio Board). I appreciate very much your taking on the CAT5e ONLY vs. CAT5e OR HIGHER issue. I've never understood why something that's designed to pass signal better is actually worse for a lower signal quality, and have only seen various kinds of hand waving to argue the opposite. I'm not persuaded by "AES 50 is not Ethernet", which is true but higher speed signals needing more efficient cable (if "efficient" is the right word for better signal passing capability) negates that argument for me. The two things that you're missing, and one needing elaboration: 1) There's a hard limit to how many couplers you can have in your signal path. It's either 3 or 4, IIRC, and when I had a DTX-1800 for a while and connected 4 or 5 cables together with couplers, I got a FAIL for too many couplings, even when using 10' or less CAT5e cables. You are using 2 couplers to adapt to Ethercon, so you're already partway to FAIL. If needing a longer cable, I never add more than one length. To test the length limit, change the add-on cable to longer lengths rather than an assortment of shorter lengths. I'm surprised the tester didn't tell you that, and can't explain why. 2) You mention variable twist rates changing the lengths of the pairs relative to each other, potentially messing up the skew. That would be true if the pairs were consistent twist rates end to end, but a quality cable would have the twist rates vary equally over the length of the cable. So if you open the cable jacket at a given point and see varying twist rates across the pairs, opening it at a different spot in the length would show the same pairs but different twist rates. I have done that and can confirm. I have never examined rock-bottom priced cables for how they implement this feature, if they bother to, so that would be one thing to watch if buying budget cable. Note that this variable twist rate is one of the ways how higher CAT cables achieve their ability to pass faster signals better. You correctly talk about individual pair shields being another way they do that. There are others, too. 3) Regarding the effectiveness of Ethercons for shielding from ESD, a) you have to have a shielded RJ45s on your cable, with the drain wire solidly connected (soldered, usually?) to that shield,. b) The simple friction fit from that to a nickel (natural metal color) Ethercon at both ends of the cable seems to be enough to give a good connection to dissipate the ESD. c) I'm not aware of anyone looking at the black Ethercon shells to see if they make as good a connection. If they are anodized, maybe that's enough? I know nothing about the conductivity of anodization. If they are painted, I wouldn't think that's enough as there'd be a gap between metal and metal. d) Brian Wynn, who was the first one I'm aware of who figured out the ESD solution, proved conclusively in his videos that you need to use Shielded cable with Ethercons at both ends of the cable. Your comment about maybe being able to use unshielded in a rack or something like that, seems to be inviting problems. Not every day, maybe, but someday. Those quibbles aside, I think you've done another great service to the community. Thanks!
I will try and do a test of couplers. Using the term coupler vs defining the type of coupler is as useful as saying cat cable vs the type of cat cable. 8 am not convinced that couplers are as big of an issue as many people make them out to be if Cat6a couples are used. As far as ESD protection, the exposure of the cable to ESD tends to be outside the rack. Once the cable hosts that rack and the RJ45 is grounded to the rack, all good. Inside the rack the exposure to ESD should be reduced. Same as inside the console. Ess vulnerability exists inside the console as well but the the exposure is greatly reduced. This of course relies on the rack and console ethercon s being grounded. But I will look into testing that as well. Best I can figure, the real issue is insertion loss. 24 dB, the spec says is ok is huge. That is like less than 1/300th of the signal is making it through. This tiny amount of signal is very volnerable and based on the cables people tend to use, 30db of loss or more I believe is typical and where AES starts to fail. That is 1/1000th of the signal at 100Mhz. Yes, couplers and ethercon shells and breathing on it wrong may cause issues with that and also create a bunch of voodoo and false beliefs as well Get the losses into the 20 dB range and all that crap goes away. Use a bunch of couplers and no ethercon shells needed is my guess. But will test.
Great detail. Nice! Minor detail: some of your graphics show 100 millihertz with the lower case ‘m’ 100mHz instead of 100 megahertz 100MHz. But LOVE your level of explanation.
This is quite definitive. If I ever need to be more than a football field away from the stage, I will know what to do. I have a 150 ft cable for my Midas setup, and shielded Cat5e ethercon cable. I feel quite comfortable about it now.
Yes, and for us, I am out at Coachella now and Sahara tent needs every bit of 100 meters to FOH. We actually have the issue quite a bit with guest engineers bringing in X32 and wanting to plug into our industry standard FOH 100 meter snake length.
I appreciate the time, effort, and research you put into your videos. You do a great job of explaining things so they make sense. Keep up the great content.
I have here in Hannover, Germany a small club that I look after that I ran 4 cat7 cable with ethercon to XLR BOB‘s, have many different guest desks here as well as the house Midas M32, run is about 30m (x4) and I’ve only ever had a problem with and I solved that by using standard CAT5e as jumpers at both ends, my link cables (BOB to desk/stagebox) are all CAT6
Great test! Would be nice to rule out connector loss as each one introduces loss. Probably would be good to test each of your extension cables on their own to make sure one is not super lossy. Most of all it would be cool to see if a single piece of 150M supercat worked/low insertion loss.
The math calc s action of the video does a good of predicting length. The couplers has such low insertion loss I could not definitively measure it. But, the extensions do cause some reflections that are visible in the return loss graphs. Also the extension cables were a lower quality than the red cable. The cable quality of the extensions are a bigger factor and with all red cable we can pass ars50 signal at 150 meters. All that said, the 24dB insertion loss is a good metric and I triangulated with 3 different methods to factor out things like a coupler being a big issue.
👍🤙👍 cool cool. And yes, most or many people do not need those length. Interestingly, on the upper levels of pro audio 100meter is the standard snake length!
@Edwin-van-der-Putten very cool and yes in the US it used to be 300 ft was the standard and in Europe was 100 m which is 328 ft but now in modern pro audio 328 ft is become the worldwide standard for the most part. It will vary from company to company but 100 m snake is the norm
Thanks again. You definitely stirred the discussion up, great! You indicated you might do some additional testing. Would be nice if you could show a possible difference between cables with T568A and T568B pinout and show what happens when a T568A terminated cable is coupled with a T568B terminated cable. Just for the fun of showing it. Thanks.
When I get a chance I'll give that a try and if something interesting happens I could do a video on it. I'm trying to think out what the possible result would be and I guess rather than having the pair with the most twists be the most twists all the way through it would attach to a different pair with less twists and same thing for the least twisted pair would attach to a pair with more twists. My guess is that the skew would reduce, and the differential between the worst and best pairs would reduce in various measurements. And the reflection caused by the coupler would show up as it always does
As an electronics engineer I wanted at one point to make an AES-50 to optical fibre converter as a side-project at some point. I never got around to actually doing it though, though I do wonder if there’d be enough interest and possible demand. But theoretically that could go to just hundreds of meters, but easily 10km or more if you ever wanted, and you have the benefit of complete electrical isolation. Basically my idea was to have converters at each end that produce raw Ethernet packets (although it uses cat5, AES isn’t at all Ethernet itself) so you can use standard gigabit Ethernet fibre transceivers. It could theoretically go through Ethernet switches then (which it can’t as raw AES-50), but you’d want to avoid it because it’d probably screw up the timing.
Holly carp sea man I heard all of the frequency sweep over my headphones and better than my Sennheiser's. I mean flutter pressure and all. I'm a bit upset really. Sure 40 BT wireless headphones I bought on a whim. Great work on the video. Great spec reveal on the red cable and thanks for the math. Loved that Fluke sir. Thanks again. I hope that new cable works out.
If you need even greater ESD protection, CAT 7a and CAT 8 have much better shielding and will provide superior protection as well as give a bit of extra length "AT" (and this is very important) the CAT 5e spec. The only downside is the cost.
Nice work Dave!. I've also wondered how AES50 works and by looking at the circuit, I can see that it uses two pairs for Ethernet data as they are connected to a LAN8720A PHY but interestingly the other two pairs are connected to LVDS trancievers (65MLVD204), this is why they wont work through a standard network switch.
I just did a gig in Turkey, where the equipment company provided a standard cat5 UTP over carpet in a conduit. Needless to say opening night was a disaster every time a human walked over the cable there was a huge bang from FOH which we traced to static discharges. Took them 3 replacements before they found a F/UTP cable, which while still not to spec solved the problem. Sucksthe AES50 doesn’t even seem to basic CRC confirmation of the transmitted frame.
AES50's spec, or, lack of spec clarity, is absolutely infuriating. I've run ethercon, but, for installed stuff, it's far easier/cheaper to just run shielded CAT6A and be careful about how you route it. ESD is a problem, and RF is a problem. No amount of shielding is going to save you if your cable is laying on top of a DMX dimmer basically, gotta use your ol'noodle and avoid problems like that. Thank you for going over the nitty gritty, and showing what actually works and what actually doesn't. That said, I've found AES50 in general to be fickle, DANTE is much, much more forgiving and predictable, because it's just using Ethernet standards which are extremely well defined and battle tested. Never had a problem with DANTE that wasn't related to actual hardware failure or egregious misconfiguration. Also, dante will run for as long as your fiber run is, so, multiple miles without issue, if you have the networking gear to handle it. Even something as simple as a cheap gigabit Ethernet switch (provided it doesn't have any power saving features) will get you to 200M without issue.
Agreed. Thing that makes AES50 so cool is the ultra low latency but that's also its weakness and that it doesn't have the error correcting and relies on a cleaner signal than other formats do. But looking at it the other way, a cable that will run AES50 over 100 m will be stellar for just about everything!
Very cool video and super useful information here! I've run into dropouts exactly once, when I had a 80m cable with me but the stage rack ended up on the wrong side of the stage and we had to extend. They used some super dodgy cable that was way too long but luckily I only had a single dropout of a split second during the show. I recently bought another Ethercon drum and talked a bit with the manufacturer and they said they actually switch to a higher grade cable for lengths 80m and above. Another cable I had didn't even work at 30m despite being Cat5e spec. I measured it and it turned out to have a DCR of twice the Cat5e spec!
Yeah I think that's a difference between the X32/M32 and the pro series is the pro series does not require shielded cable for some reason. Getting longer lengths out of unshielded cable as you see in the video is not that much of a challenge. Behringer/Midas does not make a 100 m shielded cable
@@DaveRat Funny enough, I’ve found if you use STP on a Pro series desk, you’ll end up running into weird issues with clock/sync. Sometimes I’ll work, sometimes it’ll never clock up at all, sometimes it’ll pop and click until it drops. Essentially the same happens on an M32/X32 but in reverse, and there’s the potential for permanent damage. I’ve seen the AES50 cards on the Behringer side of things get killed by using UTP a handful of times.
Yeah, UTP world better but is vulnerable and STP is unstable if it's not of very high quality or short lengths seems to be the issue. It's possible X32 m32 is lacking ESD protection the pro series has or so I have heard but not tested.
I actually did some measurements on that. One thing that can be done is to measure the difference between the cat6a and cat 5 coupler from neutrik, it actually only shows up as a fraction of a DB of loss. One way to measure the actual coupler I would need to make a cable of a certain length and then make two cables half the length and then I could measure the impact of a coupler plus two terminations. I did something similar where I measured 132 m length of cable using 100 m plus a 24 m plus an 8 m and then measured 100 m plus a 32 m. The impact of the coupler and two terminations comes out to be a fraction of a DB somewhere in the 0.2 to 0.4 dB region. Your question is of specific interest to me because I really want that SuperCat XM cable to be within spec at 120 m, as soon as I have another 10 m piece I will test it with the dual 10 m extensions to the 100 m, hopefully that reaches the spec but if not, I will look into making a single 20 m extension and see if that hits spec and possibly have those manufactured as well. Currently we are only looking at making it in 100 m, 10 m, and 60 m lengths Also I will post the data sheets from the tests on the channel members Telegram channel if you're interested in seeing those
Have you done a comparison of the insertion loss of the coupler itself? I was wondering if it would be better to have a single piece of 110 m cable rather than splicing on the extra 10 m with a coupler.
Of course I did! It's less than 0.1 dB though it does affect some of the reflections on the return loss but not enough to make a difference. It seems to add about 6 in to 12 in to the cable when the coupler is in based on the measurements The that's with the cat6 coupler. The other coupler can be problematic. The comparison between 60 m cables and one 100 m plus two 10 m extensions shows nearly identical specs.
In Ethernet the max cable length of 100m came from timing not electrical signal attenuation like most people assume. This is to help with collision avoidance and knowing how long it could be for a collision to be seen on the far end. However audio cables probably don't have to deal with multiple stations on a bus. Not sure if the Rx talks back to TX. Haven't looked at how it works from a data link framing perspective. Your ESD test confirms what I'd expect. That is you have a certain noise margin which is signal minus attenuation. If the ESD noise is greater you get a drop. Hence why you want a safety factor. I'd actually say you'd want at least 3dB or more of noise margin but some like to shave closer to the bone. Of course if you go fibre then ESD, potential differences and ground loops are not an issue. Glass is awesome apart from not many audio techs having a fusion splicer on tour. The fluke will test it with the right adaptors.
Yes, we run many miles of fiber and Cat and multiple audio formats at our festivals. This is focused on getting AES50 to work reliably without adding complexity and additional conversions, latency and potential failure points. Also, in California the maximum freeway speed is 65 MPH. No one would ever have a reason or dare figure out a way exceed that, right?
Ethernet max cable length was 100m from colision avoidance prespective. Now it is (full duplex links, switches instead of hubs) 100m but based on electrical signal attenuation.
@@DaveRat you inspired me to read some more. Looks like AES50 borrowed layer 1 and some of layer 2 of Ethernet. If I was to guess they've probably pulled the addressing from Ethernet as they designed this to be point to point. That said it might be possible to build active repeaters to increase that cable length to longer distances while blowing out the minimum latency guarantees.
Yes Klark makes a repeater called a dn9610 and a rack version as well. The issue is that for live sound, the repeater would often need to be somewhere in the audience and adds another layer of complexity and failure point. Pure quality copper is the most eloquent solution in my opinion.
There has been a lot said about the need to tie the shield to the ethercon barrel. Not sure if it's in scope of any followup video on this, but for an install (with stable power) we pulled FTP, no ethercon, and no issues at 220ft. Perhaps only severe electrical discharge scenarios would require a second conduit at the connector for the shield? Of course Ethercon is needed for touring, but not sure for install...
