VFD with small synchronous AC motor 5

[cokeguy]

I need to run a small 120VAC/60Hz/72rpm single phase synchronous motor at one third or one quarter of its nominal speed. The obvious solution that comes to my mind is using a 220 VAC VFD (one of its output phases and neutral) to run the motor, and adjust output voltage and frequency accordingly, lets say, to 40 VAC at 20 Hz. I don´t have much choices, since I can't change gearing and cannot get a slower motor. My question is, can I do it with no ill-effects other than reduced torque?

The motor is a slo-syn SS241-1003 that is part of a Honeywell Herculine series actuator. The motor is located in a tricky location, so my main concern is not damaging the motor. Thanks for your comments.

 

[cokeguy]

Update: I've been testing offline but cannot go less than 30 Hz without losing too much torque, and 30 Hz is not enough speed reduction for a process test. I paralleled another capacitor and boosted voltage, with no improvement. So now I guess my only choice for non-intrusive testing is varying duty cycle. Any simple way to make a zero crossing circuit for duty cycle adjustements? or better yet, an alternate way to get more torque from the motor at 15-20 Hz, without going 3-phase?

Without going into details about the process (which I can´t for confidentiality reasons), this is not a valve actually, we use the actuator to move slowly an electrode into "something" while monitoring its position, and as soon as we detect a certain condition we record the position and withdraw the electrode until another condition is detected, rest for a couple of seconds, and start again, about 4-5 times per minute. each cycle is basically CW movement for 5 seconds, CCW for 5 seconds, rest for 2 seconds... There are better ways to do this which we will eventually do, but in the meantime for test sake I need to be as un-obtrusive as possible, so please pardon my obstination with this seemingly trivial issue (wow, this sounds poetic)

again, any help will be greatly appreciated

 

[cokeguy]

Guys, one final question on this matter (I hope): since the VFD option doesn't look promising right now, I'll try the pulse train using a solid state relay switched maybe 10 times per second or so. Do you think the motor will suffer too much and fail quickly if worked this way? Assuming I get satisfactory results with this option, I may want to leave it working maybe 1 or 2 months while I get a slower motor, since our purchasing department is not very swift.

Thanks a lot in advance for your final comments, any opinion or educated guess will be very welcome.

 

[Mobius44]

Gunnar-
The interface I suggested basically varies the control signal duty cycle to slow down the motor movement. For example if applying the normal 120VAC control signal makes the motor move 90 degrees over 30 seconds, the interface would add "off" pulses to the signal such that the motor moves, pauses, moves, pauses, and so on. In this way the normal 30 second travel time can be changed to a longer period.

This interface is not applicable to any ordinary induction motor, only this specialized case where the end user supplies single-phase CW and CCW control signals to postion the motor.

Cokeguy-
Honeywell Herculine actuators are designed to have the control signals pulsed for precise position control, so it shouldn't hurt the motor. I think your suggested 100ms pulse widths is in the ballpark, just keep in mind if the pulses are too short then they won't move the motor.

We use Herculine's frequently in this manner to precisely position large air dampers for combustion applications, and they are some of the toughest electric actuator's we've found. We normally use PLC's to control them, and our program applies constant 120VAC power to move the actuator until it's close to the desired position, then pulses the control signal 500ms on and 1500ms off until we are within the deadband around the desired position.

Best of Luck,
Sean

 

[cokeguy]

Excellent, thanks a lot Mobius, just what I wanted to hear!!!! I´ll test it right away. And thanks everyone else for your help.

 

[waross]

The FM shows a 7.5uF capacitor for normal operation.
For 20 Hz and 40 volts try adding 60 uF for a total of 67.5 uF
You want x 3 for 1/3 frequency, and another x 3 to maintain the current at 1/3 voltage.
According to the FM, the starting current is so close to the running current that short duty cycles should not be a problem. The unit will come up to speed in about 1 1/2 cycles.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Skogsgurra]

Mobius,

So this is not a mode where you synchronize switching with periods? But rather switch the motor on and off?

That does work for setting position of a damper or a similar thing, but it doesn't reduce speed of the motor - it just makes the motor stop and then runn att full speed and so on.

Is that what the OP wanted? My take is that he wanted to reduce speed to one third and also that the motion should be smooth. If that is not necessary, the "macro" PWM technique will surely work. Done the same thing with Siemens (Landis&Gyr) actuators. But used triac output instead of relays. Relays wore out much too quickly.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Mobius44]

Gunnar-

Yes that is my understanding of the OP's application since he previously described the duty cycle as powering the motor CW for 5 seconds, CCW for 5 seconds, off for 5 seconds, and repeat.

I agree the motor while powered still moves at full speed, and that triacs are superior to relays for this type of switching.

Hopefully the OP will return and advise how it worked out.

 

[waross]

The last time I saw this type of action was to control the level in a glass melting furnace. The actuator drove down until a tungsten probe touched the surface of the molten glass and completed the circuit. The probe halted, the PID controller updated and the probe reversed out of the glass to the top of the stroke and repeated. Can you trim the controller to take care of your problem? Alternately, extend the time period at the top of the stroke.
These motors are rated to stop within 5 degrees. If you want less operations it may be easier to introduce a time delay at the top of the stroke.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[cokeguy]

Yes waross, it is similar to glass level control applications, I do need to slow down the motor close to the detection point in order to minimize overshoot. I plan to approach at normal speed, and close to the target begin a "slow speed by PWM" stage, and then back at normal speed. I'll tell you how it worked, thanks again.