analog discovery 1 pattern generator wait state

[blindgoat]

We're using an Analog Discovery 1 as a delayed-trigger oscilloscope.  We route the incoming trigger to the digital pattern generator, delay it by a user selectable amount, and send the result to the analog scope channel two as a trigger.  This allows us to sweep the 8192 sample buffer of the analog scope along a repetitive trace and capture more samples, one buffer at a time.

Our question is about the accuracy of the digital pattern generator's wait state.

1. Is the wait delay determined by counting down from a clock?
2. If so, which one?  The 100 MHz master clock?
3. Does that mean you have implemented a 36 or 37 bit counter in the system, in order to reach one day (86400 seconds) with 10 nS precision?

We'd really appreciate any information on this topic.  Right now we're not relying on the wait state out of uncertainty about how it's determined.  If we could use wait, it would make our lives and our code nicer

Thanks.

[attila]

Hello,

On Analog Discovery the Scope, Wavegen, Logic Analyzer and Pattern Generator use the same 100MHz clock.
The Scope and Analyzer have 31 bit post trigger sample counter. Acquiring at 100MHz the trigger delay can be up to 10sec, at 1MHz 1000sec...
The Wavegen and Pattern for wait and run length have 43bit counter. This ticks at 100MHz allowing up to 1day spans.

[blindgoat]

Exactly what I need to know.  Thanks for the helpful answer and the capable hardware.