100VDC Motor drive on mains volatage using PWM technique

Thread Starter

Swap_4342

Joined Jun 3, 2018
16
i am working on a project, it involves 100VDC Motor and i want to drive this motor on mains voltage, so my plan is to convert 220VAC to ≈ 308VDC using rectifier, then using mosfet control by pwm signal ≈ 40% duty cycle to drive my 100vdc motor (using capacitor to filter the DC voltage), my query was is this possible. if not what are the alternatives?
 

Ian0

Joined Aug 7, 2020
4,824
Don’t do that. It puts a big capacitive load on the MOSFET which takes massive peak currents when the MOSFET switches on. The circuit need the motor inductance.
8A motor current may be rather large for this type of circuit. You will need a lot of filtering to remove the switching interference from the mains, and you also have to accommodate the motor starting current.
 

Thread Starter

Swap_4342

Joined Jun 3, 2018
16
Don’t do that. It puts a big capacitive load on the MOSFET which takes massive peak currents when the MOSFET switches on. The circuit need the motor inductance.
8A motor current may be rather large for this type of circuit. You will need a lot of filtering to remove the switching interference from the mains, and you also have to accommodate the motor starting current.
will buck converter work?
 

LowQCab

Joined Nov 6, 2012
1,914
Do You need to run the Motor in Reverse ?

Do You need Speed-Control, ( Knob ) ?
Do You need Speed-Regulation, ( Governor with Feed-Back) ?

Do You need Current-Limiting ?

Do You need Ramp-Up, Ramp-Down capabilities ?

Do You need Isolation from the Mains for Safety ?
.
.
.
 

DickCappels

Joined Aug 21, 2008
8,674
Switching that kind of current at that kind of voltage is pretty stressful for transistors. If your DC motor can tolerate some ripple in its power supply a more reliable approach may be to put a phase controlled triac (lamp dimmer) ahead of the rectifiers and then rely on capacitors (or even better an LC filter) to smooth the voltage. I have killed too high voltage and high current transistors to count but I have yet to kill a traic or silicon controlled rectifier (SCR).

Something like the circuit below.
1641563640976.png
Yes, you can use a diac instead of a programmable unijunction transitor since diacs are more common these days.
 

Ian0

Joined Aug 7, 2020
4,824
That circuit requires a freewheel diode, because there is no oath for the inductive circulating current. If you put both thyristors in one leg, and both diodes in the other, it doesn’t. I use this circuit (with a choke, and controlled by a microcontroller and opto-triacs) for battery chargers, up to 7kW. Never had a thyristor fail.
By the way, on 415V that 470Ω resistor and zener diode will run mighty warm.
 

Thread Starter

Swap_4342

Joined Jun 3, 2018
16
what if half wave rectifier is used instead of full wave bridge rectifier, in that case i can get ≈ 99VDC that i can feed directly to motor using filter capacitor.
 

BobTPH

Joined Jun 5, 2013
4,742
A half wave rectifier with no filter capacitor would give you the equivalent if 110V and probably be okay. A filter capacitor would raise the voltage, which is probably not okay.

Bob
 

LowQCab

Joined Nov 6, 2012
1,914
What's really needed is a Paper-Clip, and some Bubble-Gum, ( ABC Bubble-Gum, already been chewed ).
.
.
.
Do You need to run the Motor in Reverse ?

Do You need Speed-Control, ( Knob ) ?
Do You need Speed-Regulation, ( Governor with Feed-Back) ?

Do You need Current-Limiting ?

Do You need Ramp-Up, Ramp-Down capabilities ?

Do You need Isolation from the Mains for Safety ?

Is this a Treadmill-Motor ?

Does it have a Tach-Wheel of some sort ?, if yes,
how many Teeth, and what does the Sensor look like, ( pictures would be a bonus ).

What do You intend to power with it ?,
or are You just messing around and want to make some smoke ?
.
.
.
 

DickCappels

Joined Aug 21, 2008
8,674
A half wave rectifier with no filter capacitor would give you the equivalent if 110V and probably be okay. A filter capacitor would raise the voltage, which is probably not okay.

Bob
This is an interesting line of investigation. If you rectify and don't use a capacitor input filter you get 311 peak-to-peak half-cycle pulses at the line frequency. Given the duty cycle of 50% and the average-to-peak ratio of a sine wave is 0.637 means that if you use an LC filter you will get your 99.1 VDC with the amount of ripple you choose (by adjusting the value of the capacitor).

However, this is still a DC load and it can lead to problems, and as @Ian0 pointed out this can get you into trouble with the authorities. If you full cycle rectification with the LC filter and drop every other cycle you would get the same voltage I don't think it would even hurt the power factor. Would that be acceptable?
 

Ian0

Joined Aug 7, 2020
4,824
However, this is still a DC load and it can lead to problems, and as @Ian0 pointed out this can get you into trouble with the authorities. If you full cycle rectification with the LC filter and drop every other cycle you would get the same voltage I don't think it would even hurt the power factor. Would that be acceptable?
The circuitry required to drop every other cycle would be not much different from your SCR-controlled rectifier in post #8, so perhaps I'd go for the phase-control so that there is speed control!
With regard to the EMC regulations, one never knows whether the TS is proposing a one-off design for home use or a commercial product, so, firstly cannot advocate anything which would be illegal, and secondly, those regulations are there for a reason: not such a great idea to make a controller which interferes with everything around it, even if it just for home use.
 

Ian0

Joined Aug 7, 2020
4,824
The triac and bridge rectifier circuit will work, but I'm not sure that the trigger circuit will.
This circuit is a light dimmer, and the timing circuit relies on current flowing through the lamp. However, you have a motor which will be generating voltage when the triac is not conducting.
If you rectify the mains then switch the motor with a thyristor that won't be a problem, or generate the trigger pulse elsewhere.
 
Top