Who Knows Servos?

[Doc Nickel]

I have an Acorn CNC setup, built off an old Logan lathe. It works, though I'll admit a few of my design choices weren't the best.

Anyway, I'm toying with the idea of adding a "C" axis- that is, making the spindle itself a positionable axis. This way I can index for cross-drilling and maybe even limited 3rd-axis milling.

This will require swapping the 3-phase motor with a servo that has a brake and encoder. The Acorn can supposedly control such an axis, but for the moment I'm looking into just the servo options.

I currently have a 2HP 3-phase, 3450 RPM motor run off a WEG VFD. 2HP is about 1.5Kw, can I just use a servo that big, or would I be better off with 2Kw or even 3? (Parts are almost always aluminum, and small enough to fit in a 5C collet. I'm more concerned with drilling than milling, and if i did mill, it'd be small features, not major hogging.)

Can anyone suggest a good brand and type to look for? The driver would have to accept step/direction signals from the CNC controllers.

And last, anyone tried this sort of thing? How did you wire it?

Doc.

[garyhlucas]

I would go larger as the servo holding position doesn’t have benefit of inertia that the spindle has. Automation Direct is a good choice, especially if you don’t know what you are doing. Best tech support going and it is free.

[Doc Nickel]

I was given to understand that servo doesn't "hold" like a stepper, which is why it needs a brake.

I doubt I can use a full two horsepower due to the limitations of the original Logan underdrive- the only reason I can think of needing more HP, is just to I have low-end grunt at slow speeds.

Doc.

[RB211]

The DMM DYN4 controllers from what I can see have three different operating modes. Position, Torque, and Velocity. Your issue is that you need it to change between Position and Velocity, and I don't believe they can do that. I'd say a brake would be a good choice for position holding.

[Sparky_NY]

The DMM DYN4 controllers from what I can see have three different operating modes. Position, Torque, and Velocity. Your issue is that you need it to change between Position and Velocity, and I don't believe they can do that. I'd say a brake would be a good choice for position holding.

That is correct. Velocity mode when using it as a spindle motor, torque or position mode when using it to position as a C axis. Actually velocity mode is also used for positioning quite often as well (my bridgeport drives operate in velocity mode but more modern cnc machines tend to use torque mode for the positioning axis) I have not seen this done with a step/direction drive and have doubts a single drive could handle operating in two modes selectable. Linuxcnc can do this because the feedback loop for positioning or velocity is in the computer rather than handled in the drive and the user can program all the various parameters that need to be changed between modes. When switching the spindle between velocity and positioning modes quite a few things have to change in the control to handle this, I highly doubt the acorn can handle it.



Best bet would be to post this on the Acorn forum or contact them inquiring if this sort of configuration is supported by the Acorn.

[elf]

Teknic makes step and dir servos. I like mine.

[Sparky_NY]

I was given to understand that servo doesn't "hold" like a stepper, which is why it needs a brake.

I doubt I can use a full two horsepower due to the limitations of the original Logan underdrive- the only reason I can think of needing more HP, is just to I have low-end grunt at slow speeds.

Doc.

No, servo's do indeed hold position. If the force is high enough it can be pulled out of position but will return to its original position (a function of the encoder). The more force applied to pull the servo out of position the harder the drive/motor will fight to maintain position. With nothing attempting to pull the motor out of position the drives output will fall to zero current, current increases proportionately to the force attempting to pull the motor from position. Servo tuning has a lot to do with how aggresive this "servo loop" responds.

[Sparky_NY]

One other thought. Using a servo for both a spindle motor and as a positioning C axis is getting out of the hobby user territory and more into the methods the "big boys" use. The acorn is intended for step/direction drives and the high end commercial machines (Haas, Mazak, etc) always use analog 0-10V control for the servo motors. When positioning there are two PID loops in play, a velocity loop and a position loop. Step direction drives generally have both of these loops in the drive itself. Analog drives (0-10V) generally have the velocity loop in the drive and the position loop in the controller, this opens a lot more possibilities for varying control parameters and changing operating modes.

