z Axis ---- Quill or Head?????? LATHEMASTER

[bill south]

I've finally, after an entire year, finished my lathemaster mill conversion Yeah............ !!!!
Unfortunately, due to my expert planning I have a very poor design on the z-axis drive. I installed a 3/4 inch ball screw and nut (warren electric) and am driving it with a electro-craft 712 motor (1600 in. oz.), 1080 encoder with a 3 to 1 reduction pully. I'm using a g320 with a 60 volt, high amperage, power supply and mach 3. I have the gecko tuned at maximum (I think) and dampened enough to prevent oscillations. However, during a cutting cycle, the gecko faults, the head drops and mach NEVER senses the chagne. In other words, the DRO's don't change to reflect the movoement of the Z axis. I have yet to cut a good part.
I was attempting to face some aluminum with a 3/8 end mill just for the heck of it using the mach 3 conversational wizard (great). During the cutting cycle, everything appeared normal but the z axis was stepping down after every pass. The gecko was not faulting but the ball screw was rotating every so slowly (like clock hands). The DRO for the z axis was not changing!!!! Is there something in mach that would cause this???
Anyway, instead of swinging a 300# head, I need a good design for a square column mill quill drive system. Does anyone have a good design that thay can share or sell me? I really thought the setup as describe above would be adequate for a z drive system. My war cry of today is HELP ME PLEASE. This thing is driving me Crazy!!!!!!
Have a good day
billyjack

[bill south]

Bump!

[fkaCarel]

Okay mr Bump. What you describe is probably noise. The servo drive has feed-back and thus should, if not operated stay, in position. Possibilities: loose encoder disc, loose drive. Stupid question: you use a timing belt? And then electric: if the axle is not operated, the puls signal should be a clear 0 or 1. If the mechanicals are okay and you dont operate the Z-axis and it moves, the pulse line is "floating" or influenced by a near signal. That's scope-time or trial and error.

Carel

[bill south]

Thanks Carel;
Food for thought and yes, I do use a timing belt. I had not thought of the encoders and they may be the culprit. Questions for those in the know! I'm useing 1080 count encoders which results in 1080 X 4 X 3 (pully ratio) X 5 (.2 screws) = 64.8K pulses / inch. The encoders are 6 wire listed as A, B, not A, not B, 5V and ground on the wiring diagram. For my input I only used the A, B, ground and 5V wires and just cut the others. Is this a mistake???? Does it leave the signals floating??? The pulses looked great on the scope.
Also, I'm running my Mach3 system at 25K. The computer is new and has plenty of guts. Would it help to up the freq????
Any help will be appreciated.
bump. aka billyjack

[fkaCarel]

If the signal are truely Not A and Not B, they are only complementary signals and redundant. A not working encoder makes a servo very confused. If it's a speed problem, then start with low speeds and go up and see where the problem comes in. This also goes for the pulse. If there is a speed where the Gecko can't make cheese anymore of the pulse, then this also needs tuning. But you described a stationary problem.

Carel

[phil burman]

Sounds like an easily identifiable problem for those in the know. Unfortunately that's not me, but it's probably dog meat for the Mach and/or Gecko user groups.

Don't even consider the quill as an alternative. Its a far worse solution and you may still have to fight the same problem. A 1600 oz-in servo with 3 to 1 reduction should be able to pull the house down (or hold it up), let alone #300.

Solve the problem you have don't trade it for another.

Regards
Phil

Originally Posted by bill south I've finally, after an entire year, finished my lathemaster mill conversion Yeah............ !!!! Unfortunately, due to my expert planning I have a very poor design on the z-axis drive. I installed a 3/4 inch ball screw and nut (warren electric) and am driving it with a electro-craft 712 motor (1600 in. oz.), 1080 encoder with a 3 to 1 reduction pully. I'm using a g320 with a 60 volt, high amperage, power supply and mach 3. I have the gecko tuned at maximum (I think) and dampened enough to prevent oscillations. However, during a cutting cycle, the gecko faults, the head drops and mach NEVER senses the chagne. In other words, the DRO's don't change to reflect the movoement of the Z axis. I have yet to cut a good part. I was attempting to face some aluminum with a 3/8 end mill just for the heck of it using the mach 3 conversational wizard (great). During the cutting cycle, everything appeared normal but the z axis was stepping down after every pass. The gecko was not faulting but the ball screw was rotating every so slowly (like clock hands). The DRO for the z axis was not changing!!!! Is there something in mach that would cause this??? Anyway, instead of swinging a 300# head, I need a good design for a square column mill quill drive system. Does anyone have a good design that thay can share or sell me? I really thought the setup as describe above would be adequate for a z drive system. My war cry of today is HELP ME PLEASE. This thing is driving me Crazy!!!!!! Have a good day billyjack

[lerman]

Originally Posted by bill south Thanks Carel; Food for thought and yes, I do use a timing belt. I had not thought of the encoders and they may be the culprit. Questions for those in the know! I'm useing 1080 count encoders which results in 1080 X 4 X 3 (pully ratio) X 5 (.2 screws) = 64.8K pulses / inch. The encoders are 6 wire listed as A, B, not A, not B, 5V and ground on the wiring diagram. For my input I only used the A, B, ground and 5V wires and just cut the others. Is this a mistake???? Does it leave the signals floating??? The pulses looked great on the scope. Also, I'm running my Mach3 system at 25K. The computer is new and has plenty of guts. Would it help to up the freq???? Any help will be appreciated. bump. aka billyjack

Your encoders are probably differential (current) mode encoders. They tend to be less sensitive to noise than single ended encoders, but must be wired to the proper type of receiver. They will NOT work with a single ended receiver.

Ken

[Milkman5547]

Hi; Using just the A and B signals from the encoder will work but by doing so you lose all your noise immunity. That’s a bad thing when your running motors and such. There are several units available that convert differential encoder outputs to single ended outputs. The unit we use at work is called a model PU-204 but I can’t think of the manufacture off hand. Mount it close to the drive and any problems with noise will clear up. Good luck and hang in there.