I haven't tested whether or not the barrel matters. But my guess is that it does not make that big of a difference whether it's there or not as long as the RJ45 has the metallic housing. The issue from ESD is related to the magnitude of the ESD versus the magnitude of the signal coming down the cable. A cable with less loss will be more impervious to ESD naturally. 24 dB of loss which the spec says is the maximum amount of loss is a lot of loss! That's like 1/3 of 1% of the signal is actually making it to the other end at 100 MHz. With this tiny little signal things like ESD can easily disrupt it With something like 14 DB of loss there will be 10 DB more signal which is 10 times the amount of signal showing up the other end. So my take is yes the barrel could be important if you're running extremely high loss cable and working right on the verge of function. Which is exactly what would happen if you were using the KT cable or most of the cables that are out there. So I'm guessing it's a combination of good shielding, low loss and to a lesser extent a tiny impact of the shell of the ethercon
Would like to see you test out the game changer cable. That is designed to allow gigabit and poe 200m. Curios to how that holds up to AES50 spec and your new cable
Just a quick look at their specs tells you most of what we need to know. It has a 16.1 dB loss at 100 megahertz at 100 m, so maybe 2 or 3 dB or so less loss than the red cable, which is good. Probably calculates out to about 10% farther runs than the red cable does with AES50. But It has solid conductor wires and it's fairly easy to get better specs with solid conductors, all good for installs and sucks for portable live sound. The red cable uses stranded wires which makes it tough to get long lengths. It's got a PVC jacket which is cheaper,, pretty much the plastic electrical tape is made out o, less durable and does not deal well with cold or hot temps. The SuperCat XM has a polyurethane jacket, more like what skateboard wheels are made out of. Looks like a decent install grade cable product line but using it for live shows and such would be less than ideal.
I come from the RF field. Each connector 2.4GHz and above has an insertion loss value. Was wondering if there is an insertion loss value for each connection when stacking multiple lengths of cables. Interesting video.
Insertion loss seems to measure out at about .1dB for the Cat6 coupler That said, My goal here is to offer useful information that parallels the real world. So having the insertion loss of the couplers is a valid configuration. If I was to test this without the couplers in an ideal situation that many are most people will not encounter, I believe that would not be as valuable of information
Dave, Thanks so much for all your very informative videos! Always amazing work! It's so interesting that my team and I never experience this problem with our Pro2 and the 100m unshielded cat5's that it came with. But when we use even 75 meter Elite Core STP cat5e with our M/X32's we get the ESD explosion from hell every other hour or so. We never get intermittent drop outs though (except when the discharge happens). I'm wondering if there are two issues at play that are somewhat unrelated: 1. intermittent connection due to user's cable not being to the spec you speak of in this video, for the length people are running and 2. the ESD that seems to freak out the word clock generator in the mixer when the discharge grounds to the circuit board. What do you think about installing an RJ45 grounding block at least at one end of the cable to redirect the ESD to the ground plug of an outlet? Just to solve the ESD and not the intermittent connection issue, of course. Thoughts?
It seems to be several things going on. Maybe take a look at this other video I did and let me know what you think. I go into much more depth and testing. th-cam.com/video/bQXOXWuxCGA/w-d-xo.htmlsi=46Ma4RwH4BCSfF0e
I believe Cat5e, Cat6, and Cat6A are rated for a max length of 100m, any longer would need a switch. Of course, at that point it just comes down to getting good cable from a reputable brand. In my opinion, If you need more than 100M of any category of ethernet, you should just switch to coaxial or fiber. Well, if possible. And, I guess at that point it might just be worth to switch off of AES50 and go for MADI. Great video, love the testing segments at the end!
Perhaps watching the video will clarify how I am pushing the 100m limit. But hey the speed limit 65 miles an hour and you never ever go faster than that correct? It's always better just to follow the rules rather than test them and determine what actually happens and prove confidently whether or not the rules can be broken! I made the super Cat XM cable specifically to prove that 100 m limit to be breakable. I did this video to show you how I gained the confidence to know that extending beyond 100 meters is safe
@@timbo19 My point still stands, If you need more than 100m you should be looking away from AES. AES is 20 years old at this point. Broadcast has mostly moved away from it, and a lot of touring acts we get have been using fiber.
Ok, so let's say you are a sound vendor supplying sound systems for large events. 100 meters is the standard main snake length. Artists often bring their own consoles to events and festivals you supply sound for. Some of those consoles are AES50 based and often, the type of console is not advanced by the artist. Do you? A) convince them to buy the proper gear to be able to connect their gear reliably during the 30 minute set change? B) carry conversion to Dante cards for the various assortment of consoles and help install the hardware during set change? 3) carry some dn9610 units 2 wall warts and run a fiber for them? 4) watch their show go down in pop and drop flames 5) make them mix from side stage 6) say, oh, use the red cable you are good to go Let's say you are the small band or pa company, what do you carry to deal with shows with long runs? Is having a cable that solves the problem less or more easy, faster and reliable than converting and converting back? It depends of course on the application, time available and other factors. I would say that those that are able to embrace simple fast and reliable solutions over complex and cumbersome solutions are more likely to succeed.
Before i watch anymore, Id say 250' shielded is safe. I used a cat 5 unshielded for a ui24 as a monitor board. logging trucks passing by outside were injecting interference on it from their CBs. Never use unshielded...... Just MyTwo Cents. You ROCK DAVE!
@DaveRat Great Vid. I know what to upgrade to now. You know the science to get things correct. I'm doing larger and larger venues, and the distance is getting critical. Running a mesh tandem wifi is helping on some of it. Rock on Dude!
Love this video! Thanks for getting to some of the finer details behind this. Can your Cat6a couplers be used/backwards compatible with your SuperCat purple cable? Love my supercat...never had a lick of trouble with it. Keep rocking the videos! They are much appreciated!
Thank you and yes the cat6 couplers are the only way to go! There's somewhat waterproof and way better. The other couplers have straight wires inside no shielding and they're junk
Given that AES50 is using the 802.3 1000Base-T PHY (ethernet layer 1)for data (plus the additional clock lines), any 1000Base-T pass/fail tests your Fluke unit has for cabling will also be applicable here. Frames are CRC protected which offers some additional noise immunity - Some Fluke etc testers can measure the on-wire bit-error-rate for ethernet frames. If your cable results in received frames with errors close to the limit (1 in 10^-12 erroneous bits or thereabouts) you're asking for trouble. If you want to know what your margin is for a given cable, such test results might be useful to you. Also, don't forget to factor in the insertion loss of connectors. Connector assemblies/patchbays not appropriately specified or assembled can give you a headache.
If I recall, weren't the early Midas Pro and Behringer X32 specs contradictory? I'd swear that initially one said use STP (shielded) and the other UTP (un-shielded). Have to look through a pile of saved PDFs to find it.
When Behringer started recommending STP after Brian Wynn's tests in about 2014, Midas still insisted that UTP was fine for its higher end consoles (before they introduced the M series), even though my buddy who had a Pro 3 and I proved that you could disrupt the signal with ESD like Dave was doing in the video. It took a few more sparks to disrupt it than on an X 32, but it was still disrupted. Don't know how long they stuck to that or if they ever changed before they discontinued the Pro series. My buddy talked to them at the time with our results, to no avail.
We were having headaches with this issue recently. 300ft Lyxpro Cat 6 shielded ethercon had losses and ESD problems, we bumped it dow to 250ft and it went away. This cable is on Amazon. Our behringer wing is connected to sd16 and s16.
We are having the ESD problem using 250’ of the blue LyxPro cable. We are in ballrooms with carpet so I am wondering if that is causing our issues? So glad to see Dave demo drop outs using 100m of the official KT cable.
Yes to combat ESD you really need double shielding if you're running longer lengths. S/FTP. And you also want low insertion loss. Insertion laws of 24 dB which is the spec means it approximately 1/300 of the signal is actually making it to the other end at 100 megahertz. An insertion loss of 30 dB means 1/1000 of the signal makes it to the other end. With such small amount of signal making it to the other end even the slightest static will cause dropouts
I believe the 24dB insertion loss lengths at 100mHz are: 75 meters SuperCat 80 meters SuperCat sound 90 meters SuperCat 7 These are calcs based on the 100 m loss measurements. Have not cut to length and confirmed yet.
Do you plan on finding the maximum safe distance with a single cable of your spec cable and produce a reel for that specific scenario? I think that would be ultimate for the people looking for that level and quality of cable as well!
As I show in the video, based on the mathematical calculation using 100 m supercatXM as the reference, The maximum cable length of SuperCat XM while maintaining a 24 dB maximum insertion loss is 120 m. When testing the cable and extending it out even with the couplers this ends up being true as well. The predicted length is 120 m When I measure 2 x 60 m pieces in a single coupler, I get the predicted losses within a fraction of a dB. And when I measure 100 m with two 10 m extensions I also get the predicted losses within a fraction of a dB Everything points to 120 m no matter how we slice the pie. I'm not seeing any significant degradation from the Cat6 couplers. And they measure less than 0.1dB of insertion loss when inserted.
I'm using a 200-foot Cat6 STP with ethercons for my X32 and S16's and I'm fine with it the way it is. It's been that way since 2013 when I got my X32. This is partially because my old analog whip is 200-feet and so are my other various lifelines between stage and FOH, so I just carried it forward. The specification for the various cables is they really don't want you to exceed 300-feet. The idea is more based around networking, and we typically will use fiber in those longer runs, but that doesn't work for AES50. While you've shown it can work, I'm gonna stay within spec since I don't have any needs to go longer. I still think your tests were interesting and worth doing. Sometimes, that 10m of cable can be what makes the event happen or not.
The touring pro standard cable length for a front of house snake is 100 m which is 328 ft. Doing festivals with a pro touring snake and having artists come in with X32 or M32 consoles is not uncommon. So having a 100 m cable capable of carrying AES 32 that they can plug into is extremely useful on the upper levels of pro audio. Ideally artists would carry higher level and more expensive consoles or not use AES50 but the reality is that they do and accommodating them is important so that they can have quality shows that don't have dropouts. That is why I designed this cable
I show a method in the video of how to test the cable and determine it safe using length. In the method doesn't cost any money at all you just need the cable to test and whatever units you want to test with
Slightly Offtopic: So we've all been "abusing" several kinds of networkcables for 4 Lines analog or 4 Lines AES runs. What about turning things around ? Why not trying to use an oldschool 4 Lines 110 Ohms AES multicore for network data ? Well ok, it will be quite difficult to crimp it into a Hirose RJ45 connector, but it should work with e.g. Neutrik NE8FAV-YK or something similar with punchdown contacts. I would be very much interested in what your Fluke certifier says about this idea 😊
What if you extend the signal using SFP+ extenders to convert the signal to fiber optic then back to RJ45 copper at the receiving end? You could even get ones which use separate fiber cables for sending and receiving data. And fiber optic cable is not affected by ESD either.
Yes, convert to fiber works. And also, we all love loads of extra dongles and a few more wall warts or maybe some extra rack gear. Or Just plug in a cable that works and no extra crap for less than the cost of the fiber cable. That said, fiber is great and we use it for nearly all of our snakes and runs to delay clusters. But for people that own consoles that cost less that $3000 that has an AES50 connection right on the back, a simple relatively low cost plug and play solution that involves buying a cable that works and plugging it in, that may be the way to go for some of them. We're almost sold out of the first run and the second round of cable will be in shortly. So far the response to the Super Cat XM has been nothing but positive. Everyone seems to love having a high quality high speed reliable table living shores there will be no issues for aes50
How much loss are you getting from each connector/coupler? Your calculated max is pretty close to observed max, but I wonder how much longer you could get from a continuous cable.
I did my best to measure the coupler loss and with that cat6 coupler I'm seeing less than 0.1 dB of insertion loss. I tried to calculate out about how much farther I could get without the couplers and it comes out to be maybe a foot or two longer
Awesome work to continue this dilemma. When we were messaging back and forth a year ago I had several cables that connected immediately but took 16 min or so to develop a glitch presumably through eventual discharge. Most all them connected right away with the exception of the purple and blue rat cables (no hate here, we’re all learning and experimenting to find our way). Which numbers explain that phenomenon, insertion loss?
Yes in this past year I've done a bunch of research and insertion loss is the magic number. Other factors will affect it but insertion loss is the big one. A 24 dB insertion loss according to the specification is acceptable and stable and that seems to align with the three different tests that I did. What's a 24DB insertion loss you have about a 20% buffer which should be more than enough to account for any degradation that happens over time. 24 DB is a lot. 10DB is 1/10th the power. 20dB is 1/100th and 24 is like 1/300 which means about 1/3rd of 1% of the signal at 100MHz. 30 dN is 1/1000th I feel very confident in the outcome of these tests that the testing method that I described by extending till failure and then reducing the length by 20% or ensuring that no more than about 24 dB is lost, Will give you reliable and dependable signal over the long term. When I designed the other sound tools cables, The design goal was durability coilability and optimization for analog over cat while maintaining shielded Cat 5e specs. For AES 50 that means they should be good to about 75 to 80 m. For this red cable I specifically focused and worked with the cable design engineers to make it optimal for AES 50 and other formats that require less loss at high frequencies in the hundred megahertz and beyond
when you just shield the positive, and on balanced so both balanced wires r shielded could you not use a resistor on the shield to overcome the capacitance?
No the capacitance is created by two wires next to each other and also created by wires next to the shield. Adding resistors will either increase your loss or reduce the quality of your shielding
Cool, so this is stranded too? I missed that part. I assume it works even better with Dante. I’m very curious how it works with the DiGiCo MADI used with the SD9 and SD11. DiGiCo’s cable maxes out at 75m. Its not network through. Would be great to get it to 150m as 1 continuous cable.
