...Sounds an interesting test, why not do a case study? :) :) It's quite interesting indeed, but unfortunately I'm not at liberty to discuss the details. It would end up a very boring case study, little more than "I needed to capture what happened before the signal went high, and here's the trigger...

Thanks Mike, the Window Dropout looks like it's just what I need.

Funny, I'm certain that I had tried it previously with no success... I guess I must've screwed it up somehow. Oh well.

Testing resumed.
Thanks again.
Tom

Mike, thanks for your help. Yes, I've been unclear. What I need is an ending trigger, that is, as soon as the voltage stays high for 5 seconds I need to capture the data that's already been recorded: the previous 30 seconds or so.

hbr, yes, thank you, that was my problem with aliasing. I didn't realize that the number of samples referred to the time scale shown in the window. With 10 s/div and sampling at 1 GS I get good resolution. I still haven't solved the end-of-sampling issue: I can't be there when the event occurs. I ne...

I've got a breakout harness in the sensor, and I know it's wired correctly. It looks like the Automotive views have aliasing issues that are not present in the regular Scope views. Here's what a cycle of Crank and Cam W look like on Ch. A, at 500 S/div (2 separate shots because I can't get to Ch. B ...

Yep, that works, thanks. I'm looking at the crank sensor voltage, and tried to use Window Dropout so it would trigger when the voltage stuck between 4.5 and 5.5 V for more than 10 seconds (indicating that the engine had stopped for that long), but nothing happens when that condition is met. So unles...

When the trace gets to the end of the screen there's about an 8-second time lag, during which no data is captured, before the second buffer screen starts populating. There's a very high probability that the event will happen during that time. Is there no FIFO buffer that elimitates holes in the data?

I'm trying to identify something that happens very infrequently, and would like to be able to trigger manually on the event and have only the last 15 seconds (or so) recorded (rather than just record for hours at a time and end up with a huge data file). Is there a way to do that?