Thanks for the heads up, that could cause a lot of confusion. You could try the CASPER FFT, at least you can fix any bugs without having to update your Xilinx install :)
We have, over the years, come across a few similar bugs. The most painful was a compiler bug which generated incorrect logic during optimisation. A fixed selection on a multiplexer would optimise the multiplexer away as expected, but would output the incorrect input. This was deep in the twiddle factor generation logic in the FFT, causing incorrect results. Locating the bug was interesting as it would mostly go away when other logic was inserted to locate it. Regards Andrew On Fri, Aug 14, 2020 at 5:55 PM Sean Mckee <[email protected]> wrote: > I'm working on a design where the input sources to the Red Pitaya are > switched at ~4Hz for calibration purposes. I was implementing functionality > to prevent data from one calibration state ending up in the accumulator > during the next calibration state, and I started getting some strange > behavior from my simulation. A bit of digging turned up this: > > > https://forums.xilinx.com/t5/AI-Engine-DSP-IP-and-Tools/Xilinx-FFT-Block-In-Out-Sync-problem/td-p/975620 > <https://forums.xilinx.com/t5/AI-Engine-DSP-IP-and-Tools/Xilinx-FFT-Block-In-Out-Sync-problem/td-p/975620> > > TL;DR: The start frame in/out on the Simulink FFT block do not behave > *quite* as advertised, and there is significant discrepancy between > simulation and hardware performance. If your design only needs 1 sync pulse > and then runs continuously, this bug does not affect it. > > Looks like this has been patched in Xilinx 2020.1, but then I would lose > support for all of the Casper goodies! Guess I'll try to get creative with > the vector accumulator triggers. > > -- > You received this message because you are subscribed to the Google Groups " > [email protected]" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/b75d7f6d-69ad-4c36-a10d-ccfff1c1cbefo%40lists.berkeley.edu > <https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/b75d7f6d-69ad-4c36-a10d-ccfff1c1cbefo%40lists.berkeley.edu?utm_medium=email&utm_source=footer> > . > -- You received this message because you are subscribed to the Google Groups "[email protected]" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/CADEwHTeQGjc0ue9CojrRc3CEHBWoHkhW6qB2PbP9zQpWobTUQw%40mail.gmail.com.
