Question about USB-GPIB

[cykhung@my-deja.com]

This is regarding the National Instruments GPIB-USB-A. If the GPIB-USB
cable is not long enough, can I attach a standard GPIB cable onto it to
extend the length? In this connection, one end of the GPIB cable is not
connected to any device and would this cause any problem?


PC GPIB-USB-A
USB port <-------------> normal GPIB cable
<------------------------> device
not
connected
to any device


Sent via Deja.com
http://www.deja.com/

[LabVIEW_Joel]

Hello,

You should not have any problems. Just hook up the USB part of the cable to the PC. Then, on the other end of the GPIB USB cable you will hook up your GPIB only cable. Finally, on the end other end of the second cable is your instrument. Remeber that it is not recommend to have your second cable be longer that 2 meters.

[sajjad]

Hello Joel,
With regards to GPIB, i'm in a bit of a fix right now. I'm having problems with cable length in my setup. Most of the specs say that it shouldnt be more than 20ft but then i've heard stories that people ahve broken the max length rules and gotten away with it.
I'm attaching a file, which i'd like you to view, and then tell me if i can get away with these lengths of gpib cables in my setup.
BTW the x are devices, otherwise the connection is just a split to a device later on. I'm concerned with the first split, that you'll see in the drawing, and with the 2 meters each to the device at the bottom part of the network. If required, i can set the bottom three devices as a star network, although i dont see how that would solve the problem.
The
box on the left is the CPU/host.

[CraigGraham]

"sajjad" wrote in message
news:50650000000500000001590000-1007855737000@exchange.ni.com...
> Hello Joel,
> With regards to GPIB, i'm in a bit of a fix right now. I'm having
> problems with cable length in my setup. Most of the specs say that it
> shouldnt be more than 20ft but then i've heard stories that people
> ahve broken the max length rules and gotten away with it.
> I'm attaching a file, which i'd like you to view, and then tell me if

Whether you can go out of spec depends on the instruments, particularly on
the worst instrument on the bus. And on data transfer rate. "Suck it and
see". You seem to be asking if this will work- is it not possible to try it
first? I've certainly got away with this kind of layout but at other t
imes
I've also had problems, and had to rearrange things. Or does this
arrangement work for you and you're just concerned with long term
reliability?

I think daisy chain is the preferred configuration, not star. Can you not
put the GPIB interface at the enf of the chain instead of hanging off a long
wire from the middle, and then run a long USB cable to the PC?

--
Craig Graham
Physicist/Labview Programmer
Lancaster University, UK