Need help getting my initial set-up to work.

[sstralkowski]

That actually clarifies a lot, man. I really appreciate you taking the time to explain it to me like that. I just retested it and while the Z axis seems to be pretty close now, the XY is still way off It tries to start the wing in the center of the lower lobe. Mach3 says those coordinates are around X5 and Y4 while when I actually measure the distance with a ruler, it's like X14Y10. What would cause that? I mean it's about a factor of 2.5. That's why I thought it was the units thing, but I've checked that over and over again.

[blinkenlight]

There are two possibilities here - the milled part might either be completely the wrong size (measured between two points that actually belong to the milled feature - ie. you get a wing, say, three times larger) or it can be the right size, just in the wrong place. The second case means there must be a problem with the way you zero your work coordinates. The first case could have two causes that I can think of - either wrong units (which should be obvious since a millimetre is nowhere near the size of an inch) OR the wrong value set as steps/mm: either Mach3 is configured with the wrong steps/mm value for your machine, or your machine is actually set up to a different steps/mm value that it is expected to have - which is most commonly caused by DIP switches on the driver PCB being set for a different microstepping setting than one would expect.

The point of which is this: forget the g-code; just move you machine at some arbitrary place, note the value and mark the point, then tell it to go (with the Mach3 manual controls), say, 30mm further and note the new value on the display and new physical position. If it really moves as far as the difference from the original value is, you can forget all about the steps/mm thing, you're fine, you just have to figure out where your 0,0 is and why can't you set it to be in the lower left corner of your material. If, however, your mill moves some other distance, that's something that needs to be solved first.

[sstralkowski]

That's my problem. The software says I moved 4mm when it really moved 10mm. How do I find the proper steps/mm for my machine?

[aarggh]

That's my problem. The software says I moved 4mm when it really moved 10mm. How do I find the proper steps/mm for my machine?

Looks to me like you have the wrong steps/pulse or mm/step setting for your machine, this link (and the other links in it) should help you, don't forget the value of any micro-step you have set working it out:


Mach 3 Setup & Configuration - Tutorials - A Quick CNC

cheers, Ian

[sstralkowski]

So based on your advise, I tried looking through the mostly Chinese user manual for the numbers to fit into the calculation. Oddly enough I found a steps/mm table. Seems to be working much better now. Thanks everyone!!!

[sstralkowski]

Ok so I got everything set up and tried etching some plastic. The bit I have is a 0.6mm end mill bit. The problem is it is fine for a few minutes and then just starts melting the plastic and making a mess. Any idea what is causing this? I mean, I'm guessing friction, but how do I rectify it?

[blinkenlight]

Short answer: you don't.

Ok, now the long version: knowing what material can be cut at what spindle speeds and what feeds is what machining is all about; I'm admittedly not much of a machinist - it's just a hobby of mine - but my understanding is that when dealing with heat, you either apply cooling (not much used with plastics as far as I can tell) or cut faster - so the chips carry away the heat. The problem with that is that my spindle only goes up to 10000RPM (much like yours I suppose) and last time I looked up some numbers for a plastic I was having similar problems with, it looked like I would need a spindle speed several times higher than that to cut it properly. So basically you can try cranking up the spindle speed as high as you can and move ahead as fast as you dare - but obviously, a tool of 0.6mm might not take going fast all that well. Which leaves us with the short answer: you don't. Anyway, don't let that discourage you - do trial runs with several different speeds and feeds: you might just find some that work better than the rest (or at least, less badly...).

[sstralkowski]

Not good news, but I appreciate the honesty. What do you think about putting the material in the freezer and popping it out just before carving? I know that may harden the material, making it harder to carve, but it may also reduce heat.

[blinkenlight]

Hmmm, I never tried that, but now that you mention it I kinda remember reading about something similar being attempted by someone (no idea if it worked) - it's certainly worth a try!

[bhurts]

What kind of plastic are you cutting? If it's acrylic use cast not extruded and that will fix most of your melting issues. As for spindle speeds and feeds they work in tandem so one must keep up with the other. Higher rpm is not going to help and will most likely make it worse. When I cut plastic on a mill at lower rpm I get much cleaner results then with a router at high rpm. Plastic likes high feedrates to get clean unmelted parts. If you are using a multi flute bit and getting melting try again with a single flute tool. If no single flutes available increase feed while leaving rpm the same. If your tool or machine limits what you can feed at reduce rpm while maintaining feedrate. It's all a experiment and results come with experience.