Yes it is stranded. I will do maximum length tests with digico soon and let you know the results. As far as lengths, we need buffer length between the maximum functional and the realistic safe working length. For this my goal is to have a sound tools guaranteed length which is 110 m for AES 50 and will be determined for the other protocols. And then have an off menu safe length which for AES50 is 120 m. Beyond that we start getting into concerned with environmental factors and wear and tear over time being a concern that could jeopardize your show. To get those links we chose 100 m Plus a 10 m extension to do the 110 And for the 120 using two 10 m extensions or 2 x 60 m cables That's the safe way to achieve. I will consider doing a 110 m continuous length cable possibly a 120 but we need some time to consider it and we'll see what demand dictates
Nice. As far as the Ethercon couplers you are using. Do they pass continuity (bonding) through as well? Are all the Ethercon connectors used in the cables in your series bonded to each other? When you used more than 2 couplers did you normalize for that? I've seen and measured cables that although they have ethercon connectors they are only connected to ground on one end, so they are not bonded to each other. Ethercon's bonded to each other helps a lot in reducing ESD. I know the Midas M32 Mixers were spec'd that they will handle a longer length of cat cable than the X32's are. (at least they were the last time I looked) Not sure why but you may be able to go a little longer using a Midas M32 with your new cable ;-) Something to keep in mind.
Here is a link to the cat6 couplers www.neutrik.com/en/product/ne8ffx6-w I didn't use bonded pairs in the design because it made the cable too rigid And I was able to achieve better specs that exceed what I was initially shooting for using stranded conductor wire and without using bonded pairs All of the cables we make are grounded on both ends. I intentionally used the X32 because I wanted worst case scenario. There are a lot of x32s out there and I set out to solve the 100 m issue and allow them to run 120 m safely and reliably with this cable.
@@DaveRat Having the Ethercon's bonded is what gives a better ESD performance. A longer length at the expense of possible less ESD performance may not be the best way to go IMHO. For clarity what does ""All of the cables we make are grounded on both ends."" mean? Couldn't the Ethercons be connected to a shield so they would be bonded to have continuity from one end to the other? If the Ethercons were bonded, I'd say you have a great design for getting 25 or so more meters out of a Cat Cable used on the X32 series. Not bonded you missed the mark for best possible ESD rejection in the design. th-cam.com/video/PJJBhSvagn0/w-d-xo.html
Can you explain in detail what you mean by bonded please. What I have seen, from taking cables apart, is the shield of the Cat cable soldered to the metal jacket of the RJ45 cable. I have not seen any cable that “bonds” the cable shield or the RJ45 to the Neutrik Ethercon. These seem to rely solely on making physical contact with the metal jacket of the RJ45 when they are assembled.
@@johnschalk1271 For me, not speaking for Dave, bonding is when there is the ability to have current flow from one Ethercon connector on one end to the one on the other end. Using a multimeter, you will get a "Beep" to show continuity when you touch one Ethercon connector and the other one on the other end.
In which test? In the hundred meter test where we mathematically calculate out the longest distance that would automatically assume no couplers. As far as the actual terminations on the end, since those have to be there anyway we might as well include them And since we're getting similar results between the three different tests looking at it from three different angles I think we can triangulate what is actual and usable
Hi Dave last week I've experienced an issue on using My 50m KT AES50 cat5e (shielded) supplied with my Midas Pro2 at time of purchase. So I needed to extend my 50m with 30m via DN9610 I also used KT AES50 cat5e but this one was nun - shielded, With struggling to get synchronisation, and testing my cables, I discovered that Extending the 2 AES50 cat5e, it only worked if both where Shielded. Question❓ Why whoud both needed to be the same , either shielded together. Or nun Shielded together and not mixed? Does the Sheilded connected with nun-Shielded then work differently?
Sounds like there could be another issue. What kind of coupler were you using? The neutrik cat6a coupler is much better. The wiring in the cheaper coupler. That said, it is an interesting thing and will test when I get a chance
1304 a belden is an unshielded cat 5e cable that has an insertion loss of 22 dB add 100 m/100 megahertz. So it meets the insertion lost spec for AES 50 but since it's unshielded it's not recommended because it doesn't have protection against electrostatic discharge issues. So yes it should work just fine at 100 m but if you're using X32 or M32 it is vulnerable in static environments and it's got about 10% more loss than the Super Cat XM cable I tested
Absolutely AES50 to box with power cord to fiber to box with power cord then back to AES50 is an option Klark makes a Dn9610 and a rack version as well. So now instead of just running a cable, you have to run a fiber cable, you need 2 ethercon jumpers, 2 power outlets, 2 conversion boxes. For us as a sound vendor doing big shows, it makes a lot more sense to just have one of the spare ethercon cables in the main snake bundle be a cable that will handle AES50 for when a band shows up with an X32/32. You can also convert to Dante with cards in the consoles and other solutions as well. But in my opinion, just having a dependable cable that works, plug and play, no dongles, no extra crap to carry and fail, may be smart move
@@DaveRat Oh, I meant using only a RJ45 connector (with shield) compared to the whole Neutrik etherCON assembly. There seems to be many points of shield continuity to be examined.
Interesting. I guess I could test the cable with the ethercon barrels removed, but what would we be looking to learn? Who travels with a 100m RJ45 snake?
@@DaveRat Looking to learn if susceptibility to electrostatic discharge (ESD) is compromised when EtherCon connectors are not used. Although the long run from FOH to stage is fully EtherCon, there might be RJ45-only jumper cables involved behind permanent rack installations. For example chaining two S16 stage boxes or bringing a unit's AES50 port to the front of a rack panel. New units come with a yellow warning tab covering the AES50 ports stating "STP only! Use shielded cable with EtherCon connectors only." That's were I find the official requirement for EtherCon connectors.
@svenglansberg I believe at their connectors provide a better ground shield but if you have an RJ45 inside the rack and the other con hit something before that that is also grounded you should be fine because the static electricity won't occur inside the rack most likely
so i can tell i have a solid sommer cable and the same works with behringer wing and x32 but the wing is less robust as the x32 as i can only using my 100 meter cable and in the venues its failing all the time..., when my x32 had never an issue
Hello Dave! I work in a college concert hall. I recently heard from touring engineers that all cables should be 2, 4, 8, 16... meters long. and not 1, 5, 10, 20. What do you say about this? Thanks!
People like to have rules. We used 100 m because it is a standard. We made a 10 m piece so you could go to 110 m cuz it was fun to go to 11 just like final tap. And we made 60 m because it's 200 ft and that is a useful length for smaller system providers. Ask your friend why he hates spinal tap
Since sound tools sells products worldwide we tried to use length that transfer well from meters to feet. Typically we do an 8 m cable that transfers to a 25 ft. A 16 m is a 50 ft, A 30 m is 100 ft, 60 is 200 ft And then either a 70 or 80 m cable to cover the 250 ft range. Sin jumps to 100 m which is 328 ft And yes there are standards for europe-ish and the US ish and we put a lot of thought into finding ways to straddle both
It really depends. If you have a cable that's installed that doesn't work and you want to try and make it work then adding a DN 9610 or 9620 set could solve the issue. Conversely if faced with the choice of carrying a 100 m cable or carrying a 100 m cable and two DN9620s and two jumper ethernet cables and supplying power to the 9620s. Of course we would go for the more complex thing because we always love having as many dongles as possible and things in the signal chain to go wrong right! Simply connecting a dependable cable is way overrated Joking aside, adding more crap in line when there's a simpler solution is probably not optimum
They do make signal boosters and since we love to have extra little dongles and wall warts and other things in line they can go wrong and ruin our show those are definitely an option. But also and alternatively developing a solution that does not require dongles and wall warts and additional complexity may be desirable as well
A big impact ist still the loss of the rj45 connectors. Every connection sums up 10m. RJ45 connectors was made for a one time connection, the industrie today ist using that for everyday pluging. A RJ45 connector can also not be better than CAT6e. Maybe we have to use better connections than standard ethernet...
I have been unable to confirm that to be true. I show the couplers induce about 1/10 of 1 dB in insertion loss. Interestingly what I did find causes issues is changing cable types like I did in the video. That tends to degrade things if your extension is a lower grade cable
I will dive into the fiber world at some point right now I'm doing research and considering whether to make our own fiber optic cable and as I travel that path I'll try and share what I learn
@@DaveRat Splicing is easy and the tooling is getting cheaper every year. The trick is handling the mechanical support at the spice in a compact durable way. I’ve got some leftover media converters and SFP modules from Internet installs, I might just daisychain into our AES50 run and see what happens. If not latency (ns?) At least I can verify L2 transport.
Had bought a 30m and had rewound like you had on the test bench and I couldn't get any signal to pass. However when I unwound the cable completely... bingo. Thoughts?
The coiling should not make a difference with quality cable but yes if it is to make a difference it will definitely make things worse and not better. Which is why I test in the coiled configuration because I want to test worst case scenario. If someone has a really long cable but only a short run that day it would be unacceptable for the cable to fail unless it was unrolled.
That I do not know. Nor do I know how long it's been there. But when I dived into this adventure it was helpful to find useful information that aligned with the test I was doing. Because inherently I don't trust their specs either unless I've tested
Hello Dave, since under the hood all audio networking protocols are based on Ethernet the usual limits that apply for computing also will apply for audio routing. Depending on the speed of the connection there will be a maximum distance that can be reached, no matter how good the cable is: More than 100m is very unlikely to work for 1GBit/s speed, which is common in digital audio. Everything above that is better served by using fiber connections. A bridge from Ethernet to fiber and back is not that expensive and then you immediately start with a comfortable 2km lenght limit. The only question is how expensive the fiber cable will be, given that it needs to be protected much better than a copper cable :)
@@DaveRat Alright did it. So yes, it differs from full Ethernet ("uses only layer 1 from the OSI model") but what I did not find is which kind of electrical encoding it uses - often called inofficial layer 0. Ethernet would use a simple thing called Manchester encoding, and that determines how the signal is attenuated over the length. There could be better encodings / signal forms / wave forms that can be sent over the wire to minimize such losses, and that would extend the notorious length limit of 100MBit/s and 1GBit/s connections.
Cool and perhaps there are ways to alter the 10s of thousands of AES50 based consoles and stage boxes out there in the field world wide, that would be a big adventure. Tha manufacturer won't even make a cable that works. Changing what exists is not gonna happen, at least in the near term to mid term Other options involve converting to fiber which adds complexity or using Dante which adds complexity and latency AES50 is awesome for its ultra low latency. So compare the feasibility of modifying, updating or replacing 1,000s and 1,000s of built in AES50 boards, backwards compatibility issues and costs Or Use a cable that works perfectly with all of the existing hardware. Plug and play, reliable, done.
@@DaveRat Oh don't get me wrong, I am not suggesting to changing anything. After all, a standard is a standard, and should not be changed so much, if it is successful. What I want to say is that there is no indication in the standards of AES50 nor Ethernet that indicate that 100m length restriction can be extended too much in a safe way, due to the restrictions of digital encoding on the wire.
Interesting, that is why I did so much testing using multiple methods. Just like with rigging, we design it for a theoretical spec and then do destructive load tests to determine the actual strength, then we use a % of the actual strength as a safe working load. If you change the materials or quality of materials, the outcomes and usable loads or distances change. By using improved materials the tests and specs show that the working length can be increased beyond the lengths achievable using lower quality materials.
Dear mr. dave i think when tested the cable should be spreed all the way because this when problem start to popup if it is raped it will cause less problem
Perhaps but also I tested all the cables on reels, And I compare them to each other. And we already know what the performance is of the KT cable and we already know that the testing of the unshielded cable aligns with the general experience at everyone has. So by establishing a baseline that is trustable we can then test the other cables using that same baseline and derive confidence in the test
@@DaveRat i totally agreed and i always follow all your videos but my concern is when we spread the cable we might have the problem in the first 100 meter , just and idea 🥰
Without extensions the cable actually works at 150 meters. So at 100 meters there is a 50% safety margin. Using most cat 5e cables stop working around 80 or 90 meters , so they use them at 75 meters, a 10 to 20% safety margin. This cable at 110 meters is way way safer than Cat5e shielded at 80 meters!
I am not sure that I get why you would want more than 100 meters (328 feet). At 200 feet from the stage I need binoculars to see detail on stage. Like seriously, if the gig is wanting the FOH to be 250 - 300 feet away from the stage they need to have a budget for a console that uses fiber!(And have a center stage camera that feeds a video monitor for FOH to be able to see.) We need to train venues, promoters, corporate clients, venue designers that cable runs between stage and FOH can not exceed 100 meters (328 feet).
Hmmm, the standard snake length for large shows and large rental companies, is 100 meters. Years ago, the US was 300 feet and Europe was 100 meters but with world touring constancy, about 20 years ago world wide standard became 100 meters. Just finished doing Coachella festival providing sound for 7 of the 8 stages, we used 100 meters snakes on all the stages, but only needed the 100 meter length on main stage, quasar, Sahara tent. Actually Sahara tent needs 110 meters for guest consoles.
@@DaveRat Yes 300 feet max. If memory serves correctly: The shops we pulled gear from for Broadway tours would have 250 foot runs and a 50 foot extension. (Masque, PRG) I would try to just use the 250 foot run. However with Broadway theatres will force the FOH to be at the rear of the orchestra seating level, so with some of the really large theatres we would have no choice but to add the extension to get 300 feet. And we had to use binoculars to be able to see the MKE2 mics on the cast. Again my concern would be someone wanting to run more than 100 meters as I saw you were testing trying to extend past 100 meters in this video. I have had to send and receive a stereo audio feed across the Mandalay Bay convention center for a UFC event and we used fiber.
@mixingjd yes, yes, smaller format shows like Broadway may well be fine with 300 feet and less. Large outdoor shows quite othen need longer runs. We used to use 300 but as we grew to be a larger company we standardized on 328. We have 20 or 30 100 meter main snakes and runs hundreds of audio channels via, Dante, AVB, and other formats. We also have numerous fiver systems and runs 1000s of feet. Our main snakes also have fiber as well. For stage coach festival a few weeks back we had 38 delay clusters covering over 100,000 people. Separate from all that stable and common stuff, AES50 format which Midas uses, struggles with lengths over 80 meters. This cable will carry AES50 reliably 120 meters. Looking beyond what is familiar, you will find many applications where long lengths are common. And the tools needed for larger format or off the beaten path gigs can be quite fun once you get outside the box of a building.
For AES 50 there is a Klark DN9610 and they make a rack mount version as well. These come in handy if you want to add a dongle and another set of devices and spend some extra money and increase complexity to reach your length that you can't reach by using a cable that gets you there
@@DaveRat. I’ve seen that unit. Since I posted my question I’ve found other same-format extenders sold primarily for camera and IP use. Would you consider testing some of these? It would be great to know there might exist some “throw $ at it” solutions.
