Yes if you looked at LSUSB it would show up when plugged in as a USB1.1 hub, possibly 3 port or 1 port. Then the other side is also a USB 1.1 hub, though it might not enumerate as it is now in a pass through mode with the floating pin. I have an old 8 port USB hub that I use as an extender, because it is perfect for slow peripherals like keyboards and printers, and as an 8 port hub it also does double duty in allowing USB powered devices like a small wood grain LED clock to plug in, and also will support slow charging of things, like NimH cells in a modified charger.
@@SeanBZA LSUSB doesn't recognize it at all. It only recognizes whatever is plugged i at the farend. I'm increasingly convinced this is just a couple of amplifiers - totally unaware of the USB protocol.
I wouldn't be surprised if that chip is literally just amplifying the incoming differential signal as hard as it can, clipping at the rails, and blasting that down the cable with no regards for SI. That would definitely give the results you're seeing. EDIT: Yep that transfer speed benchmark absolutely says it's falling back to USB 1.1 speeds, meaning this thing is absolutely trashing the signal.
*MB* is mega *_BYTES_* per second. - 11:27 Lower case *b* would be bits. MB/s = Mega bytes per second Mbps = Mega *bits* per second (Mb/s is less commonly used and can cause confusion with MB/s) I know you know this but just thought I would mention it for those that don't know or are confused by these terms.
Hey fellow beer drinking local. I did this once upon a time to run a 30' USB cable to an old USB webcam (ghetto office security camera). Worked good enough. I just soldered the ends together... Worked pretty good for a 240P webcam. 🍻
I got one with the idea of using it to extend the cable from a USB camera. I only got it to work with a powered USB hub and, like you observed, the resolution from the camera was crap when the cable was connected.
When you opened that thing up and I thought it was just splitting out the usb to ethernet, I cracked up. I kind of wish it was that. Thanks for the entertainment
I use a similar looking thing to run from my computer... over drop ceiling in my home office to my electronic bench. I have it hooked to a powered hub and use it program arduinos, esp32s, etc from my development computer on other side of the room. Probably only about 30ft of ethernet cable but it does the job perfectly. I got ripped off though I think I paid like $15 for mine on amazon. :)
You need to unroll 1/2 the cable and roll back up in the other direction, this should cancel out the electron spin! ;-) strange beer you have there as well.
Next to my NAS and router, there is a UPS. The UPS (it's old) has a USB to monitor it. A few UTP-cables were laying dormant and I used one of these extenders. The cable length is 15 meters, 16 yards, or 50 feet. It works fine and the UPS reacts instantly, but how fast the data is transferred I do not know. Did you only fry a bit of the chip(s)?
Propogation time on a long cable will create limits on data capacity especially on dynamic data like the camera. Where a handshake is needed the long propogation time makes handshakes take long. One further issue is that the USB may be dropped to the 1.2 megabit/sec. The resistor on the USB lines should set the capable baud rate. Personally I have only ever used the max15m special buffered long USB cables, not the RJ45 lan type cable extender
Just a tip when you use these passive USB extenders... Don't! If you do anyway, plug the "active" part in as close to the computer as possible, no extensions at all. @pileofstuff - I am curious what load you can get at 5v on the end of that grey extension cord - probably not enough ?
Looks like it would be faster to transfer files with a 1.44MB floppy. I wonder if there's a legitimate version of the product like all those old USB to RS232 or Parallel etc things I used 10-15 years ago no problem.
I bought the same sort of device on Amazon back at the beginning of the pandemic. I had an identical experience to yours with a camera. When I came across this video, I thought, optimistically, that you could actually get it to work better than I was able to. Alas, we were both taken in by false claims: I more than you, because I paid closer to $5 for the experience.
Bummer. I find it hard to imagine that the crossover cable fried the chip. It does still pass data accurately; just too slowly. Would have been handy....
Maybe, but I'm not convinced this thing is even aware of the format of the data it is passing. I suspect it is primarily an amplifier or line driver type thing. I'm hoping someone can have more success than I did finding a datasheet.
