STFT DUC Not works in Phaser Ch1

I am interested in STFT pulsegen to generate consecutive pulses that has carrier phases fixed to the time zero of pulse sequences. I spent a month to see outcome of STFT example included in Phaser STFT branch.

The problem I found is that STFT functionality is not enabled in Ch1.
To ask your help and share my problem, I clarify what I have setup for this test.

I should ask for help directly to the pulsegen developer Norman Krackow. But I am not sure mine is hardware problems. So I leave here, feeling sorry if this seeking help bothers community.

Environment

• Ubuntu 22 (LTS)

Phaser firmware

• QUATIQ Phaser STFT-pulsegen branch

git clone https://github.com/quartiq/phaser.git --branch stft_pulsegen

ARTIQ

• Starting from the released ARTIQ-7, I modify files:
  $(ARTIQ)/coredevice/phaser.py
  $(ARTIQ)/gateware/rtio/phy/phaser.py
  $(ARTIQ)/gateware/eem.py.
  The modification of code-works followed to works by the developer Norman Krackow. I just copied and pasted his commits in Quartiq/Artiq branch stft_pulsegen.

Installation

• "nix develop" in the modified ARTIQ-7 main directory.
• In the same nix shell, I compiled the phaser stft-pulsegen firmware by "python phaser.py"
• I flashed using vivado and flash.tcl. (and hardware Xilinx platform cable II).
• I compiled Artiq firmware as the manual's guideline using the original json description, and flashed using "artiq_flash".

Results

• $(phaser git repository pulsegen branch)/stft_pulsegen_example/spetralspass.py example code
  
• $(phaser git repository pulsegen branch)/stft_pulsegen_example/timing.py example code
  

Problem observed

The example code "spectralspass.py" seem to emit signals Ch1 and Ch2, and "timing.py" emits to Ch1.

• But the signal only appears Ch2 for "specktralspass.py",
• and "timing.py"' shows only channel DUC (not showing pulse envelope, but DUC oscillator carriers):
  For "timing.py" case, results in Ch1 are exactly the one attained after I disabled (flag, or removed lines of STFT setting) STFT.
• The desired 3-pulse figure was achieved in Ch2, which I enabled by modification:
  for ch in range(2):.

I hope this problem to remain software areas. I am afraid whether ch1 circuits is broken due to user mishandling.

Thank you very much.

rjo

I forgot to tell that I am obsessed by the Phaser hardware, fascinating firmware and its ARTIQ interface.

I now read the known issues and the STFT thesis in more details. Though realizations of low noise and generalized pulse generation are good enough for various application (Surely this can even make linear chirp pulses! Thanks much for great works and efforts), I require multiple profiles of different pulse length and different phase (X, Y, -X, -Y pulses) for my experiments. But STFT pulsegen seems to support only a single profile of window length.

I should wait for the release of Miqro firmware of phaser. I would like to join as testers, although this permission should be given up to Quartiq's business plans and decisions. I am highly interested in Miqro.