I want to mount a brushed DC motor on a bike and drive it with PWM. I know I could directly use a PWM generator IC, a 555 timer, use an Arduino or use an already made DC motor driver, but I want to build the circuit myself.

Attached is the circuit I imagined:

An RC oscillator generates a sine wave signal (output of U1) that is rectified through a full bridge rectifier and is then fed along with a fixed voltage into a comparator (U2) to generate a PWM signal. This PWM signal then actuates an opto SSR that drives the motor. R5 potentiometer would act as the "throttle" of the motor, varying the reference voltage and therefore the duty cycle value of the PWM signal.

Here is the drawn circuit

I also attached an LTSpice simulation file I made (took the oscillator part from the internet). I had to add a few things to make it kind of work, mainly resistors and a buffer opamp between the rectifier and U2. Here is the google drive link : https://drive.google.com/file/d/14ZA4uFyuyh9bzQHPsTTj34qJEv0vY_dM/view?usp=share_link

I am very new to circuit design and started dabbling in electronics only recently (I have a materials engineering background).

Here are my main questions/concerns:

  • Since both the PWM generator circuit and the motor circuit would run on batteries, would the grounding as I've drawn it on paper work? (I think common is more appropriate in that case) Grounding in LTSpice is complicated with full bridge rectifiers and I'm having a hard time understanding it.

  • Could the PWM generator circuit be run on low voltage batteries? (I was thinking about 9V batteries?) My guess is that I would have to take into account the voltage drop across the diodes as well as the voltage needed for the opto SSR to turn on.

  • I've seen on the internet that a good PWM frequency is around 25 kHz and above, as it's beyond human hearing range. Do you think an opto SSR could switch that fast? I've read that their weakness is switching speed.

  • An H-bridge is unnecessary to drive the motor since I only want forward motion, is that right?

  • All in all, do you think this is realistic and would work in real life? I have never built real circuits, so I don't have a reference when it comes to real circuits behaviours (signal distortions, losses, heat dissipation, power supplies stability etc).

  • If you think I could give this circuit a try, could you please give me the names of parts you would recommend for this job? (especially opamps and opto SSR).

I would run the motor with two 12V lead-acid batteries in series. Here are the specs of the motor: https://www.lynchmotors.co.uk/pdfs/lmc-lem-130.pdf

LEM-130 95 motor from LynchMotors:

  • 2.27 kW rated power (3 kW peak)
  • 75 A rated current (100 A peak)

This is for 65V, so I guess for 24V it would be less than half current and power?

Please be critical and give me your guidance regarding this circuit. Any advice will be greatly appreciated.

  • \$\begingroup\$ Generally em generators use sawtooth oscillators, you’ll get a much more linear response with that. \$\endgroup\$
    – Bryan
    Commented Dec 3, 2022 at 7:18
  • \$\begingroup\$ Yes, that's what I wanted to do in the beginning but couldn't figure a way to generate a sawtooth signal with my limited knowledge of circuits. PStechPaul in his answer below gave me a good solution. \$\endgroup\$
    – Solmyr999
    Commented Dec 3, 2022 at 9:06

1 Answer 1


You will have a problem driving a high current DC SSR at PWM frequencies above a few kHz. I would suggest using a proper half-bridge made from NMOS devices and a high-low half bridge driver. A good choice would be the IR2103, which can be driven by a single PWM signal for both high and low side MOSFETs.

And I think you should use a triangle wave and comparator to create your PWM signal. Here is a circuit using three op-amps and a 9V single supply, which could be derived from your 24V battery. You don't really need isolation, but be very careful to separate the high current circuit from the signal circuit. I'd suggest starting with a small 24V DC hobby motor to develop your design before graduating to the 3 HP end project.

Simple PWM using Op-amps

I have added a high-low half-bridge driver, an NMOS half-bridge, and an inductive motor load, with a 24 VDC supply. The actual hardware design should add gate resistors to the MOSFETs, and various bypass capacitors on the power supplies, but this shows, in principle, how a DC brushed motor PWM drive should work. The PWM frequency is about 13 kHz, and can be changed using different values for C1.

PWM DC half-bridge motor drive

It's important to note that I added an inverter for the PWM signal for this particular half-bridge driver. Also note that the high side MOSFET drives the motor ON, and the low side turns on to conduct the stored inductive current. It may be wise to add a commutating diode across the motor, and perhaps an R-C snubber, especially for a brushed DC motor to reduce noise.

  • \$\begingroup\$ Thank you very much for your helpful answer! In the case of the IR2103, do I simply put the low voltage PWM signal between the input pins and get the same PWM signal between the output pins but matching the higher voltage of the 24V battery ? Have I understood correctly what you meant by half bridge driver ? \$\endgroup\$
    – Solmyr999
    Commented Dec 3, 2022 at 8:56
  • \$\begingroup\$ Also could you please explain how the oscillation is generated with U1 and U2, or give me the name of the oscillator circuit so I can look it up ? \$\endgroup\$
    – Solmyr999
    Commented Dec 3, 2022 at 9:04
  • \$\begingroup\$ R4 and C1 form an integrator with U2, and the output is fed back to U1, which has positive feedback so the output toggles +/-. Have a look at the datasheet for IR2103 to see how it should be connected for a half-bridge. It will require a 10-20 VDC supply for gate voltage, and you can use that for the PWM circuit and logic inputs to Hin and /Lin. Note that /Lin is inverting, so a positive PWM drives the low side gate low, while Hin drives the high side gate high. It has a bootstrap circuit for high side gate. \$\endgroup\$
    – PStechPaul
    Commented Dec 3, 2022 at 9:53
  • 1
    \$\begingroup\$ Thank you very much for your detailed answers. I often struggle to find the specific parts needed for a real circuit since I don't have any knowledge of parts used in actual hardware; you putting the parts names in your LTSpice schematic helps me a ton! I actually looked up the half bridge circuit and understood its relation to the H-bridge (with which I am familiar) and how it drives the motor. And why it needs a dedicated driver IC. Anyways, thank you for all your invaluable advices, electronics has always been pretty occult for me and this is the help I need to wrap my mind around it. \$\endgroup\$
    – Solmyr999
    Commented Dec 3, 2022 at 13:33
  • 1
    \$\begingroup\$ Thank you for your recommendations and sharing your experience! Now I know what to ask for christmas. Hands-on projects is what I want to do. I've developed a phobia of electronics when theory was forced down my throat during preparatory classes. Only now am I coming back to it, at my own pace and from a practical angle, and actually discovering that I like it a lot. I wish you well for the future and maybe see you in another post! \$\endgroup\$
    – Solmyr999
    Commented Dec 5, 2022 at 15:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.