Ben

[aarggh]

Not good news, but I appreciate the honesty. What do you think about putting the material in the freezer and popping it out just before carving? I know that may harden the material, making it harder to carve, but it may also reduce heat.

I wouldn't even bother the freezer route, by the time you've clamped it down it will pretty much be at ambient temp. As others have pointed out, try increasing your feedrates, and if it's still struggling reduce the depth of cut per pass but keep the federate up.

One factor that really contributes badly when cutting plastics, is not clearing the debris well enough with a dust extractor, so it keeps re-cutting the waste and heating up.

cheers, Ian

[sstralkowski]

Ok so first off, thanks a ton to everyone for the pointers. I got the problem pretty much fixed. I only get melting issues when I accidentally set the Z axis too low now. I reduced the pass length to 0.2mm and it seems to be working ok. I wish I could get a jig set up so I don't have to hand zero all the axes before every attempt, but I'll work towards that later. Jigs seem like a whole other world, and I'm already learning an awful lot of info at once.

Second, I am having a new issue now. It seems my motors are off a bit. Specifically my Y motor is really noticable. It's most obvious when I try and cut a circle. I usually end up with an oval that is squished along the Y axis or a weird angled oval squished at a 30-45 degree angle to the Y axis. Maybe I need to grease the drive bar for the Y axis? Any other ideas?

[awerby]

Lubrication is always a good thing to do. If that doesn't fix the problem, you probably need to reduce the acceleration on the Y axis until you stop losing steps.

[sstralkowski]

What is the best thing to use for lubricant? I'll also try to lower the acceleration a bit and see if it fixes it. Thanks!

[awerby]

Try Vactra #2 way oil. It's formulated to stay put better than regular oil, which tends to flow off high places and into low ones.

Figure out the max velocity each axis can sustain without errors, then use an acceleration value about a tenth of that to start with. If that works, you can creep it up a little to improve performance until it starts screwing up, then run it down again until it's reliable.

[sstralkowski]

I just placed an order for the Vactra oil. Stepping down the speed by itself didn't help. Also apparently the WD40 I had lying around was a bad idea. It seems to have stripped the black grease that was already on the gear and now the y axis motor clicks like when you hit the end of the drive shaft. I'm guessing it just can't handle the amount of power required to move the bit. So frustrating.

[awerby]

WD-40 is good for cleaning things off, but it's not a lubricant you'd want to leave on the slides and screws. If your gears are running in grease, you'd want to replace the grease with grease, not oil. That's just for sliding parts where grease would be too thick and tend to capture dust and chips.

Once you've got everything lubricated sufficiently, then go back to motor tuning. Max velocity is one thing, acceleration is another. Try different values with the velocity, jogging the axis back and forth until it sounds good and doesn't skip. Once you've got that working reliably, then enter a value about 75% of that speed and hit "save axis settings". That's the speed it will execute rapids at. The acceleration value should be about 10% of that, but test it to make sure it isn't losing steps, and if it is, run the acceleration down a little further. If everything's working great and you're feeling brave, running the acceleration up a little will make it respond more swiftly.

People that cut a lot of plastic use vortex coolers to chill the air, which combats melting, but these units are pretty expensive. Just having a little shop air blowing on the tool can help.

[sstralkowski]

I just placed an order for the Vactra oil. Stepping down the speed by itself didn't help. Also apparently the WD40 I had lying around was a bad idea. It seems to have stripped the black grease that was already on the gear and now the y axis motor clicks like when you hit the end of the drive shaft. I'm guessing it just can't handle the amount of power required to move the bit. So frustrating.

Ok so I finally got around to oiling the screw drive with the Vactra. Then I reduced the axis velocities as low as they would go. No deal. I still can't carve a circle. So I was thinking that if the y axis is the one delaying, why not delay the x and see if it helps. So I set the y axis acceleration at 20% the velocity and the x at about 4%. It helped a bit but I still can't do circles. Any other advice? Do I have to have the axis velocities match or could I try dropping the x velocity so the y has a chance to catch up?