Fast ADC for 20ms from 125MHz sample rate and ext trigger

[Pestatio]

Hi,

I'm struggling to find info on the 'ACQ:SOUR1:DATA?' ? At 125MHz the rate will be 8ns per sample, meaning a buffer of 16384 float values will be filled in 0.131072ms... The trigger is external, but I am using a normal IO to detect the trigger.... Is there an easy method of streaming the date the entire 20ms and plot it in matlab? If not, the I would like to know if somebody can inform me what happens when I query 'ACQ:SOUR1:DATA?'? Does it clear out the buffer and the readings start from the beginning? I am struggling to read the SCPI command sheet it seems as some of it is unclear to me.

Kind Regards

Peet

[Nils Roos]

Hi,

the scpi-server can only record one buffer of samples (16384) per trigger. Once the buffer is filled, all calls to 'ACQ:SOUR1:DATA?' will deliver data from that buffer, so calling 'ACQ:SOUR1:DATA?' multiple times without starting a new acquisition (with 'ACQ:START' and 'ACQ:TRIG...') will always return the same samples.

[chrispwkoe]

Hi,

i'm currently working on a project using the Red Pitaya and I think I will soon encounter a similar problem.
Have You found a satisfying solution yet?

If not, try the following approach (will only work with periodic input signals): You acquire one buffer of data (0,13ms),
then you send the buffer to the PC for evaluation, after that You increase the trigger delay by 8k (or 16k) and repeat the acquire/evaluate/increase
cycle until You have 2ms of data. You will, however, need a strategy to get rid of gaps at the end of one buffer.

Best Regards
Chris

[Amraam]

Hi,
my first experience with Red Pitaya and continuous acquisition (rp_remote_acquire) was a success (thank you Nils and all community).

Now I'm looking to acquire 8192 samples at 25 MHz each external trigger that occurs every 300 microseconds.
There is a bitfile able to do that or source code as a starting point that can I study?

Thank you in advantage

[Nils Roos]

Hi,

the standard osciloscope module can do that, the only problem with it is that you can't get the samples out of there fast enough. You might want to take a look at the axi adc extension that is no longer supported (here's the 0.94 code). It can record to external RAM in a triggered fashion. Like the stock oscillosope logic it does not re-arm automatically, you'll have to do that in software - poll the flag and reset the trigger when a recording is complete.

Here's a thread that has a lot of information about the logic and the accompanying example code, in the form of a somewhat prolonged question and answer session.