Re: [CAD_CAM_EDM_DRO] Re: My stepper getting hotter

jmkasunich wrote:

> --- In CAD_CAM_EDM_DRO@y..., "mariss92705" <mariss92705@y...> wrote:
>
> > Step motors get hot for two reasons, copper losses and iron
> > losses. Copper losses are due to current passing thru a
> > resistance, (I^2times R) and are independent of power supply
> > voltage in switch-mode drive.
>
> For example, step frequency current causes losses when moving.
> The current is large, but frequency is small compared to the
> PWM carrier frequency. Carrier frequency ripple current causes
> eddy losses all the time. Are carrier frequency losses usually
> greater or less than step frequency losses?
>
> How much carrier frequency ripple is there? On our VFDs, the
> carrier is 4KHz, and the fundamental (corresponds to step
> frequency for steppers) is 60Hz. The carrier frequency
> ripple current is usually less than 10% of the 60Hz current.

Ahh, you're getting to the heart of the matter! If the current ripple is
very small, then the voltage is going to have insignificant effect when
the motor is not moving. This requires a high switching rate, however,
so may cause more heating of the transistors. One little data point
I have, though, is that the Vexta 5-phase motors with the Super Vexta
driver, allows the motor to remain stone cold, even with the idle
current reduction turned off. The motors rapidly heat up when moving
at high speed. The switching frequency is inaudible, but I haven't
measured it. I would think any decent drive should keep the current
ripple to no more than 10% for the range of motors it is designed
for.

For microstepping drives, it may be quite common for the current
ripple to be held lower than in full or half step drives.

>
> Does a similar statement hold true for steppers on Gecko power?
> Or is the ripple lower, due to much higher carrier frequency?
> BTW, what is the Gecko carrier frequency?
>
> At rest, heating is due to only to IsqR and carrier frequency
> eddy current. Which dominates, or are they somewhat balanced?
> I^R depends on the idle current, eddy current depends on DC supply
> voltage.

I suspect generalization is hard, as small motors will have a LOT more
inductance than large ones. At fixed frequency, that would make the delta-I
much less on the small, or higher voltage motors.

> (And carrier frequency, which I assume is fixed.)

Not all drives are fixed freq. Many use simple comparators with hysteresis,
so the frequency changes for different motors, and at different supply
voltages. Mariss' drives run at 25 KHz, I think.

> At speed, I would expect carrier frequency eddy losses to
> remain constant. IsqR losses would drop once the current
> begins to drop (voltage limited), and IsqR heating would
> also be distributed over all the windings. (Skin effect
> would increase R - does skin effect show up in steppers?)
> On the other hand, step frequency eddy losses will keep
> rising with speed. Do they eventually become the dominant
> heat source, or does the motor run out of torque before
> you reach that speed?

Given enough voltage, and strong enough insulation, you can
absolutely burn up a stepper at high speed, where the voltage
alone won't cause trouble! The iron losses increase
until it smokes. This is very sensitive to the motor
design, so some will be easy to smoke in what would normally
be considered a reasonable application, while others will be
pretty tolerant of high speed operation.

I hope Mariss won't poke holes in all my above statements!

Jon