⚠A FEW POINTS (And 2 Month Update)⚠ • I'm not pulling the cell tower theory out of my ass like a conspiracy theorist, look at what it has done to my audio recordings at 5:55 and 6:12 (and was solved with ferrite cores, so not like it was a ground loop) • While I'm still not fully sure whether the cable tester is correct or not about the split pair, replacing the cable with one I made myself and running in a different route (directly to the wall across the middle of the room, not behind the desk with all the other wires) DID indeed fix the problem. So really I can see 2 possible explanations. First, the cable tester is right (just inconsistent) and the cable was just badly wired. Or second, the cable tester is wrong about the split pair, the old cable is fine, but there is such a massive amount of RFI and crosstalk it overwhelmed the shielding occasionally. • If my problem is indeed caused by RFI, then this is HIGHLY unlikely to apply to you. Definitely don't think you need some massively shielded cable like the one I bought. ~2 Month Update: • Regardless of the true original problem, the new cable has indeed solved my problem, and I have not had a single internet disconnect since installing it. Several people have proposed another possible explanation which is that the original cable's shielding (or wiring itself) had imperfections such that when I coiled it back up, the internal cabling went back correctly into place, which is why it wasn't generating the same error anymore. I'm thinking that's probably the most likely scenario, since from what I've read, the shielding even in cat6e should be sufficient for most situations. So my point still stands for spending the extra on the cable though, I wanted the absolute best shielding possible, and I went with this particular brand because they actually tested each individual cable, though yes that does substantially increase the cost.
What about the wiring running through the structure? You've got a top line connection coming from the wall, then, there's the wall, everything in the wall, the box outside the home or business, and the rest of wiring leading back to your service provider. It's a fine looking cable, but a solution?
A cell tower won’t transmit at higher power than your cellphone (2Watt). The reason is that adding power to the tower would increase the reception of the towers signal on your phone but not the reception of your phone’s signal at the tower. In other words communication would only improve in one direction and it’s only useful if it improves on both directions. Your own cellphone is therefore much more likely to cause interference at home than a cell tower oven if that tower is nearby. (Signal strength will decrease with the distance squared. It will lose its strength very quickly).
I'm an audiophile, and I noticed my expensive headphones were occasionally (maybe 10% of the time) producing a static-like noise while I was listening to music. Turns out it was the power cable of the amp running parallel to the USB cable feeding the audio data to the DAC. Interference is very real!
@@certainperson reroute cables, increase spacing between cables, replace defective cables, upgrade to shielded cabling if it isn't already, add ferrite beads/toroids would be some solutions to consider.
@@tjmbv8680 A simple cat5 ethernet UTP (Unshielded) cable is perfect for audio. It is not transporting music, it is transporting data packets. But some people like to inject noise into their DAC through the Ethernet port. Best way to do that? A shielded cable! Some expensive "audio grade" ethernet cables do not conform to the standard, they inject noise, so they "sound different" (granted! there is the noise!) but to some people "different" always translates as "better" when they have put a lot of money on it.
At $90, I'm surprised they didn't call it "Monster". Here are some facts about Ethernet cables. 1) Negotiation is done at 10 Mb, so with a defective cable, it may negotiate a higher speed, but fail when it gets there 2) With 10 & 100 Mb, only the green and orange pairs are used, so a cable that works fine at 100 Mb might not work at Gb, which requires all 4 pairs. 3) Gb Ethernet was designed for plain CAT 5 cable. 5e wasn't even available at that time, so plain CAT 5 is good for 100M and 5e just has tighter tolerances. 5e has replaced plain CAT 5. 4) Anything better than CAT 5e for Gb is a waste of money and may be harder to work with. This is noticeable between 5 & 6. 5) The 4 pairs in a cable have different twist rates, to reduce cross talk between pairs. That wouldn't happen if they all had the same twist rate. 6) Shielded cable may help in situations such as yours, where there may be interference. I have installed it a few times in noisy environments, but generally it won't make much difference otherwise. 7) The shield must be connected to ground at only one point, so check the continuity to ground at each end to ensure you're not grounded at both ends.
Think you may have missed a crucial part: try meassuring the cable while bending it, or pulling it. A intermittent connection can be rather confusing, and give different readings!
I was going to add this. Great catch. I've also had issues in my home with the coax / cable modem and a microwave tower near my house. A quick replace of the main splitter in the utility box solved this problem. I was told there are interfering frequencies, but I have not measured this yet and feel it's debatable.
exactly my thoughts, i doubt the cable has ANY problems with even gig connections....when it has a good connection and isnt missing 1 pin when it bends wrong
I used to think all cables were equal and then I spent a summer working as network install technician, and I realized how difficult it could be to get a permanent link test to pass on an ethernet cable
I do ham radio and shielded ethernet cables make a huge difference for interference. On transmit it would take out my NIC until I switched to shielded, and on receive I discovered 10/100 IoT devices produce a lot of noise on the HF spectrum.
this! he bought a nice $90 cable but I could crimp a shielded cat7 or him which would work just as well... biggest issue is just getting the grounded ends.
Back when Commscope first came out with its Ion-E product, instead of coax to carry the RF around, they used Cat 6 Ethernet cable. As the company I worked for documented it's processes for installation they bought one of those Fluke testers and tested pre-made VS customer made to length cables with those fancy connectors you mentioned and there was an amazing range of quality, or lack thereof between pre-made, with the highest quality being ridiculously expensive (at the time). As most of the runs would be 100 meters or less, it was decided we would make our own and then use the tester to confirm they worked properly. One of our engineers made and tested the cables and connectors and several dozen scenarios and found that even if one cable out of the bunch isn't wrapped the same counts, or if it is wrapped too tightly, the throughput performance of the cable will vary drastically. Every one of those cables must be balanced as an individual pair, and in respect to one another. Needless to say it was pure hell finding people with the meticulous skills to install those connectors on those cables on those jobs. But after a few weeks, and lots of practice, it would finally come together. Regarding the tower, just guessing you have T-Mobile/Sprint as they use TDD with is like Wifi on steroids. But the problem is more likely caused by your phone simply due to the proximity of it to your hardware VS that tower out front. It is basic physics and the inverse square law, try your test again but this time with all cellphones and your access point in the area turned off and see if it is coming from the tower or your access point since both switch back and forth from TX to RX many times a second. Happy hunting. PS, I am in the process of transitioning away from copper and going to fiber all over the house. Lightning popped in my backyard a few years ago and wiped out my Ethernet based network so I've been going back with fiber since it isn't impacted nearly as badly by lightning, and if I want to upgrade the speed, I don't have to replace any cabling, I just replace the media converters. :)
I like how you're asking the real question. I'm running MM fiber in my home-office connecting two switches on the opposite side of the room and I'm going to put some sfp+ pcie cards into my main pc + backup and run either mm fiber or DAC to the access switch. Cat8 is like he said overkill, but unlike he said, it just cannot be justified to install it in his house unless he just wanted to have a cat 8 cable.
This is what I was gonna say. I put fiber cards in my computers for like $50 each. And the fiber cable is extremely easy to work with. Doesn't need any special shielding or bulk. The speed (ping) is much faster than electrons. And each connecter uses like 0.7 watts of electricity per port (at each end of the cable) versus 2-5 watts for copper. And it's very thin, and easy to work with -- typically the solid core copper wires he refers to in the video are used inside walls where they're not handled often and don't need to bend around tight spaces like inside a desk or entertainment center. I used point-to-point connections to my NAS until I was lucky enough to pick up a brand new ($1100) 10G fiber switch from a datacenter for $100!! They were moving to Atlanta and selling a lot of equipment so I bought a few servers for like $100 each.
@@AnastasiaValentineNAiden Actually, the signal propagation of electrons in wire, or photons in fibre are in the same ballpark at ~66% the speed of light. This is why Satellite communications are potentially lower latency than undersea fibre as in space a radio wave, or laser pulse can travel at ~99% the speed of light. Fibre may achieve lower latency due to fewer transmission errors, meaning less retransmits, so communications become more reliable due to the immunity from RFI and crosstalk fibre brings. Fibre encoding schemes are also usually simpler, meaning data doesn't need any buffering, waiting to be sent across 4 pairs with all that QAM rubbish, you just line the bits up and send them down the single fibre channel. Another way fibre can reduce latency is say in a line of 5 buildings that form an commercial centre. Each one has a switch, and those are daisy chained. with ~100m links between each one. So the two end buildings have to communicate through 4 switches to see each other. This is fine as the links are already at around the max distance, there's no other way... Enter some cheap SM fibre with 1km range so now you can run point to point links between each building instead of needing to daisy chain them, each building has a direct link to each other building. effectively cutting out the latency. On a large campus of buildings, or an office complex, this can reduce latencies MASSIVELY by taking out switch hops, while also multiplying bandwidth.
Should have just done Fiber.Pre made cabling and SFP suppliers like FS and equipment from Mikrotik make it dirt cheap.Mikrotik recently released their new range of 100Gbit switches recently as well.
I work in the cell tower field. If you’re getting interference, you need to tell the cell company. There should be an identification placard, like a site number or name and possibly a phone number. The other possible option is contacting your city planning department to let them know you’re getting interference. They may be able to help get the carrier fix it.
Do you have any suggestions for how many RF chokes per foot or per cable, or on which types of cables, etc, and what type of choke, is most effective at blocking the interference until the tower is corrected?
Folks would do well to acquaint themselves with the inverse square law relating to radio signal power levels, if they haven’t already. The available RF energy at any point along a straight line away from a radiating antenna will decay by at a minimum of 1/r², where r is the distance. In practice, atmospheric absorption and other obstacles along the path will further reduce the available RF energy. If you look up the licenses for local RF sites (with your local telecom licensing agency) the details of the site tx frequencies, power levels and antenna gain/radiation patterns will likely be documented. With the above in mind, you can get an estimate of how likely a culprit the tower a mile away is likely to be compared to, say, the emissions from your wifi router in the same room.
ferrite cores are awesome! it's a magic "interference be gone" solution. I have at least one on every analog audio cable and on all of the interference-heavy devices' power cord.