[Sparky_NY]

A few Years ago Macona did a retrofit of a lathe to a servo motor for the spindle. I believe he did a project thread on it over on cnc zone. I don't recall if he also had positioning capability. He could give some excellent input to this thread.

Edit: I recalled it was cnczone he did the build thread on, but he may have done it on more than one forum. A quick google search found a build thread of his on PM. https://www.practicalmachinist.com/v...-motor-154967/

[Doc Nickel]

Your issue is that you need it to change between Position and Velocity, and I don't believe they can do that.

-Presumably you mean the DMM servos specifically, as we know that there are some servos out there that can do just that.

The OmniTurn, for example, can optionally come with a servo drive, that can run the spindle in position mode for milling and cross-drilling. Apparently it can even work in a way that produces a functional Y axis.

I don't know how that works, or what programming steps are necessary to swap from one to the other. I doubt it's a case of having to stop the machine and flip physical switches, so presumably it's some way the controller/driver can be told which one to use.

Point in fact, I'd convert the Omni using factory-supplied components, except the driver alone is $3,000, I suspect the servo would be $2,000, and in any case, I need to work on parts long enough to require a tailstock.

Best bet would be to post this on the Acorn forum or contact them inquiring if this sort of configuration is supported by the Acorn.

-I posted this same question on the Acorn boards, but thought I'd get a wider set of viewpoints asking here, too. The Acorn board, I've also found, isn't as helpful with questions outside a fairly narrow scope of physically wiring and/or programming/configuring the board. Things like mounting encoders, sizing steppers, and just about everything else not directly related to the board itself, is "that's up to you."

Doc.

[RB211]

-Presumably you mean the DMM servos specifically, as we know that there are some servos out there that can do just that.

The OmniTurn, for example, can optionally come with a servo drive, that can run the spindle in position mode for milling and cross-drilling. Apparently it can even work in a way that produces a functional Y axis.

I don't know how that works, or what programming steps are necessary to swap from one to the other. I doubt it's a case of having to stop the machine and flip physical switches, so presumably it's some way the controller/driver can be told which one to use.

Point in fact, I'd convert the Omni using factory-supplied components, except the driver alone is $3,000, I suspect the servo would be $2,000, and in any case, I need to work on parts long enough to require a tailstock.



-I posted this same question on the Acorn boards, but thought I'd get a wider set of viewpoints asking here, too. The Acorn board, I've also found, isn't as helpful with questions outside a fairly narrow scope of physically wiring and/or programming/configuring the board. Things like mounting encoders, sizing steppers, and just about everything else not directly related to the board itself, is "that's up to you."

Doc.

Correct, DMM with current firmware. A phone call to them(they are a Canadian company, not Chinese) and inquire about any special firmware they may have to allow for this.

[Noitoen]

Could keep the original motor and mount a weaker servo for the special function. This could be easily achieved with a removable timing belt. The main motor could tag along or have the belts slackened for the servo mode

[skunkworks]

Remember that servos have error - usually a dozen or so encoder counts. You would certainly want to good belt reduction..

sam

[MaxHeadRoom]

Remember that servos have error - usually a dozen or so encoder counts. You would certainly want to good belt reduction..

sam

In my experience doing CNC retrofits using Mitsubishi systems or even home built using the Galil Motion cards, I would consider servo positioning to a dozen encoder counts a gross error and unacceptable.

[skunkworks]

In my experience doing CNC retrofits using Mitsubishi systems or even home built using the Galil Motion cards, I would consider servo positioning to a dozen encoder counts a gross error and unacceptable.

Sure - if we are talking about hitting a target position - now add side force of cutting. The servo motor deflects and pushes back. You are not going to get that down to 1 encoder count. Normally on a positioning type (ball screw) system - the dozen or so encoder counts are well within the positioning accuracy of the whole system.

Now a rotary type positioning system is harder and usually needs a bit of gear reduction to get the 'accuracy' wanted. (this is why you normally use a indexing head gear box or harmonic drive system. Just throwing a servo onto a spindle and expecting it to position to your expectation without a decent reduction is going to be disappointing..

sam