In Search of A Better G0704 Spindle

[cjtoombs]

That torque is more important than horsepower is a common misconception. If you gear a 1/4 hp motor down to 1 RPM, you will have about 1300 lb*ft of torque, but you won't be able to do any more work with that than that same motor turning 5000 rpm with about 1/4 lb*ft of torque (ie. neglecting gearing inefficiencies, your metal removal rate would be the same if the full 1/4 hp were being used). That's because horsepower is the measure of the amount of work that can be done. Torque kind of equates to voltage in electricity, it is the amount of force with which you can rotate a shaft like voltage is the amount of pressure exerted to move current. Without the RPMs for shaft rotation or the amperage for electricity, neither torque or voltage can produce work. The linear torque relationship of the VFD is what translates to the linear Hp curve, since the relationship is linear. The loss in torque does become a problem for VFDs, as large cutters run at low motor speeds tend to stall in the work at anything approaching reasonable cut depths and feed rates (I have way more experience with this and I would like). The stalling is a function of torque, since the movement stops. That's generally why it's best to err on the side of a large motor, or better yet size the motor for the minimum horsepower you want to cut at. If you need to slow a nominally 1800 RPM motor down to 900 RPM and you want 1 hp at the shaft at all times, you need a 2hp motor.

 

[spumco]

Which is why if someone is shopping for a motor, they'd be well served to look for the highest CT ratio (constant torque) possible.

3P, 1.5hp motors:
IronHorse general purpose - 4:1 CT, ($175) nameplate torque available from 1750 to 437rpm without overheating
Marathon BlackMax- 1000:1 CT, ($454) 1750 to 17.5rpm without overheating....

 

[cjtoombs]

Yea, that's the other problem with running motors at low speed, the fans don't work and they overheat. I think some of the motors specifically designed for this have fans driven by a separate motor. Not a problem for fanless motors that rely on sheer mass for cooling, but most of those are antiques.

 

[mksj]

TENV and TEBC motors (typical inverter rated motors) are designed to be used down to ~0 RPM without overheating (they run cooler at low speeds), the primary problem with the more traditional TEFC motor is that they tend to overheat at low or very high RPM. Most conventional rated 3 phase 1750 RPM motors when driven by a variable speed drive are usually limited to an operating range of around 15-90 Hz, but one still has the limitations that the Hp will be 1/4 of the rated nameplate at 15 Hz. The constant torque range is ~4:1 for standard 1750 RPM motors, inverter motors are for all intensive purposes maintain full torque down to 0 RPM (CT is usually > 1000:1 ratio). To some degree VFDs can compensate for short periods to increase the rated current/output, and inverter motor can be pushed a bit further, around 150-180% for 1 minute. Conventional motors would be something in the 120% range. I use an inverter 2 Hp motor on my lathe set to 150% overload setting, and it will still trip the VFD running at 60Hz in a deep continuous cut with large stock. The other factor that is often overlooked is the gearing/belt ratio and the applied power to the spindle. If you run an inverter motor at 120 Hz on a 2:1 ratio to the spindle speed vs. a conventional motor run at 60 Hz with a 1:1 ratio, the inverter motor will be delivering twice the spindle Hp and the spindle torque will be about the same. On a mill of this size, one would most likely not be able to hang a 2Hp 3 phase inverter motor on the head of this mill, nor do I think it is practical. If you look at small CNC mills, they all overdrive their motor usually to something like 6000 RPM. So my recommendation in this setting for this size mill would be a 1 or 1.5 Hp inverter motor with a sensorless vector VFD, or a BLDC motor with controller which would offer more Hp in a more compact package. A number of individuals have done CNC conversion of larger mill like the PM932 and PM940 and the 2 Hp Marathon Black Max motors have been more than adequate for their purposes. They have used a 1:1 drive ratio if I recall. I use a Baldor IDNM vector motor on my lathe, it is absolutely quiet and smooth at speed.

The discussion of using an encoder for spindle feedback is appropriate for full blown CNC machines, but would be an expensive and unnecessary adventure in this level of CNC conversion. The main reason for encoder feedback, is for precise spindle positioning, maintaining a 0 Hz position or very low speed control, and absolute speed accuracy which exceeds 0.01%. The accuracy of a sensorless vector VFD/Vector or BLDC motor is 0.1%, usually withing +/- 1 RPM even under varying load. Both my current lathe and mill which use sensorless vector control maintain this level of speed control. There are also other variants of feedback control, w/wo an encoder. Many VFDs do not have the ability to use encoder speed control feedback, those that do are usually 2-3X more expensive. The VFD frequency range on my lathe is 20-125Hz using a TENV inverter motor, the mill is 20-200 Hz and uses a factory TEBC inverter motor. The other possibility as other have mentioned is to use an integrated motor/controller system, some possibilities are below. The major limitation I see is the Kw motor rating and the limited top speed ability of anything over 1 hp (0.75 kW), as mentioned previously you could go with a DMM 2.5 Hp drive and use a 1:2 drive ratio which is a very nice combination.
https://www.teknic.com/products/clearpath-brushless-dc-servo-motors/
http://www.dmm-tech.com/Dyn4_main.html