I will consider it but I haven't heard any complaints or issues with them. The problem from my side is that the location of the repeater would end up somewhere underneath a cable mat with the audience trampling on it or under the stage in an awkward location attached to a extension power cord. So my opinion these are not optimal solutions
@@DaveRat I can see that my use scenario, feeding a recording truck, is far friendlier than your stage to FOH. Power supply isn’t a big issue as these repeaters are so efficient. 5-10W is negligible for battery supply.
You have signal loss with each coupler you put inline to extend the length. You should make custom cable at the desired length then perform your test. Also cables wound up on a spool will also create capacitance in the cable. Unspool it with no over lapping of wires and then test again.
Well, true except I am not looking for best case scenario. I am looking for real world and worst case scenarios. That said, the cat6 coupler shows less than .1 dB of loss My take is that knowing what will happen in the real world, a cable comes up short and needs to be extended is more valuable than the pipe dream of always having the exact perfect cable length.
Hi dave. Im from argentina, cant find any cat5 stp cable. Only utp. Can i work with 50mts utp cat5? Any member for argentina who could help and tell me where to buy stp cat5 cable? I have try mercado libre. Thanks everyone. Great channel dave, thanks for sharing you knowledge!
You can make your own. Go to an electrical wholesaler where you can buy 1000ft box of STP then you can terminate it yourself but make sure you get STP Ethernet connectors. Most computer shops can terminate for you.
Interesting, then how would you explain that the exact same extensions were put on all the cables and we had different results? If the issue was the extensions and couplers why would they have a different effect on different cables? And if they do have a different effect on different cables wouldn't the cables that they affect the most in a negative way still be undesirable?
@@DaveRat I don't consider myself the one that should explain that. The Music Tribe AES50 cable specifications that you refer to in your video indicates the maximum insertion loss (ATT) for the Point-to-Point connection (thus including all terminations, couplers patchboards etc.) to be 24dB @100MHzb per 100m. ATT might differ a bit after replugging because of differing transfer resistance of the terminations. Professional cable have their spec sheets independently measured by a certifying body. Any termination, coupler ore what else added to such a cable (and we need at least 2x RJ45) will change the ATT for the final Point-to-Point ATT. That's all I tried to indicate.
Ok, and I test their cable and all the others with a certification tester exactly as they say to do and show the results in the vid. It says in the music tribe docinemt to test exactly like I test with the fluke dsx5000 certification tester. Since music tribe does not make a 100m shielded cable, to get 100m you must use a coupler. Now there are a bunch of people that are making uninformed, exaggerated and untested assumptions about the impact of the couplers on insertion loss. I will do do a vid and tests on that and show what the couplers actually impact and how much. So far tests have shown the cat6 couplers I am using have an impact of a fraction of a dB.
I think we’re on the same page from the beginning! I don’t mind/care any cable, any coupler, any length as long as it works. For my gigs 75m is enough (and with quality cable & terminations that length will hardly ever cause any problem). For others 75m is too short and then all the variables in the AES50 and equipment specification becomes more and more important. Hence your cable becomes an interesting option.
Very cool and realistically we don't really need a hundred meters for many of the shows but a 100 m main snake is kind of the international standard for upper level pro Audio and what we deploy for most gigs. So that kind of force is us to look for AES50 compatible hundred meter cable, hence this adventure
Well if I do that then the red cable will go even farther. So yes, to be fair to the red cable it may help to use red extensions And if I used the same cable type extensions as the test cable types for all the cables, all the other cables would perform worse. So yes, this test is unfair to the red cable and it is even better than the other cables than the tests show.
Yes Dante can be very useful and the added latency is not typically an issue. The added cost is not significant. And the added complexity is not that big of a hassle. Conversely, having a simple plug and play solution of the year on hand May also be desirable. Especially when a band shows up at a gig with an X32 And they are getting dropouts. Having him use a cable at works rather than carrying spare Dante cards to mount into their console maybe a bit better solution
How about the distance and delay caused by it? I know it's electricity and it's fast, but that is a factor for the electronic signals. That was the problem with running PCIe over 200m in this video: th-cam.com/video/Iz_MPlRCbCQ/w-d-xo.html where LTT are running multiple optical thunderbolt cables - of course it's more complicated with active elements in the chain, but is it of any concern to us with AES50?
Other than the fact that I actually demonstrate in the video that the interconnect works at the various lengths. FYI the music is running back and forth down the AES50 32 times in the vid. But, the delay is called propagation delay in the cable specs and it ranges between 450 to 600 nanoseconds for the various cables. Put digital conversions in such we typically see latencies of 1 to 2 milliseconds. The propagation delay would be 0.0005 ms for 100 m cable or about 1/1000 th of the delay a simple digital conversion would take
There is are spilled coffee grinds in my kitchen Hmmm .... Bulldozer ..... I am guessing you may not be involved with deploying concert sound systems or any application where long cables need to be run and then coiled and re run show after show. Else r commending a 100 meter PVC slinky would not be the go to idea But I could be wrong, if so please send a link to a flexible easy to coil stranded conductor S/FTP Cat8, preferably without those cheap PVC jackets
You really shouldn't be testing with extensions... In computer networking those are just a temporary bodge to be used when nothing else is available. The insertion loss from those couplers is far worse than any characteristic of the cable itself.
That assumes that the goal of the test is to show the best possible outcome. Which is the opposite of what the goal is here. To go here is to show what people are likely to experience in the real world and worst possible scenarios. Also if you look at the various tests you can see that I triangulate the lengths. One of the tests I measure the loss at 100 megahertz 100 m and then mathematically calculate the maximum usable length. Which of course does not include the existence of couplers. I then test real world using couplers. So you can compare the outcomes of those two tests and arrive and informed opinion of what will happen in the real world. I believe that the test that I did are superior to doing tests with dedicated lengths for informing people of what will happen in the field Also based on comparing the two tests of calculating the length versus actually testing the length you can see that the predicted light and the length with couplers comes out to be about the same. I found this to be true in the fact that the couplers that I'm using actually measure about a 0.1 dB loss which is insignificant
esd isnt the same as zapping connectors on equipment.esd is only with static electricity not real electricity. And 100 meter is the standard for ethernet cables. It can go further but do you want to trust the gods that its getting there. The only difference between cat 5, cat 5e and cat 6 is the speed of the data. Has nothing to do with length and quality of the cable you can get cat 5 cable for 1 euro per meter or have one for 100 meter per meter. the fluke tester is the only way to test ethernet cables. If you need huge lengths like 100 meters you should use fiberoptics. I feel signal loss is a bit of a weird standard to test ethernet cable with. There are amazing test standards for ethernet cables.
Wonderful, thank you for voleteering to supply that gear and be on standby on all of our stages for all of our festivals and work out that adventure with every band that shows up with an unanounced X32, M32. Why would we add a cable that works perfectly when we could convert to another format and reprogram the artist X32 and setup for Dante during set change. That sounds way more fun! Actually if we are into the fun of adding dongles and complexity, Klark makes an aes50 to fiber converter that is more plug and play than the Dante idea. Why is it that actually using a cable that plug and play works is not just "wow, that's simple and easy, I will do that" Why is it all about "ok, if we convert it to this and program it to do that and change the setup that they are familiar with" yeah, that should be good, and we can have fun hooking it all up during set change yay! Joke ng aside, Dante is great and we run many miles of fiber with AVB. We have racks that accept and send in just about any format and will have 30 delay cluster s and 7 stages running over the next 3 weekends. And Having a cable that solves interconnecting the X32 M32 consoles that show up unanounced at our stages with FOH over 100m snake run away, is the simplest, fastest, easiest and most reliable solution.
Thank you so much Dave! I've argued this ad nauseum on a couple Facebook groups. Networking is my dayjob and I basically live with a DSX-5000 attached to my hip. I've tried to explain this and been told I'm wrong. Has to be CAT5e and you can't use better. The one thing I will say is you're increasing your loss with every connector, but really who keeps all those different lengths on hand to test. I would say you can fail the length but if is one actual cable it would probably work.... but as I typed this you talked about increasing the insertion loss because it was multiple extensions. Once again Dave, thanks for going total geek with a tester that most audio people just wouldn't buy because of the cost and need.
Thank you so much and it is so refreshing to meet people that understand and get it! I really tried to triangulate this from multiple angles and not leave much room for doubts
So glad this is pinned at the top.
It matters, as does the crimping of the cable if you make them yourself.
First thing I questioned myself: "how will you prove actual usable distance of a single run of any cable when you are increasing the loss with EACH separate connection?"
I solder the shield drain to the J-45 to ensure continuity, knowing the shell of an ethercon may not do the job itself through friction alone.
I didn't set out to show the best possible results. This test is about showing worst case scenarios when people extend cables out and what to expect and what they can trust.
Knowing what a specific length of cable can do and not having that length available to you is not of significant value when you're setting up a show
I'm not surprised you've had to argue this ad nauseum. My day job is in networking too, and I've stated the same points as you probably have. Very few people actually understand the properties of cables, yet will stand toe to toe with someone who does and argue with them. Thank you for taking the time to do this Dave. :) It's proven a point I've been trying to make to people for years.
Awesome! And thank you!
Your a true blessing in this industry 🙏 thanks Dave.
👍🤙👍 thank you and honored!
Great video! My background is in RF and audio. One thing you always want to avoid is cable extensions. If at all possible a single run of the appropriate length will work best. Cable extensions add connectors and connectors effect insertion loss far more than the cable itself. In other words two 50 meter cables connected together will always be worse than one 100 meter cable (assuming they both are made from the same type of cable). Cable extensions should only be used when there is no other option and they should never be used in an install situation.
I was interested in that and actually tested the couplers.
Adding a coupler and a short 6-in piece of cat cable showed less than a 0.1 dB differential.
I also compared a 100 m super Cat XM plus two 10 m extensions and two couplers against 2x 60 m super Cat XM cables and a single coupler.
Well the coupler was visible in graphs that showed cable lengths and reflections, the impact on insertion loss was less than .1 dB.
All that said the goal of these tests is not to get the absolute best results possible.
That's the opposite of my intent.
The goal of these tests is to give real world information that is usable and reliable.
In the real world people are going to extend cables to get their show up and running. That's where the valuable information is rather than the ideal situation that we often don't have the luxury to experience.
I came to the comments to find this same observation. It is interesting that Behringer specifies a max length of 100M, but they only sell reels of 50M. It might lead the consumer to believe that it's OK to couple 2 reels to stay within Behringer's spec. I agree that a single run of 100M should give much better performance sans couplers.
I will try and do a vid testing couplers.
Let's we how bad or not couplers really are!
@@DaveRat Consider as well that at installation scenarios they may run the cable in the wall and terminate to wall jacks... So it might be interesting to see if your supercat is plenum rated, and any impact of neutrik wall jacks? Assume one wall jack at FOH and one at the stage... e.g. 75meters in the wall and 3 meter patch at each end
@huktek The Super Cat XM cable is not designed or meant for installation purposes.
Cheaper stiffer plenum rated cable should be used for that. As that less expensive install grade cable is typically solid conductor as well so it's less expensive to make higher bandwidth cables.
It's all about selecting the right tool for the job and using a pro touring cable for installs is not optimal just like using a solid conductor rigid install cable is a bummer for pro touring
Dave! Thank you for this video! I went down the rabbit hole on AES50 and cat cables after a failure on a recent tour. Turns out it was a bad AES50 card in the console. I did hack a hundred feet off my Supercat reel just to be safe. I have major jealousy over your Fluke tester! Thanks again for all you do!
Ah yes and and running a full 100m of SuperCat is cutting it close with aes50. You are in a good safe zone at the 250 length
I priced up the fluke once, I hire it when I need it now and price it into the job
@@DaveRat Would be cool to see the rest of the cables you make go through the Fluke's tests.
@LucasWasson I will ponder that. The issue is that The fluke is focused on cable certification to specific specifications. The sound tools cables are designed for optimizing analog over cat while maintaining Cat5e or better specifications.
That said just yesterday I put together a product grid for the sound tools to cables with the various specifications and product design focuses.
That should be out soon on the sound tools site
@mikeharris6083 Yes they're not cheap
Love this nerd-out so much! I've been saying for the past decade for so that if I knew I had to be a network engineer to do live sound, I might have just gone into networking. Well, touring is mostly more fun I think!
Agreed
I would never use something longer than 10M or 20M in length at home...yet I find this video fascinating.
Awesome and yes we each have our own fun and different challenges for the audio realm we're in.
My SoundTools SuperCat 6A arrived with me in Scotland this morning! Can’t wait to see the difference this makes!
Awesome and please update me as you know more!!
@@DaveRat I’ll make a video about it!
When I installed the M32 at my church 6 years ago I made sure to do my research. I spoke with Midas/Behringer directly and was told to NOT use Cat6. They told me at the time almost every problem they see is with people using Cat6 with their consoles. So we ordered high quality Cat5e and never had a single problem. The length was about 200 ft.
Interesting. Pretty much every cable works at 200ft so you are safe. At around 250 feet many or most cables have issues.
If you watch the video I show how to test your cable and determine how far it will go.
It's important not to just make assumptions that all Cat5e cables have similar performance because they do not, they are vastly different varying from manufacturer to manufacture and even within brands.
Same with all the cat cables
@@DaveRat Exactly right Dave. If I had been closer to the max recommendation range id be testing cables. I did pull 2 spares in as well. Always a good idea to have spares. This comparison you did was very eye opening.
Actually AES50 runs much better over Cat5E than over Cat6, specially one with ferrite on the sides. UTP or FTP doesn't make a difference with AES50
I think the test that I do prove that concept incorrect if you actually watch the video.
Except for maybe cat5e unshielded, which I also show in the video.
But this video tests exactly what you're talking about on shows the results
@LaminarSound awesome and thank you
I will always like every video by this man. Dave, thanks again for the knowledge.
Thank you Brian!
Good evening Dave, I just signed up. Been meaning too for months, but chemo has slowed me down. God bless you sir. ❤heart
Welcome and heal up and overcome!
Thanks for doing this video Dave. There is a huge discussion on Behringer World about cabling and distance, ethercons, etc.. Very useful video. Also, thank you for dropping the Cat5e only bombshell at 2:56. Very useful information/video!
Awesome and thank you!
Love you Dave and the work you do, showing it by being a member for years.
