USB 6363 Task - Bandwidth Limitations

[jmaslek]

Hi NI Forums. I am currently working with the USB 6363 and I am trying to generate some funky output waves with the MatLab Data Acquisition Toolbox.

However, I am running into an issue where I am using too much memory/bandwidth?  About a week ago, I was able to generate sessions where AO 2 and 3 on my device were outputting for .25 seconds at 1Mhz rate (so 250k) points.  Then out of nowhere, I could not achieve that.

I went into MAX and created a Task, with the Generation mode set to N samples, and for my AO, if I use more than 400 points (independant of rate), I get the error:
Error -200946 occurred at DAQ Assistant

Possible Reason(s):
Task could not be started, because the driver could not write enough data to the device. This was due to system and/or bus-bandwidth limitations.
Reduce the number of programs your computer is executing concurrently. If possible, perform operations with heavy bus usage sequentially instead of in parallel. If you can't eliminate the problem, contact National Instruments support at ni.com/support.

I was hoping for some advice on how to troubleshoot this.  I went into MAX and rese my device, I have tried changing which USB port it is plugged into.  I reinstalled the Daqmx drivers and I still run into this issue 😞

[A-Guz-Man]

Hi jmaslek

The error description sounds like the hardware is pulling more information that the one is been generated at the computer by software. Do you happen to know the generation rate of the Output waves? Or the size of the buffer the signal is being saved?

Definitely sounds weird that it was working and out of nowhere it stopped, do you remember anything else that you changed?

Also, this error happens with both AO 2 and AO 3 right?

[jmaslek]

Thanks for the reply.  

As I have created an NI task, I have the generation mode set to N samples, which I can only do about 400, when I was easily reaching 250k.  This is independent of the rate.  I am not sure how to check the buffer size from that.

The only other thing we did was add a USB extension cable, but it is the same USB 2.0 that can do 12 MBps so that seems highly unlikly (as we have also switched the cable).

This issue happens on all of the AO (0-3).  Actually, I can write 400 samples on 0 and even less on 1,2 and 3 😞

[A-Guz-Man]

Are you able to generate an output signal with more that 400 points at all? I mean using other hardware if you have.

I'm trying to eliminate the possibilities and if you are able to do that then it makes me that the problem isn't the computer. We could still think of:

1) configuration of the hardware (what might have changed?)

2) this specific hardware not working or somehow the output buffer buffering too fast independent of the configuration.

The buffer size is the same to Samples to Write when the Generation Mode is set to N Samples so that configuration shouldn't be a problem either.