Topics

PCMill 100 - Motor & Encoder info


SimonSFX@...
 

Hi everyone, thanks for adding me to your group!

I have a PCMill 100 which, when I bought it, was plagued with RS485 Errors - so I reluctantly embarked on the process to convert from Fanuc to Mach3 / 4.  Reluctantly because it's a lovely, clean, almost unused machine and I love Fanuc!

I've tried to do it as sympathetically as possible, using a CS Labs Motion Controller, keeping the original 5 Phase Stepper motors and I had intended to keep the original spindle controller.  Unfortunately, while it would work with the original controller, I couldn't persuade it to work via the motion controller.

Although I have a manual for the mill including circuit diagrams, some of the bits including the servo drive are different - so there was a lot of guessing!

This week, I admitted defeat and fitted an Allen Bradley Ultra 3000 servo driver, which is very easy to configure and can use Step / Direction (or Analog).

In order to configure it, I need some details of the motor & encoder - and wondered if any of you happen to know the answers.

The Motor is an AMK DVSA 07-4-2-4HOF-5000.
On line, I've found it is 190V, 2.5A Continuous, 5A Peak, 2.6 Ohm.
What I need to know is the number of Poles, possibly the coil inductance (less important).

The encoder has no markings - but works perfectly.  It gives about 4000 counts per spindle revolution.  I believe the gear ratio is 1.5:1 - but that seems unlikely as that would make the encoder 2666.666 counts per motor revolution.
In order to use the encoder to commutate the motor (no sign of hall effect commutation), I think the number of counts should be divisable by the number of poles - so, either the number of counts or the gear ratio is wrong. 

Any ideas? even for similar machines?

When I get it up & running, I'll publish details as there's nothing much on line. 

Thanks in advance,

Si


John
 

Hi Simon.

The encoder on my mill is a Heidenhan ROD 420 D.

It is a position sensor. I think there is a 1 pulse per revolution setting as well for speed management.

The wiring diagrams for this will likely coincide with your encoder and are available online.

Hope this helps.

John


rcflyer
 

There is no position feedback because it's an induction motor. AMK still sells the DV line. Mill 100 and 125 use the same motor. It is reduced 2:1 and runs 10kRPM at 5k spindle speed. I'm not sure about the encoder, but it shouldn't be difficult to sort and 1000 lines would be common.


On Sun, Oct 6, 2019 at 1:42 PM <SimonSFX@...> wrote:
Hi everyone, thanks for adding me to your group!

I have a PCMill 100 which, when I bought it, was plagued with RS485 Errors - so I reluctantly embarked on the process to convert from Fanuc to Mach3 / 4.  Reluctantly because it's a lovely, clean, almost unused machine and I love Fanuc!

I've tried to do it as sympathetically as possible, using a CS Labs Motion Controller, keeping the original 5 Phase Stepper motors and I had intended to keep the original spindle controller.  Unfortunately, while it would work with the original controller, I couldn't persuade it to work via the motion controller.

Although I have a manual for the mill including circuit diagrams, some of the bits including the servo drive are different - so there was a lot of guessing!

This week, I admitted defeat and fitted an Allen Bradley Ultra 3000 servo driver, which is very easy to configure and can use Step / Direction (or Analog).

In order to configure it, I need some details of the motor & encoder - and wondered if any of you happen to know the answers.

The Motor is an AMK DVSA 07-4-2-4HOF-5000.
On line, I've found it is 190V, 2.5A Continuous, 5A Peak, 2.6 Ohm.
What I need to know is the number of Poles, possibly the coil inductance (less important).

The encoder has no markings - but works perfectly.  It gives about 4000 counts per spindle revolution.  I believe the gear ratio is 1.5:1 - but that seems unlikely as that would make the encoder 2666.666 counts per motor revolution.
In order to use the encoder to commutate the motor (no sign of hall effect commutation), I think the number of counts should be divisable by the number of poles - so, either the number of counts or the gear ratio is wrong. 

Any ideas? even for similar machines?

When I get it up & running, I'll publish details as there's nothing much on line. 

Thanks in advance,

Si


SimonSFX@...
 

Thanks both of you.
The gear ratio and encoder type gives me that the encoder is 2000 lines per revolution.

You can just run it as an AC induction motor - but in it's original guise it was driven with Step/Dir for toolchanging.  To use it with an industrial servo drive, it can use an encoder (with a 1 PPR Index) for it to figure out the commutation - and achieve precise positional control.

I'll give these values a go - and see what it does!

Si


rcflyer
 

It's driven as an asynchronous motor. Rotor position feedback is not used for commutation as there are no permanent poles on the rotor.

Positioning with step/dir using encoder feedback is independent of driving the motor. Assuming the drive supports AC servos, this will be no problem. I don't think Ultra3000 will work. The drive Emco uses is a VFD.

