Author Topic: capturing a trace  (Read 9816 times)

rebar

  • Member
  • ***
  • Posts: 17
capturing a trace
« on: August 07, 2014, 03:04:40 pm »
hi - I have beagle 480 and also the power-480 analyzer. I am working on a usb problem where the test is set up to run for a long time. If something fails, the test will halt. With the regular 480, I can run in circular mode, but if the test stops and I'm not around, the "useful" part of the "failure" will be overwritten with NAKs if I don't get to it soon enough. With the power-480, there is an option to off-load to the analysis PC, but evidently only to the RAM and not the hard-drive, so once the RAM is filled up, the trace will stop. I think if I could filter out NAKs, this might work, or if I could have the analyzer trigger after no-data for say 5 minutes, that would work as well. I haven't been able to get this to work and I'm curious if anyone has come up with a way to do this. I've been in touch with totalphase, but I don't see a solution in sight. Thanks for any suggestions!

Barry Twycross

  • Frequent Contributor
  • ****
  • Posts: 263
Re: capturing a trace
« Reply #1 on: August 07, 2014, 03:40:30 pm »
I've done this sort of thing, not with a Beagle though.

If I'm remembering right, one time I made the computer shut down after I detected the event. Suspending the port/computer would probably work to do that as well.

Other times I had a scope in the setup as well and managed to send a command to it to stop capture (The scope had a USB488 interface to send commands to.) I've also used scopes to trigger the analyzer through the analyzer's external input (don't know if yours has an input like that). If I tried hard enough I could probably hook up some USB hardware to send an external trigger to the analyzer. I've also used a spare GPIO on a device to trigger the analyzer.

I've also send out a unique data pattern over the USB after the event happened and had the analyzer trigger on that. A vendor specific setup command that no one else is likely to use is useful for that.


rebar

  • Member
  • ***
  • Posts: 17
Re: capturing a trace
« Reply #2 on: August 07, 2014, 05:48:48 pm »
Thanks Barry - All good ideas! I'm hoping that maybe I missed something on my analyzer setup (I can wish.). That would be the easy way, but it might not be possible...