USB interface cables?

Advertise with us
shop deals on amazon
Shop deals on ebay
GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
ForthMonkey

ForthMonkey

Well-known member
Joined
Sep 4, 2013
Messages
1,115
Location
Turkey
Hi,

I want to ask something.

There is no info about it.What about USB interface cables? I know mic & inst cables affect to sound.But i don't know anything about USB cable.I'm using stock USB cable.But there are lots of expensive USB cable.Worth to buy instead of stock cables?

Thanks!
 
ForthMonkey said:
There is no info about it.What about USB interface cables? I know mic & inst cables affect to sound.But i don't know anything about USB cable.I'm using stock USB cable.But there are lots of expensive USB cable.Worth to buy instead of stock cables?
Click to expand...

The stock cables are generally just fine.

The "audiophile" USB cables are a complete rip-off. Even the Belkin cables on sale at the Best Buy for $30 are about an order of magnitude too expensive.

NB: USB's maximum cable length spec of 5 meters is determined by the need to turn around a data packet acknowledge, and not with cable capacitance or other concerns. Thus, extending the cable beyond that limit is pretty much a guarantee of failure.

-a
 
abbey road d enfer said:
Andy Peters said:
NB: USB's maximum cable length spec of 5 meters is determined by the need to turn around a data packet acknowledge, and not with cable capacitance or other concerns.
Click to expand...
Andy, can you explain that?
Click to expand...
I'm no Andy but USB uses ACK and NAK  handshaking to acknowledge that good data transfer occurred (or not). If the sender does not receive an ACK in the timing window expected, it interprets that as a data error occurred and will probably try to resend the same data. 

I suspect this length is related to max data rate, and shortest timing window to receive a good data ACK, but this is mostly a guess. Andy has actual experience with this stuff.

JR
 
JohnRoberts said:
abbey road d enfer said:
Andy Peters said:
NB: USB's maximum cable length spec of 5 meters is determined by the need to turn around a data packet acknowledge, and not with cable capacitance or other concerns.
Click to expand...
Andy, can you explain that?
Click to expand...
I'm no Andy but USB uses ACK and NAK  handshaking to acknowledge that good data transfer occurred (or not). If the sender does not receive an ACK in the timing window expected, it interprets that as a data error occurred and will probably try to resend the same data. 

I suspect this length is related to max data rate, and shortest timing window to receive a good data ACK, but this is mostly a guess. Andy has actual experience with this stuff.

JR
Click to expand...
OK, I checked the USB standard; the maximum one-way propagation delay is 30ns, including the interface delays. So the cable delay must be less than 26ns. Standard USB cables lineic delay is around 5ns/meter, hence the 5m limit. It is not impossible to imagine a "faster" cable that would allow some length extension. But it's clear that it's a more drastic limit than that given by capacitance.
 
abbey road d enfer said:
JohnRoberts said:
abbey road d enfer said:
Andy Peters said:
NB: USB's maximum cable length spec of 5 meters is determined by the need to turn around a data packet acknowledge, and not with cable capacitance or other concerns.
Click to expand...
Andy, can you explain that?
Click to expand...
I'm no Andy but USB uses ACK and NAK  handshaking to acknowledge that good data transfer occurred (or not). If the sender does not receive an ACK in the timing window expected, it interprets that as a data error occurred and will probably try to resend the same data. 

I suspect this length is related to max data rate, and shortest timing window to receive a good data ACK, but this is mostly a guess. Andy has actual experience with this stuff.

JR
Click to expand...
OK, I checked the USB standard; the maximum one-way propagation delay is 30ns, including the interface delays. So the cable delay must be less than 26ns. Standard USB cables lineic delay is around 5ns/meter, hence the 5m limit. It is not impossible to imagine a "faster" cable that would allow some length extension. But it's clear that it's a more drastic limit than that given by capacitance.
Click to expand...

Yes, that's it (sorry, yesterday was busy so I didn't reply).

FWIW, if you manage to string together USB cables (with the illegal-according-to-the-spec extension cables) so they're longer than the 5 m limit, the bus really does fall down.

-a
 
Aaahhh, okay. I was already aware of the 5m limit. Now I see why.

However : another question, and I hope somebody here is experienced with this stuff :
USB over CAT5 cables.

I have my PC setup outside of the studioroom, so, I have bought a bunch of those USB over CAT5 converters.

2x PC> converter 15 mtrs CAT5 > USB hub > mouse& keyboard : fine, both sets work without any problem.
1x PC> converter 15 mtrs CAT5 > 2ch Midi Interface (only input to PC): no problem. However, when I send data to the Midi Interface, the timing isn't really stable, so I have simply used 5 mtr USB cable and 10 mtr midicables instead.

But now the last one , which is a bit of a problem to me:
1x PC> converter 15 mtrs CAT5 > midicontroller : the input to the PC fails after 10 minutes. The output of the PC to the controller keeps on working though. At this moment I have a 5mtr USB cable  connected directly to the PC, but I need it to be 10 mtrs....

Any tips about this stuff ? I don't believe in highgradeaudiophiliacmumbojumboextraterrestialwhatever cable for digital stuff, but if needed I'll buy something of course.

 
helterbelter said:
Aaahhh, okay. I was already aware of the 5m limit. Now I see why.

However : another question, and I hope somebody here is experienced with this stuff :
USB over CAT5 cables.

I have my PC setup outside of the studioroom, so, I have bought a bunch of those USB over CAT5 converters.

2x PC> converter 15 mtrs CAT5 > USB hub > mouse& keyboard : fine, both sets work without any problem.
1x PC> converter 15 mtrs CAT5 > 2ch Midi Interface (only input to PC): no problem. However, when I send data to the Midi Interface, the timing isn't really stable, so I have simply used 5 mtr USB cable and 10 mtr midicables instead.