Greg


On Mon, Oct 7, 2019 at 8:27 AM <SimonSFX@...> wrote:
Thanks both of you.
The gear ratio and encoder type gives me that the encoder is 2000 lines per revolution.

You can just run it as an AC induction motor - but in it's original guise it was driven with Step/Dir for toolchanging.  To use it with an industrial servo drive, it can use an encoder (with a 1 PPR Index) for it to figure out the commutation - and achieve precise positional control.

I'll give these values a go - and see what it does!

Si


rcflyer
 

Apparently, I could be entirely wrong. I'm working on a CM 155 and it looks like the spindle motor needs bearings. I pulled the top cover of the asychronous servo (spindle) motor to find a resolver. The machine uses a Lenze drive The resolver talks to the drive and there is a separate encoder for position. Now, I need to figure out how to service the bearings either without displacing the resolver or realign it after.

On Mon, Oct 7, 2019 at 9:54 AM rcflyer via Groups.Io <sclcl082=gmail.com@groups.io> wrote:
It's driven as an asynchronous motor. Rotor position feedback is not used for commutation as there are no permanent poles on the rotor.

Positioning with step/dir using encoder feedback is independent of driving the motor. Assuming the drive supports AC servos, this will be no problem. I don't think Ultra3000 will work. The drive Emco uses is a VFD.

Greg

On Mon, Oct 7, 2019 at 8:27 AM <SimonSFX@...> wrote:
Thanks both of you.
The gear ratio and encoder type gives me that the encoder is 2000 lines per revolution.

You can just run it as an AC induction motor - but in it's original guise it was driven with Step/Dir for toolchanging.  To use it with an industrial servo drive, it can use an encoder (with a 1 PPR Index) for it to figure out the commutation - and achieve precise positional control.

I'll give these values a go - and see what it does!

Si


SimonSFX@...
 

Following a bit more digging (and dismantling), The Motor is a Synchronous Servo.

AMK DYNASYN DVSA07-4-2-4HOF-5000
660W, 1.26Nm, 5000rpm, 3x190V
Fitted with a Heidenhain ERO1221B X000, 500 Line Quadrature Encoder.

The original driver - and EMCO FRC122C 0-400Hz 1.5kW

Although it's shown in the circuit diagrams, they have been simplified as there are half a dozen 'mystery' wires in addition to those shown.  I'm wondering if it has Hall Effect sensing as well as the encoder?
I've asked AMK for a legacy Datasheet - I'll post that if/when it arrives.

Si


rcflyer
 

A bit more research into my machine shows that the resolver is used only for velocity feedback to the drive. There is an encoder on the spindle, but I haven't determined if the signal goes to both the control and the drive. The documentation for the Lenze 9300 drive specifically states it supports AC asynchronous servos.

Mill 125 certainly does not have hall sensors as asynchronous motors do not use them. The spindle of Mill 100 and 125 use timing belts so they use the encoder in the motor for both position and velocity.


On Tue, Oct 8, 2019 at 4:02 AM <SimonSFX@...> wrote:
Following a bit more digging (and dismantling), The Motor is a Synchronous Servo.

AMK DYNASYN DVSA07-4-2-4HOF-5000
660W, 1.26Nm, 5000rpm, 3x190V
Fitted with a Heidenhain ERO1221B X000, 500 Line Quadrature Encoder.

The original driver - and EMCO FRC122C 0-400Hz 1.5kW

Although it's shown in the circuit diagrams, they have been simplified as there are half a dozen 'mystery' wires in addition to those shown.  I'm wondering if it has Hall Effect sensing as well as the encoder?
I've asked AMK for a legacy Datasheet - I'll post that if/when it arrives.

Si


SimonSFX@...
 

Good - and bad news!

I got the motor running pretty well with the drive with both positional and velocity control, using the encoder for commutation.  The PID Gains need a bit of playing with - but worked well enough for testing.

The bad news is the motor bearings are completely shot.  It makes a TERRIBLE noise when running - and now I've removed the motor, it feels a bit crunchy!

Since time is against me - I've decided just to replace the motor and drive with one of these:
https://www.cnc4pc.com/dmm-dyn4-ac-750w-servo-kit.html?SID=65e7215e5600dee6c3340466f9211369

It's the same frame (80mm) and shaft (19mm) size  and should be pretty much a drop-in replacement.

This motor has a top speed of 5,000RPM, compared to 10,000 for the AMK Motor.  It does develop 5x the torque though, so I plan to change the pulleys to give me 1:1 gearing with the same spindle speed range at twice the torque.


atsel02.ell@...
 

Hello Simon,

Did you manage to get the spindle working with the Allen Bradley Ultra 3000 with tool changing ,is it easy to set it up in mach?
I have a PC Mill 125 with the main board broken and I was able to make the vfd smoke while trying to figure out its pins 🤦‍♂️, and I don't  want to spend money unless I'm sure it's gonna work.

Thanks for all the information of the topic