What to do when Galvo calibration fails?

When I print the target and run the circle gcode, the result looks like this (the red ellipse, but cut off on the left and right sides):

I’ve double checked the cable connections - and they all match the images shown in the assembly instructions.

For such drastic out of shape circle, it is almost always cable/drive mismatching.

Can you double check the galvo drivers (those PCB cards.)
They look identical, but there is a specific order to them. One is only for X-Galvo, and the other is for Y-Galvo. Please check page 30th page on the galvo drivers and be sure you have them in the right order.

Yes, that was the first thing I checked (the labels match the connections). Then I double checked every cable and compared them to the assembly instructions. Everything is in the right place according to the instructions.

can you provide pictures of the internal showing connections to boards and to galvos?
as well as to the pcb. X-galvo has to be close to the front

Here you go:

The full images are high res, but you’ll probably have to download them to see the full resolution.
And no, I haven’t tied down any of the cables yet - waiting until it passes checks and calibration, but none are in the way of the beam.

Your setup is fine. There is a part that is not quite right on the Y-Galvo.

Here:

This part seemed to be bumped and half separated from the board. If you look closely, there should be metal contact points at the back of the blue square. push it back and solder the contact spots.

The run circle calibration, (don’t close off the panel yet). If it is still off, follow this guide to manually adjust galvo (while make sure your x,y size is set to 90:

I can’t believe I didn’t see that… Ok, disassembling now.

update:
Desoldered the pot, reseated it, resoldered.
I’m reading 100kOhms across the top and bottom pins, 90k across the top and middle.
Put the board back in, reconnected cables, double checked the cables, and tried the circle test again – unfortunately, the result didn’t change.

Hmm, looks like it should be a 10k pot: https://www.digikey.com/product-detail/en/bourns-inc/3299X-1-103LF/3299X-103LF-ND/1088157
So I may be reading the other components it is connected to, and it is still open (damaged inside).

And the document referenced above appears to be empty.

updated the link. The screw on the driver board is very sensitive and small amount turn leads to big changes.

Turning the pot screw had no effect – so, as I guessed, it’s broken internally. I’ll try ordering a replacement from digikey, and see if HSC electronics has comparable ones on Monday.

P.S. for silicon valley locals: Anchor Electronics was much quicker to find the exact part.

does the other one (X-galvo driver) change if you turn?

We will give you that part number Monday so you can try to find it local. Each galvo is tuned to a specific driver, that is why there is a specific order to them.

The careful tuning is why I’m afraid to change the X galvo driver :-).

OK, removed the old pot.
As I suspected, the top pin was completely disconnected. The middle pin was loose, and had pulled off part of the copper trace from the top side of the board. The bottom pin was actually intact, and the only thing really holding it in place.

I confirmed that it is a 10K pot, and the original calibration point was around 4.5k between the middle and lower pins (the only two remaining to measure).
I did find a replacement pot that fits the space at a local shop.

Unfortunately, the missing trace is going to make it difficult to fix. If you can tell me where that middle trace goes, I might be able to solder in a re-work wire, or just wire it up off board.

Can you take a picture of the replacement part? Just to confirm

you have the X galvo working,so just look at that one and see how the pot is mounted. this pot is mounted differently than the rest on the board and you can use x-galvo driver as a reference.

Mounting isn’t the problem. I need to know where that middle copper trace went so I can replace it with a wire. Right now I could only figure that out by removing more trimpots (and risking more damage to the board)…

I’ll get a shot of the trimpot I have, but I also have the exact part on order from Digikey (will take a few days to arrive).

trimpot on board, no solder

trimpots side by side (leads are straight line, not offset)

can you take a picture of the back of the driver that is borken? we like to see if there is any damage to the back

I can’t really expose the back without removing all the screws on the heatsinks. But the only change on the back is from the flux I added and haven’t washed off yet.

there are damages to the board as well. the part needs to come back. emailed you with detail

Darnit, I was hoping to get this working soon. (so frustrating to have an almost working printer)
Alright, I’ll follow up in email.