Great video. I'm at my current job now for almost 10 years and building CAT 5# and CAT 6 and 6A shielded cables is part of my job. Molex cables as well and I even messed with Fiber Optic fusing. This video is spot on accurate with everything. We sell CAT 8 shielded cables too but not the 22 AWG one you had in this video. Those shielded ends were really nice too. I mentioned this to my boss, who is also one of my best friends from childhood lol. I sent him this video to see if we can experiment with making them where I work. I got all the tools need on my desk as I type this, even the heat shrink tubing haha. Thanks for posting this!! My name is Joe too btw. Great channel!
[ edit: This is all completely wrong] The tech behind balanced audio cables isn't quite the same as twisted pair cabling. It's got the same intention (cancelling out noise) but the actual mechanics of it are different. Just to note. Twisted pair cabling is *not* balanced. Each individual wire strand is carrying different Stuff, whereas in a balanced cable you send data down one wire, and a reversed copy of the data down the other. The diagram at 3:35 is explaining a balanced cable, but not a twisted pair cable.
"Twisted pair cabling is not balanced." Nonsense. It's a differential signal, which means the two wires have opposite sides of the same signal. Ethernet has used differential signalling since the 10baseT days, with 10 Mb on CAT3 cable. From "Ethernet The Definitive Guide, by Charles Spurgeon, 1st ed. Under 1000Base-T Signal Encoding pg. 158 "The differential voltages used on the wire pair swing from approximately zero to +1 volts on the positive wire and from zero to -1 volts on the positive wire."
2:20 that sounds unlikely based on the footage you showed before. I think the insulation somewhere in the cable is broken and the 1,2 pair are connected. The reason why it doesn't have the problem after removal is that the 2 vains aren't touching each other at the broken spot anymore due to moving of the cable.
The method used to reduce interference in pro audio is Common Mode Rejection. In professional audio gear there will be a differential amplifier inside of any gear that uses standard 'balanced cables.
I would argue that common mode rejection is an attribute or desired outcome; the “method used” is differential signaling. By making your desired signal exist in the “differential mode”, everything you want to remove will be in the “common mode” (ie signals that are radiated more or less equally, or common, to the two differential conductors). Common mode rejection ratio (CMRR) is a metric for indicating the effectiveness of cancelling the common mode noise.
Thought I'd post my experience here just in case anyone hadn't thought of it, i was having the EXACT same experience of my connection randomly dying, but then I noticed it would always die if I saturated the read speed of the gigabit network (with a file transfer over LAN). I saw this video and was like, wow I should check all the cables in my network setup. So I spent the whole day swapping out cables and testing it again to see if it would fix it, but wasn't able to. Then I decided to try a different ethernet adapter (I was using a gigabit port in a USB4 hub I had recently purchased), and the problem was gone! So in my case it was the actual ethernet to USB-C adapter I was using. Just something to keep in mind. Also, it was a Realtek RTL8153 chip in a Cable Matters 201308 I'm having trouble with.
I doubt the cable had actual split pairs. What I've seen several times are cables which show as OK on a network tester. Then if the cable is slightly bent or squished the tester shows a split pair or misswire. This is because of poor winding which opens up when the cable is bent, which causes crosstalk. Those basic Ethernet testers interprets it as a split pair or misswire. Edit: For a short (
The added thickness is not just the wire alone, every pair is twisted and shielded. And then all the shielded pares are twisted and shielded again. The larger metal connector now has to accommodate all the individual shielding for each twisted pair. Way overkill but highly efficient with virtually no signal loss or degradation. Even though I would not recommend it you could probably run a cat8 cable parallel with an AC supply wire and not have any issue. But I still would refrain from doing that.
I like the idea of the connector. It's pointless for me and the use I could give it, but it looks so good that I may go and try to build one cable, just to check out that connector.
The problem with the cell tower theory is that its interference would be fairly continuous like on the microphone, not a few times a day. I would be more suspect of something in your house like a compressor (or other large motor) cycling, producing a burst of EMF. For example is the AC compressor or refrigerator on the other side of the wall? The better shielding on the cat 8 cable would help.
I have been using CAT7 for years, for a 30foot run, works great, now for 6+ years! ThioJoe, file a Complaint with the FCC, as they want to know about Problem transmitters..
That's the kind of chonky Ethernet cable we use for some POE devices as well as the controllers of the Microh LED panels of giant LED walls (but we use an ordinary CAT7 cable to connect the wall to the Nova Star controller 🤷).
Is that actual POE. or just some honking DC conductors sharing the same insulation as the STP? I've seen stuff like that before. Or run three ethernet cables. One has the real data... the other two are all just shorted at each end and crimped into a big DC lug each with a red or black heat shrink. Ethernet cable also takes a couple of amps of mains (230 VAC) pretty well, If you're in Europe you even have the right colours, Blue=Neutral, Brown=Live and Green=Earth. The Orange one could be used for something like RS485.... be VERY weary of cutting cable out of a home automation nutters house, cause that's the kind of stuff you're likely to find LOL. Just for the love of god use an ethernet cable that is a different colour to the data cables!
I have dealt with a LOT of industrial level ethernet that is dealing with interference from dozens of noisy motors and drives... I have also dealt with the single side grounding and found industrial equivalents of the cable connector you have shown. The really cool thing is they are clamp on and have special heads for pulling cable or for different plug styles... expensive, but really great stuff when you have difficulties or interference issues.
Audio tech and electronics designer here. To help with interference you want to have as well shielded XLR cable with as good connectors as possible and high quality preamp for that SM7B. SM7B is difficult, low outoput + somewhat high impedance source that many cheaper preamps struggle with. But it's also a case of how one uses the SM7B, it's very low output means that you must be RIGHT ON THE MIC for it to pick up high enough signal and then you don't need as much gain for the same SNR, if you're more than 1 inch away from the mic wind screen, you're too far away. SM7B is an artist mic, voice over mic... voice talent mic. It's not suitable for much else, thus one must know how to use it properly. You can also try fethead or similar j-fet preamplifier (that uses phantom power) to boost the signal very cleanly right at the source and then have the cable run to your interface. This helps with the low quality preamps quite a bit :)
I use a DSX5000 at work, more for industrial M12 connectors than ethernet so usually 10G cat7 cable. They're great for these random and confusing issues, as they can detect NEXT & FEXT, ACR & ACR-F, and graph everything by pairs so you can see exactly what's happening. I think part of the price is for the software, which is designed to test setups for whole buildings and large scale installations and map it all out nicely.
Finally seeing a guy on TH-cam use it. I've replaced my whole house network with cat8 for about a year now. I'd say it's as advertised, 10GB is reachable but the nic is really hot. Also, the only time it reaches 10G is when I use iperf or copy a huge file from my nas.
I've had people telling me for 30yrs now that I don't need more memory, hard drive capacity, etc. Fact of the matter is they were trying to con me into buying the minimum required at the time. I always try and go with the best highest quality that I can afford.
I didn't realize that using an older cable standard would result in intermittent problems. My experience has always been that intermittent problems were due to a bad connection, but I guess that's wrong. I'm going to get one of those cable testers just in case I run into anything like that in the future
It definitely is probably a bad connection in 99% of problems. I really think mine is a special case because of the cell tower which I already know has caused issues with my audio like I show later in the video. Of course I could be wrong about that, but I can't really test it without $10K+ equipment
@@ThioJoe It's probably better to diagnose by elimination than buy $10,000 worth of cable testing equipment. The other option is to bring in a technician who spent the aforementioned equipment in order to figure out what's going on
@@ThioJoe Bad connections are a common problem. I've been a tech for almost 50 years (50 on May 1) and I can say that with certainty. If you have a problem, check the cable and connections first.
MPTL ( modular plug terminated link) connector is what those plugs are . We use them for cat 6 or 6A shielded as well . It’s actually now an official standard for using modular plugs for direct connection to cameras , access points and other devices without having to have a jack and box plus patch cord .
I bought a shielded Cat8e cable (15 m) from a German manufacturer (inLine) that tests their cables(they even give you the test results). Apparently I live in an area close to shipping lanes (I live by the water) and their radars can interfere with my signals. So I got the 8e cables to give me more stability rather than anything to do with speed - if you have 1000 Mbps transmissions - then it won't help anyway. But I have noticed a more stable connection during the day compared to my unshielded CAt6 cable.
Industrial Ethernet is what is used to control automation devices like PLCs. Sometimes called (depending on the application) EtherCat, EtherNet/IP (industrial/protocol), POWERLINK etc.
when you work in the audio industry you become accustomed to 300ft long cat8 cables that are double shielded. someone shut one in a door a few months ago, thats $290 we are never getting back
imo if you need something greater than 10G (ie cat 6/6a) you should just make the jump to multimode fiber. You can probably get affordable used equipment for it on ebay since big companies will get rid of their old equipment when they need an upgrade, and you can reap the benefits of cheap but fast compared to commercial networking equipment
I run multimode fiber between my switches (ironically Dell gigabit models that haven't been sold new for years), including a run that is parallel to an electrical feed. I missed the part about whether the rest of the run was replaced too: Fancy cable connected to a wall jack having regular cable could still be getting interference elsewhere.