So cool and thank you and much appreciated! My long-form videos are not for everybody in this world of TikTok and such, but it makes me very happy with the imposed appreciated. Cheers to maximum nerds!
Once again another great, informative but unpretentious video from dave, thanks so much.
👍🤙👍
Thanks, for this great video, Dave. It's ALMOST the definitive video on this subject, which has been argued to death on at least two forums that I post on (Behringer World and Live Audio Board). I appreciate very much your taking on the CAT5e ONLY vs. CAT5e OR HIGHER issue. I've never understood why something that's designed to pass signal better is actually worse for a lower signal quality, and have only seen various kinds of hand waving to argue the opposite. I'm not persuaded by "AES 50 is not Ethernet", which is true but higher speed signals needing more efficient cable (if "efficient" is the right word for better signal passing capability) negates that argument for me.
The two things that you're missing, and one needing elaboration:
1) There's a hard limit to how many couplers you can have in your signal path. It's either 3 or 4, IIRC, and when I had a DTX-1800 for a while and connected 4 or 5 cables together with couplers, I got a FAIL for too many couplings, even when using 10' or less CAT5e cables. You are using 2 couplers to adapt to Ethercon, so you're already partway to FAIL. If needing a longer cable, I never add more than one length. To test the length limit, change the add-on cable to longer lengths rather than an assortment of shorter lengths. I'm surprised the tester didn't tell you that, and can't explain why.
2) You mention variable twist rates changing the lengths of the pairs relative to each other, potentially messing up the skew. That would be true if the pairs were consistent twist rates end to end, but a quality cable would have the twist rates vary equally over the length of the cable. So if you open the cable jacket at a given point and see varying twist rates across the pairs, opening it at a different spot in the length would show the same pairs but different twist rates. I have done that and can confirm. I have never examined rock-bottom priced cables for how they implement this feature, if they bother to, so that would be one thing to watch if buying budget cable. Note that this variable twist rate is one of the ways how higher CAT cables achieve their ability to pass faster signals better. You correctly talk about individual pair shields being another way they do that. There are others, too.
3) Regarding the effectiveness of Ethercons for shielding from ESD,
a) you have to have a shielded RJ45s on your cable, with the drain wire solidly connected (soldered, usually?) to that shield,.
b) The simple friction fit from that to a nickel (natural metal color) Ethercon at both ends of the cable seems to be enough to give a good connection to dissipate the ESD.
c) I'm not aware of anyone looking at the black Ethercon shells to see if they make as good a connection. If they are anodized, maybe that's enough? I know nothing about the conductivity of anodization. If they are painted, I wouldn't think that's enough as there'd be a gap between metal and metal.
d) Brian Wynn, who was the first one I'm aware of who figured out the ESD solution, proved conclusively in his videos that you need to use Shielded cable with Ethercons at both ends of the cable. Your comment about maybe being able to use unshielded in a rack or something like that, seems to be inviting problems. Not every day, maybe, but someday.
Those quibbles aside, I think you've done another great service to the community. Thanks!
I will try and do a test of couplers.
Using the term coupler vs defining the type of coupler is as useful as saying cat cable vs the type of cat cable.
8 am not convinced that couplers are as big of an issue as many people make them out to be if Cat6a couples are used.
As far as ESD protection, the exposure of the cable to ESD tends to be outside the rack. Once the cable hosts that rack and the RJ45 is grounded to the rack, all good. Inside the rack the exposure to ESD should be reduced. Same as inside the console.
Ess vulnerability exists inside the console as well but the the exposure is greatly reduced.
This of course relies on the rack and console ethercon s being grounded.
But I will look into testing that as well.
Best I can figure, the real issue is insertion loss. 24 dB, the spec says is ok is huge. That is like less than 1/300th of the signal is making it through.
This tiny amount of signal is very volnerable and based on the cables people tend to use, 30db of loss or more I believe is typical and where AES starts to fail.
That is 1/1000th of the signal at 100Mhz.
Yes, couplers and ethercon shells and breathing on it wrong may cause issues with that and also create a bunch of voodoo and false beliefs as well
Get the losses into the 20 dB range and all that crap goes away. Use a bunch of couplers and no ethercon shells needed is my guess. But will test.
I'm trying to imagine a problem so wide-spread that Music Tribe actually pays attention and responds to it.
Yeah they respond by releasing that document I show in the video but unfortunately did not address it by releasing a cable that will solve it.
Great detail. Nice! Minor detail: some of your graphics show 100 millihertz with the lower case ‘m’ 100mHz instead of 100 megahertz 100MHz. But LOVE your level of explanation.
Ha and thank you
Brilliant, Dave. When this post goes public, I'll be posting it on a few sites/forums.
Very cool and thank you Greg!
Useful, practical guidelines. Right on brand! Thanks, Dave.
Wonderful thank you!
This is quite definitive. If I ever need to be more than a football field away from the stage, I will know what to do. I have a 150 ft cable for my Midas setup, and shielded Cat5e ethercon cable. I feel quite comfortable about it now.
Yes, and for us, I am out at Coachella now and Sahara tent needs every bit of 100 meters to FOH. We actually have the issue quite a bit with guest engineers bringing in X32 and wanting to plug into our industry standard FOH 100 meter snake length.
Great video. Thanks for eliminating the guess work by using data!
Awesome and thank you so cool a long form vid gets watched in today's tic toc world.
I appreciate the time, effort, and research you put into your videos. You do a great job of explaining things so they make sense. Keep up the great content.
So cool and thank you!
Thanks as always Dave! Very informative and entertaining. Love the depth in which you covered this.
👍🤙👍
Top notch! Thanks Dave, I'll be showing this to all my peers.
Super cool and I hope the test method helps reduce show issues and helps the optimal cable selection for whatever application distance
I have here in Hannover, Germany a small club that I look after that I ran 4 cat7 cable with ethercon to XLR BOB‘s, have many different guest desks here as well as the house Midas M32, run is about 30m (x4) and I’ve only ever had a problem with and I solved that by using standard CAT5e as jumpers at both ends, my link cables (BOB to desk/stagebox) are all CAT6
Yes, just about any cheap cat cable will work for 30 Meters. Over 80 meters is where the fun begins
Great test! Would be nice to rule out connector loss as each one introduces loss. Probably would be good to test each of your extension cables on their own to make sure one is not super lossy. Most of all it would be cool to see if a single piece of 150M supercat worked/low insertion loss.
The math calc s action of the video does a good of predicting length. The couplers has such low insertion loss I could not definitively measure it.
But, the extensions do cause some reflections that are visible in the return loss graphs.
Also the extension cables were a lower quality than the red cable. The cable quality of the extensions are a bigger factor and with all red cable we can pass ars50 signal at 150 meters.
All that said, the 24dB insertion loss is a good metric and I triangulated with 3 different methods to factor out things like a coupler being a big issue.
Thanks Dave! Great! I never have to work with such lengths, but it very interesting seeing you testing!
👍🤙👍 cool cool. And yes, most or many people do not need those length.
Interestingly, on the upper levels of pro audio 100meter is the standard snake length!
@@DaveRat Wow! Didn't know that! Thanks for your answer! 🙂
@Edwin-van-der-Putten very cool and yes in the US it used to be 300 ft was the standard and in Europe was 100 m which is 328 ft but now in modern pro audio 328 ft is become the worldwide standard for the most part.
It will vary from company to company but 100 m snake is the norm
Thanks again. You definitely stirred the discussion up, great! You indicated you might do some additional testing. Would be nice if you could show a possible difference between cables with T568A and T568B pinout and show what happens when a T568A terminated cable is coupled with a T568B terminated cable. Just for the fun of showing it. Thanks.
When I get a chance I'll give that a try and if something interesting happens I could do a video on it.
I'm trying to think out what the possible result would be and I guess rather than having the pair with the most twists be the most twists all the way through it would attach to a different pair with less twists and same thing for the least twisted pair would attach to a pair with more twists.
My guess is that the skew would reduce, and the differential between the worst and best pairs would reduce in various measurements.
And the reflection caused by the coupler would show up as it always does
Aw cool shirt man! I just scored some late season pow in New york unexpectedly. Always recharges the soul. Take care!
Thank you and awesome! Love snowboarding!
7:21
"and it's red"
I'm sold!
Why would we want boring cables? Right!
As an electronics engineer I wanted at one point to make an AES-50 to optical fibre converter as a side-project at some point. I never got around to actually doing it though, though I do wonder if there’d be enough interest and possible demand. But theoretically that could go to just hundreds of meters, but easily 10km or more if you ever wanted, and you have the benefit of complete electrical isolation. Basically my idea was to have converters at each end that produce raw Ethernet packets (although it uses cat5, AES isn’t at all Ethernet itself) so you can use standard gigabit Ethernet fibre transceivers. It could theoretically go through Ethernet switches then (which it can’t as raw AES-50), but you’d want to avoid it because it’d probably screw up the timing.
Search out Klark Teknik DN9610.
Holly carp sea man I heard all of the frequency sweep over my headphones and better than my Sennheiser's. I mean flutter pressure and all. I'm a bit upset really. Sure 40 BT wireless headphones I bought on a whim.
Great work on the video. Great spec reveal on the red cable and thanks for the math.
Loved that Fluke sir. Thanks again. I hope that new cable works out.
Super cool and thank you!
If you need even greater ESD protection, CAT 7a and CAT 8 have much better shielding and will provide superior protection as well as give a bit of extra length "AT" (and this is very important) the CAT 5e spec. The only downside is the cost.
awesome video dave! learned a lot ❤
🔧👍🔧
Nice work Dave!. I've also wondered how AES50 works and by looking at the circuit, I can see that it uses two pairs for Ethernet data as they are connected to a LAN8720A PHY but interestingly the other two pairs are connected to LVDS trancievers (65MLVD204), this is why they wont work through a standard network switch.
Very interesting and I think there's some other differentials as well in that AES 50 is not packeted information it's raw data being shipped
AES50 is serial data, not packet data. That's the other reason it won't pass through a network switch.
👍🤙🤙
@@DaveRat I wonder if the differential lines are for frame synchronisation.
That I don't know
Rockin Fluke DTX cable meters!!! That’s the true test set. Not cheap either. I test all my cable on an older version of these Fluke test sets.
Super cool and fun to do! Yeah we've got an older version as well and recently upgraded as we needed a cat 6A capable
I grew up the son of a phone man... all my testing equipment is Fluke too... I've never known any thing else :)
👍🤙👍
Thanks Dave, informative as always!
🤙👍🤙
I just did a gig in Turkey, where the equipment company provided a standard cat5 UTP over carpet in a conduit. Needless to say opening night was a disaster every time a human walked over the cable there was a huge bang from FOH which we traced to static discharges. Took them 3 replacements before they found a F/UTP cable, which while still not to spec solved the problem. Sucksthe AES50 doesn’t even seem to basic CRC confirmation of the transmitted frame.
Oh my. Learning lessons of the hard way
AES50's spec, or, lack of spec clarity, is absolutely infuriating. I've run ethercon, but, for installed stuff, it's far easier/cheaper to just run shielded CAT6A and be careful about how you route it. ESD is a problem, and RF is a problem. No amount of shielding is going to save you if your cable is laying on top of a DMX dimmer basically, gotta use your ol'noodle and avoid problems like that.
Thank you for going over the nitty gritty, and showing what actually works and what actually doesn't.
That said, I've found AES50 in general to be fickle, DANTE is much, much more forgiving and predictable, because it's just using Ethernet standards which are extremely well defined and battle tested. Never had a problem with DANTE that wasn't related to actual hardware failure or egregious misconfiguration. Also, dante will run for as long as your fiber run is, so, multiple miles without issue, if you have the networking gear to handle it. Even something as simple as a cheap gigabit Ethernet switch (provided it doesn't have any power saving features) will get you to 200M without issue.
Agreed. Thing that makes AES50 so cool is the ultra low latency but that's also its weakness and that it doesn't have the error correcting and relies on a cleaner signal than other formats do.
But looking at it the other way, a cable that will run AES50 over 100 m will be stellar for just about everything!
Very cool video and super useful information here! I've run into dropouts exactly once, when I had a 80m cable with me but the stage rack ended up on the wrong side of the stage and we had to extend. They used some super dodgy cable that was way too long but luckily I only had a single dropout of a split second during the show.
I recently bought another Ethercon drum and talked a bit with the manufacturer and they said they actually switch to a higher grade cable for lengths 80m and above. Another cable I had didn't even work at 30m despite being Cat5e spec. I measured it and it turned out to have a DCR of twice the Cat5e spec!
🤙👍🤙
All of our Cat5E reels provided by Midas that came with the Pro1/Pro2 kits are 100m in length. Zero issues with them.
Yeah, are they unshielded cables?
@@DaveRat They are, as the Pro series does call for UTP
Yeah I think that's a difference between the X32/M32 and the pro series is the pro series does not require shielded cable for some reason.
Getting longer lengths out of unshielded cable as you see in the video is not that much of a challenge.
Behringer/Midas does not make a 100 m shielded cable
@@DaveRat Funny enough, I’ve found if you use STP on a Pro series desk, you’ll end up running into weird issues with clock/sync. Sometimes I’ll work, sometimes it’ll never clock up at all, sometimes it’ll pop and click until it drops.
Essentially the same happens on an M32/X32 but in reverse, and there’s the potential for permanent damage. I’ve seen the AES50 cards on the Behringer side of things get killed by using UTP a handful of times.
Yeah, UTP world better but is vulnerable and STP is unstable if it's not of very high quality or short lengths seems to be the issue.
It's possible X32 m32 is lacking ESD protection the pro series has or so I have heard but not tested.
Love this testing. Wonder what the losses are on the extending connectors?
I actually did some measurements on that. One thing that can be done is to measure the difference between the cat6a and cat 5 coupler from neutrik, it actually only shows up as a fraction of a DB of loss.
One way to measure the actual coupler I would need to make a cable of a certain length and then make two cables half the length and then I could measure the impact of a coupler plus two terminations.
I did something similar where I measured 132 m length of cable using 100 m plus a 24 m plus an 8 m and then measured 100 m plus a 32 m.
The impact of the coupler and two terminations comes out to be a fraction of a DB somewhere in the 0.2 to 0.4 dB region.
