2
\$\begingroup\$

I am making a band-pass filter with a gain of 75v/v. Center frequency of 30k. Passband of ~30k. I used the analog wizard site (https://tools.analog.com/en/filterwizard/) to generate the band-pass circuit. Pictured are the band-pass info and the proposed circuit. I am using a +-5V power supply. enter image description here enter image description here

For the first stage of the band-pass filter (Stage A), when I use an SMT resistor of 22 ohms for R3, the circuit rails to the -5V power supply. However, when I used a through-hole resistor of the same value (22 ohms), the circuit works as intended. Below is the top view of my pcb. The resistor numbers dont match, but I have highlighted R17 to be equivalent to R3 in the opamp picture. enter image description here My guess is that the SMT resistor is forming a series capacitance with the ground plane below the resistor. I am relatively new to making PCBs, so any advice on improving my design would be greatly appreciated.

\$\endgroup\$
11
  • \$\begingroup\$ Which resistor corresponds to R4? \$\endgroup\$
    – Finbarr
    Apr 25, 2022 at 15:24
  • \$\begingroup\$ You need to rule out tolerances and there is also more than one kind of surface mount resistor and through-hole. \$\endgroup\$
    – DKNguyen
    Apr 25, 2022 at 15:24
  • 3
    \$\begingroup\$ By the sound of it, the SMT resistor wasn't making proper contact with the pads or one end was shorting with 0V. A bit of probing with a scope or a meter should identify the issue. \$\endgroup\$
    – Finbarr
    Apr 25, 2022 at 15:34
  • 1
    \$\begingroup\$ @Finbarr I went back after rechecking, and I redid the solder on the pads again for R17 for the 3rd time. That fixed it miraculously. For some reason, my previous resolders did not adhere to the pad. And it is the same problem across all 5 boards for the same resistor, so there must have been some issue with this particular resistor for this particular configuration.. If you propose your comment as an answer, I will accept it. Thank you all for your help. I made the false assumption that the factory would correctly solder all the components, and I will no longer make that assumption. \$\endgroup\$
    – ScottMay
    Apr 25, 2022 at 16:07
  • 1
    \$\begingroup\$ Done. It may be that the resistor isn't quite sitting down on the pads if the solder resist doesn't have enough clearance. \$\endgroup\$
    – Finbarr
    Apr 25, 2022 at 16:38

1 Answer 1

Reset to default
3
\$\begingroup\$

For capacitance to the ground plane to affect the operation of a 22 Ohm resistor at 100 kHz it would need to be in the tens of nanoFarads and that's way, way more than you'd ever get between tracks on PCB layers. In any case, it would just reduce the gain or attenuate the signal rather than send it to a supply rail.

It's far more likely that the resistor isn't being properly soldered to the board or there's a short somewhere. Some probing with a scope or a meter should reveal the problem.

\$\endgroup\$

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.