Ask not for whom the motor turns. It turns for thee!
October 9, 2008 9:17 AM Subscribe
How do you control the speed of a 3-phase AC induction motor?
I've written software to control a 3-phase H-bridge inverter, where each of the three phases is connected to a corresponding phase of an AC motor. My software is capable of starting and running the motor (yay!) but changing the parameters that control the software does not result in the motor behaving as I expect. Therefore, either my controls are not having the result I expect on the output of the H-bridge (not likely), or I don't understand what the motor should do when I change the output (very likely).
The software currently uses no feedback: it just produces a PWM signal to turn the H-bridge switches on & off at the right times to make the equivalent of 3 sine waves. I can control the frequency of the sine waves and the dead time between "high" and "low" pulses. (The PWM period is 40us == 25kHz.) I can also control the percentage of modulation by the sine wave at a given deadtime, but I'm running it at 100% modulation now. That is, if the deadtime is 25%, then the high and low pulses vary between 0 and 30us each, and the sum of high and low pulses is always 30us out of each 40us PWM period.
At a single frequency (I'm using 60Hz) the motor runs nicely at 25% deadtime, slowly at 26% deadtime, very slowly at 27% deadtime, and not at all above that. I haven't tried less than 25% deadtime -- I think it will make the motor spin faster (and maybe faster than I want it going while it's sitting near my computer!)
I had expected motor speed to be controlled by the inverter frequency, and power to be controlled by the deadtime. That doesn't seem to be the case. Instead, it looks like inverter frequency doesn't matter very much, and deadtime controls both speed and power (???) I don't have a tachometer, so it's possible that 40Hz-50Hz-60Hz-70Hz makes a small difference in speed, but it's the deadtime that really changes things.
What am I missing / misunderstanding?
I've written software to control a 3-phase H-bridge inverter, where each of the three phases is connected to a corresponding phase of an AC motor. My software is capable of starting and running the motor (yay!) but changing the parameters that control the software does not result in the motor behaving as I expect. Therefore, either my controls are not having the result I expect on the output of the H-bridge (not likely), or I don't understand what the motor should do when I change the output (very likely).
The software currently uses no feedback: it just produces a PWM signal to turn the H-bridge switches on & off at the right times to make the equivalent of 3 sine waves. I can control the frequency of the sine waves and the dead time between "high" and "low" pulses. (The PWM period is 40us == 25kHz.) I can also control the percentage of modulation by the sine wave at a given deadtime, but I'm running it at 100% modulation now. That is, if the deadtime is 25%, then the high and low pulses vary between 0 and 30us each, and the sum of high and low pulses is always 30us out of each 40us PWM period.
At a single frequency (I'm using 60Hz) the motor runs nicely at 25% deadtime, slowly at 26% deadtime, very slowly at 27% deadtime, and not at all above that. I haven't tried less than 25% deadtime -- I think it will make the motor spin faster (and maybe faster than I want it going while it's sitting near my computer!)
I had expected motor speed to be controlled by the inverter frequency, and power to be controlled by the deadtime. That doesn't seem to be the case. Instead, it looks like inverter frequency doesn't matter very much, and deadtime controls both speed and power (???) I don't have a tachometer, so it's possible that 40Hz-50Hz-60Hz-70Hz makes a small difference in speed, but it's the deadtime that really changes things.
What am I missing / misunderstanding?
Response by poster: Thanks, that's an excellent reference! Not too basic, not too mathy.
posted by spacewrench at 10:28 AM on October 9, 2008
posted by spacewrench at 10:28 AM on October 9, 2008
This thread is closed to new comments.
But I do know that you need to control both the frequency and voltage your feeding the motor. Your trying to keep the mV/hz value constant inside the windings. Failing to control this will burn out the insulation on your windings in short order.
Essentially you need to control the PWM over different parts of your waveform.
This (pdf) should get you started.
posted by Confess, Fletch at 10:12 AM on October 9, 2008 [1 favorite]