Universal motor PWM control

Thread Starter

florian_1034

Joined Apr 2, 2022
14
Hello everyone,
I want to control a universal 750W washing machine motor in PWM.
The requirements are:
- maximum current 3A, average 1A
- power supply: direct connection to the grid, only half way sine is used (I don't need full power) 230V, 50Hz
- control mode: PWM, chopping the half sine.
- optocoupler isolated

I know that phase fired controller is the most used option here (and I tried that too) but I want to keep it simple and don't use any zero cross detection circuit. The motor speed is controlled in a PID closed loop.
This module is to replace my old module that I used for 4 years where the ground for the logic circuit was the same with the ground for power circuit (logic part not electrically isolated from the motor). And that cause my mcu to reset sometimes.

I want to design a new power module that is simple and up to the task.

Till now I have a working prototype that uses a mosfet as a main switching device. See the attached picture.

Now I'm wandering if I can use a BJT in place of the mosfet. Driving a BJT base is simple when working with voltages up to 24V. But setting the right amount of base current when all you have is 230V ac is a real challenge. You can see in the attached schematic that the base resistor need to deliver 200mA to transistor and will have to dissipate around 60W!
Also, I don't want to use a dedicated power supply just to drive this transistor.

If anyone have an idea how to drive this BJT in a simple manner or advice to improve the mosfet approach, please help. I will design and print a pcb after.
 

Attachments

Thread Starter

florian_1034

Joined Apr 2, 2022
14

Ian0

Joined Aug 7, 2020
9,680
In most 230V countries half-sine wave control (half wave rectified) is illegal under EN61000-3-2 (IEC1000).
 

Thread Starter

florian_1034

Joined Apr 2, 2022
14
I imagine you could with a PWM controller, you would need to design a closed loop for the speed control, however.
I have done this already and is working very well. In the old design I have the taco generator of the washing machine motor which feeds in a half way rectifier, a capacitor and a parallel resistor. This signal is going to the ADC of an ATmega 328 that use a PID (PI) algorithm to output the pwm for the switching. The control worked surprisingly well, being able to control the motor even at 100 RPM when loaded.
 

MaxHeadRoom

Joined Jul 18, 2013
28,621
If working well with the old system, what is the reason for change?
Why not use a Opto Isolator with the present method, see Fairchild AN-3006.
 

Thread Starter

florian_1034

Joined Apr 2, 2022
14
If working well with the old system, what is the reason for change?
Why not use a Opto Isolator with the present method, see Fairchild AN-3006.
I attached a picture for the old design. As you can see, power mosfet ground and mcu ground are the same. Also they share the 12V rail. This is not a good practice because when brushes have a bad contact or when there was humidity there it reset the mcu. Also I designed a new control board with bigger ground plane.

This is a washing machine motor that opens a car access gate. The gate is opened with an acceleration ramp to be fast.

It's a simple schematic in the picture, in reality I have switches, remote module, EMI filter etc.
 

Attachments

MaxHeadRoom

Joined Jul 18, 2013
28,621
I would have used a opto on the MCU output into a Mosfet driver such as MIC4427 etc., you can get them in non-inverting or inverting.
 

Thread Starter

florian_1034

Joined Apr 2, 2022
14
Also this darlington configuration seems to work in simulator. But placing 4 diodes in series with 3 amps is a no go.
Any other ideas of how I can feed the base of the power BJT with the required current and no additional psu?
 

Attachments

Thread Starter

florian_1034

Joined Apr 2, 2022
14
Thyristor?
Thyristors are notorious for not turning off when you want to. Need to wait the sine to go to zero or force a reverse current through them to shut off. Possible but not simple in my vision. Phase fired controller was already discussed.
 

MaxHeadRoom

Joined Jul 18, 2013
28,621
This module is to replace my old module that I used for 4 years where the ground for the logic circuit was the same with the ground for power circuit (logic part not electrically isolated from the motor). And that cause my mcu to reset sometimes.
It may be worth exploring a decent ground plane that includes Earth GND and a star point GND, many of the DIY CNC crowd have had similar experiences when neglecting to use this method.
 

Thread Starter

florian_1034

Joined Apr 2, 2022
14

Ian0

Joined Aug 7, 2020
9,680
Thyristors are notorious for not turning off when you want to. Need to wait the sine to go to zero or force a reverse current through them to shut off. Possible but not simple in my vision. Phase fired controller was already discussed.
Then definitely a job for an IGBT
 

shortbus

Joined Sep 30, 2009
10,045
Interesting device. Thanks.
For a home user is hard to find. All I can find are between 500A to 1500 amps. Not suitable for my application. Also they seems to be pretty hard to control.
They are out there in a TO-247 package and lower amperage. I just don't have a part number right now.
 
Top