Your question is of specific interest to me because I really want that SuperCat XM cable to be within spec at 120 m, as soon as I have another 10 m piece I will test it with the dual 10 m extensions to the 100 m, hopefully that reaches the spec but if not, I will look into making a single 20 m extension and see if that hits spec and possibly have those manufactured as well.
Currently we are only looking at making it in 100 m, 10 m, and 60 m lengths
Also I will post the data sheets from the tests on the channel members Telegram channel if you're interested in seeing those
Amazing. Thank you so much!!!
👍🤙👍
Have you done a comparison of the insertion loss of the coupler itself? I was wondering if it would be better to have a single piece of 110 m cable rather than splicing on the extra 10 m with a coupler.
Of course I did! It's less than 0.1 dB though it does affect some of the reflections on the return loss but not enough to make a difference. It seems to add about 6 in to 12 in to the cable when the coupler is in based on the measurements
The that's with the cat6 coupler. The other coupler can be problematic.
The comparison between 60 m cables and one 100 m plus two 10 m extensions shows nearly identical specs.
In Ethernet the max cable length of 100m came from timing not electrical signal attenuation like most people assume. This is to help with collision avoidance and knowing how long it could be for a collision to be seen on the far end. However audio cables probably don't have to deal with multiple stations on a bus. Not sure if the Rx talks back to TX. Haven't looked at how it works from a data link framing perspective. Your ESD test confirms what I'd expect. That is you have a certain noise margin which is signal minus attenuation. If the ESD noise is greater you get a drop. Hence why you want a safety factor. I'd actually say you'd want at least 3dB or more of noise margin but some like to shave closer to the bone. Of course if you go fibre then ESD, potential differences and ground loops are not an issue. Glass is awesome apart from not many audio techs having a fusion splicer on tour. The fluke will test it with the right adaptors.
Yes, we run many miles of fiber and Cat and multiple audio formats at our festivals.
This is focused on getting AES50 to work reliably without adding complexity and additional conversions, latency and potential failure points.
Also, in California the maximum freeway speed is 65 MPH. No one would ever have a reason or dare figure out a way exceed that, right?
Ethernet max cable length was 100m from colision avoidance prespective. Now it is (full duplex links, switches instead of hubs) 100m but based on electrical signal attenuation.
@volodumurkalunyak4651 And also AES50 is a completely different protocol that does not go through switches or hubs and is incompatible with them.
@@DaveRat you inspired me to read some more. Looks like AES50 borrowed layer 1 and some of layer 2 of Ethernet. If I was to guess they've probably pulled the addressing from Ethernet as they designed this to be point to point. That said it might be possible to build active repeaters to increase that cable length to longer distances while blowing out the minimum latency guarantees.
Yes Klark makes a repeater called a dn9610 and a rack version as well.
The issue is that for live sound, the repeater would often need to be somewhere in the audience and adds another layer of complexity and failure point.
Pure quality copper is the most eloquent solution in my opinion.
There has been a lot said about the need to tie the shield to the ethercon barrel. Not sure if it's in scope of any followup video on this, but for an install (with stable power) we pulled FTP, no ethercon, and no issues at 220ft. Perhaps only severe electrical discharge scenarios would require a second conduit at the connector for the shield? Of course Ethercon is needed for touring, but not sure for install...
I haven't tested whether or not the barrel matters.
But my guess is that it does not make that big of a difference whether it's there or not as long as the RJ45 has the metallic housing.
The issue from ESD is related to the magnitude of the ESD versus the magnitude of the signal coming down the cable.
A cable with less loss will be more impervious to ESD naturally.
24 dB of loss which the spec says is the maximum amount of loss is a lot of loss!
That's like 1/3 of 1% of the signal is actually making it to the other end at 100 MHz.
With this tiny little signal things like ESD can easily disrupt it
With something like 14 DB of loss there will be 10 DB more signal which is 10 times the amount of signal showing up the other end.
So my take is yes the barrel could be important if you're running extremely high loss cable and working right on the verge of function.
Which is exactly what would happen if you were using the KT cable or most of the cables that are out there.
So I'm guessing it's a combination of good shielding, low loss and to a lesser extent a tiny impact of the shell of the ethercon
@@DaveRat Excellent breakdown - thank you!
@huktek 🤙🔧🤙
Would like to see you test out the game changer cable. That is designed to allow gigabit and poe 200m. Curios to how that holds up to AES50 spec and your new cable
Just a quick look at their specs tells you most of what we need to know.
It has a 16.1 dB loss at 100 megahertz at 100 m, so maybe 2 or 3 dB or so less loss than the red cable, which is good. Probably calculates out to about 10% farther runs than the red cable does with AES50.
But
It has solid conductor wires and it's fairly easy to get better specs with solid conductors, all good for installs and sucks for portable live sound.
The red cable uses stranded wires which makes it tough to get long lengths.
It's got a PVC jacket which is cheaper,, pretty much the plastic electrical tape is made out o, less durable and does not deal well with cold or hot temps.
The SuperCat XM has a polyurethane jacket, more like what skateboard wheels are made out of.
Looks like a decent install grade cable product line but using it for live shows and such would be less than ideal.
Thank you sir you are a blessing from God.
🤙👍🤙
I come from the RF field. Each connector 2.4GHz and above has an insertion loss value. Was wondering if there is an insertion loss value for each connection when stacking multiple lengths of cables. Interesting video.
Insertion loss seems to measure out at about .1dB for the Cat6 coupler
That said, My goal here is to offer useful information that parallels the real world. So having the insertion loss of the couplers is a valid configuration.
If I was to test this without the couplers in an ideal situation that many are most people will not encounter, I believe that would not be as valuable of information
Dave, Thanks so much for all your very informative videos! Always amazing work! It's so interesting that my team and I never experience this problem with our Pro2 and the 100m unshielded cat5's that it came with. But when we use even 75 meter Elite Core STP cat5e with our M/X32's we get the ESD explosion from hell every other hour or so. We never get intermittent drop outs though (except when the discharge happens). I'm wondering if there are two issues at play that are somewhat unrelated: 1. intermittent connection due to user's cable not being to the spec you speak of in this video, for the length people are running and 2. the ESD that seems to freak out the word clock generator in the mixer when the discharge grounds to the circuit board. What do you think about installing an RJ45 grounding block at least at one end of the cable to redirect the ESD to the ground plug of an outlet? Just to solve the ESD and not the intermittent connection issue, of course. Thoughts?
It seems to be several things going on.
Maybe take a look at this other video I did and let me know what you think.
I go into much more depth and testing.
th-cam.com/video/bQXOXWuxCGA/w-d-xo.htmlsi=46Ma4RwH4BCSfF0e
I believe Cat5e, Cat6, and Cat6A are rated for a max length of 100m, any longer would need a switch. Of course, at that point it just comes down to getting good cable from a reputable brand. In my opinion, If you need more than 100M of any category of ethernet, you should just switch to coaxial or fiber. Well, if possible. And, I guess at that point it might just be worth to switch off of AES50 and go for MADI. Great video, love the testing segments at the end!
Perhaps watching the video will clarify how I am pushing the 100m limit.
But hey the speed limit 65 miles an hour and you never ever go faster than that correct?
It's always better just to follow the rules rather than test them and determine what actually happens and prove confidently whether or not the rules can be broken!
I made the super Cat XM cable specifically to prove that 100 m limit to be breakable.
I did this video to show you how I gained the confidence to know that extending beyond 100 meters is safe
A switch is not applicable to AES50 due to it not being a network protocol. It just happens to use CAT cable but not in the same way as Ethernet.
Agreed
@@timbo19 My point still stands, If you need more than 100m you should be looking away from AES. AES is 20 years old at this point. Broadcast has mostly moved away from it, and a lot of touring acts we get have been using fiber.
Ok, so let's say you are a sound vendor supplying sound systems for large events. 100 meters is the standard main snake length.
Artists often bring their own consoles to events and festivals you supply sound for. Some of those consoles are AES50 based and often, the type of console is not advanced by the artist.
Do you?
A) convince them to buy the proper gear to be able to connect their gear reliably during the 30 minute set change?
B) carry conversion to Dante cards for the various assortment of consoles and help install the hardware during set change?
3) carry some dn9610 units 2 wall warts and run a fiber for them?
4) watch their show go down in pop and drop flames
5) make them mix from side stage
6) say, oh, use the red cable you are good to go
Let's say you are the small band or pa company, what do you carry to deal with shows with long runs?
Is having a cable that solves the problem less or more easy, faster and reliable than converting and converting back?
It depends of course on the application, time available and other factors.
I would say that those that are able to embrace simple fast and reliable solutions over complex and cumbersome solutions are more likely to succeed.
Before i watch anymore, Id say 250' shielded is safe. I used a cat 5 unshielded for a ui24 as a monitor board. logging trucks passing by outside were injecting interference on it from their CBs. Never use unshielded...... Just MyTwo Cents. You ROCK DAVE!
Let me know what you think after you watch the whole vid.
@DaveRat Great Vid. I know what to upgrade to now. You know the science to get things correct. I'm doing larger and larger venues, and the distance is getting critical. Running a mesh tandem wifi is helping on some of it. Rock on Dude!
@soundgod3336 👍🤙👍👍
Love this video! Thanks for getting to some of the finer details behind this. Can your Cat6a couplers be used/backwards compatible with your SuperCat purple cable? Love my supercat...never had a lick of trouble with it. Keep rocking the videos! They are much appreciated!
Thank you and yes the cat6 couplers are the only way to go! There's somewhat waterproof and way better. The other couplers have straight wires inside no shielding and they're junk
Thanks again dave! Amazing data
Awesome thank you Steve!
Given that AES50 is using the 802.3 1000Base-T PHY (ethernet layer 1)for data (plus the additional clock lines), any 1000Base-T pass/fail tests your Fluke unit has for cabling will also be applicable here. Frames are CRC protected which offers some additional noise immunity - Some Fluke etc testers can measure the on-wire bit-error-rate for ethernet frames. If your cable results in received frames with errors close to the limit (1 in 10^-12 erroneous bits or thereabouts) you're asking for trouble. If you want to know what your margin is for a given cable, such test results might be useful to you.
Also, don't forget to factor in the insertion loss of connectors. Connector assemblies/patchbays not appropriately specified or assembled can give you a headache.
Very cool and thank you for the useful information much appreciated!
Great job.
🤙👍🤙
If I recall, weren't the early Midas Pro and Behringer X32 specs contradictory? I'd swear that initially one said use STP (shielded) and the other UTP (un-shielded). Have to look through a pile of saved PDFs to find it.
Well the X32M32 initially was okay with UTP and they changed it to STP after people had issues with static electricity
When Behringer started recommending STP after Brian Wynn's tests in about 2014, Midas still insisted that UTP was fine for its higher end consoles (before they introduced the M series), even though my buddy who had a Pro 3 and I proved that you could disrupt the signal with ESD like Dave was doing in the video. It took a few more sparks to disrupt it than on an X 32, but it was still disrupted. Don't know how long they stuck to that or if they ever changed before they discontinued the Pro series. My buddy talked to them at the time with our results, to no avail.
Yeah they don't want to admit issues as it will create liability for them. They're taking the safe route of ignoring as much as possible
We were having headaches with this issue recently. 300ft Lyxpro Cat 6 shielded ethercon had losses and ESD problems, we bumped it dow to 250ft and it went away. This cable is on Amazon. Our behringer wing is connected to sd16 and s16.
We are having the ESD problem using 250’ of the blue LyxPro cable. We are in ballrooms with carpet so I am wondering if that is causing our issues? So glad to see Dave demo drop outs using 100m of the official KT cable.
Yes to combat ESD you really need double shielding if you're running longer lengths. S/FTP. And you also want low insertion loss. Insertion laws of 24 dB which is the spec means it approximately 1/300 of the signal is actually making it to the other end at 100 megahertz. An insertion loss of 30 dB means 1/1000 of the signal makes it to the other end. With such small amount of signal making it to the other end even the slightest static will cause dropouts
Are the insertion loss specs at 100m for original SuperCat, SuperCat Sound, and SuperCat 7 available?
I believe the 24dB insertion loss lengths at 100mHz are:
75 meters SuperCat
80 meters SuperCat sound
90 meters SuperCat 7
These are calcs based on the 100 m loss measurements.
Have not cut to length and confirmed yet.
Do you plan on finding the maximum safe distance with a single cable of your spec cable and produce a reel for that specific scenario? I think that would be ultimate for the people looking for that level and quality of cable as well!
As I show in the video, based on the mathematical calculation using 100 m supercatXM as the reference, The maximum cable length of SuperCat XM while maintaining a 24 dB maximum insertion loss is 120 m.
When testing the cable and extending it out even with the couplers this ends up being true as well.
The predicted length is 120 m
When I measure 2 x 60 m pieces in a single coupler, I get the predicted losses within a fraction of a dB.
And when I measure 100 m with two 10 m extensions I also get the predicted losses within a fraction of a dB
Everything points to 120 m no matter how we slice the pie.
I'm not seeing any significant degradation from the Cat6 couplers. And they measure less than 0.1dB of insertion loss when inserted.
I'm using a 200-foot Cat6 STP with ethercons for my X32 and S16's and I'm fine with it the way it is. It's been that way since 2013 when I got my X32. This is partially because my old analog whip is 200-feet and so are my other various lifelines between stage and FOH, so I just carried it forward. The specification for the various cables is they really don't want you to exceed 300-feet. The idea is more based around networking, and we typically will use fiber in those longer runs, but that doesn't work for AES50. While you've shown it can work, I'm gonna stay within spec since I don't have any needs to go longer. I still think your tests were interesting and worth doing. Sometimes, that 10m of cable can be what makes the event happen or not.
The touring pro standard cable length for a front of house snake is 100 m which is 328 ft.
Doing festivals with a pro touring snake and having artists come in with X32 or M32 consoles is not uncommon.
So having a 100 m cable capable of carrying AES 32 that they can plug into is extremely useful on the upper levels of pro audio.
Ideally artists would carry higher level and more expensive consoles or not use AES50 but the reality is that they do and accommodating them is important so that they can have quality shows that don't have dropouts.
