I am receiving PWM signal at 7.8V from a given controller, with which we need to control the speed of a DC motor at 18 V. I implemented the following circuit for this. enter image description here However, the output to motor signal (blue) is not following the original PWM (yellow) and duty cycle is increased by 50%. What could be the culprit. I dont think it is the awkward taper off on the downleg of the input signal enter image description here.

What could be a remedy?

How about this solution?

How to reduce MOSFET turn-off delay

  • 1
    \$\begingroup\$ Tyr to reduce 10k resistor to 1k or more. Or add a voltage follower i.stack.imgur.com/ry2GB.png \$\endgroup\$
    – G36
    Dec 9, 2018 at 12:36
  • \$\begingroup\$ What is the driver for the "awkward signal"? That is your problem. The NPN is an inverting current sink with a current gain of hFE and it's output is not shown. The result can work if there is a fast turn off of the base current., such as with 10K to gnd on PWM input. The FET bias looks ok. Also show cable layout \$\endgroup\$ Dec 9, 2018 at 15:16
  • \$\begingroup\$ You probably want to use a commercial mosfet driver instead of building one yourself. A mosfet driver will take a logical input signal (you may have to reduce the 7.8V to 5V), and will produce a high current output sufficient to charge the mosfet's gate capacitance. \$\endgroup\$
    – C_Elegans
    Dec 9, 2018 at 16:04
  • \$\begingroup\$ @TonyEErocketscientist The source is another micro-controller with proprietary software/hardware. We are constrained to use it. \$\endgroup\$
    – Bilal
    Dec 10, 2018 at 3:57
  • \$\begingroup\$ @C_Elegans can you recommend a good diver? What do you think about IRF9540? \$\endgroup\$
    – Bilal
    Dec 10, 2018 at 3:58

1 Answer 1


Thanks to the comments/suggestions provided here, we found the solution. We modeled the circuit in Proteus and were able to recreate the output signal. Proteus recreation of Original Circuit Then, as suggested by @G36 the 10k ohm resistor was reduced to 820 ohm, and few minor modifications. Modified Circuit Simulation environment predicted better results: Simulated output of modified circuit On oscilloscope too results were verified as true to form. Actual output of modified circuit

Thanks to all contributors!


Your Answer

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

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