Glad to hear you like the vids David. I did the same thing at first trying to remove all the extra stuff. Then it hit me that I simply didn't have to use all of them to setup cuts - was a major ahha moment. Sponz actually sent me the "raw" dxf files for the pietenpol since they have all the different types of cuts on separate layers and I figured that would make it quicker and easier. It was kind of nice, but really didn't make it that much easier for me. I actually found it somewhat harder to work with since the different types of cuts were harder to tell apart unless I viewed it only one layer at a time (or took the time to color layers differently or something.)
And yeah, I was really worried when making that video because my needle has NEVER sounded THAT bad before. I actually don't have a problem tightening the grub screws as I can tighten them to the point that they squish the motor shaft and start to bind the motor. I've just never had good luck with that style of mount even on planes. They seem to go from too loose and will come apart to too tight and will bind the motor with no "just right" zone for me.
I've actually been changing my workflow the past few days to try and improve things a bit more. When I built my test cub I noticed that things weren't quite 100% square. So I printed a set of these:
https://www.thingiverse.com/thing:1755510 to help square the machine before each cut.
Those ALMOST work...on my pre-525 parts they don't quite work on the X axis because the center part hits the left Y roller before the X rollers hit the alignment tools. Would probably work to align to the right...but that would mean doing a lot of moving the machine with the steppers engaged and that's kind of annoying since I run untethered and have to use the LCD to do those moves which I don't really enjoy. I'm kind of tempted to grab an arduino and some buttons and make up a controller I can plug into the RAMPS to send my most commonly used commands without messing with the LCD - will probably do it at some point but lower priority for now.
Anyway - alignment. It turns out my X axis is pretty good. I just added a spacer to the alignment tools to move them out a bit further and my X axis is almost always square. But my Y...not so great. It's always about 1/4"-1/2" off. So that explains a few issues I've had.
But I'm starting to think this WP foam or the cub have it out for me. My DTFB version went just fine - other than one bad cut due to me forgetting to zero the machine between sheets. My own mistake no doubt. But with the WP foam...every other sheet I cut is failing
That first failure was totally my fault with the short Z wires. Fixed that. Cut another sheet...no problem. Then went to cut the next sheet...that went ok. Then tried to cut the 3rd sheet...and...ugh:
I'm not 100% sure what happened here but I have a theory. I think it's a bug in Marlin where the command to reset the machine position doesn't always work properly. I've heard reports of this and been told it's why people like to reset the RAMPS instead of giving a command to set zero. But with the alignment stuff I need to keep the steppers powered. The way I do the alignment is:
1) Pull the machine against the "stops"
2) Use the Prepare-Move Axis to power up an axis and position the machine at my workpiece origin
3) Print before the steppers timeout so the machine stays square
So doing a reset on the RAPS in #2 would cause the steppers to power down and loose lock/square so isn't a viable option.
But it seems doing that zero command more than once without powering down can often result in Marlin's coordinate system getting confused. (at least in the old RC6 version I'm still running on this machine.. I should update it this week but want to get this cub cut before I go messing with it that much.)
So what happened there? Well. I first started the cut but had forgot to reset Marlin's coordinates to 0 on XYZ before doing #3...so it immediately started moving off the workpiece and I stopped it before it even touched the foam. I then did the square/home setup procedure again making sure to zero the axes this time. And the cutting started ok.
First two cuts were the score cuts in the lower left piece. Just fine. Then it went to cut the piece above that...and seemed to do fine...but...cut it way too short. I figured it must have been a mistake I made in the gcode and let it keep cutting. But then it went to cut the bevel in the horizontal stab to the right of that...and again they came out way short? Now I was curious what was going on and let it keep cutting...and for a bit it did just fine. Seemed like as long as it stayed below a certain Y coordinate all was fine.
Until it go over on the long horizontal cuts on the wing...when it did the same thing in X. If it went past a certain point it just suddenly decided it was done. But I still let it keep going out of curiosity. Then it got to stuff above the Y coordinates it didn't like...and it just went crazy doing those triangles. I let it keep going thinking it may get back on track after the cut it was doing (should have been the outside of the wing) but it just kept doing goofier and goofier triangles. Finally I decided it really was a lost cause and stopped it.
All I can figure is that it was the resetting of the zero point that got the machine confused. I called it a night at that point. Well...almost. I saw Ryan's video of etching aluminum and had to swap to my laser and give that a try. Grabbed and SVG of the rotor riot logo and setup a 50mm/min burn on the back of my wattmeter. Nothing
Must be some TOUGH anodizing on there! Put my trusty accucell6 charger under there instead, figured it's black anodizing may be easier to etch. And...got results but not nearly as nice as what Ryan got...I really need to swap to a 3 element lens, I think the larger spot of the G2 is really hurting me on this kind of stuff:
(BTW hard to tell scale but the whole logo is less than 1" across)
So...today...gave sheet 3 of the cub another go. Didn't have the odd problems from last night...but something else went horribly wrong:
The cut started in the lower left and went clockwise around the part. I know exactly what happened...but before I reveal it I'm curious if anyone can guess. Big hint - it had nothing to do with the gcode, or MPCNC.