Funny that this issue occurred on ps5. For almost a year I’ve been having this issue and I couldn’t find any thing. And last week I’ve been digging in some information about lan cables I found out there is a cat8 and decided to buy one. And GUESS WHAT! The issue is gone. I now get at most 1ms ping during online game play 🙌🏼🙌🏼
Cat 8 is worth the money. Old, small capacity cables that some of us tend to use as patch cables between desktop devices, reduce the capabilities of every device they are connected to, including internet speeds. I use a single piece of 33 meter Cat 8 cable from my 1 Gig fibre gateway running through my attic to my primary router at my desktop iMac in my home office, and there is no noticeable loss of signal strength.
I have used Supra CAT8 for my hifi system and wired to the TV, to shield the systems from electronic noise. Just ordered another to replace a faulty cable from the front to the back of my house
I'd be worried that solid core cable would put a lot of strain on the Ethernet ports due to the lack of flexibility and extra weight pulling down on the cable. Maybe it could be alright if you have a way to secure the cable to the wall or desk near the ports in such a way that it isn't putting force on the ports.
We use industrial Ethernet in our factory. Mostly we have Cat5e and Cat 6, mostly shielded. The distinction of "industrial Etherne" is more about the protocol used and the type of managed switches. Everybody seems to be relabeling some garbage Cisco switches that take all day and require lengthy reboots. I haven't purchased one in a while, so maybe they have made improvements. 1) Ethernet/IP is a poorly named protocol from Allen Bradley and typically used for cyclic and deterministic communications. Messages are sent as fast as 1ms for most I/O functions, but the engineer balances the needs for each I/O device and may set the packet interval at one second, or larger in a huge network. Most of mine are around 10ms to keep traffic down on the master ( , which is called a PLC). For motion applications another protocol is used. There is clock-time synchronization so that all devices will start their work at the same time. For example a master device will command a bunch of servo drives to perform coordinated motions. Since the servo drives are relatively simple, they reply with the motor position and an accurate time stamp of what time the position was taken. The master then performs math and sends new velocity commands to the servo drives with a begin time. These exchanges typically occur at 250 microseconds. On these systems s managed switch must include the feature to add it's own processing time to the time given by the master. On return traffic it does the same thing with data from the servo drives that have a time stamp. 2) Profinet is a Siemens brand protocol used for the same purpose, with some pretty advanced features. 3) EtherCAT is a protocol by Beckhoff that uses a ring network that can be configured also as a straight line where the last device in the string sends data back the way it came. This protocol is not compatible with TCP/IP and therefore a separate network is required. This protocol has two distinctions. It is very fast. And a non-related equipment manufacturer can purchase protocol chips from Beckhoff and instantly they have a certified EtherCAT device without any testing. And the chips are pretty cheap. So a lot of third party companies produce compatible equipment. But you need an EtherCAT master controller. People like the controllers and languages they have previously learned. The cost of trying something different is immense in a short calendar time project. I used a bunch of Beckhoff hardware in the past because they have the largest selection of I/O in the world. And they provide protocol adapters for just about every Ethernet and Serial network type out there. I used their controllers on cost sensitive projects, but the learning curve was fairly big.
The Plugs aren't even CAT7, not to mention CAT8. On the RJ45 Plug the pins are to close and not shielded pairwise to meet CAT7 standards. A solution to this is the GG45 plug, but the tech we use today didn't implement it GG45 ist back compatible to RJ45, it has pairs of pins on its "shoulders" shielded against each other. I am using a Fluke CAT7 verifier btw
@@ThioJoe you're right... I did my homework and there are 2 CAT8 Standarts: 8.1 with RJ45 (compatible with 6A) and 8.2 with GG45 / TERA (compatible with 7A) Thanks
As a first measure I would keep data cables apart from those providing electrical power, especially those providing 110V or 230V (in Europe). As an alternative to expensive cables there are metalized cable housings available which can contain (and tidy up) several data cables. There is no need to continue this kind of housing from port to port because if the major length of the cable in question is protected a short way between the connectors and the housing do not catch enough RFI to get distorted.
I had the same problem with an old cat 4 cable that comes from my router modem upstairs, I replaced it with cat 7 and now it doesn't come down a conduit with 240V ac leads and it's been fine.
We confused our IT cabling folks one time we asked them to test an Ethernet cable and it was stranded. It was used on a robot and the stranded cables can handle the constant bending I learned. I’ve also learned a little about Ethernet Industrial Protocol and those beefy cables are used on factory floors where cost is next to nothing compared to system they’re used on.
Question: I see you made all the measurements on your network infrastructure (cable to wall, the wiring, to patch panel). Did you made a measurement of the signal from the nearby cell tower antennas, if they were within the boundaries of regulation for your location? It would be an interesting follow up.
Had similar interference problems with my TV setup. The combination of mess of wires, wireless signals, newer cell signals, cheap low shielded antenna(with extra long attached coax cable), amplifier on top of noise, movement from people, and the use of regular dual shielded coax cable made it extremely random and irritating when watching TV. I bought a higher range antenna that had the coax connector on it instead of those long, cheap, low shielded ones attached. I paid extra for two extra short quad shielded cables for most of the setup(and cut out at least eight ft of unnecessary cable). I bought a cell signal filter for LTE and 5G signals and put it before the amplifier. I tested the different configurations but only together did I get a perfect signal on all channels. The only thing I didn't do was use a heavily shielded USB cable for the amplifier which has been known to cause problems especially with cheaper cables and power supplies.
This is almost certainly an intermittent physical short in the ethernet cable due to chemical degradation of the insulation between the conductors. The "RFI" you are getting in your audio is not from a cell tower, which is specifically what lead me to conclude your cable problem was just physical conductors shorting. Cell tower signals are in the order of millions of hertz, usually 400 - 700MHz range - well outwith the range of human hearing. Your spectrogram of the audio signal looks a lot more like switching noise generated from switched mode power supply units. This is a common problem for external audio interfaces sadly that isn't unique to high impeadance mics like the SM7B that need a lot of signal gain to drive them. If you disconnect the mic and whack the gain you'll get the same broadband noise with no cable attached to the interface and if you gain up the output volume and amp for your speakers you will hear the noise audibly. It's usually worse if you are on the same circuit as high current draw switching devices, so if you have a beefy GPU and spin up some GPU heavy task you will hear the noise increase in intensity. Ferrites are one solution for this, isolation transformers (DI boxes) are another, expensive USB isolation boxes are another. YMMV.
I use cat8 for my WAN connection because it's likely more durable and gives me a cable to use on a faster connection if need be. It actually wasn't that expensive, though it helps that it was a simple flat cable. For a while, I had weird issues with my WAN connection randomly going to 100Mbps on my old flat cat6 cable. It gave me some weird readings on my cable tester and that's when I replaced it with cat8.
Category 7 and category 8 cables are for shorter distances. Cat5e and Cat6/a will work for longer distances I have a question regarding that cell phone tower have you had any headaches recently
I kept getting network disconnects, same as you mention, with my desktop that is my NAS/Media Server. I have a small fridge about 8ft from it. I eventually figured out that every time the Fridge compressor kicks on or off, the NIC reports disconnect. I had to not only replace the cat6a STP cable with a cat8 S/FTP cable (better shielding), but also, put the fridge on a surge protector that has high EMI/Noise filtering spec (high dbi). Those changes fixed the issue.
Yeah that's a fifth-generation cell tower and the ultra high frequencies they use not only cause issues like this but other stuff like thinning the blood-brain barrier
Guys, this was a nice video and all, but please don't listen to his advice and don't waste your money on Cat8 cables. For distances of 100m or less in 99% of non-datacenter environments, Cat5e or Cat6 will do just fine. Even if there is a cell tower in your backyard.
3:22 You are mixing between twisted pair and differential pair, Twisted pair cancels 99% of the interfence from itself to other cables and itself it the cable was looped, but differential pair is to modulate the signal to two wires in a way that the difference between the voltage will equal to the original signal at the receiving end, thus removing noise.
If you're just going to be using it as a patch cable, I suggest stranded wire... in a wall I'd recommend solid wire. Solid wire is more fragile, as you mentioned
22AWG is very unnecessary for indoor patch cable use. Most cables are 26AWG, with some 24AWG ones. Thin run 30AWG cables are also suitable for in home use, and can handle 10GIG fine. Unshielded (UTP) cables are fine for home use, but I recommend Shielded (STP, SSTP, S/FTP) if you are running any cables near power cords, in-wall electrical, or have anything with a motor, compressor, transformer near the cable. I have used Monoprice brand cables for 15yrs, never had a cable go bad without any sort of obvious clue as to why (physical damage etc). Those CAT8 are meant for data centers 40GB+ networks.
The reason you get that intermittent split pair error is due to the fact that it's really hard to detect. You can't find it by testing for continuity, as each pin is wired to the corresponding pin at the other end. So, you'll need to detect it by the crosstalk introduced by the mismatched pairs. This is essentially extremely short range radio interference. So I'd wager you rather simple tester (in the grand scheme of things, a 'real' CAT-tester is around 5000 dollars) does some simplified version of a split pair test by injecting a pulse and seeing if it get's it back on a conductor it didn't expect. So RFI from an external source could also theoretically trip that detection. Or you cable might actually have a split pair, and your tester is a bit unsensitive, who knows.
