What is the purpose of the transistors in the feedback arrangement of this OP-AMP?

enter image description here

I am mostly interested in the purpose and general use of this kind of "trick", and perhaps also what it does in the circuit where I saw it.

The circuit occurs in a bass-boosting audio circuit "DBX 100" of which the schematics are found here:


enter image description here

There's a similar arrangement discussed here, Op amp with transistor in feedback, but I'd like to better understand the specific role of the transistor.

Maybe hysteresis? Diode in positive feedback?

  • \$\begingroup\$ Can you elaborate please, @TonyStewartSunnyskyguyEE75? Log circuits that I am familiar with have diodes or grounded-base BJTs in the NEGATIVE feedback. And why 4 transistors? Do multiple breakpoints better model a desired log-curve? \$\endgroup\$
    – P2000
    Commented Jan 9, 2021 at 19:18
  • \$\begingroup\$ This is far outside my area of professional expertise. But at the basic level, those transistors are diode connected. So their behavior will be like a diode. That is just basic electronics. The V-I curve of a diode has a logarithmic relationship. Transistors may sometimes make better diodes than diodes. That may be why transistors were used. I have never really used or studied log converters but it makes some sense. \$\endgroup\$
    – user57037
    Commented Jan 9, 2021 at 19:25
  • \$\begingroup\$ @mkeith yes you are right, those BJTs are diode connected. I edited my post: maybe hysteresis? \$\endgroup\$
    – P2000
    Commented Jan 9, 2021 at 19:39
  • \$\begingroup\$ what I don't understand in this circuit is, it seems like any slight positive signal on the non-inverting terminal will instantly send that particular op-amp to the positive rail. Maybe the overall circuit prevents that and keeps that terminal from every going positive. \$\endgroup\$
    – user57037
    Commented Jan 9, 2021 at 19:47

1 Answer 1


I spoke too soon. It is a schematic reverse engineering error. This cct will saturate with open loop gain and positive feedback.

But as a soft limiter followed another soft limiter with an active LPF, it converts it into a limited soft fuzzy bass.

enter image description here enter image description here

  • \$\begingroup\$ Aha. So you are saying the schematic is wrong, and after correction it's a soft limited bass. And that part controls the bass amplitude in a soft controlled VCG way at Q4? The transistors provide a piecewise soft saturation curve? \$\endgroup\$
    – P2000
    Commented Jan 9, 2021 at 20:13
  • \$\begingroup\$ Yuppers, won’t work guaranteed as was. \$\endgroup\$ Commented Jan 9, 2021 at 20:21
  • \$\begingroup\$ Well that error is a brilliant find! \$\endgroup\$
    – P2000
    Commented Jan 9, 2021 at 21:01
  • \$\begingroup\$ You might say, but it only comes from LOTS of experience \$\endgroup\$ Commented Jan 9, 2021 at 21:03
  • \$\begingroup\$ The -in pin is always next to the output so that 1pF stray capacitance won't cause unstable gain so the pin numbers might be right but the polarity is wrong.. Improve your skills with this knowledge and see if you can find at least 3 more schematic errors, then distrust that "site" in future \$\endgroup\$ Commented Jan 9, 2021 at 21:07

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.