[time-nuts] Timelab, two SR620s and losing samples
john at miles.io
Sat Jan 16 21:33:54 UTC 2016
Agreed with Magnus that there are a lot of possible variables in your setup that need to be ruled out.
Are you using the SR620 driver in TimeLab, or did you find a way to get it to emit data continuously via the RS232 port for use with the talk-only driver? I've seen occasional instances where a counter has ignored every other trigger when used in addressable mode by TimeLab. It's not always readily reproducible, but the counter drivers in the current beta at http://www.miles.io/timelab/beta.htm seem to be a bit less prone to that behavior. If you aren't already using the beta, try that.
It's always safest to use counters in talk-only mode when possible, since that rules out any timing problems that might arise in a two-way GPIB or serial conversation where each individual reading is requested by the software as done by the 5313xA, Philips, DTS2070-series, and SR620 drivers. Some of those counters can be used in both talk-only and addressable modes, but I don't believe that's true of the SR620, unfortunately. All of the programming examples in the SR620 manual work by requesting each reading individually.
-- john, KE5FX
Miles Design LLC
> > I have here a setup with four (FPGA) nodes that produce synchronized pulses
> > with a 20kHz rate. I have two SR620s two measure those pulses.
More information about the Time-nuts_lists.febo.com