If you look close at a piece of category cable you will actually notice that each pair is twisted differently. This is so that the cross points overlap on a non crosspoint in a adjacent pair. Also there is a huge misconception about shielding. A piece of cable that has all four pairs with an outer shield is not shielded that is what they call screened. A shielded category cable has each individual pair shielded. When you start to deal with high frequencies a new phenomenon called reflection begins to happen. If all the pairs are under a common metallic barrier the frequency can reflect off that barrier and induce itself into an adjacent pair. Anything above cat6 will be shielded anything below it is screened. In the United States screened category cable below cat 6 is actually very uncommon. This is due to the FCC and the fcc's regulation on part 15 this requirement limits how much RF is allowed to be induced by the piece of equipment. Regardless if the device is a piece of radio equipment or not it must be tested and comply with the standards. Europe does not have the standards therefore they tend to have a lot more RF present in their buildings. Sometimes radio equipment can have issues and start to put noise into the air which can interfere with equipment. The lowest frequency that the cell phone company uses is the 800 MHz band. Therefore you should not have any interference on your category cable below cat 6 if you are experiencing this it could be that the equipment has a problem or it's something else that is interfering with your cable. If you are in a environment where a radio tower is present a shielded cable may be necessary or screened. Back in the day I did some work at a radio station and we had to run category 5 screen cabling because of the transmitters presents. FM radio waves are within the category 5 spectrum. Category 6 cable is screened category 8 cable is shielded each individual pair has its own shield around it. This eliminates reflection on adjacent pairs. Another purpose for using a screen cable is industrial environments where they use a 600 volt rated category cable in the rap a metallic foil over it. This is for the purpose of preventing ac voltage to be induced on the pairs and potentially damaging Network equipment.
I have 6 ft 26AWG Cat 8 cable from my router to my switch. It was pretty cheap on Amazon. The router is a WiFi repeater for my mobile data hotspot on my phone. I don't have any interference issues.
If the cable you are using has proper color coded wires, and clear RJ-45 connectors, you can visually check the colors are in the correct locations per T568A or T568B spec by looking through the back side of the clear connector. This won't tell you if the color pairs are in the wrong place for the spec but still wired correctly by pairs for the standard, (for that you have to compare both ends to each other), but split pairs should be visually evident. This also won't usually make a substandard custom cable crimp evident unless it's REALLY BAD, for that you will need a tester.
I had a different problem with my apple tv randomly disconnecting from the tv, it turned out to be the hdmi cable, i bought another one and the apple tv still would randomly disconnect telling me "the content can't be played on this tv" which was infuriating. Until i got myself an hdmi cable they use in television studios for connecting high end gear together, i got this cable from a friend, it's like your ethernet cable, very THICC and not bendy at all, but man, we never encountered that error again, like you said, it might be interference from other devices around the house or influences from the outside, but this 'industrial' hdmi cable does the trick for me! Good video
I had a nicked piece of Cat6 in the roof from a roofing nail. Worked 90% of the time. A basic tester showed no issues. Then I borrowed a friends super expensive fluke cable tester and sure enough, it indicated a problem with the insulation and resistance on one of the pairs. If money was no issue, I'd buy one up.
Yes you are looking at a G5 tower. G5 is very good, faster, more signal, but much shorter distance, ergo far more cell connections. Every block in town, every floor in new buildings etc.
That would be a more viable cable to use in a data center/server room environment for sure, but once you get up to 1gb/S data rates most connections are done with single mode fiber optic connections ( SC or LC) because the electrical signal ( I meant "data signal", sorry) is easier to get from point to point using a fiber optic connection. I have installed and turned up equipment (handling over 8.8 terabits/S, or, 88-100 gigabit/S Dense Wave Division Multiplexed channels, or, "DWDM") where all of the high speed data connections within the shelf were connected to the front with fiber optic jumpers instead of trying to move them from slot to slot on a shelf with the plug in card backplane connections. Good info on the cable though, always nice to see options available
I am confused by your tester B roll. When the "split pairs" is illuminated, both 1 and 2 are illuminated. Usually on a tester like that it means that the two pins are shorted together. Either way it's pretty clear to me you have a marginal cable. I would never go solid for a patch cable - even if you don't think it will be moving it's just asking for hard-to-diagnose problems in the future.
These type of errors are far more likely to occur at the weakest point which is the plug & jack connection. I have seen instances where the plug has been over crimped and pins 1 & 2 of the plugs are lower than the other pins. This makes the connection between the too low pin and the corresponding pin in the jack suspect. The intermittent nature may be a "loose" connection that when subjected to a temperature change like the one caused by load on a PC. Cat 8 will cause you more issues than it will solve. The cable will break the jack it plugs into, its heavy and inflexible and surface mounted ports were not engineered to cope with the weight of the cable. If you feel limited by cat 6 (which will do 10GBaseT @ 55m) OM4 fibre is cheaper / and far more mature technology than Cat8 and associated plugs (which are marvels of engineering - and prices generally reflect this $25 - $200+ per connector). Given fibre is more complex to terminate Id just purchase LC - LC patch leads and avoid the cost / hassle / expense of field terminations. One other tip is if you need 1 point run two (or even four), I have never once regretted running spare network points (copper or fibre) in an installation...
The audio interference shown actually WAS with Mogami Gold Studio cables 🧐. It does actually make a big difference, the cheaper cables I tried had way more interference.
Ive been having this same damn issue now for months! EDIT: Ive been troubleshooting my wiring paths with this new knowledge in hand and I think I have solved my issue. Thanks for this, its been driving me nuts trying to diagnose this with my ISP who has been anything other than helpful and not offering any real solutions other than the standard "did you turn it off then on again?" nonsense.
RFI from a local cell tower is an unlikely source. You would need to know the frequency of the RFI that you are getting to be able to even start allocating blame. You could also find the RFI is the result of a non linear mixer. These could take the form of a rusted bolt somewhere that effectively mixes two RF signals and reradiates at another frequency which is causing you grief. More likely sources of RFI are poorly designed switch mode power supplies, solar inverters and alike. It will normally be quite close to the appliance having the problem. Cell towers usually do a very good job of their RF.
Cat8 cable is almost a dead standard on arrival. If you are in really high interference environments or if you require the higher data speed (up to 40gig for cat 8) you are better served with a fibre cable. The cable will be totally immune to electro magnetic interference. Or go for a certified cat 6a cable for the compatibility with you existing gear. The only reason I can think of to use a cat 8 cable is for a future wifi-ap that requires the higher data rate, and can also be powered over that same cat8 cable with PoE.
I had to run an ethernet cable parallel next to power wire which I know isn't ideal, and it was an underground run, as well as a long run, so I figured getting an uprated category 7 cable with outside rated cable was smart. Not one problem. Gigabit speeds with no renegotiations. And no, I did not want to use a fibre media converter.
Good video on CAT8. I believe you are right in that if you don't have a need for the extra shielding, it is probably not worth opening the wallet. The only other reason I can think if is if you are into Hi-Fi and want to eliminate noise/interference as much as possible. But then you would probably have a network switch for audio (with a linear power supply) ensuring that only healthy packages are being sent anyway.
A broken wire, leading to intermittent faults would seem like a more normal reason for what you are seeing. In that case you would only see the error depening on how the broken part of the cable was bent.
You could have likely replaced this with any Cat5 Shielded cable, more likely the original cable had an imperfection at a bend. You can buy a new Porsche if your car won't start but, I typically suggest replacing your battery first. I am a Network Engineer, noisy environments can cause interference on cable but the upgrade Cat standard was just flushing money down the drain. Your wall jack lets in way more interference than the cable. Noise in an audio cable is much more sensitive than Ethernet, especially at that distance.
⚠A FEW POINTS (And 2 Month Update)⚠
• I'm not pulling the cell tower theory out of my ass like a conspiracy theorist, look at what it has done to my audio recordings at 5:55 and 6:12 (and was solved with ferrite cores, so not like it was a ground loop)
• While I'm still not fully sure whether the cable tester is correct or not about the split pair, replacing the cable with one I made myself and running in a different route (directly to the wall across the middle of the room, not behind the desk with all the other wires) DID indeed fix the problem. So really I can see 2 possible explanations. First, the cable tester is right (just inconsistent) and the cable was just badly wired. Or second, the cable tester is wrong about the split pair, the old cable is fine, but there is such a massive amount of RFI and crosstalk it overwhelmed the shielding occasionally.
• If my problem is indeed caused by RFI, then this is HIGHLY unlikely to apply to you. Definitely don't think you need some massively shielded cable like the one I bought.
~2 Month Update:
• Regardless of the true original problem, the new cable has indeed solved my problem, and I have not had a single internet disconnect since installing it. Several people have proposed another possible explanation which is that the original cable's shielding (or wiring itself) had imperfections such that when I coiled it back up, the internal cabling went back correctly into place, which is why it wasn't generating the same error anymore. I'm thinking that's probably the most likely scenario, since from what I've read, the shielding even in cat6e should be sufficient for most situations. So my point still stands for spending the extra on the cable though, I wanted the absolute best shielding possible, and I went with this particular brand because they actually tested each individual cable, though yes that does substantially increase the cost.
probably a 5g node since the higher bands cant travel rough stuff very well and thus has to be set up as a more dense cell grid.
@@kakurerud7516 Yep it's definitely 5G based on it's placement and shape
Ns
What about the wiring running through the structure? You've got a top line connection coming from the wall, then, there's the wall, everything in the wall, the box outside the home or business, and the rest of wiring leading back to your service provider. It's a fine looking cable, but a solution?
A cell tower won’t transmit at higher power than your cellphone (2Watt). The reason is that adding power to the tower would increase the reception of the towers signal on your phone but not the reception of your phone’s signal at the tower.
In other words communication would only improve in one direction and it’s only useful if it improves on both directions.
Your own cellphone is therefore much more likely to cause interference at home than a cell tower oven if that tower is nearby. (Signal strength will decrease with the distance squared. It will lose its strength very quickly).
I'm an audiophile, and I noticed my expensive headphones were occasionally (maybe 10% of the time) producing a static-like noise while I was listening to music. Turns out it was the power cable of the amp running parallel to the USB cable feeding the audio data to the DAC. Interference is very real!
How'd you fix this?
I was just about to comment this is an audiophile level cable, honestly I want one just to go with the thick cables I have on my system now.
