Slant-Bed CNC Lathe Build

[macardoso]

These are built solid as a brick sh*t house

Love it!

 

[barnbwt]

So my controller showed up, and now I'm going to try to get it energized this weekend, and hopefully not myself in the process. First order of business is a bunch of document research to double-check-re-verify that it runs on 24VDC the way I think it does. Apparently the servo & stepper drive boards can be configured to share power with the main board so only one power connection is necessary, but at 48V rather than 24V.

Board looks really good, btw; even came with some expensive-looking tape connectors that combine all the Dsub encoder & i/o plugs back into a single unhelpful 50-pin bus plug (I think it's the same style used in the big Galil breakout panels, but I won't be using them). Price was very good, so if the boards aren't toasted, I assume it's because the seller didn't quite know what he had (the daughter boards cover up most of the markings)

I've been looking back into motors, trying to remember what all was lost. These guys look to be sized about right for use with a 3A, 48V driver (though mine goes up to 60V);
https://www.ebay.com/itm/NEMA23-381...2100-35-4B&_from=R40&rt=nc&_trksid=m570.l1313

I would be only running them to about 700rpm before their fall-off; this gets me minimum torque of 10.6 in-lbs per my calculations and their torque-chart. That rpm would traverse the Z-axis in 5 seconds at a 1:1 ratio, and the X-axis in the same 5 seconds at a 3:2 ratio. More phoney-baloney calculations suggest this would only apply around 300lbs in the Z axis; I'm not sure that's quite enough to be rigid, even on a machine this light.

So now I'm looking at servos, since they appear to generate about twice the torque at that rapid speed, and still a good 50% more even at low speeds. Plus I know they're generally more efficient all around. What I'm still unsure about are my options in the 200W 60V range, and exactly what types of servos I can run. It seems like everyone is calling every kind of digitally-controlled motor a servo now, which isn't helping.

It looks like Animatics makes a line of NEMA23-sized 48V servo motors that seem like a logical (if not economical) choice. Any other brands or product lines out there I should look into?

 

[JimDawson]

Exactly what Galil board did you get?

 

[barnbwt]

Tis a DMC-2183, with SDM-20640 and AMP-20440 daughter cards

 

[macardoso]

Here are my thoughts on the stepper/servo debate.

Steppers remain king in the hobby world because they are cheap and easy to apply. Their moderate speeds suit us hobbyists who will likely crash a machine moving 1000+ ipm (at least i would ).

However, servos are truly better in every aspect except for price. And when I say servos, I am talking specifically about modern AC servo motors. They maintain a fairly flat torque curve out to their rated speed and have a peak output of somewhere around 300% of the rates torque. What this means for you is that you only need a servo that produces a rated torque of 1/2 or 1/3 of the size of stepper you would buy. I would aim for the 200-400W range of servo for a mini mill to a bench top mill or lathe, and 750W for a machine the size of a tormach, however you can push that in either direction.

There are DC fed drives for AC servos (Look at the DMM technology DYN2 drive) but a vast majority are fed from the AC line. Expect single phase input up to 2kW and 3 phase input after that. I’ve used the DYN2 before and was quite happy with it. Not too if the line, but the price can’t be beat. Remember you need a beefy DC supply to feed them at at least 60VDC.

I personally would stay away from DC servos unless you want an electronics project. Closed loop steppers sound like a nice mid point but I’ve never tried them. They cost nearly as much as a low end AC servo.

 

[Karl_T]

You most certainly want servos. Don't skimp on size. Jim bought from a really good servo vendor on his recent build. I am a fan of DC servos on ebay - they go for a song if you watch a while. Then use AMC servo amps which also go for nothing.

You might consider more I/O with a DB 28040 card. An operator panel with lots of I/O really makes the machine

 

[barnbwt]

Here are my thoughts on the stepper/servo debate.

Steppers remain king in the hobby world because they are cheap and easy to apply. Their moderate speeds suit us hobbyists who will likely crash a machine moving 1000+ ipm (at least i would ).

