Hardinge Lathe CNC Retrofit/Conversion

[Muzzer]

Haha my Tree is described as "ultraprecision". That just seems to mean it has good quality bearings and ballscrews (all HSK) and the bed / frame is solid cast iron but still there is no means of compensating for any backlash after the ballscrews other than rely on the quality of the parts to minimise it.

I swapped the original 4000rpm 5HP DC spindle motor and drive on the Tree for a Yaskawa VFD and 3000rpm 4kW TEFC induction motor which work nicely with the Acorn using the 0-10V analog output. My Shizuoka mill came with a nice 6000rpm Hitachi induction machine, an ancient VFD and a 2 speed gearbox. With a modern Yaskawa VFD, this works fine up to 7500rpm with the Acorn. And despite being intended for industrial use with a wide speed range, the motor has no separate cooling fan. unlike the Tree which has a little blower on the side.

[glbreil]

Another bit of a worry, is that the DC servos on the machine now are 1.3NM and Centroid recommended the Leadshine ELM2 750 watt AC servo motors to go with the Hickory control. The new motors are rated at 2.39NM so, but looking at the physical size they are tiny in comparison to the Siemens DC motors. I have never had anything with AC servos, I guess that is the norm?

Thanks Gary

[m_c]

In short, it's unlikely to get external encoders that will work with the Hickory.

In full, Hickory is Centroid's EtherCAT system, which means technically you should be able to use any EtherCAT drives.
However the choice of EtherCAT drives is quite limited, and most EtherCAT drives are newer designs which make use of absolute/serial output encoders, not conventional quadrature output encoders.

If you could find a suitable external encoder (Leadshine EL7's are using 23-bit encoders), although you could connect it to the servo drive instead of the motor encoder, there may be a couple issues.
First is if the drive is using the encoder to establish the servo motors rotor position, then if you move the encoder to a random position, the motor could run roughly, or just not run at all.
Second is the potential issue that the drive may not handle any additional backlash between the motor and the encoder, leading to performance issues.

If you really wanted to make use of the additional encoders, you'd have to consider something like a Dynomotion Kflop or Kogna, as they can close the loop between the drive and additional encoders.

[m_c]

Another bit of a worry, is that the DC servos on the machine now are 1.3NM and Centroid recommended the Leadshine ELM2 750 watt AC servo motors to go with the Hickory control. The new motors are rated at 2.39NM so, but looking at the physical size they are tiny in comparison to the Siemens DC motors. I have never had anything with AC servos, I guess that is the norm?

Thanks Gary

Not entirely normal, but some older servos were quite large things, but there were still some fairly compact ones.
I'd be more interested in how the torque/speed/inertia compares.

Inertia selection isn't an area I know much about, but I know when I asked somebody who was knowledgeable about it when looking for servos for my big mill, they advised to go for the larger 1250W servos, as despite the smaller 1000W servos having comparable performance to the original 1000W DC SEM servos, they said the larger frame size would be a better match and easier to tune.

[glbreil]

Thanks, I probably should have started a new thread, but when I asked the first question I had. I idea it would lead to so many others.

I believe I have worked out how the Hickory control will give the command for the Fanuc AC spindle drive to run and control speed.

What I can’t understand of how the drive will give feed back to the control. The Fanuc drive gets feedback from the spindle encoder, but how is the position information transmitted to the control?

The drive has a Speed Arrival output which I think must transmit the position, but I don’t see where that would be received and coordinated with the axis drives for threading and such.

Gary

[m_c]

There will need to be some form of encoder signal to the controller from the spindle.

[glbreil]

Yes that is what I thought, I believe the Fanuc drive must put out the signal through the Speed arrival output, but I don't see anywhere on the Hickory controller to accept it. Fanuc Spindle IO.pdf I may need to contact Centroid.

Thanks Gary

[m_c]

There is most likely an independent spindle encoder, which will connect directly to the controller.

If you look at the Hickory wiring diagram, it makes mention of a spindle encoder cable (DB9 connector at lower left of the hickory board)

[glbreil]

The Fanuc spindle encoder connects directly back to the drive through the CN2 cable mentioned on the Centroid/Fanuc Spindle I/O. If it requires a separate encoder to get feedback to the controller, it might be easier to just use a simple VFD rather than the Fanuc drive.

[Muzzer]

If you don't plan to use ethercat, it looks to me as if the Acornsix would do the closed loop control of both the axes and the spindle. It's less expensive than the Hickory and that also applies to the associated software.upgrades.

If you prefer to use the motor encoders for axis position feedback, rather than use linear encoders, you can usually take an "encoder output" from the servo driver back to the system controller. This encoder signal can be a different resolution to the actual motor encoder itself - you can set the resolution in the driver setup

Most spindle VFDs are intended for speed control, rather than position control (which would be a servo's job) and they tend to be sensorless. This means you need to provide your own position encoder unless you have a fairly exotic VFD with position sensing. Furthermore, for cnc control, the spindle encoder needs to be directly geared to the spindle 1:1, which isn't usually the case if the encoder is mounted in spindle motor - unless the motor is part of the spindle itself.