cbaugher has shared is RP2035 version which used remora-core. A really neat implementation that has both ethernet and spi communication options. Uses PIO for the SPI and hardware QEI (Quadrature Encoder Interface).

github.com/cbaugher/Remora-RP2350-PIO/tree/main

Hard t tell but it does look the the velocity loop is unstabley

Just as a test have you tried adding some D term to the PID?

It was originally set up with a virtual tach signal coming from the Fanuc controller. The original manuals showed a tach voltage of 3V/1000RPM which I should have. I ran the motor on a bench power supply, bypassed the drive, calculated the motor speed with an oscilloscope connected to one of the encoder pins, then measured the output voltage on analogout3. my output seems to be scaled correctly and the voltages match the original drive documentation.

I believe I have the scaling and direction correct, but am wondering if I am running into some type of resonance between the virtual tach and commanded velocity.

Does that seem possible or correct?

Method 34 is using only the Z pulse of the encoder, no homing sensor is needed for this homing method.
The homing sensor is connected to any of the digitals input pins of the drive. It is not assigned by default, i. e. it is configurable.
I just tested homing using a homing sensor with homing method 19. But still, the drive behaves the same way as in the video - starts looking for the sensor at high velocity, when rising edge is triggerred the motors reverse at slow velocity, when falling edge is triggered the motor starts moving quite fast and position deviation errkr is triggered.

Did you connect the origin switch to the servo?
It's usually called the HOME switch.
The VECON manual is really terrible.
It doesn't say anywhere where to connect the limit switches and the origin (home) switch.

@leandrocontarini Could you share the bit files you are using?

The original part was a NE5230, which is now obsolete, a suitable replacement is a THS4281D or a OPA197IDR
 

I achieved some kind of success in homing the servo motors using the CiA402 home component. There is still a problem when the homing is executed in a certain situation. Basically, I am using homing method 34. Then using the halshow I set  the 'request-custom-homing' pin, the homing procedure starts but when the servo find its Z pulse it start moving quite fast. Perhaps it tries to reach a certain position but LinuxCNC reports 'joint 0 amplifier fault' and the servo drive error is 'Position deviation is too large'. You can see videos of what is happening - one video in normal velocity and one in slow motion.  Videos
This behavior of the drives can be observed the the actual position is above 35 000 000 pulses. When the homing is started below this pulse number, the motor finds it Z pulse and the move quite fast to the position where the homing was started.
Attached are the configuration files.

Thanks again for help. I swapped both analogoutput0 and analogoutput3 to RG58 cable that I had laying around and had the same results. I also double checked the ground and shielding on the original wiring. i had disconnected one of the grounds but reconnecting it also did not change anything.It looks like the oscillation is at about 33Hz so I don't think it is from the main power. Another note, I change the base period on the servothred from 1000000ns to 800000ns to see if that changed anything. it seems to change the frequency slightly but it no affect on the amplitude. 

Yes. First thing I checked is the HAL. This has been a machine that runs daily. no changes whatsoever.

That is exactly what I measured. Disconnected the analog out from the spindle drive, then measured. It's wildly non-linear. What Can I do? I live far away from US and don't really have a way to get a new one too. Can you provide the op amp model so that I can try to fix it?

Spindle drive did emit a bad EMI. I had to put filters in the power line of the PC. But didn't expect it to affect the analog out. 10V power is fed through a separate isolated 10V supply. spindle drive is a 11KW SZGH drive. This worked untouched from 2021 and did run daily without a hiccup. This was the only issue that I ever came across.

I suspect its something with the tachometer emulation or connections to the Fanuc drive

How did the original drive get velocity feedback?