1.8 or 0.9 degree stepper?

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

ray_parkhurst
Posts: 3416
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

mawyatt wrote:
ray_parkhurst wrote:
mawyatt wrote: ...
This is an interesting result that the unloaded moving motor power is actually lower than the peak stationary motor power at certain angles. Of course this doesn't consider any other motor losses, non-ideal sinusoidal current, or friction, or load dynamics.

I'll try and run some tests when I get some "free" time to see if these motors behave this way, certainly an interesting possibility IMO.

Has anyone experience this result?
I'd look at it more in terms of energy than power. The actual movements don't take very long, so not much energy is dissipated in the windings, but the stationary power is nearly continous, so makes up the bulk of the heat generation.
Ray,

Power is the right metric to use since power is energy divided by time and heat (dissipated) is the parameter we are trying to manage. 100 Joules energy dissipated over 1000 seconds wouldn't make much of a temperature change in our motors, however 100 watts certainly would!

Thermal impedance is expressed as temperature change/power, or C/W. A motor has a certain thermal impedance to the ambient Ta, called Rt.

The motor temperature is Ta + Rt*Pm, where Pm is power dissipated in the motor.

So a motor with a Rt of 5 Degrees C /Watt will have a temperature rise of 50 degrees C when dissipating 10 watts, or about 75 degrees with Ta~25C.
Well, I agree that power is the right metric for thermal resistance, since thermal resistance is calculated with continuous dissipation. So if you're doing the calcs using the hold currents, that works. But for the peak operating currents, energy dissipated is the better metric, which can then be converted into an average power through a duty cycle calculation to determine rise temps.

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

ray_parkhurst wrote:
mawyatt wrote:
ray_parkhurst wrote:
mawyatt wrote: ...
This is an interesting result that the unloaded moving motor power is actually lower than the peak stationary motor power at certain angles. Of course this doesn't consider any other motor losses, non-ideal sinusoidal current, or friction, or load dynamics.

I'll try and run some tests when I get some "free" time to see if these motors behave this way, certainly an interesting possibility IMO.

Has anyone experience this result?
I'd look at it more in terms of energy than power. The actual movements don't take very long, so not much energy is dissipated in the windings, but the stationary power is nearly continous, so makes up the bulk of the heat generation.
Ray,

Power is the right metric to use since power is energy divided by time and heat (dissipated) is the parameter we are trying to manage. 100 Joules energy dissipated over 1000 seconds wouldn't make much of a temperature change in our motors, however 100 watts certainly would!

Thermal impedance is expressed as temperature change/power, or C/W. A motor has a certain thermal impedance to the ambient Ta, called Rt.

The motor temperature is Ta + Rt*Pm, where Pm is power dissipated in the motor.

So a motor with a Rt of 5 Degrees C /Watt will have a temperature rise of 50 degrees C when dissipating 10 watts, or about 75 degrees with Ta~25C.
Well, I agree that power is the right metric for thermal resistance, since thermal resistance is calculated with continuous dissipation. So if you're doing the calcs using the hold currents, that works. But for the peak operating currents, energy dissipated is the better metric, which can then be converted into an average power through a duty cycle calculation to determine rise temps.
From earlier, if Motor Power is:

Motor Power = R(I^2)*{1 + 2*Sin(x)*Cos(x)}

From Trig Sin(x)*Cos(x) = (Sin(2x))/2

Motor Power = R(I^2)*{1 + Sin(2x)}

If we consider that the only power dissipation element is the motor resistance (neglecting friction, load and such)

let x become a time variable wt, where w is radian shaft frequency (2*pi/rotation/sec) and t is time,

Motor Power (t) = R(I^2)*{1 + Sin(2wt)}

So the motor time varying power is an offset sinusoidal and ranges from R(I^2) to 2R(I^2) at twice the rate of the motor shaft rotation with an average of 1.5R(I^2). This assumes of course that the thermal motor time constant is much longer that the electrical motor time constant LR, which is usually the case.

Note that the Motor Power (t) minimum, maximum, and average are independent rate of shaft rotation w under the conditions described.

So at light loads and relatively slow speeds the motor potentially dissipates less average power than at the peak static motor power which occurs at odd multiples of 45 degrees. Interesting indeed!!!

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

ray_parkhurst
Posts: 3416
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

mawyatt wrote:
...

Note that the Motor Power (t) minimum, maximum, and average are independent rate of shaft rotation w under the conditions described.

So at light loads and relatively slow speeds the motor potentially dissipates less average power than at the peak static motor power which occurs at odd multiples of 45 degrees. Interesting indeed!!!

Best,
Mike,

Let me try again, as you seem to not be getting my point...

