I agree, managing the sled’s location is important.
1 Like
Hmm, OK, I’ll try those zip files tomorrow. I downloaded and ran something this morning. Would you have expected this code to affect the calibration? I seemed to get a skew so a square was a quadrilateral when I did a test cut. Anyway I’ll try the proper code tomorrow.
1 Like
No, I would not expect that. I am certainly interest if you are seeing something like that.
OK, it looks like the calibration problem may have been a chain jumping a link. I put the firmware back and there was no change so I then reset the chain lengths and the problem went away. I didn’t hear or see a jumping link, but I do remember a ping, but it didn’t sound serious enough for a chain jump. I’m thinking of rotating my motors a bit so that the sprocket is in line with the chain a bit more, does the standard frame do this? I may have to bring my motors forward too, when I start cutting thicker material (which I want to do).
Anyway, I put the krkeegan movement fixes on and cut a test square. I got this:
https://photos.app.goo.gl/EhHYNXP66LYLlDuy1
The Z axis still moves after the XY has started a move, it runs on for about 4 seconds after the XY starts a movement. You can see it in the picture as there are sloping cuts in there. The other flat cuts are a mis-set zero for the Z axis, my fault.
There’s also a strange distortion which looks a bit like a barrel distortion, I’m not sure where that’s coming from.
One thing I really enjoyed with this firmware was the faster jog speed. That was nice. I can see from these test cuts that the Maslow will be useful for me, but without a stable and accurate enough calibration it’s not yet usable.
Does my smaller frame affect the maths that much? I’d have thought not.
1 Like
Wow that is way off. Okay so your ZAxis most be moving at a much slower rate than expected. Let me look into that.
As for the distortion, that isn’t something that would be affected. So I think you have a calibration issue, something isn’t correct in their.
2 Likes
My Z Axis pitch is -1.819. I don’t know how that compares with other routers.
1 Like
Yup, that I think that is the issue. It was my fault, we were supposed to remove a hardcoded value for the pitch a while ago.
When you have a chance try this firmware, it should solve the issue:
https://github.com/krkeegan/Firmware/archive/Fix_Z_Pitch.zip
No need to change your GC version.
2 Likes
Hmm, I have put that code on and it seems to be worse. The XY and z axis moves now seem to start simultaneously. I’ve put standard 0.92 back on and that has the old 4s run-on. So the code does indeed seem to be worse. It’s like there’s no wait at all on Z.
what is the g-code that misbehaves?
There’s a link to the original circular file that had a problem in my original post for this problem. It’s from Kiri:Moto, which looks to be useful, so I suspect any file from there would have the same behaviour. For recent tests I used a different test cut, which was also created on Kiri:Moto, it is a 100mm square with tabs. I don’t think I have access to that file until tomorrow morning, but the circular file definitely exhibited the same problem. I’ll upload the square file tomorrow, or you could create a 100mm square 10mm high in openscad and then kiri:moto it with tabs.
it’s also worth noting that every Z axis move in this file is separate and had no X/Y movement
@kirkeegan my guess is that you missed some path in the interpreter that redirects Z axis only moves to a single-axis function call and doesn’t wait for the result
This is all singleaxismove.
@jetrock, i noticed you keep referring to the pitch as negative. Do you raise your router by turning clockwise?
Hmm, it depends on what you mean as clockwise I suppose. I have to use negative pitch to get it to turn the correct way as indicated by gc. Looking towards the router with the end of the motor pointing towards me then to raise the router the motor turns anti-clockwise.
Got it, ok I think this is hopefully the last issue then.
I don’t know if anyone actually considered a negative pitch when writing the code. It sounds like it largely works and it isn’t a bad solution. I know if one location that has to be fixed that might be causing your an issue. I will look for anything else and get your another patch this evening.
Thanks for that, I’ll try it tomorrow morning. I’d have thought all routers had clockwise turning to drop the bit into the work, does yours have an anti clockwise rotation to do that then?
Yeah I think so, never really thought about it.
OK, I pushed a commit that will hopefully solve it for you 
https://github.com/krkeegan/Firmware/archive/Fix_Z_Pitch.zip
3 Likes
Thanks for that code. The Z axis moves complete and then the XY move starts when I use that code. You’ve fixed it. Thanks again.
2 Likes
Awesome, thanks for being the guinea pig and testing for me.
1 Like
I think there’s more thanks due to you, personally. For your enjoyment, here’s the cut before the fix:
and after:
OK, it’s different thickness ply, but it definitely works now.
So, I’ve now made my first ‘thing’ with the Maslow: A filament reel for my 3D printers. Or at least, I’ve made the parts for one. It was a lot quicker than doing it by hand too.
4 Likes