PSU's cutting out on motor reverse, now dead

Ever since I first made my rig over a year ago, the PSU's have tended to lose power when the motion of the motors became more violent... mostly due to quick motor direction changes (afaict). I believe this to be one of the protections offered by these psu's. I have two of these 24v 25a psu's driving 2x IBT2's, driving 2x 24v 75rpm 250w wheelchair motors from ali. CTC was around 55mm.



Very rarely if ever did I experience these conditions in flight however. Only when using the manual slider in SMC3, which is kinda jumpy if you know what I mean, or using the "motion" mode, which is quite violent.

If you've seen my other thread about bumpy motion, I've be redesigning the drive system to a rack and pinion system to try to smooth out the motion by allowing the motor to use more degrees of rotation. However, even without load, I'm noticing that the PSU's have been experiencing this power loss much more often. One motor did it just in the SMC3 sine wave, the other motor did it with the sine wave if I pushed down on the rack (creating a load) while it was descending. Strangely it tended to cut out only when I pushed in the direction of the movement, and not against.

EDIT: I also noticed when I ran a sine wave in Flypt, the motors would be fine if the sine wave was very slow and shallow. However, if I decreased the period of the wave, at some point, one of the psu's would start losing power and result in motion in the video posted at the bottom of this post.

My initial thought was I'm pulling more amps than the psu can provide, so I bought a TRMS clamp meter. The "poorer" motor was showing no more than 5 amps with me pushing against the motor with my hand. The "better" motor is only showing 3 amps. The motors are too powerful for me to stall them with my hand so I don't know the stall current. But considering these PSU's are 25a, and the power-off is during motion, I can't imagine that's the actual problem.

I think (not know) the problem might be worse now because the motor is running at a higher rpm (much shorter "CTC" of about 17mm but using a gear instead of a lever) than before and takes longer to stop and reverse direction.

Yesterday I was trying to get different amp readings and ended up frying both PSU's (not at the same time). I had already ordered 50a versions that will be here Tuesday, but after seeing how little current is even under load, I'm thinking these won't change anything.

How likely is it that this is back EMF? Or could my ibt2's be damaged? Or the motors? How would I know? I swapped one ibt2 out with an extra one I had but no change. Not sure how to proceed. My electronic knowledge is insufficient unfortunately. I've been reading a bit about the basics but not sure it's helping me in a practical way yet.

This is a video of the worse motor that was pulling 5a before the psu failed. This is supposed to be a sine wave in smc3, but notice it gets to the ends of the wave and lose power and backs up a ways, coasting to a stop, before regaining power and continuing. I'd not seen this behavior before. The other motor wasn't behaving this way yet, but did at the end right before its psu died.

https://youtube.com/shorts/_4x6mbRrKhI

EDIT: Also, I purchased one of the 100A ebay controllers mentioned on another thread and it'll be here at some point. I'm just trying to prevent damaging more electronics!