@@certainperson reroute cables, increase spacing between cables, replace defective cables, upgrade to shielded cabling if it isn't already, add ferrite beads/toroids would be some solutions to consider.
@@certainperson I put distance between the cables (a few inches did the trick).
@@tjmbv8680 A simple cat5 ethernet UTP (Unshielded) cable is perfect for audio. It is not transporting music, it is transporting data packets. But some people like to inject noise into their DAC through the Ethernet port. Best way to do that? A shielded cable! Some expensive "audio grade" ethernet cables do not conform to the standard, they inject noise, so they "sound different" (granted! there is the noise!) but to some people "different" always translates as "better" when they have put a lot of money on it.
going overkill is great when you got the ching ching
Hey! Anti-Chinese racism is not welcome here! 😜
@@SpykerSpeed XD
Specially you have fiber.
this sounds mad racist
@@bootlxgching ching is just the sound used to represent money XD
At $90, I'm surprised they didn't call it "Monster".
Here are some facts about Ethernet cables.
1) Negotiation is done at 10 Mb, so with a defective cable, it may negotiate a higher speed, but fail when it gets there
2) With 10 & 100 Mb, only the green and orange pairs are used, so a cable that works fine at 100 Mb might not work at Gb, which requires all 4 pairs.
3) Gb Ethernet was designed for plain CAT 5 cable. 5e wasn't even available at that time, so plain CAT 5 is good for 100M and 5e just has tighter tolerances. 5e has replaced plain CAT 5.
4) Anything better than CAT 5e for Gb is a waste of money and may be harder to work with. This is noticeable between 5 & 6.
5) The 4 pairs in a cable have different twist rates, to reduce cross talk between pairs. That wouldn't happen if they all had the same twist rate.
6) Shielded cable may help in situations such as yours, where there may be interference. I have installed it a few times in noisy environments, but generally it won't make much difference otherwise.
7) The shield must be connected to ground at only one point, so check the continuity to ground at each end to ensure you're not grounded at both ends.
@Huss Ranneman Must be a cheap knock off then. ;-)
I think the shield for cat8 testing is connected at both ends. Not sure what the equipment at each end does. Are you thinking ground loops?
@@James_Knott Thanks for sharing!
Nr 7 is wrong though. It depends on the situation.
@@anullhandle Yes. If the shield is grounded at both ends, you may get ground loops.
Think you may have missed a crucial part: try meassuring the cable while bending it, or pulling it.
A intermittent connection can be rather confusing, and give different readings!
I was going to add this. Great catch. I've also had issues in my home with the coax / cable modem and a microwave tower near my house. A quick replace of the main splitter in the utility box solved this problem. I was told there are interfering frequencies, but I have not measured this yet and feel it's debatable.
Hi
exactly my thoughts, i doubt the cable has ANY problems with even gig connections....when it has a good connection and isnt missing 1 pin when it bends wrong
True
I used to think all cables were equal and then I spent a summer working as network install technician, and I realized how difficult it could be to get a permanent link test to pass on an ethernet cable
Yea, it can be hard explaining that to people especially when they are technical as well.
IT techs can be the worst at understanding this
I do ham radio and shielded ethernet cables make a huge difference for interference. On transmit it would take out my NIC until I switched to shielded, and on receive I discovered 10/100 IoT devices produce a lot of noise on the HF spectrum.
this! he bought a nice $90 cable but I could crimp a shielded cat7 or him which would work just as well... biggest issue is just getting the grounded ends.
Back when Commscope first came out with its Ion-E product, instead of coax to carry the RF around, they used Cat 6 Ethernet cable. As the company I worked for documented it's processes for installation they bought one of those Fluke testers and tested pre-made VS customer made to length cables with those fancy connectors you mentioned and there was an amazing range of quality, or lack thereof between pre-made, with the highest quality being ridiculously expensive (at the time). As most of the runs would be 100 meters or less, it was decided we would make our own and then use the tester to confirm they worked properly. One of our engineers made and tested the cables and connectors and several dozen scenarios and found that even if one cable out of the bunch isn't wrapped the same counts, or if it is wrapped too tightly, the throughput performance of the cable will vary drastically. Every one of those cables must be balanced as an individual pair, and in respect to one another. Needless to say it was pure hell finding people with the meticulous skills to install those connectors on those cables on those jobs. But after a few weeks, and lots of practice, it would finally come together.
Regarding the tower, just guessing you have T-Mobile/Sprint as they use TDD with is like Wifi on steroids. But the problem is more likely caused by your phone simply due to the proximity of it to your hardware VS that tower out front. It is basic physics and the inverse square law, try your test again but this time with all cellphones and your access point in the area turned off and see if it is coming from the tower or your access point since both switch back and forth from TX to RX many times a second. Happy hunting. PS, I am in the process of transitioning away from copper and going to fiber all over the house. Lightning popped in my backyard a few years ago and wiped out my Ethernet based network so I've been going back with fiber since it isn't impacted nearly as badly by lightning, and if I want to upgrade the speed, I don't have to replace any cabling, I just replace the media converters. :)
Nice
If you were concerned with cell interference, why not just run fiber and install an SFP+ card in your PC?
I like how you're asking the real question. I'm running MM fiber in my home-office connecting two switches on the opposite side of the room and I'm going to put some sfp+ pcie cards into my main pc + backup and run either mm fiber or DAC to the access switch. Cat8 is like he said overkill, but unlike he said, it just cannot be justified to install it in his house unless he just wanted to have a cat 8 cable.
This is what I was gonna say. I put fiber cards in my computers for like $50 each. And the fiber cable is extremely easy to work with. Doesn't need any special shielding or bulk. The speed (ping) is much faster than electrons. And each connecter uses like 0.7 watts of electricity per port (at each end of the cable) versus 2-5 watts for copper. And it's very thin, and easy to work with -- typically the solid core copper wires he refers to in the video are used inside walls where they're not handled often and don't need to bend around tight spaces like inside a desk or entertainment center.
I used point-to-point connections to my NAS until I was lucky enough to pick up a brand new ($1100) 10G fiber switch from a datacenter for $100!! They were moving to Atlanta and selling a lot of equipment so I bought a few servers for like $100 each.
@@AnastasiaValentineNAiden Actually, the signal propagation of electrons in wire, or photons in fibre are in the same ballpark at ~66% the speed of light. This is why Satellite communications are potentially lower latency than undersea fibre as in space a radio wave, or laser pulse can travel at ~99% the speed of light.
Fibre may achieve lower latency due to fewer transmission errors, meaning less retransmits, so communications become more reliable due to the immunity from RFI and crosstalk fibre brings.
Fibre encoding schemes are also usually simpler, meaning data doesn't need any buffering, waiting to be sent across 4 pairs with all that QAM rubbish, you just line the bits up and send them down the single fibre channel.
Another way fibre can reduce latency is say in a line of 5 buildings that form an commercial centre. Each one has a switch, and those are daisy chained. with ~100m links between each one. So the two end buildings have to communicate through 4 switches to see each other. This is fine as the links are already at around the max distance, there's no other way... Enter some cheap SM fibre with 1km range so now you can run point to point links between each building instead of needing to daisy chain them, each building has a direct link to each other building. effectively cutting out the latency.
On a large campus of buildings, or an office complex, this can reduce latencies MASSIVELY by taking out switch hops, while also multiplying bandwidth.
Should have just done Fiber.Pre made cabling and SFP suppliers like FS and equipment from Mikrotik make it dirt cheap.Mikrotik recently released their new range of 100Gbit switches recently as well.
I work in the cell tower field. If you’re getting interference, you need to tell the cell company. There should be an identification placard, like a site number or name and possibly a phone number. The other possible option is contacting your city planning department to let them know you’re getting interference. They may be able to help get the carrier fix it.
Do you have any suggestions for how many RF chokes per foot or per cable, or on which types of cables, etc, and what type of choke, is most effective at blocking the interference until the tower is corrected?
@@nataliegrn17 sorry. I’m not a RF engineer, I’m a real estate and permitting guy. I just know the legal and regulatory remedies.
Folks would do well to acquaint themselves with the inverse square law relating to radio signal power levels, if they haven’t already.
The available RF energy at any point along a straight line away from a radiating antenna will decay by at a minimum of 1/r², where r is the distance. In practice, atmospheric absorption and other obstacles along the path will further reduce the available RF energy.
If you look up the licenses for local RF sites (with your local telecom licensing agency) the details of the site tx frequencies, power levels and antenna gain/radiation patterns will likely be documented. With the above in mind, you can get an estimate of how likely a culprit the tower a mile away is likely to be compared to, say, the emissions from your wifi router in the same room.
ferrite cores are awesome!
it's a magic "interference be gone" solution.
I have at least one on every analog audio cable and on all of the interference-heavy devices' power cord.
One thing with ferrite cores is they are not all the same depending on the material it can make it better for blocking 1 frequency over another.
Great video. I'm at my current job now for almost 10 years and building CAT 5# and CAT 6 and 6A shielded cables is part of my job. Molex cables as well and I even messed with Fiber Optic fusing. This video is spot on accurate with everything. We sell CAT 8 shielded cables too but not the 22 AWG one you had in this video. Those shielded ends were really nice too. I mentioned this to my boss, who is also one of my best friends from childhood lol. I sent him this video to see if we can experiment with making them where I work. I got all the tools need on my desk as I type this, even the heat shrink tubing haha. Thanks for posting this!! My name is Joe too btw. Great channel!
[ edit: This is all completely wrong] The tech behind balanced audio cables isn't quite the same as twisted pair cabling. It's got the same intention (cancelling out noise) but the actual mechanics of it are different. Just to note. Twisted pair cabling is *not* balanced. Each individual wire strand is carrying different Stuff, whereas in a balanced cable you send data down one wire, and a reversed copy of the data down the other.
