Z axis servo gets agonisingly hot

[Stewey]

HI Folks,

I've had interesting issues with our router, stemming from a z axis servo getting incredibly hot, then a burned out drive. It was a Pittman 14203 24 voltDC, and 27 oz-in torque.

I dealt with the drive, and got a new servo, a Pittman 14206, 24Vdc, 37 oz-in torque and fitted that. It lifts the router via a timing belt drive with a 1:1.6 reduction, connected to an acme screw. IN the end it takes about 3.15 revolutions of the servo per 1mm of Z axis travel. I don't feel it's overloaded or underpowered at all.

I ran a file yesterday, and again today- cutting a lithophane into 3mm perspex sheet with a 1/16" ballnose endmill and 66% stepover. Again, a tiny load, moving at 80mm/sec X & Y speed when it can, and 30mm/sec Z/plunge speed.

Even with the new servo, it got incredibly (or agonisingly) hot running this 3D file which took over an hour. After a while, I start to get position errors, due to the heat, I guess & it grooved the workpiece where you don't want it to briefly.

Does anyone have ideas on possibly why it would be getting so hot so quickly, when it's only zipping up & down a couple of mm inside a 3mm perspex sheet? It seems it's applying a high braking force, perhaps, and unnecessarily. I'm stuck for ideas on why, and how to cure it.

I'm wondering if altering the PID settings could have an affect?

I don't think it's stiff bearings, as I can raise the z axis by hand readily enough, turning the acme screw. The weight of the 3 hp Perske spindle & head & stuff might be 5 or 6 kg.

Any ideas would be appreciated, thanks!

[ger21]

What voltage are you running the servo at?

[Al_The_Man]

Generally with a servo it is either the back feeding effect of a ball screw or the friction problem with a acme screw, also if a low friction acme screw is fitted it could also be back feed effect.
You need to check the current flow at the end of a position, i.e. when the servo should be at rest.
Does it get warm or hot even when no motion is commanded?
This will be a clue.
Also when moving the Z in both directions, the difference in current should not be that large, if it is, it is a sign that counterbalance is required.
Al.

[Stewey]

Thanks for the replies.
The supply voltage is rectified 30Vac to the drive, which is about 42 v DC, but the drive doesn't supply that full voltage unless it's driving the servo at fullspeed, which is isn't. When I tested it at a slow speed, it was sending out about 4 to 12 V DC.

Essentially it's a more powerful servo of the type that was supplied with the machine's retrofit 9 years ago.

Al, there's a nylon type of nut, which runs in the acme screw, enabling rise & fall. It depends on gravity to keep the downpressure on. A counterbalance might be OK, but we'd need to fix the nylon nut in place, so it's forced downwards then. At the moment, there'a a bit of a safety factor in the fact that it can be lifted if it has to be, and gravity keeps it in place.

I'll recheck it tomorrow, but I am certain it's fine at rest, and it's fine doing 2D cut-out files all day long, which have negligible z axis effort except to go to a certain level & stay there till a perimerer has been cut. It's just the constant variations in the Z axis positioning in 3D files that seem to heat it up- even negligible/no-load files.

Thanks,
Ian

[Runner4404spd]

I don't want to hijack this thread but my mill is using 3/4 inch ball screws on the z axis. the motor seems to get really hot just trying to hold position. i'm running a g320.

my mach settings are

x acceleration 9 top speed 120
y acceleration 9 top speed 120
z acceleration 7 top speed 100

on my geckos i have the limits on x and y set to 75% and the z limit is set to 100%.

my power supply puts out 72 VDC so that shouldn;t overload the geckos.

motor specs
Continuous Stall Torque Tcs 0.847 (7.50) Nm (lb-in)
Peak Stall Torque Tps 4.237 37.50 Nm (lb-in)
No Load Max. Speed w nl 6000 RPM
Continuous Rated Torque Tcr 0.585 (5.18) Nm (lb-in)
Peak Rated Torque Tpr 1.172 (10.37) Nm (lb-in)
Rated Speed w rated 4000 RPM
Rated Power W (HP) 245 (0.33) Watts (HP)
Inductance L-L L 4.35 mH
Resistance L-L R 1.33 ohms
Max. Terminal Voltage V dc 90 Vdc

[Al_The_Man]

Originally Posted by Runner4404spd I don't want to hijack this thread but my mill is using 3/4 inch ball screws on the z axis. the motor seems to get really hot just trying to hold position. i'm running a g320.

