Hi,
I might have a comparable if not similar problem. Here is more or less the information that I already posted in another forum:
Recently I decided to defork my CPC1100 and use the C11 tube on the AM5N. My total setup weight is about 17kg (37.5lbs), using a 5kg (11lbs) counterweight.

I read about RA spike/jump issues of a "couple of people" even before I used the AM5N, but did not dig into this any deeper as not everyone seemed to be affected by this problem including myself. Everything went fine for me with this setup, but those huge RA spikes/jumps started in my last two sessions. RA gets suddenly and repeatedly far off for no obvious reason and PHD2 needs a couple of seconds (if at all) to get guiding numbers back under 2".
I recognized a lot complaints about this and assumptions about the potential root cause. Some people reported that tensioning the belts solved the issue for them (maybe upfront: not for me). But the information that this could lead to a solution and a thread where somebody teared down an AM5 gave me confidence and encouraged me to continue this investigation.
In my case (and I want to emphasize at this point that I do not want to give the impression that this could also be the root cause of the RA jumps in all the other cases) I could recognize kind of a knocking noise, especially amplified due to the relatively large C11 tube, with an interval somewhere around 5-30s. It was not present in every session, but as soon as it started to occur it did not go away and at the same time I could see the large RA spikes/jumps in PHD2. I did alot to rule out things and in my case it has nothing to do with power supply, USB, guiding or associated settings, belt tension or belt slip. Btw: I'm directly connected to the mount with a notebook via USB and I do not use ASIAIR.
Long story short - I decided to remove the front and counter weight covers, run my setup, provoke the knocking sound and have a look what exactly is happening inside the mount. With "run my setup" I mean: Mount supplied with separate 12V/5A power supply, direct connection to my computer via USB cable and sideral tracking started in the ASIMount ASCOM window (knocking noise is not present when tracking is set to off). What shall I say. It looks like the RA motor is snapping back in a regular interval. There is definitely no belt slipping. And as there is no other piece of software interfering, the issue must have its root cause definitely in the AM5N mount itself.
I recorded a video of this (hopefully this link works): https://drive.google.com/file/d/1BzbL5wN4N1btif_eidnZ-RBd0Ae4y37B/view?usp=drive_link
I already wrote a ZWO support ticket and attached the video in it. No response until now.
Furthermore I recognized that the issue does not appear when my setup is on the west side of the pier (but probably also just a matter of time). So my assumption is here, that the RA motor does not seem to have a problem if the weight/payload drives it in direction of rotation, but it snaps back the other way around. Not sure what this means at this stage, but my gut feeling tells me that this might be a hardware issue (insufficient RA motor power, RA motor electronics/mechanics failure) or simply poor motor driver firmware. "Hopefully" the last case and ZWO can come up with a firmware update.
Andreas