The diagram at 3:35 is explaining a balanced cable, but not a twisted pair cable.
Hm good to know. I was reading about twisted pairs and it said it 'could' be used as a balanced line, and just kind of assumed it was for ethernet.
"Twisted pair cabling is not balanced."
Nonsense. It's a differential signal, which means the two wires have opposite sides of the same signal. Ethernet has used differential signalling since the 10baseT days, with 10 Mb on CAT3 cable.
From "Ethernet The Definitive Guide, by Charles Spurgeon, 1st ed.
Under 1000Base-T Signal Encoding pg. 158
"The differential voltages used on the wire pair swing from approximately zero to +1 volts on the positive wire and from zero to -1 volts on the positive wire."
@@ThioJoe It is balanced. The person you replied to is incorrect. See my reply to him.
@@James_Knott Huh, I did not do nearly enough research here. I'm gonna amend my original post in case anyone sees that but not your correction.
When the Ethernet cable is thicc 😳
2thicc4u
2:20 that sounds unlikely based on the footage you showed before. I think the insulation somewhere in the cable is broken and the 1,2 pair are connected. The reason why it doesn't have the problem after removal is that the 2 vains aren't touching each other at the broken spot anymore due to moving of the cable.
Very nice Ethernet cable, it sure is Thicc and looks quality made. I have to pick one up soon so I'll keep this in mind. Thanks for the info👍
The method used to reduce interference in pro audio is Common Mode Rejection. In professional audio gear there will be a differential amplifier inside of any gear that uses standard 'balanced cables.
twisted pair
I would argue that common mode rejection is an attribute or desired outcome; the “method used” is differential signaling. By making your desired signal exist in the “differential mode”, everything you want to remove will be in the “common mode” (ie signals that are radiated more or less equally, or common, to the two differential conductors). Common mode rejection ratio (CMRR) is a metric for indicating the effectiveness of cancelling the common mode noise.
Thought I'd post my experience here just in case anyone hadn't thought of it, i was having the EXACT same experience of my connection randomly dying, but then I noticed it would always die if I saturated the read speed of the gigabit network (with a file transfer over LAN). I saw this video and was like, wow I should check all the cables in my network setup. So I spent the whole day swapping out cables and testing it again to see if it would fix it, but wasn't able to. Then I decided to try a different ethernet adapter (I was using a gigabit port in a USB4 hub I had recently purchased), and the problem was gone! So in my case it was the actual ethernet to USB-C adapter I was using. Just something to keep in mind. Also, it was a Realtek RTL8153 chip in a Cable Matters 201308 I'm having trouble with.
here i was thinking there was up to cat5 lol
I doubt the cable had actual split pairs. What I've seen several times are cables which show as OK on a network tester. Then if the cable is slightly bent or squished the tester shows a split pair or misswire. This is because of poor winding which opens up when the cable is bent, which causes crosstalk. Those basic Ethernet testers interprets it as a split pair or misswire.
Edit: For a short (
*"Marginally improve your internet speed! NOT for free"*
I think it would be pretty educative if you've measured these pairs with oscilloscope. I'm really curious what it would look like.
🌊🌊🌊🌊🌊💦🌊💦💦💦💦💦💦
The added thickness is not just the wire alone, every pair is twisted and shielded. And then all the shielded pares are twisted and shielded again. The larger metal connector now has to accommodate all the individual shielding for each twisted pair. Way overkill but highly efficient with virtually no signal loss or degradation. Even though I would not recommend it you could probably run a cat8 cable parallel with an AC supply wire and not have any issue. But I still would refrain from doing that.
I like the idea of the connector. It's pointless for me and the use I could give it, but it looks so good that I may go and try to build one cable, just to check out that connector.
The problem with the cell tower theory is that its interference would be fairly continuous like on the microphone, not a few times a day. I would be more suspect of something in your house like a compressor (or other large motor) cycling, producing a burst of EMF. For example is the AC compressor or refrigerator on the other side of the wall? The better shielding on the cat 8 cable would help.
Dirty power is a big deal.
I have been using CAT7 for years, for a 30foot run, works great, now for 6+ years! ThioJoe, file a Complaint with the FCC, as they want to know about Problem transmitters..
That's the kind of chonky Ethernet cable we use for some POE devices as well as the controllers of the Microh LED panels of giant LED walls (but we use an ordinary CAT7 cable to connect the wall to the Nova Star controller 🤷).
Thank you for mentioning Cat7... Just because it doesn't have TIA/EIA recognition doesn't mean it doesn't exist.
Is that actual POE. or just some honking DC conductors sharing the same insulation as the STP? I've seen stuff like that before. Or run three ethernet cables. One has the real data... the other two are all just shorted at each end and crimped into a big DC lug each with a red or black heat shrink.
Ethernet cable also takes a couple of amps of mains (230 VAC) pretty well, If you're in Europe you even have the right colours, Blue=Neutral, Brown=Live and Green=Earth. The Orange one could be used for something like RS485.... be VERY weary of cutting cable out of a home automation nutters house, cause that's the kind of stuff you're likely to find LOL. Just for the love of god use an ethernet cable that is a different colour to the data cables!
I have dealt with a LOT of industrial level ethernet that is dealing with interference from dozens of noisy motors and drives... I have also dealt with the single side grounding and found industrial equivalents of the cable connector you have shown. The really cool thing is they are clamp on and have special heads for pulling cable or for different plug styles... expensive, but really great stuff when you have difficulties or interference issues.
why did i just watch a 10 minute video about an ethernet cable
Thanks for making this video, its incredibly re-assuring when someone else has had the same issue for just as long.
Audio tech and electronics designer here.
To help with interference you want to have as well shielded XLR cable with as good connectors as possible and high quality preamp for that SM7B. SM7B is difficult, low outoput + somewhat high impedance source that many cheaper preamps struggle with. But it's also a case of how one uses the SM7B, it's very low output means that you must be RIGHT ON THE MIC for it to pick up high enough signal and then you don't need as much gain for the same SNR, if you're more than 1 inch away from the mic wind screen, you're too far away.
SM7B is an artist mic, voice over mic... voice talent mic. It's not suitable for much else, thus one must know how to use it properly.
You can also try fethead or similar j-fet preamplifier (that uses phantom power) to boost the signal very cleanly right at the source and then have the cable run to your interface. This helps with the low quality preamps quite a bit :)
I use a DSX5000 at work, more for industrial M12 connectors than ethernet so usually 10G cat7 cable.
They're great for these random and confusing issues, as they can detect NEXT & FEXT, ACR & ACR-F, and graph everything by pairs so you can see exactly what's happening.
I think part of the price is for the software, which is designed to test setups for whole buildings and large scale installations and map it all out nicely.
Summary- it’s not worth it
07:31 Yeah same, I bought 2 cables (25cm and 50cm) cat8.1 S/FTP 2 weeks ago, they were delivered with their own measurement certificates.
Finally seeing a guy on TH-cam use it.
I've replaced my whole house network with cat8 for about a year now.
I'd say it's as advertised, 10GB is reachable but the nic is really hot. Also, the only time it reaches 10G is when I use iperf or copy a huge file from my nas.
Btw, yeah 22awg makes it really hard to bend it, I've to do a bunch of roller coaster style loops.
I've had people telling me for 30yrs now that I don't need more memory, hard drive capacity, etc. Fact of the matter is they were trying to con me into buying the minimum required at the time. I always try and go with the best highest quality that I can afford.
I didn't realize that using an older cable standard would result in intermittent problems. My experience has always been that intermittent problems were due to a bad connection, but I guess that's wrong. I'm going to get one of those cable testers just in case I run into anything like that in the future
It definitely is probably a bad connection in 99% of problems. I really think mine is a special case because of the cell tower which I already know has caused issues with my audio like I show later in the video. Of course I could be wrong about that, but I can't really test it without $10K+ equipment
@@ThioJoe It's probably better to diagnose by elimination than buy $10,000 worth of cable testing equipment. The other option is to bring in a technician who spent the aforementioned equipment in order to figure out what's going on
@@ThioJoe Bad connections are a common problem. I've been a tech for almost 50 years (50 on May 1) and I can say that with certainty. If you have a problem, check the cable and connections first.
MPTL ( modular plug terminated link) connector is what those plugs are . We use them for cat 6 or 6A shielded as well . It’s actually now an official standard for using modular plugs for direct connection to cameras , access points and other devices without having to have a jack and box plus patch cord .
Awesome thanks I couldn't find official the name of them 👍
@@ThioJoe welcome . Great video, and channel
You can get EMI/RF resistant window film to help with external EMI/RF. I had to get it for one side of my home and it worked well.
Unless you want to have wifi in your front/backyard
Joe: Cat8 is overkill...
Linus: it's my home setup for casual gaming and movies)
I bought a shielded Cat8e cable (15 m) from a German manufacturer (inLine) that tests their cables(they even give you the test results). Apparently I live in an area close to shipping lanes (I live by the water) and their radars can interfere with my signals. So I got the 8e cables to give me more stability rather than anything to do with speed - if you have 1000 Mbps transmissions - then it won't help anyway. But I have noticed a more stable connection during the day compared to my unshielded CAt6 cable.
I would love to see the FCC cert for that cell tower equipment
Industrial Ethernet is what is used to control automation devices like PLCs. Sometimes called (depending on the application) EtherCat, EtherNet/IP (industrial/protocol), POWERLINK etc.
when you work in the audio industry you become accustomed to 300ft long cat8 cables that are double shielded. someone shut one in a door a few months ago, thats $290 we are never getting back
imo if you need something greater than 10G (ie cat 6/6a) you should just make the jump to multimode fiber. You can probably get affordable used equipment for it on ebay since big companies will get rid of their old equipment when they need an upgrade, and you can reap the benefits of cheap but fast compared to commercial networking equipment
I run multimode fiber between my switches (ironically Dell gigabit models that haven't been sold new for years), including a run that is parallel to an electrical feed. I missed the part about whether the rest of the run was replaced too: Fancy cable connected to a wall jack having regular cable could still be getting interference elsewhere.
