Lathe Variable Spindle motor options? VFD? Servo? Stepper? DC?

[whiteriver]

I am building my second cnc lathe and this one has a 6 speed belt and
pully setup. Its a real pain to be pulling the belt of one pully and
placing it on another all the time.
Questions are. What is the best setup for having a variable speed?
VFD and 3 phase motor? I would like to run 110V ac to the machine.
Trying to avoid 220 and 3 phase is not a option.
DC motor with DC speed Control? Can Mach contol a DC speed control? I
think it can do some VFD's?
Or since I would like to add a "C" axis to this lathe would it be
better to add a 1700 oz nema 34 stepper or some size of servo with a
encoder. The current size motor is a 3/4 hp AC cap start. So I need to
find a setup with comapariable power and a speed range from 60 to 2000
rpm.
Whatever I put on must be avalible for replacement. Can't rely on ebay
or surplus center for this. Commercial product and all.

Donny

[Karl_T]

The best option is VFD and three phase motor. You'll need 220 single phase to run this. You'll have to increase your motor size considerably to have the same torque at low RPMs as you do today. A VFD motor run at 1/2 speed has half the horse power, torque is constant and horsepower is RPM * torque. If you try to run all speeds with no belt changes, you could try running a two pole (1800 RPM) motor at 3600 to get 2000 RPM. The motor would then run 108 RPM at a spindle speed of 60. You'd have about 6% of name plate horse power at that speed or a 10 horse motor to get back to what you have today. That's why CNC machines have such huge motors. If you have a backgear(2 speed gearbox) for low speeds, you could cut your motor size considerably.

Karl

[Al_The_Man]

If you want to avoid 240v then, I would consider DC or AC PM motor & drive, from the likes of Baldor etc, It wont be the cheapest way however.
I believe there some VFD's that are 120 1ph in, but you may have problems getting a 3ph 120v motor.
Personally I would get a high grade flux vector VFD 240v 1ph in with a 3ph 4 pole motor and encoder input, they can be obtained with torque capability down to zero speed, The only thing is, if you want C axis control from it and it is not direct drive to the chuck or at least zero backlash from motor to chuck, you will have to put the encoder on the final chuck shaft, and scale it accordingly.
You will also need a motor cooling fan for low speed operation.
Al.

[whiteriver]

Looking at VFD's from Automation Direct they have a GS2 that takes 120v 1phase power and produces 3 phase 230vac. The GS2 is not Vector control. The GS3 is but its only takes 230 3 phase input. If I have to I will go with 240 1 phase input.
AL your saying if I can find a good (flux vector?) VFD with encoder input I can use that for my C spindle? I notice they have breaking. Will it hold enough and index accuretly enough to rotate my spindle to a set point and use live tooling on it from the cross slide? This would act like a servo drive I would think. All new to me. I'm a stepper man.
So with the right VFD and a 1hp 4 pole motor 3 phase motor it would be equal to the 3/4 hp on the lathe now? I would not need a 10 hp motor? Plus I could interface with Mach3 and control the speed and position? Am I following along on this or did I go down the wrong channel in the river without my oars?

Donny

[Al_The_Man]

There are VFD's out there that can closely compare to DC servo's, you have to buy the proper unit. I don't know why Auto.Dir limit their vector unit to 3ph as the motor is not that large.
There are many out there that go up to around 2hp and above for 1ph input.
I have seen units used on Gantry crane hoists, that will hold a stationary load, that was previously done by a brake on the AC motor.
Mitsubishi is one of the better makes of VFD's.
I don't know wether Mach3 has the ability to feed the analogue signal required I believe someone has a convertor to interface the two.
Al.

[gotis]

Originally Posted by Al_The_Man There are VFD's out there that can closely compare to DC servo's, you have to buy the proper unit. I don't know why Auto.Dir limit their vector unit to 3ph as the motor is not that large. There are many out there that go up to around 2hp and above for 1ph input. I have seen units used on Gantry crane hoists, that will hold a stationary load, that was previously done by a brake on the AC motor. Mitsubishi is one of the better makes of VFD's. I don't know wether Mach3 has the ability to feed the analogue signal required I believe someone has a convertor to interface the two. Al.

Is´nt that the pixiedrive?

[RotarySMP]

You do not use an encoder with the VFD. Flux Vector VFD are looking at the motor windings directly and working out whats needed.

[whiteriver]

Why do the Flux Vector units offer a encoder input option if they don't need it? The option is almost as much as the Vfd base price. So I got the answer above about holding power. What about posiontal accuracy? Will they be a accurate as a servo with an encoder? I seen where AL says they will closly compare. How close? Would not want to invest $1200 and find out it won't do the job.
Why not get a rutex or pushing it a gecko servo drive and get a large 90vdc pm motor add a encoder and be ahead of the game on price? Do servos not have the holding torque? Seems like other than brushes that would be the cheaper way to go. Am I way off on this one? How many hours of run time could one get before he needed to change brushes? This will be a commerical machine in a small job shop or school envoirment.

Donny

[Al_The_Man]

Flux vector use feedback electronically, the drive takes an electronic picture of the drive, usually be running through self tuning and this enables the processor to calculate where the rotor is at any time, the problem is that when the rpm drops, it has a harder time of controlling the motor. The alternative is to fit an encoder so the processor know where the rotor is at any given time.
Although I have not used one for C axis, I have used Mitsubishi for mill applications for rigid tapping etc, but there the Z axis syncs to the spindle encoder.
It is a slightly different application than C axis for a lathe, but Mitsubishi claim max torque down to zero rpm.
What is done in some commercial lathe applications, is a servo is used in tandem with the main spindle motor to position when used as a C axis.
Within your motor size, it might pay to get one of the larger DC servo's to use as a spindle and then only use one encoder on the final drive shaft, this you will need for CSF as well as threading, in any case.
Al.

[vladdy]

Not sure how fancy you want, but a PWM Dc treadmill motor of around 2 hp are available fairly inexpensively.. I recall on the Gemini site that qty 5 was just over $100 each with controller.. should give approximately same or slightly better performance than your 3/4 hp ac motor that you have now..
An add-on tach circuit [less than $10] would give output to Mach3 easily, these controls use pot to vary speed, not sure how easy that would be to integrate, never looked at that part of mach3..
just a 'cheapy' idea...
enjoy.

[whiteriver]

The VFD idea is nice but out of my price range. Now i'm looking at the DC side of things.
I believe you can take a dcpm motor like a treadmill for instance . Say 180Vdc. and add a encoder and drive it with a rutex drive and you have a servo. Am I thinking right??
Now one needs to come up with a 180 Vdc power supply. Why cant you take a cheap dc speed control that works with 180vdc motors and use its output as the power for the rutex drive? Since the rutex is a relay does it matter where the 180 vdc comes from? Or can you build a 180 vdc powersupply for less than $70? Would you need to put some large caps between the dc speed control(powersupply) and the dc motor with encoder(servo) to get a nice smooth constant voltage? If so what size?

Donny

[vladdy]

I had picked up a couple of treadmill motors and controllers at the local scrap yard, one was rated 90vdc and the other was 130vdc...both were labeled as 2hp units , but I believe this may be just a bit optomistic...
When I looked at the specs for the controllers both seemed useable up to maybe 10 amp output, after that they both recommended forced air cooling and / or additional heat sinks, neither of the treadmills had these installed ..
Both of the controllers were rather simple, just had adjustments for torque compensation, and adjustment for high speed limit, i assume that was just a voltage limit as neither had feedback visible, just the pair of power wires [and ground]
The larger of the two will wind up [hopefully] on my old south bend lathe, it' more than sufficient..

enjoy..