Discussion on Small CNC Mill Spindle and Axis Motor Performance

[macardoso]

In spindle motor news... I've run in to a bummer with those Siemens servos.

I'm unable to get them to spin faster than about 4200rpm. Looks like I didn't do enough homework on the datasheet and my bus voltage (~340vac) just isn't enough to get them to rated speed. That, or I'm completely lost on the drive tuning and motor parameters.

Unless someone has some suggestions on drive tuning, looks like I may be back to the BLDC plan.

If you want, shoot me a PM and lets chat. It might depend on the motor rated voltage and the DC bus voltage you have available. The the drive's DC bus voltage is too low, there will come a speed at which the drive output voltage minus the motor back EMF is insufficient to saturate the windings of the motor (based on winding inductance) before the rotor has passed and the next winding must start its commutation cycle.

The motor will perform with full torque up to around this critical speed and then performance quickly falls off. I can run Allen Bradley 400V class servos on a 200V drive system and get about 60% of the top speed of the motor. Likewise I can run a 200V system on 120VAC input to the drive and again get about 60% of the top speed of the motor.

This is somewhat the same phenomena that is seen in stepper motors and why running a stepper drive at a higher DC input voltage yields higher motor top speeds.

 

[macardoso]

Not according to the datasheet. Ke is 50v/Krpm, so it should go faster. But it doesn't.

I'm not an expert by any stretch, thus I'm hoping its a drive setting I can fiddle with.

Curves from the datasheet. "Vmot" is defined in the obscure "general data" brochure as the bus voltage:

View attachment 364781

OK, so those motors seems to be rated for a 380V system which would be common overseas. Given that information, I'd expect you to be able to run them at 63-75% of their rated speed on the "b" curve, which would give around 5000-6000 rpm, still very respectable but sadly not the 9100 rpm you were hoping for. This was all assuming you were powering that Copley with 240VAC 1P (US residential) giving a DC bus of 325VDC.

 

[spumco]

OK, so those motors seems to be rated for a 380V system

What did I miss on the torque curve sheet that indicates they're designed for 380VAC line voltage? I didn't think I'd get 9kRPM out of them, but they simply die at about 4200rpm. 4000rpm sounds great, very responsive, etc. Bump it up a hair and they saturate and won't go any faster.

And yes, you've got it right - I'm trying to use the Copley at 240vac. According to the drive, bus voltage is 330-340vdc (depending on supply variations).

Datasheet for your review (these are the -6AK7). Maybe you can help me identify what I should have been looking for.

Back to the BLDC, or keep searching for a 200v unicorn motor. Good news is that I can return them to HGR (one has dodgy bearings anyway).

 

[macardoso]

I was basing them on this info in your graph.



Reference speeds [a] through [c]. DC Link = DC Bus which is ~ 1.35*ACrms Line Voltage. So 240VACrms = 325VDC.

In the Siemens motors, 540VDC/1.35 = 400VACrms and 600VDC/1.35 = 444VACrms.

Also the motor voltages (Vmot) are given in a range of 340Vrms - 425Vrms. This is roughly centered around 380Vrms

Bit of rough math, but this puts you in the 380VAC market (Asia) or the 400VAC market (Germany????). These motors would also run fine on a 480VAC (650VDC) system common in the USA. Either way, you're a bit under that voltage and I would expect the behavior you are seeing.

I got to your speed by taking the ratio of real to rated DC bus voltages: 325VDC/540VDC = 0.6. Then multiplying that by the limiting speed from reference line [a] (7000rpm) to get 4213 rpm.

I guess redoing the math with different assumptions got me even closer to the speed you are seeing - cool.

Remember this is the max speed with zero torque available, so any applied torque will pull you back from this maximum.


EDIT: My initial analysis was based on reference curve, this analysis was based on reference curve [a]

 

[macardoso]

As you commented, high speed motors in a 200V class flavor are rare. It is difficult for motor manufacturers to design motors with sufficiently high inductance to generate the magnetic field strength at the stator/rotor interface to create optimal torque while simultaneously keeping the inductance low enough that the windings can be driven to the desired current levels with minimum applied voltage. Higher motor speeds require faster changes in the winding currents.

