2
\$\begingroup\$

I would like to design a circuit to drive a resistive load (around 3 to 4 ohms) with a maximum current of 2 amps. I want to monitor both voltage and current delivered to the resistive load.

I have designed a first version of the circuit with PWM (20kHz) and a Mosfet. I use a voltage divider to monitor the voltage and an INA169 for the current.

enter image description here

However, voltage and current measurements contain the PWM signal. I tried to filter the outputs, but the result is not very accurate.

Does someone have a better solution to do the monitoring ? Is it possible not to use PWM without dissipating too much energy ?

\$\endgroup\$
  • \$\begingroup\$ If the PWM is from a microcontroller, and your load can tolerate small bursts of full power, you can implement pulse stretching: Periodically set the duty cycle to 100%, quickly do your measurements, then set the duty cycle back to whatever it's supposed to be. \$\endgroup\$ – Dampmaskin Jan 26 '17 at 12:27
  • \$\begingroup\$ Thank you for your answer ! I would have to synchronise PWM with analog reading of the microcontroller. It don't feel confortable implementing this solution. \$\endgroup\$ – Jb.ca Jan 26 '17 at 13:55
2
\$\begingroup\$

What you have is the basics of a buck converter so, by adding a capacitor, an inductor and a fly-back diode you can significantly reduce the size of the pulses delivered to your load down to a few milli volts peak to peak: -

enter image description here

Choose L and C so that resonance occurs at a low frequency compared to your switching frequency (maybe 100 Hz).

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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