Have you considered using the 120 Hz for PWM frequency, somewhat like original SCR but using logic control of FET gate for 0 to 8 msec. delayed turn on ??
Two 555's could be the pattern for logic control. Un filtered rectified V minimum would trigger first 555, adjustable 0 to 8 msec., output C coupled to second 555, output drives gate. Leading edge of # 1 555 is reset to pin 6 of 2 nd 555.

 

Thought about that, but I need to use a PLC logic to control the PWM. I also thought about using the PCL with 120Hz but then I would need another signal to synchronize the beginning of the wave to the pulse. Otherwise the pulse could be starting at any point during the half wave, which would give different results. If it's at 10% duty and the pulse starts at the top of the half wave the result will be different than 10% duty at the beginning of the half wave. Is it correct?

Have you considered using the 120 Hz for PWM frequency, somewhat like original SCR but using logic control of FET gate for 0 to 8 msec. delayed turn on ??
Two 555's could be the pattern for logic control. Un filtered rectified V minimum would trigger first 555, adjustable 0 to 8 msec., output C coupled to second 555, output drives gate. Leading edge of # 1 555 is reset to pin 6 of 2 nd 555.

 

Take a look at the gate components diode and resistor in parallel in Tahmids blog http://tahmidmc.blogspot.ca/2013/01/using-high-low-side-driver-ir2110-with.html
fig.6.
The test I did on a similar motor, the average current was no more than 2amps off load from 0 to 2000 rpm.
Max.

Tried the diode in parallel with the gate resistor, but there were no changes. I don't have access to an oscilloscope but measured the voltage at the diode with a voltimeter. It was always negative, at different frequencies and duties. Voltage changed with duty but always remained negative. Not sure if that measure is valid or I need to use an oscilloscope to be sure?

 

Does the PLC not accept a start signal ?

 

Does the PLC not accept a start signal ?

Yes, it can accept a start signal but I would need to add more electronic components and use another PLC input. It wouldn't be necessary if the PWM works by modulating the rectified unfiltered current. That is what I would like to get in end, and if it doesn't work I would like to understand why...
Thanks for your comments!

 

I got the circuit to work without getting too hot, just by using a transformer as suggested. Reducing the voltage reduced also the current and most of the noise, as well as the heat. I'm keeping the photo coupler, the lowered resistors (pull down and gate) and the twisted cables.

The transformer has a 24VAC output, and the PWM is generated with that output without filtering. MOSFET's heat is barely noticeable and the DC motor runs smooth.

Now I'll start making some tests to analize the motor's torque and check if this method is good enough or more adjustments are needed.

Thanks to everyone for your suggestions and guidance!

 

Sorry, I forgot to mention that the transformer output is rectified (but unfiltered)