New Alpha Chatter (Resolved?)

[weslambe]

I'm sure that most of you remember my issue with chatter and my new prs alpha in the late May early June timeframe. I showed pictures of the chatter marks and so did Ted Hall. Shopbot sent out Dirk Hazeleger and Ryan on two separate dates to help with this issue. We tried everything and couldn't resolve the chatter. Ted offered me a refund.

I started building a new machine for myself right away. It's a stout machine that should be able to take what is dished out with ease but that is not the point. Stan Holt is another shopbotter who builds excellent control boxes for chemical processing plants etc... He built a box for my new machine and took a couple of days while on vacation to help me set the new box and motors up. Thanks again Stan.

Keep reading cause it gets interesting.

I am using 960 oz steppers with a 5 to 1 reduction on all but the z. I am also using Mach 3, Gekko 2 drivers, pmdx power and processor boards. When we hooked up the motors at my brothers shop on his computer (older Dell) all worked well.

When I got home and hooked up the controller and all of the other stuff on my Dell 3100 the motors bounced while moving at speed during the bench test. I took the software and hardware and hooked it all up to my older Dell 2400 and the bouncing went away.

Dell had me wipe the 3100 machine and start from scratch but the problem continued. They are sending me a new motherboard and we will see what happens then.

The bouncing that I mentioned above is what I was describing to Shopbot in my earlier threads. I only wish that someone would have had the idea to switch computers out before I send the shopbot back. I can't say for sure, but I'd bet that I would have been a happy customer If I'd have only known. Who knows, I may have even kept the machine that I owned before the shopbot.

Word to the wise. Try an older computer if you have chatter.

btw, here's my new machine being built.

[cnc_works]

Wes, it was brave and generous of you to share your latest findings.

Sound like you are going to have a dynamite machine when you are done though. Post a good link so we can see.

Donn

[richards]

Wes,

Interesting post.

If I drill down to the bottom of all of this, is it your conclusion that the computer that you were using with the PRS-Alpha was sending a 'jittery' pulse train, such that the Oriental Motor Alpha stepper motors/drivers were possibly using their closed-loop function to overcome the 'jitter'? (I might be drawing the wrong conclusion - but that's how I understood your post.) If that's your conclusion, then would it be safe to say that the PRS-Alpha had very little or nothing to do with the 'chatter' problem?

Now, on to the machine that you're building. I find it very interesting that you're using such large motors with a 5:1 gearbox. 960 oz*in X 5 = 4,800 oz*in. 4,800 oz*in = 300 lb*in. That must be one tough gearbox. The Oriental Motor gear box on the PK296AxA-SG3.6 motors is limited to about 1/10th that value. I'm not disputing the fact that more torque is better than less torque, but I am wondering how 30-tooth pinion gears and a rack can deal with that much torque without breaking or bending gear or rack teeth. (As a side note, I've found it hard to tame the harmonics in a motor that large, even when I use the G203v stepper driver. The 300 oz*in to 600 oz*in motors have run much smoother for me over their full range of speeds. The larger motors have run a little rough in the 50 RPM to 150 RPM range when I adjusted the trim pot to give good response in the 1,000 RPM range.)

Also, have you had 'jitter' problems with Mach 3? When I loaded the latest 100Khz program, I got lots and lots of 'jitter' at pulse rates over 50K. Of course that is more a function of my computer than of the software, but Windows XP seems to interfere with the higher pulse rates. (I'm able to get beautiful pulse rates, up to 250K, from my G100 pulse generator, but not from a parallel port.)

One other interesting observation (at least interesting to me). A few weeks ago, I visited a factory that does nothing but make 5-piece doors. They make at least 2,400 doors a day. Their CNC router is a big heavy Busellato machine with all the goodies. It was amazing to watch the machine work, but one thing really bothered me. The feed speed was only a little more than 2-ips. I thought that surely a machine that large and that heavy could plow through the wood at much higher speeds. When I asked my guide about the speed, he just smiled and said that they had hoped to be able to run at 3X that speed, but quality dictated the slow speed. He handed me a few samples that they had cut at faster speeds to show me what he meant. Those cuts were terrible with chatter marks as bad as any that I've ever seen. It surprised me that a machine that weighed tons, a machine that had a massive 16hp spindle, a machine that had heavy-duty metal clamps to hold the work-piece, would have chatter problems, but it did. When I left the factory, I had new respect for my 'light weight' PRT-Alpha and little or no desire to ever move to the 'big iron' machines.

[weslambe]

Dell came and changed the motherboard today with no effect on the chatter. There has to be something wrong with the 3100 series motherboard since all of the same components (mach3, pci parallel port, cables) worked on the 2400 series.

Another try at the pic of my new machine. The x and z motors are not on yet. The gantry weighs about 350 to 450 lbs (best guess) Linear guides in all directions. Ball screw z. Speed should be about 12 ips max and resolution of .0003 Not bad for a home built. Gantry plates were water jet cut and all linear guides were aligned using K&E transit. Within .002 of parallel over 10'. Have a way to go but I think it will be a great machine when it's done.

Mike, Why do I need 300 lbs of force? Because I designed this machine to be a belt drive machine but used rack and pinion for now to get back in business. I actually have 600 lbs since i have two motors on the x. I would like to be able to stop and go without losing steps since my gantry is heavy. Besides, I don't have to push the machine to the max because I have big steppers. I can ramp slowly into cuts like any other machine can.
3817

[richards]

Good luck, Wes! You've got some interesting concepts going on with your new machine. I'm sure that a lot of us will be watching to see how everything turns out.

[richards]

Wes,
My comment about the 300 lb*in may have been misunderstood. In normal operation, as you ramp up the speed and as the cutter cuts the material, you would only be using a fraction of the available torque of your geared motors. My concern is about the possible damage to the spur gear, the rack or the gear box when something out of the ordinary happens. Let's say, for example, that something in the material stops the cutter from cutting. If that happened and the cutter didn't snap, then the spur gear, the rack and the gearbox have to be able to handle the torque until the motor stalls. I have no idea how much torque each of those components can handle. I do know, however, that when I installed 3:1 belt drives on my PRT-Alpha, which gave me about 1,500 to 1,800 oz*in of holding torque or about 100 lb*in, that I had to cut my jog speed back to 12-ips instead of 30-ips so that I had time to stop the machine if I had given it the wrong jog distance. (I only have the prox. switches installed on the 0,0 end of the machine. If I get careless and jog past the 120-inch or 60-inch table limits, the spur gear(s) will hit the aluminum stops.)

[weslambe]

I'm not worried about hitting anything. I have snapped 1/2" bits before and it didn't do a thing. I don't expect to be running off of the rack and pinion for too long since I do have other thoughts on belt drive systems. (zero backlash and no rack chatter at all)