usb-6251 works differently than PCI 6250 for multiplexed analog inputs

[Rafi2003]

HI,

I have a system that works fine with PCI 6250.  We purchased the USB-6251 in order to run the application from a laptop as well.

I'm sampling a 500 us pulses running at 14 hz.  I get good results with my PCI card when the sampling rate is 58000 hz.  However, with the USB card the pulse width I'm measureing is much lower than the right value (132 usec).   The value is getting better when I decrease the sampling rate.  For exaple at 15000 the pulse widht is OK.

I have 4 channes with the above described pulses, and i'm multiplexing 2 channels at a time.  My application subtract one ch from the other in order to measure the voltage between the 2 channels.

I read abaout the gohsting issues when multiplexing channels.  The fact is that the same hardware is giving good results with the PCI 6250.

1.  Is it possible that the gohsting issues is much more severe withthe USB 6251?

2.  Any other ideas why I'm getting different results?

Thanks

RK 

[Brandon_G.]

Hi RK,

Thank you for your question about the differences between the USB-6251 and the PCI 6250.  Both devices have the same spec on the settling time for multi-channel input readings so I am surprised you are seeing different results.  Are you using the same 4 channels on both devices?

Can you explain a little more your 4 channels with multiplexing 2 channels and then subtracting them?  Technically, you can only measure voltage on one channel at a time, unless you are doing a differential measurement.  Are you performing a differential measurement when you say, "My application subtract one ch from the other in order to measure the voltage between the 2 channels"?

Also, as a possible workaround, have you considered setting up a counter task to measure the pulse width?  The counter inputs aren't subject to ghosting like the ai inputs so if you are interested in the timing parameters, I would consider a counter task.

Lastly, can you post some screenshots of your 6251 waveforms vs your 6250 waveforms?

Thanks RK!

Brandon G