@@pileofstuff Based on the benchmark results you were getting I am guessing this extender only supports USB Low Speed and USB Full Speed and you were trying to use it with devices designed for USB High Speed. Since there is an active IC in the middle of the connection that IC would be what both the USB device and PC are negotiating with and would determine the USB link speeds supported. Low Speed goes up to around 1.5 Mbps (or 187.5 KB/s) and Full Speed goes up to 12 Mbps or 1.5 MB/s. These are maximum rates that do not account for any overhead - real world speeds for Full Speed is around 1 MB/s which is almost exactly what you got on your benchmark. You should be able to see what USB link speed the devices are using. On most Linux distros lsusb will give you that information. In my experience it is fairly rare for USB extenders to support High Speed USB mode - particularly for a low price; and even rarer to have any support for any of the more modern SuperSpeed modes. These devices are perfectly fine for serial-based communication such as management ports for enterprise network equipment or managed AV equipment which often don't support anything faster than full speed anyway. (Earlier generations of such devices used RS232 ports for configuration so raw speed is almost never a big concern.) They also work fairly well for USB input devices such as Mice and Keyboards which are often low speed devices operating at 1.5 Mbps.
@@pileofstuff I have some usb isolators that limit to usb 1.0 but otherwise don't understand the data. It is possible (likely?) these are very similar ICs. I wouldn't be surprised if the only thing these chips do is the initial usb handshake (which sets the speed) and maybe some sort of basic amplification. BTW, if you did want to investigate more, ben eater recently published a couple videos on USB that explain handshake and the protocol used by keyboards well enough to probe the ethernet cable with an oscilloscope and get an idea what is happening.
These specific ones probably aren't widely available, but I bet there are some local microbreweries near you that have something interesting and different.
I tried it with the long cable and the test timed out after 30 seconds during the first sample every time. Didn't think it was interesting enough to include when the short cable already shows how bad it is.
What did you expect ? The full 144 MBit ? .. Still... to get some temperature (or whatever couple digit number) data from a distant micro controller that has USB but no ehternet/wifi these would do the job ? Would be cool if you try that out with the long cable.. like controlling 4 servos dangling from the ceiling or something ;)
For a microcontroller it would be better (and cheaper) to just use a RS485 serial interface, you can get a USB converter for a few dollars and a transceiver for the microcontroller, and it's far more reliable
@@pileofstuff *even if* the cat5e cable was CCA (copper clad alu), the voltage would not have dropped this much as the load on the other side is only a couple miliamps.
I’m almost betting it’s trying to use RS-485 for the distance comms. But it might be a timing issue. I have run into timing problems when using RS-232 ports and RS-485 null modems
I suppose that would depend on the impedance. that the mystery chips are designed for. I'm still holding out hope that someone will have better luck finding a datashet than I did.
Wondering if the trace difference between the PCBs is telling the chip which end it's functioning at... pulling pin 6 to ground signals it's working on the device side vs the host side (or the other way around, don't remember which had pin 3 grounded..)
so, any chance the rj45 cable was too short ie not enough resistance for the line drivers to work with the receivers properly? looks like it works in an identical manner to the chips datasheet you DID find, just a different pin out?
I'm not convinced it is sophisticated enough for too short a line length to cause problems. Maybe if it was operating in RF frequencies, but that isn't what's happening here.
@@pileofstuff While USB does care about line resistance, it does not care this much. The computer's root hub will have mitigations for a lot of bad stuff.
I BOUGHT IT. THAT'S WHY I WAS WATCHING THIS. BUT DIDN'T WORK ATALL. MAY BE I ALSO USED A CROSSED CABLE. I WILL CHECK THAT. BUT PLS LET ME KNOW A GOOD BRAND
It didn't work for anything. remember this is the cheapest thing I could find on ebay. There probably are other models that actually work. Check further down the comments. Someone suggested one that they use.
@@pileofstuff May I ask you this. I want to extend the cable of usb wifi for my laptop and put it higher and farther. Do you know any good solution to do that? . Thanks
also , skip all this chip nonsence and just use a cat5 cable ,use a twisted pair for ground, 5v ,data and data. 2 wires will double the wire gauge and you will have less voltage dropover the distance
I would recommend only using one pair of twisted wire per line - as Bob suggested. You might think that using one wire from each pair would be better as the twists cancel noise, right? Well; USB is differential. Doing that would also cancel itself out. No bueno!
I still need them occasionally at work on some of the legacy equipment. Normally I have my crossover cables labelled. Not sure how this one slipped thru the cracks.
This is obviously a cheapass Chinese knockoff from a decent product unless the crossover cable made all the difference. I wish you had tried it with a standard network cable. They come with modems and over the years I have gathered quite a lot of them, even if they are only 1m long.
The crossover cable is used for computer to computer traffic to avoid using a network switch, because you are using a standard USB device then the data cables are switched and the computer is unable to recognise the device. It would still work if you only needed to power the device. That being said I doubt that you have done any permanent damage to the USB extenders so I would simply try a normal network cable and see if it will recognise your UBB stick.
