Milling Wax Adventure
The year is running out fast. I will soon have to say good-bye to 2011. This is the year (last Saturday actually) that I became 65 years old. Time enough for pondering that event, now to keep moving. Ha!
I did some machining of wax on the Taig CNC mill Saturday. I love machining wax (the hard blue kind). The wax is not good for creating real finished useful items but it makes excellent models for making molds and doing casting.
I spent the morning doing the design work in VECTRIC Aspire software. I was creating a Christmas ornament as a negative. I learned how to mirror and reverse the lettering and other “thinking inside out” processes in 3D design.
The afternoon was spent machining the design into the wax, using the Taig CNC mill. That was an adventure.
I haven’t run that machine for some time. The milling started out OK, but I got a couple of stalls and lost steps in the Z axis. The mill has never done that before so it was a surprise. I wasn’t pushing rapid travel or cutting speeds that hard either. Lifting the spindle and motor is fairly hard work.
The rapids are only set at 65 IPM and I was running about a third of that. So I ran through the software set-up and MACH3 warned me I was pushing the limit on pulses for the pulse generation frequency for which I had MACH 3 set. The Taig has twenty TPI screws so it takes a lot of very short pulses with 1/8th stepping (32,000 per inch) to get any speed from the drives.
That’s 32,000 per inch so at 60 IPM travel (keeping the math simple) that’s 1 inch per second requiring 32,000 Hz (32 kHz) just in the pulse train. Not one pulse is allowed to be missed. That is per axis and of course there are three axis to keep on track.
That’s when I wondered why I was set at 8th stepping. I should use 1/4 stepping as it will still give me great accuracy at 0.0000625” per step! Yipes! 8th stepping is 0.0000312” per step! Even full stepping (4000 steps per inch) is 0.00025” per step. That is still a very small movement. That is four steps to move 0.001”, a thousandth of an inch.
Note: This is theoretical accuracy as backlash, screw accuracy and several other factors reduce this precision considerably in real operation.
So I reset down to 1/4 step or 0.0000625 per step. That is still way overkill at 16,000 pulses per inch but the motors like that rate and sound good when running at quarter step. I boosted rapids up to 70 IPM. A tiny mill like the Taig doesn’t NEED to go faster than that. It could go way faster with 5 TPI screws (>300 IPM) but that is another machine for another time. That is getting into the realm where stronger faster servos and encoders perform best to overcome system dynamic loads.
I have to smile thinking a 3 inch rapid move at 300 IPM would only take 0.6 seconds if the mill could instantly accelerate and decelerate, which it can’t. The Taig takes 3 seconds at 60 IPM with the same conditions. Good coding doesn’t or shouldn’t have a lot of long rapid moves, so the rapid speed performance isn’t worth the cost.
The only excuse I can think of is I must have been experimenting with pushing the limits on the Taig and left the setup at 8th stepping. Stepping changes require changing jumpers in the controller and reconfiguring the MACH 3 software. I certainly don’t need 8th stepping, as you can read above, for normal operation. However, it WAS running good for quite some time until it dropped the steps. That’s probably why I left it there.
I think I hit a point where some combination in the G-Code I ran out of pulse clock cycles or the lube on the ways (oiling) wasn’t as good as it should have been. After the long operational layoff the drag was possibly a bit higher than standard.
In any case, it’s running fine now.
Now I am thinking I need to start taking notes and placing them on the controller computer to remind me of when I make set-up changes and why. I need all the help I can get with this aging 65 year old memory system… Humans are the weak link. 🙂
Afterthought: The fun of owning your own system is you are free to experiment and try things like push limits. I don’t have to explain the down time to the floor boss.