Baldor 1 Hp inverter motor $200-300. A decent VFD will be about $250.
http://www.ebay.com/itm/Baldor-Inverter-Drive-Motor-Inventory-Clearance-/182739811235
http://www.ebay.com/itm/BALDOR-1-HP-3-PH-INVERTER-DRIVE-MOTOR-NEW-/191731639588

If one looks at the application in practice, one needs to look at the purpose of the machine. In general, a CNC mill conversion is looking at higher speeds, and the need for low RPM is not a priority, nor would you expect to use it to be used to bore large holes with a large drill bit. If you need precise control to do say tapping, then you are looking at a completely different scenario/cost factor. I do not see the practicality of trying to do a CNC conversion and also expecting to drill holes at 100 RPM. The flip side is that you need speed, so you are looking as something like a usable speed range of 600-6000 RPM, maybe a bit lower with a 2 speed head. Standard gear head nor the bearings will live very long at these speeds, so you are looking at a belt drive.

 

[MontanaAardvark]

TENV and TEBC motors (typical inverter rated motors) are designed to be used down to ~0 RPM without overheating (they run cooler at low speeds), the primary problem with the more traditional TEFC motor is that they tend to overheat at low or very high RPM. Most conventional rated 3 phase 1750 RPM motors when driven by a variable speed drive are usually limited to an operating range of around 15-90 Hz, but one still has the limitations that the Hp will be 1/4 of the rated nameplate at 15 Hz. The constant torque range is ~4:1 for standard 1750 RPM motors, inverter motors are for all intensive purposes maintain full torque down to 0 RPM (CT is usually > 1000:1 ratio). To some degree VFDs can compensate for short periods to increase the rated current/output, and inverter motor can be pushed a bit further, around 150-180% for 1 minute. Conventional motors would be something in the 120% range. I use an inverter 2 Hp motor on my lathe set to 150% overload setting, and it will still trip the VFD running at 60Hz in a deep continuous cut with large stock. The other factor that is often overlooked is the gearing/belt ratio and the applied power to the spindle. If you run an inverter motor at 120 Hz on a 2:1 ratio to the spindle speed vs. a conventional motor run at 60 Hz with a 1:1 ratio, the inverter motor will be delivering twice the spindle Hp and the spindle torque will be about the same. On a mill of this size, one would most likely not be able to hang a 2Hp 3 phase inverter motor on the head of this mill, nor do I think it is practical. If you look at small CNC mills, they all overdrive their motor usually to something like 6000 RPM. So my recommendation in this setting for this size mill would be a 1 or 1.5 Hp inverter motor with a sensorless vector VFD, or a BLDC motor with controller which would offer more Hp in a more compact package. A number of individuals have done CNC conversion of larger mill like the PM932 and PM940 and the 2 Hp Marathon Black Max motors have been more than adequate for their purposes. They have used a 1:1 drive ratio if I recall. I use a Baldor IDNM vector motor on my lathe, it is absolutely quiet and smooth at speed.

The discussion of using an encoder for spindle feedback is appropriate for full blown CNC machines, but would be an expensive and unnecessary adventure in this level of CNC conversion. The main reason for encoder feedback, is for precise spindle positioning, maintaining a 0 Hz position or very low speed control, and absolute speed accuracy which exceeds 0.01%. The accuracy of a sensorless vector VFD/Vector or BLDC motor is 0.1%, usually withing +/- 1 RPM even under varying load. Both my current lathe and mill which use sensorless vector control maintain this level of speed control. There are also other variants of feedback control, w/wo an encoder. Many VFDs do not have the ability to use encoder speed control feedback, those that do are usually 2-3X more expensive. The VFD frequency range on my lathe is 20-125Hz using a TENV inverter motor, the mill is 20-200 Hz and uses a factory TEBC inverter motor. The other possibility as other have mentioned is to use an integrated motor/controller system, some possibilities are below. The major limitation I see is the Kw motor rating and the limited top speed ability of anything over 1 hp (0.75 kW), as mentioned previously you could go with a DMM 2.5 Hp drive and use a 1:2 drive ratio which is a very nice combination.
https://www.teknic.com/products/clearpath-brushless-dc-servo-motors/
http://www.dmm-tech.com/Dyn4_main.html