This is usb 1.1 .... good for sensors, keyboards and other low bandwidth applications. Thanks for great content !
It claimed to be USB2.0, but it clearly isn't.
I know this is usb 1.1 max not usb 2.0 . Only for mouse or keyboard. Test with a long wire it is ok
Yes if you looked at LSUSB it would show up when plugged in as a USB1.1 hub, possibly 3 port or 1 port. Then the other side is also a USB 1.1 hub, though it might not enumerate as it is now in a pass through mode with the floating pin.
I have an old 8 port USB hub that I use as an extender, because it is perfect for slow peripherals like keyboards and printers, and as an 8 port hub it also does double duty in allowing USB powered devices like a small wood grain LED clock to plug in, and also will support slow charging of things, like NimH cells in a modified charger.
@@SeanBZA LSUSB doesn't recognize it at all.
It only recognizes whatever is plugged i at the farend.
I'm increasingly convinced this is just a couple of amplifiers - totally unaware of the USB protocol.
LOVE the pointer!
I wouldn't be surprised if that chip is literally just amplifying the incoming differential signal as hard as it can, clipping at the rails, and blasting that down the cable with no regards for SI. That would definitely give the results you're seeing. EDIT: Yep that transfer speed benchmark absolutely says it's falling back to USB 1.1 speeds, meaning this thing is absolutely trashing the signal.
*MB* is mega *_BYTES_* per second. - 11:27
Lower case *b* would be bits.
MB/s = Mega bytes per second
Mbps = Mega *bits* per second (Mb/s is less commonly used and can cause confusion with MB/s)
I know you know this but just thought I would mention it for those that don't know or are confused by these terms.
I introduce you to the world of microsoft, where MiB's are called MB's. Microsoft singlehandedly caused half the world's binary-base10 confusion
I was suprised how well those work for not too intensive purposes. I work at a radio station, and we use a crap load of those!
Hey fellow beer drinking local. I did this once upon a time to run a 30' USB cable to an old USB webcam (ghetto office security camera). Worked good enough.
I just soldered the ends together... Worked pretty good for a 240P webcam.
🍻
Didn't check voltage drop because I didn't have to check for any problems. =D. Win!
what can I say I never really expected it to work, but failing on a metre of cable is just well astonishing how many good reviews on ebay did it get??
I'm glad the beer worked out, the connectors did'nt. LOL.
I got one with the idea of using it to extend the cable from a USB camera. I only got it to work with a powered USB hub and, like you observed, the resolution from the camera was crap when the cable was connected.
When you opened that thing up and I thought it was just splitting out the usb to ethernet, I cracked up. I kind of wish it was that. Thanks for the entertainment
It probably wouldn't work any worse if that's what it was.
I use a similar looking thing to run from my computer... over drop ceiling in my home office to my electronic bench. I have it hooked to a powered hub and use it program arduinos, esp32s, etc from my development computer on other side of the room. Probably only about 30ft of ethernet cable but it does the job perfectly. I got ripped off though I think I paid like $15 for mine on amazon. :)
At least your works properly...
I have used these things but only for printers which use a relativity slow bit rate. They are useless for file transfers or USB memory sticks.
You need to unroll 1/2 the cable and roll back up in the other direction, this should cancel out the electron spin! ;-) strange beer you have there as well.
Next to my NAS and router, there is a UPS. The UPS (it's old) has a USB to monitor it. A few UTP-cables were laying dormant and I used one of these extenders. The cable length is 15 meters, 16 yards, or 50 feet. It works fine and the UPS reacts instantly, but how fast the data is transferred I do not know. Did you only fry a bit of the chip(s)?
I'm still waiting for the test with the 75 meter cable ?
It 100% didn't work.
I decided it wasn't interesting showing nothing happening, so it got edited out.
Propogation time on a long cable will create limits on data capacity especially on dynamic data like the camera. Where a handshake is needed the long propogation time makes handshakes take long.
One further issue is that the USB may be dropped to the 1.2 megabit/sec. The resistor on the USB lines should set the capable baud rate.
Personally I have only ever used the max15m special buffered long USB cables, not the RJ45 lan type cable extender
Just a tip when you use these passive USB extenders... Don't!
If you do anyway, plug the "active" part in as close to the computer as possible, no extensions at all.
@pileofstuff - I am curious what load you can get at 5v on the end of that grey extension cord - probably not enough ?
Looks like it would be faster to transfer files with a 1.44MB floppy. I wonder if there's a legitimate version of the product like all those old USB to RS232 or Parallel etc things I used 10-15 years ago no problem.
