14x40 Lathe Retrofit

I'm starting this thread to ask for some help and guidance updating the retrofit on my Takisawa 14"x40" tool room lathe. It was previously retrofit with Centroid controls, ball screws, linear scales, and DC servos. That control went up in smoke a couple years ago, which is when I was given the lathe in trade. I did a lot of research and planned out an update but the project fell off the radar for various reasons, but one was my indecision on how to proceed.

I'm now at a point where I want to get the lathe up and running. I want to ditch the linear scales and current motors, but obviously will keep the ballscrews. Everything else will have to be replaced. I also have a red cap Fanuc servo (model 10, serial A06B-0501-B201, 12Nm 2000rpm with Pulse Coder and brake) that I would really like to use as the spindle motor - it would be replacing a 2.2KW DC motor.

I like the Mesa cards and prefer to stick with those over the Pico systems.

I'm eyeing these servos to replace the Z and X axis motors: www.ebay.com/itm/IIS-Industrial-Indexing...&hash=item565677f144. Thoughts? I can't find a lot of info on them but the size and price is right.

My understanding is that from Mesa I would need:

5I25
7I77
8I20 (replaces the servo driver for the big Fanuc, right?)

I would also need smaller servo drivers, but here is where I am a bit confused. Does Mesa have a card that would drive the servos listed above (or any other servo from Yaskawa, Mitsubishi, etc)? Could I use a generic driver from AMC, or do I need matching drivers?

I believe the 5I25 has all the I/O I would currently need (mainly limit switches).

Lastly I really need some help understanding the best way to power these servos.

Any help is appreciated! A few years ago I had a much better understanding of the CNC world and now I'm feeling a little lost. This is my first project and I'm excited to get the machine running after walking past it for 2 years just idle.

andypugh wrote:

giz wrote:

I would also need smaller servo drivers, but here is where I am a bit confused. Does Mesa have a card that would drive the servos listed above (or any other servo from Yaskawa, Mitsubishi, etc)? Could I use a generic driver from AMC, or do I need matching drivers?

It rather depends on the nature of the encoder that those motors have. If it is a proprietary interface then there might be problems, if it is quadrature then it is very easy. If it is BiSS or one of the other absolute encoder formats supported by Mesa then it might be ok too.

I am running a set of 750W servos very similar to those on my mill using Mesa 8i20 boards. They may be a bit over-specified for the job, but work nicely.
AMC drives should work. It might take some effort to sort out the commutation if those motors you are looking at don't have Hall sensors. (and the connector doesn't look to have enough pins to have both encoders and hall sensors).

I think those motors would be a bit of a gamble without a bit more information about them

Have you tried contacting the URL on the motor plates? I had a look at their web-site but nothing seemed to match those servos. I suspect that they might actually be Panasonic.

As you have some perfectly good DC servos and drives and power supply why not start off using them? The 7i77 will drive them perfectly happily. If they are inadequate in some way, then upgrade then.

Thanks for the response Andy, I was hoping you'd chime in.

I hear you on the uncertainty with those motors. I looked up some servos I was interested in years ago from DMM Tech today and learned that I could pair them with Mesa boards so now I'm leaning that route. The 750W servos with the new Dyn2 drives would fit well on this machine and the Mesa boards would allow me to run the Fanuc servo as the spindle.

The reason I'm avoiding using the DC motors is somewhat of the same objection you had for the ebay motors - I don't know anything about them and could not find any info online. But I also didn't consider the possibility of using the same control boards for both motors...

So it sounds like I could pick up a 5I25, 7I77, some power supplies and maybe get up and moving fairly quickly. I'd have to figure out the pinouts my DC servos and tune them, as well as figure out how the spindle encoder works (its 4 or 5 wires). Am I missing a component here - the servo drivers?

I just checked an all the encoders cables are 6 wire. The servos have 4 more wires, 2 pair of black and red.

The alternative is to go all-in with the DMM motors (self tuning and well-documented) and the Fanuc servo as a spindle. Does you BLDC config work for the Model 10 (not the 10S)? This sounds like it could be the most difficult part of the job, but a live C-axis is important to me down the road. I also have no idea what power supply the Fanuc would need.

I'm starting to remember why I never made any progress on this before! Too many options.. haha

The old Dyn2 was a bad choice as it was 0-5V + direction with the rather unpleasant feature that the 0-5V input floated high (max speed) when disconnected. The new DYN2 looks OK in that regard except that it does not appear to offer encoder output to the controller.
If you retained the glass scales and ran the motors in velocity-servo mode then this would probably be OK. Or you could run the servos as steppers (though not with the 7i77, you would want 7i76)

The DMM motors can take step/dir or +/-10v, but the loop is closed at the drive. Does the 7I77 require that I connect the encoders and close the loop in the control? Am I losing anything by using the 7I76 and step/dir in terms of performance?

I would expect two big red and black cables to be servo power and two thinner ones to be the output from a Tachometer.
The original Centroid control used velocity-mode servo drives, where the drive output is (approximately) the command voltage in - the tachometer voltage output. The Centroid controller then sent velocity commands to the drives to achieve the required position on the scales. (Or that is my understanding from what I have read here).
So, I am a little unsure which encoders you mean?
To run the servos + tachs you need an old-school serve drive, but the machine probably has perfectly good ones, and the right power supply.

Ahhh, you are jogging my memory here. I remember now thinking these were run in velocity mode with the tachometer and that was a big reason I planned on moving away from them. I only have the Z scale (and the "reader" is disassembled, unknown if I can fix it) and I don't have anything from the original control box (servo drives, power supplies, etc).

Sorry about the confusion on the encoders, clearly I'm still a little confused myself! But I believe the DC motors are fitted with resolvers and that is what the 6 wire cable is for. Since it's the same cable, I think the spindle may have a resolver as well.

I don't know. Do you know what the encoder output is? If it is 4-wire Halls then that is handled.

I don't know what the encoder output is. I can pop the top off tonight and see if I can get more information on the encoder. I do know its labeled as a Pulse Coder.

Knowing what I know about the dc motors and their resolvers, I think I'll stick to the plan to update the X and Z motors. And for the DMM Tech, it sounds like I need to go with the 7I76 because the loop is closed at the drive.

The spindle resolver is setup in a way that would easily allow me to swap the DC inudction motor for the Fanuc servo and keep the resolver. My Fanuc servo has a 17 pin encoder and I'm certain it is from the same family as the one jmelson talks about here: www.cnc-arena.com/en/forum/fanuc-abs-pulse-coder--172614.html

It looks like he figured it out but I don't want to throw hardware at a problem if I don't need to, plus still I want to stick with Mesa cards. I could replace this servo with another if needed - I have to weigh my options. The current resolver will need an additional Mesa card, the current Fanuc servo might need Pico hardware, etc.

They have a data sheet available for the old 14 bit encoders - it says they use SPI. It looks to me, from just reading the website, that the Mesa 5I25 can handle that. I'd have to contact DMM about the new encoders but if they are the same it seems like I could potentially bypass the drives, but I'm still not certain the benefit of closing the loop in the control vs. the drives.

I'm still weighing my options but the DMM motors/drives look to be a sure thing. There is still lsome uncertainty about the Fanuc motor on my part, and I'm starting to wonder if I would be better off trying to find another ~2kW motor ebay.