search

lneuhaus / pyrpl

pyrpl turns your RedPitaya into a powerful DSP device, especially suitable as a lockbox in quantum optics experiments.
http://lneuhaus.github.io/pyrpl/
MIT License
139 stars 109 forks source link

0 Hz beatnote/non-offset phase lock? 2 input CORDIC request #509

Open EsbenMadsen opened 6 months ago

EsbenMadsen commented 6 months ago

I am an electrical engineer writing my master thesis, so i apologize for any misunderstandings in the following:

As far as i can tell the only phase lock implemented in PYRPL is the offset phase lock, where a frequency difference in light generates a beat note on a photo detector due to the periodic change in relative phase (which i feel stretches the definition of a phase lock, but i digress) and this beat note is then demodulated into IQ components. For my setup the beat note should ideally be DC/0 Hz, so demodulating at any frequency is not an option. Instead i have two balanced photo detectors measuring the output of a 2x4 MMI, corresponding to the I and Q of the relative phase of the input lasers. From this i want to calculate the angle of the resulting phasor in order separate phase noise from power noise by using arctan(Q/I), which i can see from the documentation should be possible through the CORDIC algorithm. So here's my problem:

The IQ modules only allow you to apply CORDIC to one input signal and a local oscillator, whereas i want to apply CORDIC to 2 input signals (I and Q) and extract the phase information for PID control. Is this possible?

Todaycrrc commented 5 months ago

hi guys Do you have any ideas yet?

EsbenMadsen commented 4 months ago

hi guys Do you have any ideas yet?

I tried looking into the FPGA firmware to identify if the CORDIC algorithm is implemented as a separate module that takes two inputs, but i didn't find anything i could use with my limited time. If the CORDIC implementation does exist as a module, then it should be easy to change both inputs of the module to be the analog inputs.

michaelcroquette commented 3 months ago

Hello,

Currently the cordic algorithm is only implemented as a part of the IQ module and it's not possible to perform it using two separates I and Q inputs. This would require some modification of the FPGA code. Can't you use only the I signal as your error signal ? I'm not familiar with MMI.

EsbenMadsen commented 3 months ago

Hi Michael, thank you for answering and for your part in creating PYRPL.

My project has turned another direction so i haven't explored the issue further, but i believe it should be possible with relatively minor changes to the FPGA code. An MMI is an integrated version of a 90 degree hybrid, which decomposes light into I and Q phase signals using interference in a photonic chip. This is important when you want to measure the phase of light separately from its intensity. By calculating the phasor from the I and Q signals, you get phase as the angle (arctan[I/Q]) and intensity/power as the magnitude. PYRPL is an ideal candidate for implementing this calculation fast and precisely since it can also plot the phase/magnitude and even route the signals to a PID controller in a all-in-one commercial FPGA. So for instance if i were to use the phase of light in a control loop, I'd be able to separate phase changes from intensity fluctuations very easily.