I have designed this circuit and math calculation works but looking back now I feel the cap is too small may be I am making any mistake here? Can some one explain how to chose the capacitor for AC coupling the relay driver. Please do not engage in to side discussion, why I would want to drive relay with AC signal.

enter image description here

Note: RLY_RUN is PWM signal, which is changed to 5V level by U63 and 100R & 47 nF acts as AC coupling to drive the U23 MOSFET. the relay used here is 24 V DC relay .

I couldn't find any design or application note on this anywhere on the Internet so far.

  • \$\begingroup\$ Probably, your math is bad. \$\endgroup\$ Sep 30, 2021 at 18:47
  • \$\begingroup\$ probably you have to implement a push-pull configuration to energise the coil \$\endgroup\$
    – gino
    Sep 30, 2021 at 19:18
  • 2
    \$\begingroup\$ 1) "math calculation works": Would you post your calculation? 2) Why does it need to be AC coupled, DC coupling should work as well? 3) What is the PWM period and duty? \$\endgroup\$
    – jay
    Sep 30, 2021 at 19:47
  • \$\begingroup\$ @MarkoBuršič , Explain what you mean by this. \$\endgroup\$ Oct 4, 2021 at 4:14
  • \$\begingroup\$ Is that you were doing wrong calculation, this circuit won't work as is. \$\endgroup\$ Oct 4, 2021 at 6:26

1 Answer 1


As Gino says in the comments, you need a push-pull drive. You’ve chosen a 74AHCT17 which is open collector/drain. It will only pull the signal low. Your circuit wont work as there’s no means to charge the capacitor. Adding a pullup resistor will help but may cause the relay to glitch on power up. Or chose a different device for U63.

  • \$\begingroup\$ 74AHCT17APWJ is non inverting buffer , this buffer is driven at 20Khz , 80% Duty . 5V is output of this buffer. The PWM is passed through the capacitor and output of capacitor is bipolar waveform (i.e between +/- 2.5V). For +ve 2.5V it will trigger NUD3124 as NUD3124 has 2V threshold. now relay will turn ON. For PWM OFF duration the cap voltage is -2.5V , which will be clamped to ground by NUD diodes and on driver side, the 100ohm will take the potential difference. \$\endgroup\$ Oct 4, 2021 at 4:13
  • \$\begingroup\$ Ah, the ahct17 is a weird ass part. The 7417 is an open collector buffer. Ok, what measurements have you done? \$\endgroup\$
    – Kartman
    Oct 4, 2021 at 5:41
  • \$\begingroup\$ Lol.. what!...Open collector buffer... you cracked me up buddy. Why you keep referring this part as open collector. there is the part , take a look at datasheet.. for the last time open collector has nothing to with it. \$\endgroup\$ Oct 4, 2021 at 19:00
  • \$\begingroup\$ Thats what I acknowledged. What measurements have you done? \$\endgroup\$
    – Kartman
    Oct 4, 2021 at 21:23
  • \$\begingroup\$ Yes this circuit is tested and in use. I have done measurements at Vds of mosfet , Relay operation was tested. My question is about long term reliability of this circuit \$\endgroup\$ Oct 5, 2021 at 19:51

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.