Baldor 1 Hp inverter motor $200-300. A decent VFD will be about $250.
http://www.ebay.com/itm/Baldor-Inverter-Drive-Motor-Inventory-Clearance-/182739811235
http://www.ebay.com/itm/BALDOR-1-HP-3-PH-INVERTER-DRIVE-MOTOR-NEW-/191731639588

If one looks at the application in practice, one needs to look at the purpose of the machine. In general, a CNC mill conversion is looking at higher speeds, and the need for low RPM is not a priority, nor would you expect to use it to be used to bore large holes with a large drill bit. If you need precise control to do say tapping, then you are looking at a completely different scenario/cost factor. I do not see the practicality of trying to do a CNC conversion and also expecting to drill holes at 100 RPM. The flip side is that you need speed, so you are looking as something like a usable speed range of 600-6000 RPM, maybe a bit lower with a 2 speed head. Standard gear head nor the bearings will live very long at these speeds, so you are looking at a belt drive.

Thanks for this. Lots to think about.

You talk a lot about a machine this size and that's welcome. I know that G0704 is a relatively small machine, while bigger than the more common SIEG X2-size mills. It's not something like a Haas or Tormach, and one of the things I think about is that just because I have CNC control of four axes doesn't fundamentally change the machine. I don't think adding ballscrews turns a $1400 G0704 into a $10,000 Tormach or $40,000 Haas.

Ultimately, it all comes down to "one needs to look at the purpose of the machine" as you say. I'm not likely to be doing large pieces, exotic geometries or exotic materials. If I get one to do, I'll muddle through as best as I can. I'm not thinking I need to do tapping under CNC in the spindle, although I could see where that could be a really cool capability. To be honest, the machine is pretty good as it is, and it's a good compromise. I'm not in business, I'm a hobbyist. I think that means I'm sort of like the small job shops that might work on custom pieces for different companies, a few of several different parts every week.

I noticed that the closest equivalent-sized Tormach models use 5000 and 10,000 RPM spindles, of about the same 1 HP size. That's what made me think I should upgrade the motor to run 5000 RPMs. 10k would be better for engraving, but I don't do much of that.

All I'm looking for is:
About 1HP. I'm not going to throw away a clearly better 1-1/4 or 1-1/2 HP motor, but the 1 HP range is around what this was designed for and it makes sense to not get too far from there. As a bonus, CNCCookbook's G Wizard speeds and feeds program can run free for life if you limit it to 1 HP.
I'd like to get to 5000 RPM, or close.
I'd like to turn the motor on and off with the SW (Mach3) and set the spindle RPMs from inside the G-Code, too.
It would be very, very convenient to run the motor on 240V. I have high current service into that corner of my shop, more than a motor this size would want. (I think it's 220 with a 50A breaker). Right now, I'm running my computer, my CNC Control box (motor drivers with a 50V, 10A supply), the spindle, and some lights (LED) all on a 120V 15A circuit. I have this deep fear that if everything should hit its max current at the worst possible moment, that the breaker may go.

That's it. Positional accuracy enough to turn a tap and back it out is interesting, but not a strong requirement.

You said: "So my recommendation in this setting for this size mill would be a 1 or 1.5 Hp inverter motor with a sensorless vector VFD, or a BLDC motor with controller which would offer more Hp in a more compact package." I think my emphasis will be the BLDC motor.

 

[spumco]

For the price of a BLDC and driver from a known source, you can get the DMM servo package in 1-1.5hp.

On a mill of this size, one would most likely not be able to hang a 2Hp 3 phase inverter motor on the head of this mill, nor do I think it is practical.

Very, very true. Mine's in the almost-PM932 size, and my 2hp BlackMax is - frankly - too big. Mine is the cast iron frame motor, and the sheetmetal version of the same motor would probably be a bit more reasonable.

There are a couple 1hp sheet metal BlackMax motors on ebay for about $200 shipped. Sheet metal frame would be smaller/lighter than the Baldor above (or other cast-iron motor). Size 56 frame would probably work OK for the G0704 head.

Automation Direct GS3 sensorless vector VFD for $242. And an encoder input card is only another $50. Very nice drive with instructions in English.

 

[MontanaAardvark]

Automation Direct GS3 sensorless vector VFD for $242. And an encoder input card is only another $50. Very nice drive with instructions in English.

That's this one, right? The VFD, without a motor?

I was just looking at Automation Technology because that's where I got my Steppers and Stepper controllers for the CNC conversion. A stepper is a BLDC motor as well, and my only issue is converting stepper motor specs into HP. These NEMA42 steppers are spec'ed at 4230 in-oz. In a stepper, the torque is ordinarily maximum down to zero RPM (you'll see it called holding torque) and and stays constant as RPMs go up. After some RPM threshold torque goes down as the RPMs go up.