Especially if interference is actually a problem
Industrial Ethernet is generally designed for flexibility and robustness. We actually use non-shielded cable because the shield can cause issues.
Funny that this issue occurred on ps5. For almost a year I’ve been having this issue and I couldn’t find any thing. And last week I’ve been digging in some information about lan cables I found out there is a cat8 and decided to buy one. And GUESS WHAT! The issue is gone. I now get at most 1ms ping during online game play 🙌🏼🙌🏼
Cat 8 is worth the money. Old, small capacity cables that some of us tend to use as patch cables between desktop devices, reduce the capabilities of every device they are connected to, including internet speeds. I use a single piece of 33 meter Cat 8 cable from my 1 Gig fibre gateway running through my attic to my primary router at my desktop iMac in my home office, and there is no noticeable loss of signal strength.
I have used Supra CAT8 for my hifi system and wired to the TV, to shield the systems from electronic noise. Just ordered another to replace a faulty cable from the front to the back of my house
I'd be worried that solid core cable would put a lot of strain on the Ethernet ports due to the lack of flexibility and extra weight pulling down on the cable. Maybe it could be alright if you have a way to secure the cable to the wall or desk near the ports in such a way that it isn't putting force on the ports.
We use industrial Ethernet in our factory. Mostly we have Cat5e and Cat 6, mostly shielded. The distinction of "industrial Etherne" is more about the protocol used and the type of managed switches. Everybody seems to be relabeling some garbage Cisco switches that take all day and require lengthy reboots. I haven't purchased one in a while, so maybe they have made improvements.
1) Ethernet/IP is a poorly named protocol from Allen Bradley and typically used for cyclic and deterministic communications. Messages are sent as fast as 1ms for most I/O functions, but the engineer balances the needs for each I/O device and may set the packet interval at one second, or larger in a huge network. Most of mine are around 10ms to keep traffic down on the master ( , which is called a PLC). For motion applications another protocol is used. There is clock-time synchronization so that all devices will start their work at the same time. For example a master device will command a bunch of servo drives to perform coordinated motions. Since the servo drives are relatively simple, they reply with the motor position and an accurate time stamp of what time the position was taken. The master then performs math and sends new velocity commands to the servo drives with a begin time. These exchanges typically occur at 250 microseconds. On these systems s managed switch must include the feature to add it's own processing time to the time given by the master. On return traffic it does the same thing with data from the servo drives that have a time stamp.
2) Profinet is a Siemens brand protocol used for the same purpose, with some pretty advanced features.
3) EtherCAT is a protocol by Beckhoff that uses a ring network that can be configured also as a straight line where the last device in the string sends data back the way it came. This protocol is not compatible with TCP/IP and therefore a separate network is required. This protocol has two distinctions. It is very fast. And a non-related equipment manufacturer can purchase protocol chips from Beckhoff and instantly they have a certified EtherCAT device without any testing. And the chips are pretty cheap. So a lot of third party companies produce compatible equipment. But you need an EtherCAT master controller. People like the controllers and languages they have previously learned. The cost of trying something different is immense in a short calendar time project. I used a bunch of Beckhoff hardware in the past because they have the largest selection of I/O in the world. And they provide protocol adapters for just about every Ethernet and Serial network type out there. I used their controllers on cost sensitive projects, but the learning curve was fairly big.
The Plugs aren't even CAT7, not to mention CAT8. On the RJ45 Plug the pins are to close and not shielded pairwise to meet CAT7 standards.
A solution to this is the GG45 plug, but the tech we use today didn't implement it
GG45 ist back compatible to RJ45, it has pairs of pins on its "shoulders" shielded against each other.
I am using a Fluke CAT7 verifier btw
Yes the cat7 standard calls for special connectors like GG45 or TERA. But Cat8 does allow RJ45 from my understanding
@@ThioJoe you're right... I did my homework and there are 2 CAT8 Standarts: 8.1 with RJ45 (compatible with 6A) and 8.2 with GG45 / TERA (compatible with 7A)
Thanks
As a first measure I would keep data cables apart from those providing electrical power, especially those providing 110V or 230V (in Europe).
As an alternative to expensive cables there are metalized cable housings available which can contain (and tidy up) several data cables. There is no need to continue this kind of housing from port to port because if the major length of the cable in question is protected a short way between the connectors and the housing do not catch enough RFI to get distorted.
2 views and 8 likes THX TH-cam!
Just use fiber
While that would fix it, probably wouldn't be practical for my setup
I had the same problem with an old cat 4 cable that comes from my router modem upstairs, I replaced it with cat 7 and now it doesn't come down a conduit with 240V ac leads and it's been fine.
CAT 4? Never heard of it. CAT 3 was good enough for 10 Mb, but CAT 5 is required for 100 Mb & Gb.
@@James_Knott well whatever it was, the 100mb cable
@@DogsBAwesome That would be CAT5.
We confused our IT cabling folks one time we asked them to test an Ethernet cable and it was stranded. It was used on a robot and the stranded cables can handle the constant bending I learned. I’ve also learned a little about Ethernet Industrial Protocol and those beefy cables are used on factory floors where cost is next to nothing compared to system they’re used on.
Question: I see you made all the measurements on your network infrastructure (cable to wall, the wiring, to patch panel).
Did you made a measurement of the signal from the nearby cell tower antennas, if they were within the boundaries of regulation for your location?
It would be an interesting follow up.
Had similar interference problems with my TV setup. The combination of mess of wires, wireless signals, newer cell signals, cheap low shielded antenna(with extra long attached coax cable), amplifier on top of noise, movement from people, and the use of regular dual shielded coax cable made it extremely random and irritating when watching TV.
I bought a higher range antenna that had the coax connector on it instead of those long, cheap, low shielded ones attached. I paid extra for two extra short quad shielded cables for most of the setup(and cut out at least eight ft of unnecessary cable). I bought a cell signal filter for LTE and 5G signals and put it before the amplifier.
I tested the different configurations but only together did I get a perfect signal on all channels. The only thing I didn't do was use a heavily shielded USB cable for the amplifier which has been known to cause problems especially with cheaper cables and power supplies.
OMG I've been having this issue all year, and would just happen at absolute random. Thank you so much for this vid
Linkup makes great cables. I could make my own, but, naw, I'm happy to have it tested and certified. I have a bunch of Cat6 and Cat7 from them.
there was a time when you would have said the cat8 would increase your speed lol.
This is almost certainly an intermittent physical short in the ethernet cable due to chemical degradation of the insulation between the conductors.
The "RFI" you are getting in your audio is not from a cell tower, which is specifically what lead me to conclude your cable problem was just physical conductors shorting.
Cell tower signals are in the order of millions of hertz, usually 400 - 700MHz range - well outwith the range of human hearing.
Your spectrogram of the audio signal looks a lot more like switching noise generated from switched mode power supply units. This is a common problem for external audio interfaces sadly that isn't unique to high impeadance mics like the SM7B that need a lot of signal gain to drive them. If you disconnect the mic and whack the gain you'll get the same broadband noise with no cable attached to the interface and if you gain up the output volume and amp for your speakers you will hear the noise audibly. It's usually worse if you are on the same circuit as high current draw switching devices, so if you have a beefy GPU and spin up some GPU heavy task you will hear the noise increase in intensity.
Ferrites are one solution for this, isolation transformers (DI boxes) are another, expensive USB isolation boxes are another. YMMV.
And here I am running my home 10gige network over standard Cat5e just fine.
I use cat8 for my WAN connection because it's likely more durable and gives me a cable to use on a faster connection if need be. It actually wasn't that expensive, though it helps that it was a simple flat cable. For a while, I had weird issues with my WAN connection randomly going to 100Mbps on my old flat cat6 cable. It gave me some weird readings on my cable tester and that's when I replaced it with cat8.
Stuff like this is why I come to your channel!
Category 7 and category 8 cables are for shorter distances. Cat5e and Cat6/a will work for longer distances I have a question regarding that cell phone tower have you had any headaches recently
I kept getting network disconnects, same as you mention, with my desktop that is my NAS/Media Server. I have a small fridge about 8ft from it. I eventually figured out that every time the Fridge compressor kicks on or off, the NIC reports disconnect. I had to not only replace the cat6a STP cable with a cat8 S/FTP cable (better shielding), but also, put the fridge on a surge protector that has high EMI/Noise filtering spec (high dbi). Those changes fixed the issue.
Yeah that's a fifth-generation cell tower and the ultra high frequencies they use not only cause issues like this but other stuff like thinning the blood-brain barrier
Guys, this was a nice video and all, but please don't listen to his advice and don't waste your money on Cat8 cables. For distances of 100m or less in 99% of non-datacenter environments, Cat5e or Cat6 will do just fine. Even if there is a cell tower in your backyard.
If brute force isn't working, it's because you aren't using enough of it.
3:22 You are mixing between twisted pair and differential pair, Twisted pair cancels 99% of the interfence from itself to other cables and itself it the cable was looped, but differential pair is to modulate the signal to two wires in a way that the difference between the voltage will equal to the original signal at the receiving end, thus removing noise.
If you're just going to be using it as a patch cable, I suggest stranded wire... in a wall I'd recommend solid wire. Solid wire is more fragile, as you mentioned
you could have just gotten a STP cable instead of a $90 cat 8 cable...
I got a 100m outdoor rated cat6 cable for the same price. Works like a charm. You paid way too much for that one
did you need all 100m? 100m is the limit of ethernet.