As mentioned, this is generally a sign that backfeeding is occurring and motor current is necessary to hold position.
Ideally a servo should pass zero current at rest.
Al.

[Runner4404spd]

ok, so if backfeeding is a problem does this mean the gib is too loose? i've got dovetail ways on my mill. initially i though the gibbs were too tight, maybe i need to tighten them up since they are too loose now. i should mention that the program when it was running was not cutting anything but air. so there really shouldn't have been any cutting forces acting on the head. although maybe it was trying to drop so the motor had to keep pulling it up?

[Al_The_Man]

Tightening the Z axis ways should not be used to correct backfeeding.
If the servo is passing a high current to keep the Z stationary and/or to raise it, then some kind of counter balance is required.
Al.

[Stewey]

It's been a few days since I've been able to get back to the router. (Did some rewiring of power points etc in the shed, and other priorities, instead.)

Results of observations:

I ran a 2d file- routing letters 1/2" deep in hardwood timber- about 3 hours solid continuous routing- and the Z axis servo was a comfortable lukewarm temp. but I didn't set it to travel any faster than 15mm/sec. (abt 38 ipm)

I then ran a 3D file in the same timber- cutting a section almost 2" deep with a 1/4" ballnose cutter- but again, I set the Z axis speed at 10 mm/sec- and it was fine, for the 2 hours it took- no uncomfortable heat.

I should try a 1/8" deep lithophane again- that was the last file I had in which it got really hot- but I had the Z axis speed set for 25 or 30 mm/sec (roughly 67 ipm) and it was horribly hot.

It might seem that it can't handle 'high speeds' above say 20mm/sec (45 ipm) without excessive heating.

It's not getting hot at rest.

I'll try a couple of files, with varying Z speds, but I'm wondering if the 'fault' is likely to be in the servo (under powered- I don't think so) , then, or the PID tuning, or the driver settings- any ideas appreciated?

Thanks, Al for your input so far. I'm not sure about a counter balance, but maybe a spring...

[Stewey]

P.S. I'm not sure I can read the servo current when in action...I could detour one brush lead via a multimeter, as long as it's under 10 amps. I don't have a clamp meter.

[Al_The_Man]

If the drive does not have a current monitoring output I would monitor any current in the +vdc to the drive.
A DC motor takes alot more time than its DCBL equivalent to indicate overcurrent by overheat symptom due to all the heat being in the armature which has to radiate through to the outer stator.
Often by the time the case has heated the armature has exceeded safe temperatures.
Al.

[Stewey]

It's been a while, but I've tried a few options that I could try- but not read the current while in action (yet).

The servo has a happy max speed of around 3000 rpm- and at 3.2 revs per 1 mm Z travel, equals almost 16mm/sec top speed.

Whether I rout at 15mm/sec for the Z speed, or 7mm/sec, it still gets awfully hot at both speeds after about 1/2 hr of routing a lithophane, which is varying in depth by 3mm, in acrylic/perspex, with a 1/16"cutter- very minimal load.

If I rout timber, in a 3D file, say 2" thick, with a far greater load on the router, it stays lukewarm. It's as if the minute Z height variations are giving it problems.

It doesn't get hot when at rest between files- even after several hours standing there. It does act as if the brakes are on if you try to move it. At times, it 'grumbles' a bit while supposedly stationary.