I found a site that tells me 4230 in-oz = 29.9 Watt-seconds, and I think to convert to total watts, you divide by the number of seconds in a revolution, then divide watts by 750 to find the number of HP. That says the HP depends on the RPM that motor is running, but like I said, steppers tend to have constant torque vs RPM over much of their range. I don't know if it's done at the RPM value where the torque starts dropping or if everyone agrees to solve it at some number of RPMs. I can turn 29.9 W*s into just about any number for HP depending on how I set up the problem.

I'll send an email to AT to see if they have helpful numbers.

 

[spumco]

That's this one, right? The VFD, without a motor?

Yep, that's the bunny.

A stepper is a BLDC motor as well,

I'm not a motor expert, but for your and my (layman's) purposes - no. [I know there are some really bright electrical people on this forum. If you can explain this more simply, or correct anything - please do so.]

A BLDC motor has a feedback signal to the drive/amplifier. It's usually a 3-position (120 degrees apart) Hall effect trigger. Without brushes and slip ring, the only way to commutate the rotor poles is with a feedback devise, such as an encoder (expensive) or a Hall-effect trigger (cheaper). The drive receives the position/rpm signal back from the motor and can adjust the commutation timing and amperage to achieve & maintain commanded RPM.

A stepper does not have a separate feedback device to permit the drive to adjust its output to match reality (ignore 'closed-loop' stepper products for the moment). There are a bunch of individual winding pairs inside the housing, and the rotor has a bunch of 'teeth' around the circumference. Opposite pairs of poles/windings are energized sequentially, and because the rotor 'tooth' is drawn to the energized pole, the rotor turns to the newly-energized winding. If you fiddle with the voltage/amps so that one winding has a bit of juice, and the next one has a bit more, the rotor tooth is pulled partway to the second one. Result - microstepping. The big difference is that the rotor teeth are, simplified for my small brain, what is commutating the stepper motor and not a feedback device.

Steppers lose torque as RPM goes up. Gecko drive systems has an excellent series of articles on how steppers work, but essentially it takes time for the windings to energize, and it takes time for the rotor teeth to react and move to the magnetic field, and at some RPM it simply can't move fast enough to keep up with each sequence of winding pulses. RPM maxes out, and there's no torque available. Figure about 1000RPM is the upper limit for run-of-the-mill steppers and drives before they can't do any useful work beyond just turning themselves or an unloaded axis ball-screw.

BLDC motors don't have 'teeth' like steppers, so they're more like AC servos than steppers in that they have a few rare earth magnets on the rotor pose and a series of windings. The difference you and I need to know about BLDC and AC servos is that BLDC, while having feedback, are 'dumb' and can't position. The 120 degree rotor sensor pulses are simply too far apart for positioning, and the drives aren't designed for anything other than RPM control. AC servos have encoders with many more pulses per rev (thousands in some cases) so they can be positioned.

Here's the other part - grown-up servo/drive combinations can also be programmed to run in RPM mode and not just step/direction. You command 1000rpm, and the drive spins at 1000rpm, having done the math to know how many feedback pulses it should receive per time unit. You don't have to keep sending it step and direction signals - it just runs like a 'normal' motor, but with the ability to sense & increase current if the RPM dips due to load. When you're ready to position the spindle (alignment, tapping, whatever) the computer puts the servo drive in position mode and then issues a series of step/direction moves. Spin end mill, stop, rotate to index the tool holder drive dogs, hold position rigidly, tool change, back to RPM mode and making chips. This cuts down on the number of signals needing to be sent to the motor, especially at higher RPM (high frequency of step pulses).

Pretty cool.

So Steppers can be used as spindle motors, but that would be a bad idea due to the rapidly dropping torque and comparatively low top RPM limit. A BLDC motor is way better, but it's crippled when compared to AC servos. If reliable BLDC motors & drives (think Tormach 440) were available at 1/10 the cost of AC servos it'd be a no-brainer, but with the cost of decent servos being so low, and much more reliable than 'dodgy import' BLDC drives, the only advantage BLDC motors have is that they generally have higher RPM limits.

You want 10kRPM? BLDC, cause a 10kRPM servo is going to be outrageously expensive. You can live with 5kRPM and a pulley system to get you 10k? AC servo all the way.

-S

 

[markba633csi]

Isn't a BLDC motor similar to a stepper without the "teeth"? Don't they both have permanent magnets?
Mark S.

 

[MontanaAardvark]

Isn't a BLDC motor similar to a stepper without the "teeth"? Don't they both have permanent magnets?
Mark S.

Yup. If you look up stepper motor on Wikipedia (for example), it says:

A stepper motor or step motor or stepping motor is a brushless DC electric motor that divides a full rotation into a number of equal steps.
​

It's one of cases where this specialty has adopted a specific meaning for BLDC.