@@estusflask982 about 80m, so close
22AWG is very unnecessary for indoor patch cable use. Most cables are 26AWG, with some 24AWG ones. Thin run 30AWG cables are also suitable for in home use, and can handle 10GIG fine. Unshielded (UTP) cables are fine for home use, but I recommend Shielded (STP, SSTP, S/FTP) if you are running any cables near power cords, in-wall electrical, or have anything with a motor, compressor, transformer near the cable. I have used Monoprice brand cables for 15yrs, never had a cable go bad without any sort of obvious clue as to why (physical damage etc). Those CAT8 are meant for data centers 40GB+ networks.
Thio I had this exact same problem and I blamed my crappy internet. Thanks for the info :)
The reason you get that intermittent split pair error is due to the fact that it's really hard to detect. You can't find it by testing for continuity, as each pin is wired to the corresponding pin at the other end. So, you'll need to detect it by the crosstalk introduced by the mismatched pairs. This is essentially extremely short range radio interference. So I'd wager you rather simple tester (in the grand scheme of things, a 'real' CAT-tester is around 5000 dollars) does some simplified version of a split pair test by injecting a pulse and seeing if it get's it back on a conductor it didn't expect. So RFI from an external source could also theoretically trip that detection. Or you cable might actually have a split pair, and your tester is a bit unsensitive, who knows.
If you look close at a piece of category cable you will actually notice that each pair is twisted differently. This is so that the cross points overlap on a non crosspoint in a adjacent pair. Also there is a huge misconception about shielding. A piece of cable that has all four pairs with an outer shield is not shielded that is what they call screened. A shielded category cable has each individual pair shielded. When you start to deal with high frequencies a new phenomenon called reflection begins to happen. If all the pairs are under a common metallic barrier the frequency can reflect off that barrier and induce itself into an adjacent pair. Anything above cat6 will be shielded anything below it is screened. In the United States screened category cable below cat 6 is actually very uncommon. This is due to the FCC and the fcc's regulation on part 15 this requirement limits how much RF is allowed to be induced by the piece of equipment. Regardless if the device is a piece of radio equipment or not it must be tested and comply with the standards. Europe does not have the standards therefore they tend to have a lot more RF present in their buildings. Sometimes radio equipment can have issues and start to put noise into the air which can interfere with equipment. The lowest frequency that the cell phone company uses is the 800 MHz band. Therefore you should not have any interference on your category cable below cat 6 if you are experiencing this it could be that the equipment has a problem or it's something else that is interfering with your cable. If you are in a environment where a radio tower is present a shielded cable may be necessary or screened. Back in the day I did some work at a radio station and we had to run category 5 screen cabling because of the transmitters presents. FM radio waves are within the category 5 spectrum. Category 6 cable is screened category 8 cable is shielded each individual pair has its own shield around it. This eliminates reflection on adjacent pairs. Another purpose for using a screen cable is industrial environments where they use a 600 volt rated category cable in the rap a metallic foil over it. This is for the purpose of preventing ac voltage to be induced on the pairs and potentially damaging Network equipment.
do you read comments on any of your older posts? cause some people get your old posts in the recent searches.
I have 6 ft 26AWG Cat 8 cable from my router to my switch. It was pretty cheap on Amazon. The router is a WiFi repeater for my mobile data hotspot on my phone. I don't have any interference issues.
If the cable you are using has proper color coded wires, and clear RJ-45 connectors, you can visually check the colors are in the correct locations per T568A or T568B spec by looking through the back side of the clear connector. This won't tell you if the color pairs are in the wrong place for the spec but still wired correctly by pairs for the standard, (for that you have to compare both ends to each other), but split pairs should be visually evident. This also won't usually make a substandard custom cable crimp evident unless it's REALLY BAD, for that you will need a tester.
I have this same exact issue with my computer. Glad you made this video, I might try some of what you said in this video.
I had a different problem with my apple tv randomly disconnecting from the tv, it turned out to be the hdmi cable, i bought another one and the apple tv still would randomly disconnect telling me "the content can't be played on this tv" which was infuriating. Until i got myself an hdmi cable they use in television studios for connecting high end gear together, i got this cable from a friend, it's like your ethernet cable, very THICC and not bendy at all, but man, we never encountered that error again, like you said, it might be interference from other devices around the house or influences from the outside, but this 'industrial' hdmi cable does the trick for me! Good video
I had a nicked piece of Cat6 in the roof from a roofing nail. Worked 90% of the time. A basic tester showed no issues. Then I borrowed a friends super expensive fluke cable tester and sure enough, it indicated a problem with the insulation and resistance on one of the pairs. If money was no issue, I'd buy one up.
You can get double shielded Cat6a, SSTP, S/FTP. CAT8 just mandates the higher shielding.
Yes you are looking at a G5 tower. G5 is very good, faster, more signal, but much shorter distance, ergo far more cell connections. Every block in town, every floor in new buildings etc.
You should send the cable to Linus so he can use his cable tester on it. Im sure he'd love the opportunity to make a video about it
Was about to say same thing. Be interesting to see how it specs out on that $10k tester.
That would be a more viable cable to use in a data center/server room environment for sure, but once you get up to 1gb/S data rates most connections are done with single mode fiber optic connections ( SC or LC) because the electrical signal ( I meant "data signal", sorry) is easier to get from point to point using a fiber optic connection. I have installed and turned up equipment (handling over 8.8 terabits/S, or, 88-100 gigabit/S Dense Wave Division Multiplexed channels, or, "DWDM") where all of the high speed data connections within the shelf were connected to the front with fiber optic jumpers instead of trying to move them from slot to slot on a shelf with the plug in card backplane connections. Good info on the cable though, always nice to see options available
fiber doesn't transmit electricity, not sure if you meant that
@@estusflask982 You are correct, I should have said "Data signal"
I am confused by your tester B roll. When the "split pairs" is illuminated, both 1 and 2 are illuminated. Usually on a tester like that it means that the two pins are shorted together. Either way it's pretty clear to me you have a marginal cable. I would never go solid for a patch cable - even if you don't think it will be moving it's just asking for hard-to-diagnose problems in the future.
These type of errors are far more likely to occur at the weakest point which is the plug & jack connection. I have seen instances where the plug has been over crimped and pins 1 & 2 of the plugs are lower than the other pins. This makes the connection between the too low pin and the corresponding pin in the jack suspect. The intermittent nature may be a "loose" connection that when subjected to a temperature change like the one caused by load on a PC.
Cat 8 will cause you more issues than it will solve. The cable will break the jack it plugs into, its heavy and inflexible and surface mounted ports were not engineered to cope with the weight of the cable. If you feel limited by cat 6 (which will do 10GBaseT @ 55m) OM4 fibre is cheaper / and far more mature technology than Cat8 and associated plugs (which are marvels of engineering - and prices generally reflect this $25 - $200+ per connector). Given fibre is more complex to terminate Id just purchase LC - LC patch leads and avoid the cost / hassle / expense of field terminations. One other tip is if you need 1 point run two (or even four), I have never once regretted running spare network points (copper or fibre) in an installation...
Now I want to know if something like a Mogami Gold Studio xlr cable would have solved your audio issue. That could make a good video too.
The audio interference shown actually WAS with Mogami Gold Studio cables 🧐. It does actually make a big difference, the cheaper cables I tried had way more interference.
Ive been having this same damn issue now for months!
EDIT: Ive been troubleshooting my wiring paths with this new knowledge in hand and I think I have solved my issue. Thanks for this, its been driving me nuts trying to diagnose this with my ISP who has been anything other than helpful and not offering any real solutions other than the standard "did you turn it off then on again?" nonsense.
RFI from a local cell tower is an unlikely source. You would need to know the frequency of the RFI that you are getting to be able to even start allocating blame. You could also find the RFI is the result of a non linear mixer. These could take the form of a rusted bolt somewhere that effectively mixes two RF signals and reradiates at another frequency which is causing you grief. More likely sources of RFI are poorly designed switch mode power supplies, solar inverters and alike. It will normally be quite close to the appliance having the problem. Cell towers usually do a very good job of their RF.
Cat8 cable is almost a dead standard on arrival. If you are in really high interference environments or if you require the higher data speed (up to 40gig for cat 8) you are better served with a fibre cable. The cable will be totally immune to electro magnetic interference. Or go for a certified cat 6a cable for the compatibility with you existing gear.
The only reason I can think of to use a cat 8 cable is for a future wifi-ap that requires the higher data rate, and can also be powered over that same cat8 cable with PoE.
I had to run an ethernet cable parallel next to power wire which I know isn't ideal, and it was an underground run, as well as a long run, so I figured getting an uprated category 7 cable with outside rated cable was smart.
Not one problem. Gigabit speeds with no renegotiations.
And no, I did not want to use a fibre media converter.
Good video on CAT8. I believe you are right in that if you don't have a need for the extra shielding, it is probably not worth opening the wallet. The only other reason I can think if is if you are into Hi-Fi and want to eliminate noise/interference as much as possible. But then you would probably have a network switch for audio (with a linear power supply) ensuring that only healthy packages are being sent anyway.
I've used a cat8 cable for my point to point wireless link between building because a lot of the cable is outside. I wanted the extra shielding.
A broken wire, leading to intermittent faults would seem like a more normal reason for what you are seeing.
In that case you would only see the error depening on how the broken part of the cable was bent.
You could have likely replaced this with any Cat5 Shielded cable, more likely the original cable had an imperfection at a bend. You can buy a new Porsche if your car won't start but, I typically suggest replacing your battery first. I am a Network Engineer, noisy environments can cause interference on cable but the upgrade Cat standard was just flushing money down the drain. Your wall jack lets in way more interference than the cable. Noise in an audio cable is much more sensitive than Ethernet, especially at that distance.