I bought the same sort of device on Amazon back at the beginning of the pandemic. I had an identical experience to yours with a camera. When I came across this video, I thought, optimistically, that you could actually get it to work better than I was able to. Alas, we were both taken in by false claims: I more than you, because I paid closer to $5 for the experience.
Bummer. I find it hard to imagine that the crossover cable fried the chip. It does still pass data accurately; just too slowly. Would have been handy....
Useful for connecting a mouse to DVR/NVR
Sounds like USB 1.0 speed support only. I wonder how the other chipset would perform?
Maybe, but I'm not convinced this thing is even aware of the format of the data it is passing. I suspect it is primarily an amplifier or line driver type thing.
I'm hoping someone can have more success than I did finding a datasheet.
@@pileofstuff Based on the benchmark results you were getting I am guessing this extender only supports USB Low Speed and USB Full Speed and you were trying to use it with devices designed for USB High Speed.
Since there is an active IC in the middle of the connection that IC would be what both the USB device and PC are negotiating with and would determine the USB link speeds supported.
Low Speed goes up to around 1.5 Mbps (or 187.5 KB/s) and Full Speed goes up to 12 Mbps or 1.5 MB/s. These are maximum rates that do not account for any overhead - real world speeds for Full Speed is around 1 MB/s which is almost exactly what you got on your benchmark.
You should be able to see what USB link speed the devices are using. On most Linux distros lsusb will give you that information.
In my experience it is fairly rare for USB extenders to support High Speed USB mode - particularly for a low price; and even rarer to have any support for any of the more modern SuperSpeed modes.
These devices are perfectly fine for serial-based communication such as management ports for enterprise network equipment or managed AV equipment which often don't support anything faster than full speed anyway. (Earlier generations of such devices used RS232 ports for configuration so raw speed is almost never a big concern.) They also work fairly well for USB input devices such as Mice and Keyboards which are often low speed devices operating at 1.5 Mbps.
@@pileofstuff I have some usb isolators that limit to usb 1.0 but otherwise don't understand the data. It is possible (likely?) these are very similar ICs. I wouldn't be surprised if the only thing these chips do is the initial usb handshake (which sets the speed) and maybe some sort of basic amplification.
BTW, if you did want to investigate more, ben eater recently published a couple videos on USB that explain handshake and the protocol used by keyboards well enough to probe the ethernet cable with an oscilloscope and get an idea what is happening.
@@circadianrebel These don't seem to be USB aware at all. They appear to be dumb pass-thru amplifiers at best.
Chip probably only supports USB Low Speed.
I honestly didnt even know there were that many types of beer, as they certainly arent sold here
These specific ones probably aren't widely available, but I bet there are some local microbreweries near you that have something interesting and different.
@@pileofstuff people here aren't choosy about beer, but have ton of hard liquor!
Cheers for sharing, made me wonder if it would maintain equal slowness over the long Ethernet run? It might be worth it if it’ll do a really long run?
I tried it with the long cable and the test timed out after 30 seconds during the first sample every time.
Didn't think it was interesting enough to include when the short cable already shows how bad it is.
What did you expect ? The full 144 MBit ? .. Still... to get some temperature (or whatever couple digit number) data from a distant micro controller that has USB but no ehternet/wifi these would do the job ?
Would be cool if you try that out with the long cable.. like controlling 4 servos dangling from the ceiling or something ;)
For a microcontroller it would be better (and cheaper) to just use a RS485 serial interface, you can get a USB converter for a few dollars and a transceiver for the microcontroller, and it's far more reliable
Voltage drop. There is no provision for restoring the voltage at the device end. Heavy loads may pull the voltage below 4.5v
Probably with the long cable, but I don't think that was the issue on the 1 meter cable.
@@pileofstuff *even if* the cat5e cable was CCA (copper clad alu), the voltage would not have dropped this much as the load on the other side is only a couple miliamps.
I’m almost betting it’s trying to use RS-485 for the distance comms. But it might be a timing issue. I have run into timing problems when using RS-232 ports and RS-485 null modems
It doesn’t look differential.
I’m wondering if there was some termination resistors on both ends, the data throughput would increase? Say 120 ohms.
I suppose that would depend on the impedance. that the mystery chips are designed for.
I'm still holding out hope that someone will have better luck finding a datashet than I did.
Wondering if the trace difference between the PCBs is telling the chip which end it's functioning at... pulling pin 6 to ground signals it's working on the device side vs the host side (or the other way around, don't remember which had pin 3 grounded..)
I assume that's what is happening.
