Can Centroid Control this?

[Chaz]

Hello,

I've seen some DC motors with both tachogenerator feedback and encoder feedback. I see a lot of references to "measuring system" inputs on the controller and they sure look a lot like encoder inputs to me. The language in the manual suggests that they are linear encoders on the axes and not rotary encoders on the motor.

It would be worth asking the seller if there are encoders going from the motors to the Heidenhain control - even if they are not differential quadrature output it will at least make it easier to install the appropriate type. If the machine only has linear scales then you may want to ask Marc or someone else more familiar with Oak systems about that kind of setup.

And yes, you can add more axes to the Oak with OpticDirect expansion boards.

And I would assume that 'encoder' feedback could come from any suitable device, in other words, doesnt have to come off the motor shaft, can be off the ballscrew / scales etc, anything that offers feedback (in a suitable format), correct?

[cncsnw]

You can probably run that system with Oak (+OpticDirect) in analog velocity mode.

The drives need the tachogenerator feedback from the motors, to do their speed control.

The Oak needs quadrature position feedback to know where the axis is. No, there is no way to usefully generate a quadrature encoder signal from the tachs. That would be like trying to determine your car's exact location at all times, by reading only the speedometer. Theoretically possible, but not practical.

If the position control comes from linear scales, and not from motor-mounted encoders, then stable control will only be possible if you can eliminate all backlash and static friction. A change in motor speed/position needs to immediately result in a corresponding change at the feedback device.

[Chaz]

You can probably run that system with Oak (+OpticDirect) in analog velocity mode.

The drives need the tachogenerator feedback from the motors, to do their speed control.

The Oak needs quadrature position feedback to know where the axis is. No, there is no way to usefully generate a quadrature encoder signal from the tachs. That would be like trying to determine your car's exact location at all times, by reading only the speedometer. Theoretically possible, but not practical.

If the position control comes from linear scales, and not from motor-mounted encoders, then stable control will only be possible if you can eliminate all backlash and static friction. A change in motor speed/position needs to immediately result in a corresponding change at the feedback device.

Thanks. Understood. Didnt expect Tacho to replace Quad. I have a CS Labs setup running IP-A, same idea (but crap Mach 4).

[aamir]

Hello Chaz
Just Try a DC drive of Deckel Old machine with the linear scale .

https://youtube.com/shorts/9bRd3Ve8pOg? ... i-EoVSRif2

Linear Scale used in this is 1vpp and I use EXE601 for conversation.

Attached is the pic of the DC drive

As it's in testing phase that's y positioning need more tunning .Also due to linear scale tunning is tricky..
For testing I use mpu11 gpio4d as it's easy to use for testing.
For final project I use oak board .
Aamir

[Chaz]

Just taken a look at the machine again.

Interesting observations.

The DC motors all look identical, even for the 4th and 5th Axis but run smaller drives (assumably due to lower load).

There is however no other feedback than Tacho. We couldnt find any linear / glass scales or any other form of feedback. Thus, it must be Tacho only. Fitting encoders wont be easy, there isnt enough space, certainly not on the 4th and 5th Axis. You might be able to on the 3 main axis.

The 4th and 5th axis looks like it has something on the rear that looks larger than tacho but found a wiring connector and it had T+ and T- written on it and there's no enough wires for for quad encoder feedback.

Can Oak run without this encoder feedback?

[cncsnw]

Can Oak run without this encoder feedback?

No.

Neither can Heidenhain, unless all it does is move back and forth between limit switches. What exactly is a "Gate machine"?

If the Heidenhain control had a DRO position display, and the ability to move to programmed positions, then there is position feedback somewhere.

If there are not encoders on the motors, and there are not linear scales anywhere on the axes (sometimes they are well hidden), then there may be rotary encoders on the ends of the ballscrews.

[Chaz]

Can Oak run without this encoder feedback?

No.

Neither can Heidenhain, unless all it does is move back and forth between limit switches. What exactly is a "Gate machine"?

If the Heidenhain control had a DRO position display, and the ability to move to programmed positions, then there is position feedback somewhere.

If there are not encoders on the motors, and there are not linear scales anywhere on the axes (sometimes they are well hidden), then there may be rotary encoders on the ends of the ballscrews.

This is the machine in question.

https://www.ebay.co.uk/itm/115955029471

[Chaz]

And I took this pic, wanted to see if there was anything that looked like encoder wires coming in. Pity I didnt take more of the plugs.

[Chaz]

Suppose thinking through, its not step / direction, so it cant pulse a distance, so yes, there must be some feedback somewhere.

[cncsnw]

Hmm. Five-axis machine, and there in your picture are five connectors that look like Heidenhain M23 12-pin feedback plugs. Each plug bringing in nine wires, including four twisted pairs consistent with Heidenhain position feedback (BLU+RED, YEL+GRN, GRY+PNK, WHT+BRN).

Because it is Heidenhain, there is a good possibility that the signals on those pairs are analog sine/cosine signals, and not the quadrature square wave that you want. In any case, the first step is to find the sources of those five cables.