Running hot, really, really hot

[imagineer]

Recently I built a device to turn a stepping motor a pre-set number of revolutions and stop when the desired count was reached. The device can cycled repeatedly by pressing a start button. Run cycle time is less than 12 seconds. The average time between operation cycles is approximately 30 seconds but can go over one minute.

I used a 24v, SloSyn stepping motor (p/n SM-24-0065-RY) with 24 steps per revolution. (I chose that type of motor because I need high RPMs.) I’m using a 24vdc power supply. I made an adjustable stepping motor driver based on this kit http://www.kitsrus.com/pdf/k179.pdf.

There is an optical encoder on the end of the motor shaft that sends a signal to a settable countdown counter. When the countdown reaches zero, it’s output interrupts the step signal from the motor driver. The motor stops spinning and is held locked in position until the start button is pressed again.

The problem is heat. After only 5 or 6 cycles, the stepping motor is dangerously hot. So is the stepping motor control board. It gets so hot that the driver board stops working until the unit cools down. I have the MOSFETs mounted to a large heat sink. There are a bunch of ventilation holes drilled in the enclosure and a PC fan pulling air out.

Any ideas why it’s getting so hot? I don’t need high torque for this application. Would going with a smaller power supply, say 12vdc, decrease the amount of heat? Would using a motor with more steps per revolution generate less heat?

One thing I’ve been thinking about trying is to incorporate a power-off timer that will keep the motor coil energized for 3 seconds or so (after the step signal is interrupted), then cut power to the driver board. That should give the operator enough time to remove the product from the output shaft.

Thanks in advance.

[smarbaga]

the only thing that can make heat is amps.
amp through your transistors and amps into your motor coils.
maybe the motor is really 2..4 volts and the decimal point is wore off.
in any case, u will have to chop or drop the voltage.
u will loose holding power if you cut the voltage to the motor coils.

[jeffs555]

I agree. That driver shows IRFZ44 mosfets which should handle close to 50 amps with a good heat sink. If they are overheating, your motor is probably a 2.4 volt motor, or else it has an internal short.

[rippersoft]

These stepper drivers need resistors for current limiting. These are not chopper drivers. You will need to calculate how much additional resistance to add to the stepper phases to keep the amps as printed on the stepper's label.

For the given resistance of the coil, the more voltage you add, the higher the current draw. To limit the current, you increase the resistance.

RipperSoftware

[imagineer]

Originally Posted by rippersoft These stepper drivers need resistors for current limiting. These are not chopper drivers. You will need to calculate how much additional resistance to add to the stepper phases to keep the amps as printed on the stepper's label. For the given resistance of the coil, the more voltage you add, the higher the current draw. To limit the current, you increase the resistance. RipperSoftware

Thanks, I'll have to check the actual coil resistance.
The motor I used is a 24V, 2 amp motor. Since it's a 5 wire motor, should I put a single current limiting resistor on the common power lead or have 4 resistors between the MOSFETs and the motor coils?

[smarbaga]

either way will work equally.
a single resistor will mean a much larger watage then 4 individual ones.
smaller watt resistors are easier to come by, sometimes are less than 1/4 the cost of 1 larger one,
i think 100 watts of heat is hotter that 4 x 25 watts of heat. ( not in theory but in practice )
4 resistors will disapate heat better than 1 large one.

[rippersoft]

Are the coil resistance values on the tag? I am inclined to believe that the motor is 2.4v and not 24v with as much heat as it is generating.

Amps (I) x Coil Resistance (R) = Voltage (E). Voltage (E) / Amps (I) = Coil Resistance (R).

So, 24 volts / 2 amps = 12 ohms of coil resistance.

What does the tag say?

RipperSoftware

[imagineer]

It's a 24 volt motor, rated at 2 amps, the coil resistance per phase is shown at 9-11 ohms. The label is not damaged and is easy to read. If I understand the math, that means I'll need a 1 ohm resistor.

I haven't taken the motor out yet to check the actual coil resistance or to see if there might be a short. The enclosure is really tight so I will be pulling everything out and mounting it on a project board to test other motors and power supplies.

[smarbaga]

i think in a pinch, you could use 60 watt lightbulbs in series with your coils, for now, unless you have a collection of WATT resistors lying around.
it ain't gonna look pritty but they will do the trick, for current testing.
they are about 2-3 ohms, use those round construction ( basement type) screw in light fixtures.
this will give you a visual indication of what is going on as well.
once a good resistance is found ( changing bulb wattages ), you can seek real resistors.

[imagineer]

[quote=smarbaga;226233]i think in a pinch, you could use 60 watt lightbulbs in series with your coils quote]

Thanks, I was already thinking of that. I used a 60watt bulb as an inline current reducer for a home made degaussing coil (accidentally placed a car speaker on top of my shop TV and the colors on the screen went all funky).

Even with a current reducer, I will still have to incorporate some type of power-off-timer to cut power to the coils during the time between machine cycles.

In normal operation, the counter is reset before each cycle by pressing the reset/run momentary contact switch. Once the counter is reset, the step signal is no longer being interrupted and the motor starts. I like the idea of having one switch to do both functions. This whole time, there is power to the motor coils.

Last night, I put a SPST switch inline on the power lead to the motor driver board. Using the switch instead of the reset/run button, I was able to cycle the machine indefinitely without it overheating. The motor still got warm, but not too hot to touch and because the PC fan in the enclosure does not switch off, the heat sink for the MOSFETs stayed cool.

The problem with cycling the machine by switching the power to the coils is that there are some shared grounds and the spike from the SPST switch resets the digital counter and starts the motor. I suspect it will eventually damage one or both circuit boards.

I'll have the whole thing torn apart and mounted to a project board this weekend to work the bugs out. I'm thinking of building a logic latching circuit (basically a solid state push on/push off circuit) that will use the pressing of the reset/run button as the "push on" thus (via a relay) powering the coils AND (via an opto isolator) resetting the counter (which un-interrupts the step signal) AND starts the motor. The output from the counter (that interrupts the step signal and stops the motor) will also start the power-off timer. The output from the power-off-timer will be the "push off" and will be used to cut the power to the motor coils.

[mugabe]

If you can have 2 voltage supply driver, it will be possible to switch between supplies when no activity on buttons , or you can put one mosfet from power supply to motors given mosfet is controlled by square wave freq with 50 percent duty. That will reduce heat for sure.