However, servos are truly better in every aspect except for price. And when I say servos, I am talking specifically about modern AC servo motors. They maintain a fairly flat torque curve out to their rated speed and have a peak output of somewhere around 300% of the rates torque. What this means for you is that you only need a servo that produces a rated torque of 1/2 or 1/3 of the size of stepper you would buy. I would aim for the 200-400W range of servo for a mini mill to a bench top mill or lathe, and 750W for a machine the size of a tormach, however you can push that in either direction.

There are DC fed drives for AC servos (Look at the DMM technology DYN2 drive) but a vast majority are fed from the AC line. Expect single phase input up to 2kW and 3 phase input after that. I’ve used the DYN2 before and was quite happy with it. Not too if the line, but the price can’t be beat. Remember you need a beefy DC supply to feed them at at least 60VDC.

I personally would stay away from DC servos unless you want an electronics project. Closed loop steppers sound like a nice mid point but I’ve never tried them. They cost nearly as much as a low end AC servo.

Agreed, servos are definitely the more optimum solution. I maintain that if they were simply easier to deal with they'd be far more popular in this retrofit/diy/hobby field; the cost is honestly a tiny mole-hill compared with the trouble of implementation (and the difficulty in implementation makes the cost a far riskier gamble)

Well after some more research, it looks like my servo amplifier board (I'm not going to go down the rabbit hole of dealing with more external drives at this time) is feeding power to the motion controller, and since it appears my unit has the 24V-range voltage converter installed for that job, it means my servos are limited to 36V maximum* (@3.3A, therefore ~100W max output). This board also only drives 2-conductor brushed motors. So I think that puts the kibosh on going with servos for now; I'd have to invest in additional drivers & power supplies to run them, and that's beyond my scope for the moment. Hopefully I learn enough from the servo spindle aspect to implement servos at some later date. For now, I was simply looking into it since the amplifier board came 'free' which may have helped offset the cost of the servos. It sounds like biigger, AC servos are what I'd want in any case.

*Now, if that shared-power supply feature can be disabled --and it perhaps can be by removing a jumper-- the servo board can be fed up to 60V, for the full 200W at 3.3A. That's *barely* enough for it to be worthwhile from what I've seen, though I'd be limited to a rather limited and expensive selection of used motors (mostly Fanuc or similarly high-end)

 

[JimDawson]

So at least for now you are planning on using stepper motors with the onboard amps?

 

[barnbwt]

You most certainly want servos. Don't skimp on size. Jim bought from a really good servo vendor on his recent build. I am a fan of DC servos on ebay - they go for a song if you watch a while. Then use AMC servo amps which also go for nothing.

You might consider more I/O with a DB 28040 card. An operator panel with lots of I/O really makes the machine

Impressive, those used breakout boards are more expensive than the controller assembly, lol. I'll keep an eye out, though; I agree that whichever --servo or stepper-- board I end up not using would be better replaced with an I/O panel.

I did find what *appears* to be a series of quality servo motors that fit the power & voltage envelope (assuming I can decouple the power supply of the servo board from the controller)
http://www.animatics.com/products/smartmotor
These run at 48V as best I can tell, and when geared down 10:1 or so for the rapid speeds I'm looking at the torque values are very powerful. Now they're kind of a fully-integrated setup designed to be run by RS232 if I read correctly, but it does still appear to have the usual motor & encoder outputs to run it like a 'dumb' motor/encoder pair synched up by the driver board.

 

[barnbwt]

So at least for now you are planning on using stepper motor with the onboard amps?

It's certainly the most clear-cut path. I will do some more research on servos to see if I can pull them off at this time, but I suspect it won't be practical. For sure I have to find out if that 24V pass through from the servo board to the controller can be disconnected without help from Galil (assuming they haven't already blocked my phone number, lol)

The path forward with servos would be;
1) Energizing the servo board with 48V without frying anything
2) A good deal on a used ~200W, 3A servo that can be geared down to about 30 in-lbs @ 500RPM (lets say <150$ per motor)
3) Nema23 interface and a <8" length envelope
4) The ability to tune this particular set of servos (I have no idea how realistic this is for the Galil driver board)

The path for steppers is;
1) Figure out how much slower my rapids have to be to swing a 60V stepper @ 3A (though it's not like rapids have to be 'rigid' in practice)
2) Find a Chinese stepper variant that matches these specs & buy one from one of a jillion identical vendors