Strange sled jumps when moving left to right

I’m seeing a weird issue which maybe points to a faulty XY motor:

  • I did a pencil test for line straightness and 90 degree corners.
  • In the top central area, the sled jumps up producing visible notches. This happens only in the top middle of the work area, other places are OK.
  • The jumps happen only when sled is moving left to right. When it is moving right to left the line is perfectly straight.

From my initial research, it could have been a chain skip, sled being too heavy, or issues with motor controller. I’ve checked for it and chain is definitely not skipping. The issue also happens without bricks on the sled so weight doesn’t seem to be a problem. Guess it’s the motor?

I’m attaching a few photos. Have anyone seen anything like that before? Any recommendations?

Nope. Not seen this pattern. Is it always repeatable on the same spot?
Does the sled have rounded bottom corners? Are any of those branch pattern sticking out of the sheet (even slightly)?
You have the first Maslow oscilloscope, though i hope it’s not spikes in your power supply. (joking here)

I see 4 of them now… strange indeed. If it was some obstacle on the wood it would not be that uniform.
I have a doubt. Perhaps one chain is attempting to jump because your chains are not full parallel to the worksheet. Could it be be that one sprocket is slightly lifting the chain with the tip of a tooth and then the chain slides back to the sprocket?

If this is the case, the most common solution (including wear and tear) is to make chain-guides to ensure the chain is feeding to the sprocket without having any angle.
The other solution (involves some effort) is to have a balanced sled to begin with and then adjust how far the motors come out, the same distance of the chains to sheet, close to the sled, so no angle on the chains to the sprocket.

@KirillN, can you give us the distance between the first 2 top spikes and the height over the centre of the sheet this draw was made? G-code perhaps?

Bahrain .nc forensics is usually only checking g-code, but in home office, bored and else, this looks so interesting that the staff has involved.

Calling for help at Maslow-Interpol to solve the case. The findings are interesting enough to raise a question that can not be solved on this island.

  • if one tooth on a sprocket had a slight dent, how far would the sled travel horizontally with one rotation of the sprocket (given we know the height) to repeat a pattern like that?
    Hypothesis only…