That is why I designed this cable
Any recommendations for a budget tester to verify signal not just continuity on the cables? Even used the tested music tribe recommends is $2800
I show a method in the video of how to test the cable and determine it safe using length. In the method doesn't cost any money at all you just need the cable to test and whatever units you want to test with
Slightly Offtopic: So we've all been "abusing" several kinds of networkcables for 4 Lines analog or 4 Lines AES runs. What about turning things around ? Why not trying to use an oldschool 4 Lines 110 Ohms AES multicore for network data ? Well ok, it will be quite difficult to crimp it into a Hirose RJ45 connector, but it should work with e.g. Neutrik NE8FAV-YK or something similar with punchdown contacts. I would be very much interested in what your Fluke certifier says about this idea 😊
What if you extend the signal using SFP+ extenders to convert the signal to fiber optic then back to RJ45 copper at the receiving end? You could even get ones which use separate fiber cables for sending and receiving data. And fiber optic cable is not affected by ESD either.
Yes, convert to fiber works.
And also, we all love loads of extra dongles and a few more wall warts or maybe some extra rack gear.
Or
Just plug in a cable that works and no extra crap for less than the cost of the fiber cable.
That said, fiber is great and we use it for nearly all of our snakes and runs to delay clusters.
But for people that own consoles that cost less that $3000 that has an AES50 connection right on the back, a simple relatively low cost plug and play solution that involves buying a cable that works and plugging it in, that may be the way to go for some of them.
We're almost sold out of the first run and the second round of cable will be in shortly. So far the response to the Super Cat XM has been nothing but positive. Everyone seems to love having a high quality high speed reliable table living shores there will be no issues for aes50
How much loss are you getting from each connector/coupler? Your calculated max is pretty close to observed max, but I wonder how much longer you could get from a continuous cable.
I did my best to measure the coupler loss and with that cat6 coupler I'm seeing less than 0.1 dB of insertion loss.
I tried to calculate out about how much farther I could get without the couplers and it comes out to be maybe a foot or two longer
Awesome work to continue this dilemma. When we were messaging back and forth a year ago I had several cables that connected immediately but took 16 min or so to develop a glitch presumably through eventual discharge. Most all them connected right away with the exception of the purple and blue rat cables (no hate here, we’re all learning and experimenting to find our way). Which numbers explain that phenomenon, insertion loss?
Yes in this past year I've done a bunch of research and insertion loss is the magic number. Other factors will affect it but insertion loss is the big one.
A 24 dB insertion loss according to the specification is acceptable and stable and that seems to align with the three different tests that I did.
What's a 24DB insertion loss you have about a 20% buffer which should be more than enough to account for any degradation that happens over time.
24 DB is a lot. 10DB is 1/10th the power. 20dB is 1/100th and 24 is like 1/300 which means about 1/3rd of 1% of the signal at 100MHz.
30 dN is 1/1000th
I feel very confident in the outcome of these tests that the testing method that I described by extending till failure and then reducing the length by 20% or ensuring that no more than about 24 dB is lost, Will give you reliable and dependable signal over the long term.
When I designed the other sound tools cables, The design goal was durability coilability and optimization for analog over cat while maintaining shielded Cat 5e specs.
For AES 50 that means they should be good to about 75 to 80 m.
For this red cable I specifically focused and worked with the cable design engineers to make it optimal for AES 50 and other formats that require less loss at high frequencies in the hundred megahertz and beyond
when you just shield the positive, and on balanced so both balanced wires r shielded could you not use a resistor on the shield to overcome the capacitance?
No the capacitance is created by two wires next to each other and also created by wires next to the shield. Adding resistors will either increase your loss or reduce the quality of your shielding
Any idea if some manufacturer sells in line powered repeaters that are reliable? If not, is this a good product to design, do you think?
Yes there is a klark technic one
Dn9610 and a they have a rack mount one as well
Cool, so this is stranded too? I missed that part. I assume it works even better with Dante. I’m very curious how it works with the DiGiCo MADI used with the SD9 and SD11. DiGiCo’s cable maxes out at 75m. Its not network through. Would be great to get it to 150m as 1 continuous cable.
Yes it is stranded. I will do maximum length tests with digico soon and let you know the results.
As far as lengths, we need buffer length between the maximum functional and the realistic safe working length.
For this my goal is to have a sound tools guaranteed length which is 110 m for AES 50 and will be determined for the other protocols.
And then have an off menu safe length which for AES50 is 120 m.
Beyond that we start getting into concerned with environmental factors and wear and tear over time being a concern that could jeopardize your show.
To get those links we chose 100 m Plus a 10 m extension to do the 110
And for the 120 using two 10 m extensions or 2 x 60 m cables That's the safe way to achieve.
I will consider doing a 110 m continuous length cable possibly a 120 but we need some time to consider it and we'll see what demand dictates
Nice. As far as the Ethercon couplers you are using. Do they pass continuity (bonding) through as well? Are all the Ethercon connectors used in the cables in your series bonded to each other? When you used more than 2 couplers did you normalize for that? I've seen and measured cables that although they have ethercon connectors they are only connected to ground on one end, so they are not bonded to each other. Ethercon's bonded to each other helps a lot in reducing ESD. I know the Midas M32 Mixers were spec'd that they will handle a longer length of cat cable than the X32's are. (at least they were the last time I looked) Not sure why but you may be able to go a little longer using a Midas M32 with your new cable ;-) Something to keep in mind.
Here is a link to the cat6 couplers
www.neutrik.com/en/product/ne8ffx6-w
I didn't use bonded pairs in the design because it made the cable too rigid And I was able to achieve better specs that exceed what I was initially shooting for using stranded conductor wire and without using bonded pairs
All of the cables we make are grounded on both ends.
I intentionally used the X32 because I wanted worst case scenario.
There are a lot of x32s out there and I set out to solve the 100 m issue and allow them to run 120 m safely and reliably with this cable.
🤙👍🤙
@@DaveRat Having the Ethercon's bonded is what gives a better ESD performance. A longer length at the expense of possible less ESD performance may not be the best way to go IMHO. For clarity what does ""All of the cables we make are grounded on both ends."" mean? Couldn't the Ethercons be connected to a shield so they would be bonded to have continuity from one end to the other? If the Ethercons were bonded, I'd say you have a great design for getting 25 or so more meters out of a Cat Cable used on the X32 series. Not bonded you missed the mark for best possible ESD rejection in the design. th-cam.com/video/PJJBhSvagn0/w-d-xo.html
Can you explain in detail what you mean by bonded please. What I have seen, from taking cables apart, is the shield of the Cat cable soldered to the metal jacket of the RJ45 cable. I have not seen any cable that “bonds” the cable shield or the RJ45 to the Neutrik Ethercon. These seem to rely solely on making physical contact with the metal jacket of the RJ45 when they are assembled.
@@johnschalk1271 For me, not speaking for Dave, bonding is when there is the ability to have current flow from one Ethercon connector on one end to the one on the other end. Using a multimeter, you will get a "Beep" to show continuity when you touch one Ethercon connector and the other one on the other end.
Are you factoring in db loss per termination? What’s the longest continuous piece of cable you’ve tried?
In which test? In the hundred meter test where we mathematically calculate out the longest distance that would automatically assume no couplers.
As far as the actual terminations on the end, since those have to be there anyway we might as well include them
And since we're getting similar results between the three different tests looking at it from three different angles I think we can triangulate what is actual and usable
Hi Dave last week I've experienced an issue on using My 50m KT AES50 cat5e (shielded) supplied with my Midas Pro2 at time of purchase.
So I needed to extend my 50m with 30m via DN9610 I also used KT AES50 cat5e but this one was nun - shielded,
With struggling to get synchronisation, and testing my cables, I discovered that Extending the 2 AES50 cat5e, it only worked if both where Shielded.
Question❓ Why whoud both needed to be the same , either shielded together. Or nun Shielded together and not mixed? Does the Sheilded connected with nun-Shielded then work differently?
Sounds like there could be another issue. What kind of coupler were you using? The neutrik cat6a coupler is much better. The wiring in the cheaper coupler.
That said, it is an interesting thing and will test when I get a chance
Could you measure Belden 1304A ? I have seen best performance (length) with this specific cable
1304 a belden is an unshielded cat 5e cable that has an insertion loss of 22 dB add 100 m/100 megahertz. So it meets the insertion lost spec for AES 50 but since it's unshielded it's not recommended because it doesn't have protection against electrostatic discharge issues.
So yes it should work just fine at 100 m but if you're using X32 or M32 it is vulnerable in static environments and it's got about 10% more loss than the Super Cat XM cable I tested
Will Ethernet to optical fiber converters work with AES50?
Absolutely
AES50 to box with power cord to fiber to box with power cord then back to AES50 is an option Klark makes a Dn9610 and a rack version as well.
So now instead of just running a cable, you have to run a fiber cable, you need 2 ethercon jumpers, 2 power outlets, 2 conversion boxes.
For us as a sound vendor doing big shows, it makes a lot more sense to just have one of the spare ethercon cables in the main snake bundle be a cable that will handle AES50 for when a band shows up with an X32/32.
You can also convert to Dante with cards in the consoles and other solutions as well.
But in my opinion, just having a dependable cable that works, plug and play, no dongles, no extra crap to carry and fail, may be smart move
Would be great to also examine the requirement for etherCON barrels.
Those Neutrik Cat6 ethercon barrels measure less than 0.1 dB of insertion loss on the Fluke tester
@@DaveRat Oh, I meant using only a RJ45 connector (with shield) compared to the whole Neutrik etherCON assembly. There seems to be many points of shield continuity to be examined.
Interesting. I guess I could test the cable with the ethercon barrels removed, but what would we be looking to learn?
Who travels with a 100m RJ45 snake?
@@DaveRat
Looking to learn if susceptibility to electrostatic discharge (ESD) is compromised when EtherCon connectors are not used. Although the long run from FOH to stage is fully EtherCon, there might be RJ45-only jumper cables involved behind permanent rack installations. For example chaining two S16 stage boxes or bringing a unit's AES50 port to the front of a rack panel.
New units come with a yellow warning tab covering the AES50 ports stating "STP only! Use shielded cable with EtherCon connectors only." That's were I find the official requirement for EtherCon connectors.
@svenglansberg I believe at their connectors provide a better ground shield but if you have an RJ45 inside the rack and the other con hit something before that that is also grounded you should be fine because the static electricity won't occur inside the rack most likely
so i can tell i have a solid sommer cable and the same works with behringer wing and x32 but the wing is less robust as the x32 as i can only using my 100 meter cable and in the venues its failing all the time..., when my x32 had never an issue
Interesting.
Hello Dave!
I work in a college concert hall. I recently heard from touring engineers that all cables should be 2, 4, 8, 16... meters long. and not 1, 5, 10, 20. What do you say about this?
Thanks!
People like to have rules. We used 100 m because it is a standard. We made a 10 m piece so you could go to 110 m cuz it was fun to go to 11 just like final tap.
And we made 60 m because it's 200 ft and that is a useful length for smaller system providers.
Ask your friend why he hates spinal tap
@@DaveRat Got it! Thanks!
Since sound tools sells products worldwide we tried to use length that transfer well from meters to feet.
Typically we do an 8 m cable that transfers to a 25 ft. A 16 m is a 50 ft, A 30 m is 100 ft, 60 is 200 ft And then either a 70 or 80 m cable to cover the 250 ft range. Sin jumps to 100 m which is 328 ft
And yes there are standards for europe-ish and the US ish and we put a lot of thought into finding ways to straddle both
Dave, What are your thoughts on the Klark DN9620 AES extender?
It really depends. If you have a cable that's installed that doesn't work and you want to try and make it work then adding a DN 9610 or 9620 set could solve the issue.
Conversely if faced with the choice of carrying a 100 m cable or carrying a 100 m cable and two DN9620s and two jumper ethernet cables and supplying power to the 9620s.
Of course we would go for the more complex thing because we always love having as many dongles as possible and things in the signal chain to go wrong right!
Simply connecting a dependable cable is way overrated
Joking aside, adding more crap in line when there's a simpler solution is probably not optimum
Thanks Dave.
@binny911 🔧👍🔧
Should've tried cat3 cable for the giggles 😂
If I had one I would have done so.
Someone needs to make a signal booster
They do make signal boosters and since we love to have extra little dongles and wall warts and other things in line they can go wrong and ruin our show those are definitely an option.
But also and alternatively developing a solution that does not require dongles and wall warts and additional complexity may be desirable as well
A big impact ist still the loss of the rj45 connectors. Every connection sums up 10m.
RJ45 connectors was made for a one time connection, the industrie today ist using that for everyday pluging. A RJ45 connector can also not be better than CAT6e. Maybe we have to use better connections than standard ethernet...
I have been unable to confirm that to be true. I show the couplers induce about 1/10 of 1 dB in insertion loss. Interestingly what I did find causes issues is changing cable types like I did in the video. That tends to degrade things if your extension is a lower grade cable
Can you also test simple SMF fiber media converters and if it works check for latency?
I will dive into the fiber world at some point right now I'm doing research and considering whether to make our own fiber optic cable and as I travel that path I'll try and share what I learn
@@DaveRat Splicing is easy and the tooling is getting cheaper every year. The trick is handling the mechanical support at the spice in a compact durable way.
I’ve got some leftover media converters and SFP modules from Internet installs, I might just daisychain into our AES50 run and see what happens.
If not latency (ns?) At least I can verify L2 transport.
Had bought a 30m and had rewound like you had on the test bench and I couldn't get any signal to pass. However when I unwound the cable completely... bingo. Thoughts?
The coiling should not make a difference with quality cable but yes if it is to make a difference it will definitely make things worse and not better.
Which is why I test in the coiled configuration because I want to test worst case scenario.
If someone has a really long cable but only a short run that day it would be unacceptable for the cable to fail unless it was unrolled.
@DaveRat ok, thanks Dave. 😊
👍🤙👍
Legend :)
👍🤙👍
Thanks Dave, so why has it taken Music tribe so F**king long to produce this spec?
That I do not know. Nor do I know how long it's been there. But when I dived into this adventure it was helpful to find useful information that aligned with the test I was doing. Because inherently I don't trust their specs either unless I've tested
Hello Dave, since under the hood all audio networking protocols are based on Ethernet the usual limits that apply for computing also will apply for audio routing. Depending on the speed of the connection there will be a maximum distance that can be reached, no matter how good the cable is: More than 100m is very unlikely to work for 1GBit/s speed, which is common in digital audio. Everything above that is better served by using fiber connections.