Higher inductance resists change in current and higher driving voltage is required to change the current more quickly. This is the reason you see higher speeds in higher voltage motors and why small motors are rated for high speed while big motors must spin slowly (big motors = large windings = more inductance).

Conversely, low inductance motors exist and can run quickly, but lack the torque of slower motors as the lower inductance results in lower magnetic field strength and lower torque.

Kind of cool. You always play this tradeoff game between torque and speed. The only thing that can improve both factors is a higher driving voltage.

The best motors I've seen for the application are the Fanuc asynchronous spindle motors (seen on RoboDrills and other small high speed machines).

 

[spumco]

Thanks for the analysis - looks like I misunderstood the curves. Siemens defines Vmot (in another document) as bus voltage, so I mentally glossed over the "540VDC" and focused on the "Vmot = 340V" bit. My bad.

And yes, I get the need for higher voltage when you want more speed - that's why I'm running a 100vdc PS and fairly low inductance steppers.

Anyone have a Fanuc A06B-0852 laying around they want to donate?

 

[macardoso]

Thanks for the analysis - looks like I misunderstood the curves. Siemens defines Vmot (in another document) as bus voltage, so I mentally glossed over the "540VDC" and focused on the "Vmot = 340V" bit. My bad.

And yes, I get the need for higher voltage when you want more speed - that's why I'm running a 100vdc PS and fairly low inductance steppers.

Anyone have a Fanuc A06B-0852 laying around they want to donate?

I take Vmot to be Volts AC/PWM (RMS).

If they do, I'm gonna grab that sucker before you can

 

[spumco]

HGR has about a hundred A06B-0235's right now. I'm wondering if I can over-speed it a bit and get away with a 1:1.5 belt.

 

[Yukonjack]

I have just started the rebuild of my 12 yr old G0704 CNC conversion spindle for the umteenth time. I’m almost ready to pull the trigger on one of these spindles. https://www.cncdepot.net/product-page/fm-series-spindles. They really nice and about 2/3rds the weight of the 3ph/belt drive G0704 head conversion, with a PDB. These use air ISO BT30 tool holders, 12k rpms @ 3hp. they just fly through metal. Check outClough42 youtube channel on his G0704 CNC conversion.

I have probably spend this much over the life on my G0704 project.
Just thinking.
CH

Anyone else try this? Lowest practical RPM? Drilling performance in steel?
Is there any ATC spindle that is capable of real milling in steel (sub $5,000)?

 

[spumco]

Is there any ATC spindle that is capable of real milling in steel (sub $5,000)?

Assuming you mean a unitized cartridge/motor spindle, and 'real' milling in steel = 1/2" end mill or larger - then no. Assuming they can be safely run at 50% rpm, you aren't going to use a 1/2" drill in steel at 6kRPM. I'm sure it'd be fine to use a 1/4" carbide end mill to do hole interpolation, but you're SOL for reamers, big drills, ect.

If you want a package deal, you can get a BT30 spindle and a 2.2kW servo spindle with an 8 or 10k RPM top end from Automation Technologies or direct from China. With a 1:1 drive that should do just fine in steel.

The servo above may be a bit too large for the G0704. I think it's a 150mm frame, whereas the DMM and other similar sized are 130mm servos. Wider isn't really the problem... but the extra frame size may interfere with a PDB cylinder over the spindle.

I've been mulling over spindles & motors for 3 years now, and there's no magic answer. I'll stand by my statement I've made a few times/places:

The most flexible spindle arrangement I can imagine (for a hobby-sized mill) is the following:
- two-speed poly-v belt drive (preferably with auto belt engagement a-la Fadal pozi-drive)
- around 1.5:1 and 1:1.5 ratios
- 2 or 3hp PMSM or AC servo (to cut down on weight/size vs an inducton motor, and easy spindle homing/orientation)
- BT30 cartridge spindle, with 8k to 12k bearings
- Air or air over hydraulic PDB
- Direct spindle encoder (depends output/control system), and/or spindle index trigger signal