Tom Van Baak said: > A sampling rate of 1 Hz would normally be enough, but when the CBT has been > off for so long there are interesting pulses of CBT oven and Ion current and > so maybe 10 Hz or even 50 Hz would be better to capture that detail. One approach I've had good luck with... Sample fast but only log data if something changes. You probably will want a filter on "changes" to ignore tiny wiggles in the bottom bits. That probably won't work as well for your one-shot case since you won't have an opportunity to calibrate the filters. -- These are my opinions. I hate spam.