BMC25 - 4th axis

[gthlm]

Cliff, at last I have the rotary hooked up and all is working except the rotary parameters were lost when the batteries died and the encoder counts are wrong. Do you happen to know the correct encoder resolution and the max speed?

Cheers, Gregor

[bloke]

Should be 2000 counts / degree and I think 6 RPM.

Gregor, go to bed!
G'night, fella!

[gthlm]

Haha yeah it was a late one last night, still I could say the same to you!

Anyhow I was back to messing with it this morning. By the trial/error/calculator method, it turns out the correct resolution is 1426 counts/degree which seems a bit random but I took it four full turns in one direction and it seems accurate and correct. Also I have the max rpm at 4, more than that causes motion errors during rapids, but I really don't know what that one should be at all. When calibrating I get an a axis servo alarm if it has to travel too far (more than quarter of a revolution - ish) to the calibration marker. After it's cal'd though it seems to work fine.

Another minor curiosity is that it does not move very smoothly using the mpg. You have to watch closely but you can see turning a bit irregularly when you are turning the mpg quite steadily. Seems to work fine on auto though.

Any thoughts on what I've done so far, the servo fault when cal'ing or how to set the max speed correctly?

Cheers mate!

[bloke]

It sounds like you have a dirty encoder. If it's run really slowly in program, does it jump at all? It sjould always find a marker in less than 4 deg.

[gthlm]

Well Cliff, I followed your advice and opened up the encoder, but found it not dirty but broken in several pieces! How it was working at all I have no idea...



The encoder is made by Litton and I'm going to look for a replacement but I don't have particularly high hopes of finding one, more likely a new motor will be called for. Do you have any leads for finding an encoder?

Thanks, Gregor

[bloke]

Gregor, I've sent a mail to your fastmail account.

[fasto]

Gregor I don't know what Bloke's got up his sleeve, so I will just mention that all encoders are about the same. If you can source a 2000 PPR encoder (if that's the proper resolution) with a marker pulse and get it to fit on the motor's shaft you will be fine.
I'd imagine that any number of outfits catering to the CNC conversion crowd would have such a device available.

One other possibility that I ran into is a program that will print the encoder graduations using a laser printer. I don't exactly recall where I saw it; I will look if you would like.
Here it is, Codewheel Generator, though I'm not sure that it will do 2000 PPR as it seems more geared towards the jog wheel type encoders.

[gthlm]

Aaron, thanks for your suggestions. I was unable to find a suitable replacement (1000ppr, 1/2" hollow shaft mount) but with Cliff's help we got a hold of one and he helped me set it up.

Here's the first test piece:

[bloke]

Gregor,
Glad to see you got it all sorted in the end. Looks to be cutting good there!

[fasto]

Very nice Gregor. Looks like an "Archimedian Screw" for raising water up hills.

[gthlm]

Thanks, it's not anything particular, just three grooves wrapped two full turns over a random distance based on how far I had it sticking out of the chuck. Just wanted to prove that I'd got it working properly! May have some stainless marine waste pump screws to make later in the week though if the customer can get me one to copy before I leave.

But it's working nicely. I also replaced one of the screens (dying tube) and hooked it up to a new 20Hp digital phase converter (dad has ran all his machines on an old rotary converter for years, and none of the cnc's ran very well on it - for years he has ignored my pleas for him to replace it, now he's kicking himself for not doing it sooner).

Now I must prepare for the thankless task of showing him how to use it...

Aaron, have you had any success with your own 4th?

[fasto]

Gregor I'll have an update tomorrow evening EDT.