Questions regarding the implementation of Digital Discovery Logic Analyzers with a Raspberry Pi 4

[Arin George]

Our team at the moment is trying to determine whether a single, remotely-accessible headless Raspberry Pi 4 could possibly drive two Digital Discovery logic analyzers in parallel(both sampling at the same time) allowing for a total of 64 sampling channels. We would like to receive some insight on whether this implementation could be conducted in a stable and efficient manner based on the hardware and software capabilities of the logic analyzer and Raspberry Pi in combination. 

If this implementation is possible, would it require two instances of the Waveforms software or could both analyzers operate under a single instance of the software?

Finally, in order to utilize the full functionality of the Waveforms software, or at least that which pertains to adjusting and automating the sampling and capturing of signals, is it assumed that the Pi will need to have a GUI in effect? Initially we were looking for an analyzer which could be driven via console commands exclusively, but we can set up the GUI if it is necessary for the Waveforms operation.

Any help and clarification would be greatly appreciated. Thank you!

 

Edited June 1, 2022 by Arin George

[attila]

Hi @Arin George

See the following post:

 

[Arin George]

Hi @attila

Thank you for the response. We were able to run some testing on our Raspberry Pi with two logic analyzers sampling at the same time. We ran into a few questions along the way. 

• We wanted to know what exactly happens when an analyzer's sample buffer becomes completely full? Is there a brief delay in the sampling and acquisition of the data in order for the data to be sent to the PC, displayed through Waveforms, and to clear the buffer? Or does the analyzer continually capture data with some other buffer or process?
• Also regarding the buffers, if the user designates a certain number of capture buffers and they all become filled, what happens if the mode is "repeated" and the sampling is not stopped? Are the previously filled buffers overwritten with new data? And if so, are only certain buffers changed?

Any help would be greatly appreciated.

 

[attila]

Hi @Arin George

The Record (Sync) and Scan Shift/Screen modes continuously stream data (more accurately transfer in chunks) from the device. The device memory is used as a FIFO. In the WF app record mode displays warning when samples are lost (fifo overflow), in the SDK/API FDwfDigitalInStatusRecord lost and corrupt arguments.
The Single mode captures in device memory (buffer) and the data is transferred to the host only after the capture is done, then a new capture is started (automatically). So the gaps between repeated captures are not sampled.

[Arin George]

Great, thank you for the clarification!

[Arin George]

Hi @attila

After some discussion, we have a few more questions which we hope you can answer.

1. For testing purposes, are there specific conditions which would cause a FIFO overflow? How often could an overflow occur?

2. Is there any way to control the size of each chunk when the device captures in Record, Sync mode, or Scan mode? For the sake of clarity, are these "chunks" the same as the "buffers" mentioned in Waveforms?

3. How is the number of chunks determined? Is it configurable by the user interface or fixed? I was thinking the max amount is 999.

4. Does this analyzer in question support USB 3.0?

Our goal with the analyzer is to achieve something similar to a data logger(no gaps between captures) without losing too much in sampling performance, if this is possible.

[Arin George]

Is there anyone who can assist with this?

[attila]

Hi @Arin George

1. If the device buffer of 256MiB for Digital Discovery is overflown, when capturing above this size and the USB transfer can't keep up.
2. Chunks are the new samples in each status/read iteration.
3. Automatic.
4. USB 2.0 HS 480Mbps 

This is a 32b 256Mi 1GB capture, 4X the device buffer size is working up to ~10MHz (~45MBps) (pic1) but at 20MHz buffer gets overflown (pic2).
Note that for RPi the USB and processing rate specially with 2 devices will be lower than with a computer.