GRBL Loosing Zero Point between Passes

[Rob H]

Hi,

I've got my DIY CNC router up and running. I've cut simple circles and squares with single passes and adjusted the XY calibration so that they are dimensionally & geometrically correct. However I'm encountering an issue when doing multiple passes. Before running the g-code I set the zero position and it makes the first pass perfectly, but when it goes to do the second pass it effectively moves the zero position so that the second pass is cut in a different location and it gets progressively worse the more passes I take. I've tried different g-code files both ones I've generated myself in Vertic and also ones that I've down loaded, I've also tried both Universal G-Code Sender and gSender and I suffer from the zero point moving between passes, so I assume that the problem lies with GRBL. Anyone encounter this before and if so any idea what's causing the issue?

Thanks

Rob

[m_c]

How much of a difference is there between passes?
If it's small, check you have the step/dir polarity and timing is set correctly. If it's wrong you loose two steps every time you change direction.

[AndyUK]

Hi Rob,

Try the cut in air. If it keeps the zero that will tell you it's not GRBL. If it still does it I'd suggest it's potentially an electrical interference issue, otherwise it's potentially just pushing the machine too hard and dropping steps / mechanical slippage eg a grub screw not being tight enough (I noticed you have a number of pulleys and belts - make sure everything is appropriately tight but don't over tension the belts). Interference is still a possibility (the current in the spindle and motor wires will be higher when actually pushing through material) but it's the less likely bet if the in air cut works.

By the way - just seen your Instagram page from another thread. Some beautiful work!

[Rob H]

The first pass is really good, the second is off by about 3mm or 4mm the third pass is off by that again.

I'm pretty sure I've got the step direction / polarity and timing correct but I'll double check as it could loosing two steps on each change of direction would explain the drift between passes.

[Rob H]

That's a really good point about trying to cut air, I'll give it a go and see what happens. I've previously not considered electrical interference, but currently I have no shielding what so ever, so probably inviting problems.

[JohnHaine]

What is the distance moved per step? Losing only 2 steps at a reversal is unlikely to give that much error surely?

[m_c]

That big a step isn't likely to be a configuration issue.

Are you 100% sure that all pulleys/couplers are tight on their respective shafts/screws?
It's probably worth marking a line on all of them with marker pen, just to be sure nothing is slipping.