A bridge from Ethernet to fiber and back is not that expensive and then you immediately start with a comfortable 2km lenght limit. The only question is how expensive the fiber cable will be, given that it needs to be protected much better than a copper cable :)
Hmmm, maybe look up the specs and information on AES 50.
@@DaveRat Alright did it. So yes, it differs from full Ethernet ("uses only layer 1 from the OSI model") but what I did not find is which kind of electrical encoding it uses - often called inofficial layer 0. Ethernet would use a simple thing called Manchester encoding, and that determines how the signal is attenuated over the length. There could be better encodings / signal forms / wave forms that can be sent over the wire to minimize such losses, and that would extend the notorious length limit of 100MBit/s and 1GBit/s connections.
Cool and perhaps there are ways to alter the 10s of thousands of AES50 based consoles and stage boxes out there in the field world wide, that would be a big adventure.
Tha manufacturer won't even make a cable that works.
Changing what exists is not gonna happen, at least in the near term to mid term
Other options involve converting to fiber which adds complexity or using Dante which adds complexity and latency
AES50 is awesome for its ultra low latency.
So compare the feasibility of modifying, updating or replacing 1,000s and 1,000s of built in AES50 boards, backwards compatibility issues and costs
Or
Use a cable that works perfectly with all of the existing hardware. Plug and play, reliable, done.
@@DaveRat Oh don't get me wrong, I am not suggesting to changing anything. After all, a standard is a standard, and should not be changed so much, if it is successful. What I want to say is that there is no indication in the standards of AES50 nor Ethernet that indicate that 100m length restriction can be extended too much in a safe way, due to the restrictions of digital encoding on the wire.
Interesting, that is why I did so much testing using multiple methods.
Just like with rigging, we design it for a theoretical spec and then do destructive load tests to determine the actual strength, then we use a % of the actual strength as a safe working load.
If you change the materials or quality of materials, the outcomes and usable loads or distances change.
By using improved materials the tests and specs show that the working length can be increased beyond the lengths achievable using lower quality materials.
Dear mr. dave i think when tested the cable should be spreed all the way because this when problem start to popup if it is raped it will cause less problem
Perhaps but also I tested all the cables on reels, And I compare them to each other. And we already know what the performance is of the KT cable and we already know that the testing of the unshielded cable aligns with the general experience at everyone has.
So by establishing a baseline that is trustable we can then test the other cables using that same baseline and derive confidence in the test
@@DaveRat i totally agreed and i always follow all your videos but my concern is when we spread the cable we might have the problem in the first 100 meter , just and idea 🥰
Without extensions the cable actually works at 150 meters. So at 100 meters there is a 50% safety margin. Using most cat 5e cables stop working around 80 or 90 meters , so they use them at 75 meters, a 10 to 20% safety margin.
This cable at 110 meters is way way safer than Cat5e shielded at 80 meters!
Mister Hand would be proud.
Friend of yours? Club you are in? Does not sound relevant
I am not sure that I get why you would want more than 100 meters (328 feet). At 200 feet from the stage I need binoculars to see detail on stage. Like seriously, if the gig is wanting the FOH to be 250 - 300 feet away from the stage they need to have a budget for a console that uses fiber!(And have a center stage camera that feeds a video monitor for FOH to be able to see.) We need to train venues, promoters, corporate clients, venue designers that cable runs between stage and FOH can not exceed 100 meters (328 feet).
Hmmm, the standard snake length for large shows and large rental companies, is 100 meters.
Years ago, the US was 300 feet and Europe was 100 meters but with world touring constancy, about 20 years ago world wide standard became 100 meters.
Just finished doing Coachella festival providing sound for 7 of the 8 stages, we used 100 meters snakes on all the stages, but only needed the 100 meter length on main stage, quasar, Sahara tent. Actually Sahara tent needs 110 meters for guest consoles.
@@DaveRat Yes 300 feet max. If memory serves correctly: The shops we pulled gear from for Broadway tours would have 250 foot runs and a 50 foot extension. (Masque, PRG) I would try to just use the 250 foot run. However with Broadway theatres will force the FOH to be at the rear of the orchestra seating level, so with some of the really large theatres we would have no choice but to add the extension to get 300 feet. And we had to use binoculars to be able to see the MKE2 mics on the cast. Again my concern would be someone wanting to run more than 100 meters as I saw you were testing trying to extend past 100 meters in this video. I have had to send and receive a stereo audio feed across the Mandalay Bay convention center for a UFC event and we used fiber.
@mixingjd yes, yes, smaller format shows like Broadway may well be fine with 300 feet and less. Large outdoor shows quite othen need longer runs.
We used to use 300 but as we grew to be a larger company we standardized on 328. We have 20 or 30 100 meter main snakes and runs hundreds of audio channels via, Dante, AVB, and other formats. We also have numerous fiver systems and runs 1000s of feet. Our main snakes also have fiber as well.
For stage coach festival a few weeks back we had 38 delay clusters covering over 100,000 people.
Separate from all that stable and common stuff, AES50 format which Midas uses, struggles with lengths over 80 meters. This cable will carry AES50 reliably 120 meters.
Looking beyond what is familiar, you will find many applications where long lengths are common.
And the tools needed for larger format or off the beaten path gigs can be quite fun once you get outside the box of a building.
Does anyone sell a dedicated regen/repeater for CAT? I know there are TP to Fiber converters but not everyone can go that route.
For AES 50 there is a Klark DN9610 and they make a rack mount version as well.
These come in handy if you want to add a dongle and another set of devices and spend some extra money and increase complexity to reach your length that you can't reach by using a cable that gets you there
@@DaveRat. I’ve seen that unit. Since I posted my question I’ve found other same-format extenders sold primarily for camera and IP use. Would you consider testing some of these? It would be great to know there might exist some “throw $ at it” solutions.
I will consider it but I haven't heard any complaints or issues with them. The problem from my side is that the location of the repeater would end up somewhere underneath a cable mat with the audience trampling on it or under the stage in an awkward location attached to a extension power cord.
So my opinion these are not optimal solutions
I did not realize how affordable the Klark portable version is. That’s an excellent solution.
@@DaveRat I can see that my use scenario, feeding a recording truck, is far friendlier than your stage to FOH. Power supply isn’t a big issue as these repeaters are so efficient. 5-10W is negligible for battery supply.
You have signal loss with each coupler you put inline to extend the length. You should make custom cable at the desired length then perform your test. Also cables wound up on a spool will also create capacitance in the cable. Unspool it with no over lapping of wires and then test again.
Well, true except I am not looking for best case scenario. I am looking for real world and worst case scenarios.
That said, the cat6 coupler shows less than .1 dB of loss
My take is that knowing what will happen in the real world, a cable comes up short and needs to be extended is more valuable than the pipe dream of always having the exact perfect cable length.
Hi dave. Im from argentina, cant find any cat5 stp cable. Only utp. Can i work with 50mts utp cat5? Any member for argentina who could help and tell me where to buy stp cat5 cable? I have try mercado libre. Thanks everyone. Great channel dave, thanks for sharing you knowledge!
Ok, be sure to get Cat5e or better, not Cat5. Cat6a, cat6a or cat 7 will work.
@@DaveRat thanks dave, didnt kmow that. Ill try cat6 or 7
👍🔧👍
You can make your own. Go to an electrical wholesaler where you can buy 1000ft box of STP then you can terminate it yourself but make sure you get STP Ethernet connectors. Most computer shops can terminate for you.
@@46I37 thank you very much!
you doin coachella this brutha?
Yes headed there now
Some cables might have just failed because of the insertion loss generated by the cable couplers used for the test.
Interesting, then how would you explain that the exact same extensions were put on all the cables and we had different results?
If the issue was the extensions and couplers why would they have a different effect on different cables?
And if they do have a different effect on different cables wouldn't the cables that they affect the most in a negative way still be undesirable?
@@DaveRat I don't consider myself the one that should explain that. The Music Tribe AES50 cable specifications that you refer to in your video indicates the maximum insertion loss (ATT) for the Point-to-Point connection (thus including all terminations, couplers patchboards etc.) to be 24dB @100MHzb per 100m. ATT might differ a bit after replugging because of differing transfer resistance of the terminations. Professional cable have their spec sheets independently measured by a certifying body. Any termination, coupler ore what else added to such a cable (and we need at least 2x RJ45) will change the ATT for the final Point-to-Point ATT. That's all I tried to indicate.
Ok, and I test their cable and all the others with a certification tester exactly as they say to do and show the results in the vid.
It says in the music tribe docinemt to test exactly like I test with the fluke dsx5000 certification tester.
Since music tribe does not make a 100m shielded cable, to get 100m you must use a coupler.
Now there are a bunch of people that are making uninformed, exaggerated and untested assumptions about the impact of the couplers on insertion loss.
I will do do a vid and tests on that and show what the couplers actually impact and how much.
So far tests have shown the cat6 couplers I am using have an impact of a fraction of a dB.
I think we’re on the same page from the beginning! I don’t mind/care any cable, any coupler, any length as long as it works. For my gigs 75m is enough (and with quality cable & terminations that length will hardly ever cause any problem). For others 75m is too short and then all the variables in the AES50 and equipment specification becomes more and more important. Hence your cable becomes an interesting option.
Very cool and realistically we don't really need a hundred meters for many of the shows but a 100 m main snake is kind of the international standard for upper level pro Audio and what we deploy for most gigs.
So that kind of force is us to look for AES50 compatible hundred meter cable, hence this adventure
But when you extend the RED 6A cable you need to be using the same red cable to extend past the 100M point to be fare.
Well if I do that then the red cable will go even farther.
So yes, to be fair to the red cable it may help to use red extensions
And if I used the same cable type extensions as the test cable types for all the cables, all the other cables would perform worse.
So yes, this test is unfair to the red cable and it is even better than the other cables than the tests show.
It's worth the money to just go with Dante cards on stage and FOH to eliminate the use of aes50.
Yes Dante can be very useful and the added latency is not typically an issue. The added cost is not significant. And the added complexity is not that big of a hassle.
Conversely, having a simple plug and play solution of the year on hand May also be desirable.
Especially when a band shows up at a gig with an X32 And they are getting dropouts. Having him use a cable at works rather than carrying spare Dante cards to mount into their console maybe a bit better solution
How about the distance and delay caused by it? I know it's electricity and it's fast, but that is a factor for the electronic signals. That was the problem with running PCIe over 200m in this video: th-cam.com/video/Iz_MPlRCbCQ/w-d-xo.html where LTT are running multiple optical thunderbolt cables - of course it's more complicated with active elements in the chain, but is it of any concern to us with AES50?
Other than the fact that I actually demonstrate in the video that the interconnect works at the various lengths. FYI the music is running back and forth down the AES50 32 times in the vid.
But, the delay is called propagation delay in the cable specs and it ranges between 450 to 600 nanoseconds for the various cables.
Put digital conversions in such we typically see latencies of 1 to 2 milliseconds.
The propagation delay would be 0.0005 ms for 100 m cable or about 1/1000 th of the delay a simple digital conversion would take
hmm... SFTP cat8......
There is are spilled coffee grinds in my kitchen
Hmmm .... Bulldozer .....
I am guessing you may not be involved with deploying concert sound systems or any application where long cables need to be run and then coiled and re run show after show. Else r commending a 100 meter PVC slinky would not be the go to idea
But I could be wrong, if so please send a link to a flexible easy to coil stranded conductor S/FTP Cat8, preferably without those cheap PVC jackets
You really shouldn't be testing with extensions... In computer networking those are just a temporary bodge to be used when nothing else is available. The insertion loss from those couplers is far worse than any characteristic of the cable itself.
That assumes that the goal of the test is to show the best possible outcome.
Which is the opposite of what the goal is here.
To go here is to show what people are likely to experience in the real world and worst possible scenarios.
Also if you look at the various tests you can see that I triangulate the lengths. One of the tests I measure the loss at 100 megahertz 100 m and then mathematically calculate the maximum usable length. Which of course does not include the existence of couplers.
I then test real world using couplers.
So you can compare the outcomes of those two tests and arrive and informed opinion of what will happen in the real world.
I believe that the test that I did are superior to doing tests with dedicated lengths for informing people of what will happen in the field
Also based on comparing the two tests of calculating the length versus actually testing the length you can see that the predicted light and the length with couplers comes out to be about the same. I found this to be true in the fact that the couplers that I'm using actually measure about a 0.1 dB loss which is insignificant
esd isnt the same as zapping connectors on equipment.esd is only with static electricity not real electricity. And 100 meter is the standard for ethernet cables. It can go further but do you want to trust the gods that its getting there. The only difference between cat 5, cat 5e and cat 6 is the speed of the data. Has nothing to do with length and quality of the cable you can get cat 5 cable for 1 euro per meter or have one for 100 meter per meter. the fluke tester is the only way to test ethernet cables. If you need huge lengths like 100 meters you should use fiberoptics. I feel signal loss is a bit of a weird standard to test ethernet cable with. There are amazing test standards for ethernet cables.
Interesting and after you actually watch the video, let me know if you still hold those beliefs.
if you really need more than 100M, just run Ethernet/Dante over fiber and call it a day.
Wonderful, thank you for voleteering to supply that gear and be on standby on all of our stages for all of our festivals and work out that adventure with every band that shows up with an unanounced X32, M32.
Why would we add a cable that works perfectly when we could convert to another format and reprogram the artist X32 and setup for Dante during set change. That sounds way more fun!
Actually if we are into the fun of adding dongles and complexity, Klark makes an aes50 to fiber converter that is more plug and play than the Dante idea.
Why is it that actually using a cable that plug and play works is not just "wow, that's simple and easy, I will do that"
Why is it all about "ok, if we convert it to this and program it to do that and change the setup that they are familiar with" yeah, that should be good, and we can have fun hooking it all up during set change yay!
Joke ng aside, Dante is great and we run many miles of fiber with AVB. We have racks that accept and send in just about any format and will have 30 delay cluster s and 7 stages running over the next 3 weekends.
And
Having a cable that solves interconnecting the X32 M32 consoles that show up unanounced at our stages with FOH over 100m snake run away, is the simplest, fastest, easiest and most reliable solution.
@@DaveRat Thanks for the reply, it gave me a chuckle. I love your videos :D