Still haevn't managed to find a datasheet for the chip to confirm.
id like to try these out and do a review of them
The link I used to buy them is in the description of Mailbag Monday 120 th-cam.com/video/wdNqhTX7Be4/w-d-xo.html
have a pair of these somplace.these dont work for arduino usb serial purposes
why doenst a usb over ip wireless solution exist yet?
There is a way of doing USB over WIFI (just found this video yesterday)
th-cam.com/video/I5zA1lU5Tw0/w-d-xo.html
Do you think using STP cable would make a difference?
Over 1 meter, probably not.
But when I tried the long cable (off camera) which is STP, it timed out on the benchmark test.
so, any chance the rj45 cable was too short ie not enough resistance for the line drivers to work with the receivers properly? looks like it works in an identical manner to the chips datasheet you DID find, just a different pin out?
I'm not convinced it is sophisticated enough for too short a line length to cause problems.
Maybe if it was operating in RF frequencies, but that isn't what's happening here.
@@pileofstuff While USB does care about line resistance, it does not care this much. The computer's root hub will have mitigations for a lot of bad stuff.
could be the white usb extension. I have issues running my usb cam through them. I wanted a cam on my laser but the extender messed it up.
laser printer or cutter ?
That extension is the part of my test I trust the most. It can pass 1 amp with negligible voltage drop.
I wish I could find another one like it.
ANDI WOULD BE GREATFUL TO YOU
IF YOU CAN MAKE A VDO ON HOW TO EXTEND USB CABLES PROPERLY
The longer the Cat 5 the smaller the picture?
I assume the camera handshakes with the computer to decide on a max data rate.
@@pileofstuff All I visualized was a one-pixel picture on the screen if you hooked the full coil of wire.
Your testing passive devices. How about powered devices?
Parts ONLY.
maybe it needs a certain category of ethernet cable ?
Possibly. I was using CAT5E, so better than the bare minimum. Plus at the length I was using, it shouldn't make a significant differences.
My guess here is that this is great for a keyboard or mouse, but that’s it. It likely ends up looking like a low speed device.
Quite possibly. I can't see it working for anything more.
The chip on the extender may have been damaged by the wrong ethernet cable.
That is a possibility, but I suspect it would just be fully dead if that was the case.
I BOUGHT IT. THAT'S WHY I WAS WATCHING THIS.
BUT DIDN'T WORK ATALL.
MAY BE I ALSO USED A CROSSED CABLE.
I WILL CHECK THAT.
BUT PLS LET ME KNOW A GOOD BRAND
I guess there is no need to carry on watching this because i was here for a USB extender video not a Stout video.
Sorry if 20 seconds of a 14 minute video wasn't to your liking.
@@pileofstuff Dude it was a joke... I watched the whole video, I was just messing with you. :-)
@@DeanoEssex Cool. It's hard to tell with only text some times.
does it work if i use to extend a usb wifi ?
It didn't work for anything.
remember this is the cheapest thing I could find on ebay.
There probably are other models that actually work.
Check further down the comments. Someone suggested one that they use.
@@pileofstuff May I ask you this. I want to extend the cable of usb wifi for my laptop and put it higher and farther. Do you know any good solution to do that? . Thanks
also , skip all this chip nonsence and just use a cat5 cable ,use a twisted pair for ground, 5v ,data and data. 2 wires will double the wire gauge and you will have less voltage dropover the distance
I would recommend only using one pair of twisted wire per line - as Bob suggested. You might think that using one wire from each pair would be better as the twists cancel noise, right? Well; USB is differential. Doing that would also cancel itself out. No bueno!
Do yourself a favor, and either cut off the crossover end of that cable or just throw it away. They're essentially useless on all modern equipment.
I still need them occasionally at work on some of the legacy equipment.
Normally I have my crossover cables labelled. Not sure how this one slipped thru the cracks.
its for max. 50m not 75m distance
It doesn't even work over 1 meter...
This is obviously a cheapass Chinese knockoff from a decent product unless the crossover cable made all the difference. I wish you had tried it with a standard network cable. They come with modems and over the years I have gathered quite a lot of them, even if they are only 1m long.
Yeah, I wish I tried the standard cable first, but these things happen...
The crossover cable is used for computer to computer traffic to avoid using a network switch, because you are using a standard USB device then the data cables are switched and the computer is unable to recognise the device. It would still work if you only needed to power the device.
That being said I doubt that you have done any permanent damage to the USB extenders so I would simply try a normal network cable and see if it will recognise your UBB stick.
Everything after I said "oh Shit" was done with a normal straight thru cable.
Dont trust people who drink stout.