But now the last one , which is a bit of a problem to me:
1x PC> converter 15 mtrs CAT5 > midicontroller : the input to the PC fails after 10 minutes. The output of the PC to the controller keeps on working though. At this moment I have a 5mtr USB cable  connected directly to the PC, but I need it to be 10 mtrs....

Any tips about this stuff ? I don't believe in highgradeaudiophiliacmumbojumboextraterrestialwhatever cable for digital stuff, but if needed I'll buy something of course.
Click to expand...

The USB-to-Cat5 converters look like hubs to the host computer. They extract the data from the USB packets and serialize them in some proprietary format and send it all across the wire. On the far side, the data are deserialized and the USB packets are reformed and so the connected USB devices think they're talking to a host.

But there's some penalty involved. There is likely to be some latency involved in the format conversions, and MIDI timing is tight it could suffer from the latency. I wouldn't expect cable length to matter over such short lengths but I've never done the tests.

I don't know what causes the failures.

But there are limits to the legal cable lengths, and unfortunately working around them for real-time isochronous systems might not be possible.

You might have to move the computer closer to the devices.

-a
 
Hi Andy,

Ah, okay. Thanks for explaining that!

I was already afraid I have to use an ordinary 5m USB cable, now I know there's no other way. Fortunately, I don't have to put the PC in my studioroom, with a bit of rewiring a 5mtr USB should be just long enough. But only "just".

By the way :
Afaik, midi interfaces don't work properly when connected to a hub.

But again, thanks for explaining !
 
Note the distance/timing limitation is based on an arbitrary interval for acknowledgment. A different interface protocol could support longer distances.  If there was enough demand, USB could be converted to some intermediate protocol, then back again at the far end. The delay or lag should be small.

JR

[edit- a longer ACK interval could reduce the through put rate so there would be consequences using similar media. [/edit]
 
helterbelter said:
Afaik, midi interfaces don't work properly when connected to a hub.
Click to expand...

I have a USB-to-MIDI converter (I forget the manufacturer) talking to a Mackie Control Universal and an M-Audio USB keyboard (which does MIDI-over-USB) and both work fine on the hub to which they are connected.

-a
 
I'm also thinking that Signal Integrity is also important for perhaps good reason.

Have a simple 3Mp USB camera on my trinocular microscope for SMD/SMT work that I wanted to get "into the box" for image capture.

Scope is physically less than 5m away but mfg supplied USB cable would not reach the PC

used a brand new modern day USB extension cable but total length was still *less* than 5m

Windows could not recognize the camera.

Removed the camera physically from the scope to get it closer to the PC as such to Plug the camera with mfg supplied cable DIRECTLY into PC

It worked!

Found a much older USB extension cable from when USB "was young" and _perhaps_ allegedly before manufacturers of USB extension cables learned how to cut corners and save micro-pennies per unit and say it's "good enough" for the average consumer....

It worked too! ....  With the older-school USB extension cable!

Although to be fair, I also tried a hub-as-an-extension cord (all with legal length of course) and it did not work. Older hub so maybe a speed capability issue?



 
Twenty Log said:
I'm also thinking that Signal Integrity is also important for perhaps good reason.

Have a simple 3Mp USB camera on my trinocular microscope for SMD/SMT work that I wanted to get "into the box" for image capture.

Scope is physically less than 5m away but mfg supplied USB cable would not reach the PC

used a brand new modern day USB extension cable but total length was still *less* than 5m

Windows could not recognize the camera.
Click to expand...

Since extension cables are explicitly disallowed by the spec, there's never a guarantee that they'll work.

Removed the camera physically from the scope to get it closer to the PC as such to Plug the camera with mfg supplied cable DIRECTLY into PC

It worked!
Click to expand...

As it should.

Found a much older USB extension cable from when USB "was young" and _perhaps_ allegedly before manufacturers of USB extension cables learned how to cut corners and save micro-pennies per unit and say it's "good enough" for the average consumer....

It worked too! ....  With the older-school USB extension cable!
Click to expand...

Lots of possible explanations. But whatever.

Although to be fair, I also tried a hub-as-an-extension cord (all with legal length of course) and it did not work. Older hub so maybe a speed capability issue?
Click to expand...

Did you try a powered hub -- meaning it uses a wall wart -- or not? Because without the power supply, the bus-powered hub won't provide the full 500 mA allowed by the spec to the downstream device. Now, you should've gotten an error message about that, but I've seen instances where a device didn't report its actual current consumption requirements so the operating system is lied to and as such doesn't throw an error message.

-a
 
Indeed this is interesting... the powered hub (powered with wall wart and power LED illuminated) also renders the camera "Unknown"...

However, the "50MHz" (100Msa) USB oscilloscope works fine on these *new* USB extension cables..

the camera does not...

I have 3 of these new USB extension cables and the camera will not work on any of them....

only on the older USB extension cable circa the early 2000s...

Things that are curious indeed...

 

Similar threads

RuudNL
Change the display name of two identical USB devices(?)
Replies
10
Views
563
mhelin
M
Bo Deadly
long usb-c cables with audio interfaces
Replies
12
Views
2K
Newmarket
Newmarket
B
Sold Motu 24Ao Interface
Replies
1
Views
208
bjosephs
B
B
PC audio test set
2 3 4
Replies
62
Views
2K
abbey road d enfer
A
B
Design of a sound card "Front end" for audio test set
2
Replies
38
Views
1K
blue_luke
B

Latest posts

Back
Top