During SnS operation, the motors spend most of their time in a quasi-idle state. This state is whatever the final position demands from perspective of the microstepping voltage. In my SnS system, the XY motors spend perhaps 95% of the time in this quasi-idle state, while the Y motor spends >95% of its time in this quasi-idle state. The motors are only moving perhaps 5% of the time, so the dynamic power makes only a small difference to the overall heat generation, and indeed I'd say could almost be ignored. Only the quasi-idle/static dissipation really matters. However, if you do want to factor this in, you would need to determine the actual duty cycle (or some more complex integration) and calculate the average dissipation due to the motor movements.

Now, the actual static state where the motors spend 95% of their time in theory could be at any point in the microstepping curve. It is this static "hold" state which determines the heat generation and thermal rise of the system. In the mjkzz system, this state is only held for a short time, and then reverts to a selectable minimum hold current which needs to be sufficient to give holding torque so the rail does not slip. I'm not sure if this hold current is then held continously while the software is running, but I suspect it is (Peter, if you're reading this can you confirm?).

In one of your SnS system threads, I asked about how you deal with this. Holding forever at an accurate microstepping position is a serious waste of power and could result in overheating, depending on the final position of the motor. I'm not sure what the best way to handle it is, but there are many options that might be worth discussing.

mawyatt
Posts: 2497
Joined: Thu Aug 22, 2013 6:54 pm
Location: Clearwater, Florida

Post by mawyatt »

ray_parkhurst wrote:
mawyatt wrote:
...

Note that the Motor Power (t) minimum, maximum, and average are independent rate of shaft rotation w under the conditions described.

So at light loads and relatively slow speeds the motor potentially dissipates less average power than at the peak static motor power which occurs at odd multiples of 45 degrees. Interesting indeed!!!

Best,
Mike,

Let me try again, as you seem to not be getting my point...

During SnS operation, the motors spend most of their time in a quasi-idle state. This state is whatever the final position demands from perspective of the microstepping voltage. In my SnS system, the XY motors spend perhaps 95% of the time in this quasi-idle state, while the Y motor spends >95% of its time in this quasi-idle state. The motors are only moving perhaps 5% of the time, so the dynamic power makes only a small difference to the overall heat generation, and indeed I'd say could almost be ignored. Only the quasi-idle/static dissipation really matters. However, if you do want to factor this in, you would need to determine the actual duty cycle (or some more complex integration) and calculate the average dissipation due to the motor movements.

Now, the actual static state where the motors spend 95% of their time in theory could be at any point in the microstepping curve. It is this static "hold" state which determines the heat generation and thermal rise of the system. In the mjkzz system, this state is only held for a short time, and then reverts to a selectable minimum hold current which needs to be sufficient to give holding torque so the rail does not slip. I'm not sure if this hold current is then held continously while the software is running, but I suspect it is (Peter, if you're reading this can you confirm?).

In one of your SnS system threads, I asked about how you deal with this. Holding forever at an accurate microstepping position is a serious waste of power and could result in overheating, depending on the final position of the motor. I'm not sure what the best way to handle it is, but there are many options that might be worth discussing.
Ray,

Don't think I missed your point, my note was about the apparent moving motor drawing less supply current (less overall power) than the static case where the angle was near odd multiples of 45 degrees, to me this was not expected and an interesting result.

For sure any stacking operation, regular stack, or S&S, the motor(s) will spend most of the time in the static case.

Evidently the motors you use and recommend have higher internal resistance, and thus the RI^2 effect which must be paid close attention to, as well as requiring higher supply voltages. This is where gating the motor currents becomes very useful, as is the case with the USA KR-15 NEMA 11 motors I have which have a 5.7 ohm coil resistance, although I limit the max current to ~600ma. The other motors I have are ~2 ohms and have much higher current limits.

The motor control setup I have uses a method to recirculate some of the motor inductive current, so the supply current remains reasonable. However, this recirculating current does still flow thru the coil resistance thus the motor incurs a power dissipation penalty. I usually keep a selectable smaller "holding current" in place when not sacking, but don't bother during a full S&S session since the dissipation is low enough to not cause significant motor temperature rise.

So I was just curious about the result of lower dissipation when moving vs. static around certain angles, and had thought about how much power dissipation was actually taking place with the 2 coils and sinusoidal current excitation.

I'm sure you and others may have already known this, so sorry for the long post and boring math, but now I have a much better understanding on what affects actual motor power dissipation and can tailor my setups to accommodate such.

Best,
Research is like a treasure hunt, you don't know where to look or what you'll find!
~Mike

Post Reply Previous topicNext topic