Davej said:
Well, it is interesting that while I can google and find various
inexpensive products which claim to extend USB 2.0 it is much more
difficult to find such products on a respectable site such as NewEgg.
Here is one on amazon, and although it is not the extra-long Cat5
type, there are some positive reviews regarding use with some webcams.
http://www.amazon.com/PTC-Meters-Extension-Repeater-Supports/dp/B0040IASMS/ref=pd_vtp_e_5
Unfortunately, that is not a Cat5 style device.
There are two kinds of USB extenders. The one on the Amazon page is a
"Native" extender. The wires carry standard USB electrical signals
at all times. The one in particular you have selected, could be
one that "cheats". The USB spec says an active repeater can be up
to 5 meters long. This is a non-cheating one, with one "lump".
It is 5 meters long.
http://www.newegg.com/Product/Product.aspx?Item=N82E16812120087
This cable is 10 meters long, and nobody seems to notice an
extra "lump" on this one. For that to work, the time of flight
of the electrical signal, would need to be half of what it was on the
previous one. I can't understand how this is possible.
http://www.newegg.com/Product/Product.aspx?Item=N82E16812156026
And this company, sells some different ones. You can connect four of
their 5 meter products, four of their 10 meter products (one lump),
two of their 20 meter products (two lumps each?), or use one
of their 50 meter products (five lumps ?). Adding an active hub
to the end of the 50 meter product would not work, so only
a 100mA load could be put at the end of the cable (or, you
could arrange to power the device yourself).
http://www.vpi.us/usb-extender.html
*******
To detect cheating, you look for "lumps" in the cable. The lump
has a one port USB hub chip inside it. And that is used for
regeneration of the signal (plus the usual hub thru-delay).
In this example, the lumps do digital regeneration. If you
had a program which listed the USB devices, each of the
lumps would register as a "hub".
USB USB USB USB
Host ----- lump ------ lump ------ lump -------- device
In the Amazon picture, I see a lump in the middle of the cable,
plus a lump on the end. Since the cable is coiled up,
I cannot determine if there are more lumps or not. It could
be two 10 meter designs placed end to end, like the Sabrent one.
The 5 meter limit on USB, is a time domain limit. It assumes
a certain dielectric constant and time of flight. With an exotic
dielectric, perhaps the distance can be changed a bit. But I doubt
you could change things by a factor of four. And therefore, the
cable in question is undoubtedly cheating by having multiple
regenerators.
The Cat5 solution, changes transmission methods, and gets the
extra reach that way. I'm not sure how they manage to beat
the latency requirement.
USB Ethernet_over_cat5 USB
Host ------ Cat5_Box -------------------- Cat5_Box -------- device
You can buy 5 meter active repeater solutions, with one "lump" per cable.
Since the regenerator is on the end of the cable, you can have 5 meters
more of passive cable to the webcam.
USB
Host ------ lump ----X <---- Plug in another passive cable plus webcam, here
The thing is, bus powered devices have a current limit on them, so
once you start using extenders, you have a 100mA limit. This is not
sufficient for any good USB webcams (ones with motors and autofocus).
To solve that problem, you need an active hub down at the end. The
active hub, has its own wall adapter. That changes the current limit
on the active hub, to 500mA.
USB USB USB USB
Host ------ lump ----X ----- (passive cable) ----- Active_hub ------- webcam
| 500mA
|
Wall adapter
5V @ 500mA or greater
Now, the other detail, is the maximum number of hubs/regenerators, allowed
between a host and a peripheral. And this is further complicated, by
certain recent Intel Southbridges, where the idiots burned up a hub
inside the Southbridge, cutting down the max number of repeater cables
by one.
http://www.usb.org/results?q=faq&submit=Search&siteurl=www.usb.org/home&ref=
"I really need to put a USB device more than 30 meters away from my PC.
What should I do?
A: Build a USB bridge that acts as a USB device on one side and has a USB host
controller at the other end. Use a long-haul signaling protocol like Ethernet
or RS-485 in the middle. Using cables or short-haul fiber, you can get ranges
upwards of a kilometer, though there's no reason why the long-haul link in the
middle of the bridge couldn't be a pair of radio transceivers or satellite modems.
Embedded host solutions capable of doing this already exist. Also, two PCs
connected via USB Ethernet adapters are essentially a slave/slave version of
this master/slave bridge."
So that tells you, the max reach (according to the USB.org site) using regenerators,
is 30 meters. If you want more than 30 meters, then it's the "Cat5 bridge" solution,
which doesn't use USB to make the extra distance. If we use those "double-reach"
native USB cables, it would seem we could hit 60 meters.
To do the 30 meters, you'd need four repeater cables, two passive cables, an
active hub with 5V power adapter, and the web cam. I'm drawing this, to match the
30 meter claim on USB.org.
Host ----- lump ----- lump ----- lump ----- lump --------- active --------- webcam
hub 500mA
<--------> <--------> <--------> <--------> <--------> <------->
5 meter 5 meter 5 meter 5 meter 5 meter up to 5
active active active active cable meter
repeater repeater repeater repeater cable
If you have the Intel chipset with an "internal" hub in the Southbridge, you
have to remove one of the repeater cables, and the total reach is 25 meters
instead of 30 meters. Note that this is a "feature" on recent chipsets. If
you had a motherboard from the LGA775 era, chances are you could build
out the full 30 meters. On AMD chipsets, they might not have done
the same thing as Intel did, so they might be 30 meters as well.
By comparison, the Cat5 bridge, doesn't have these limits. But may still
have practical limits, if the device has its own response latency
requirements (buffer overflow). So doing it this way, for $300 or so, removes
the 25 meter or 30 meter limit, and takes you to 100 meters over copper. (And a lot
further with a $900 fiber setup.)
USB Ethernet_over_cat5 USB
Host ------ Cat5_Box -------------------- Cat5_Box -------- device
electrical signals
There are a number of aspects of this question, I cannot adequately
answer. Time of flight is proportional to SQRT(Er) and looking at
some velocity factor values, I don't see a span of a factor of two,
between the best and worst cable dielectrics. So I don't understand
how they can make a 10 meter length with one regenerator on it. And
on the Cat5 device case, if the left hand Cat5 box terminates the
USB protocol, on behalf of the right hand end, aren't there going
to be situations where that will foul up ? I can't claim to understand
even half of the details here.
What I can say, is reading the product reviews is important. I have
seen cases, for the 5 meter products, where the product fails when
you concatenate them. So check the reviews, and if you see lots of
complaints, it could be the +5V and GND wires are too thin in
the cable.
HTH,
Paul
