I'm building a dual rail tracking power supply, for the negative supply because the current is not going to exceed ~200 mA I decided to use LM337 both for simplicity and stability of the circuit (so I thought!).

I have two problems with below circuit:

  1. The output is oscillating and is unstable, with C1 and C2 I managed to make it stable but then again it goes unstable under load.
  • How can I make the output stable with load or no load?
  1. On the breadboard when I short the ADJ pin (which also shorts the output of integrator) to simulate Q1 activating and current limiting mode, the current draw is tens of mA depending on R4 value (increasing R4 to higher values causes the circuit to go out of regulation). while ADJ pin current is in uA region according to the datasheet.
  • That amount of current draw in this circuit is unacceptable, how can I reduce ADJ pin/op-amp output current draw to 1 mA or less when the circuit goes to current limiting mode?


simulate this circuit – Schematic created using CircuitLab

I'm just a hobbyist, I'd appreciate it if your answer be simple and understandable for me, thank you.

  • \$\begingroup\$ If R5 is connected to the output, where is R6 connected to? R3 should go to the output, as a bleeder, not on the output of the opamp. Also, adding a series resistor with C1 might help; it adds a zero (not just a pole). \$\endgroup\$ Commented Nov 2, 2020 at 10:16
  • \$\begingroup\$ @aconcernedcitizen R6 is connected to the positive rail of PSU, it's a tracking PSU. for R3 refer to the datasheet for more info. I tried adding a 100 ohm resistor and other values but haven't had any luck... \$\endgroup\$ Commented Nov 2, 2020 at 10:21
  • \$\begingroup\$ You were right about R3, I thought it's a bleeder (i.e. to ensure current at the output). Can the 5532 work with +5/-30 V? The datasheet does say +/-20 V max, but that's 20 V around the ground point. As for the feedback try 1k...10k + 10n...1u. \$\endgroup\$ Commented Nov 2, 2020 at 12:28
  • \$\begingroup\$ @aconcernedcitizen +/-20 V is 40 V, the absolute maximum is 44V. I'm supplying 35V so it's working just fine. \$\endgroup\$ Commented Nov 2, 2020 at 14:22

1 Answer 1


I solved the problems with the help of a friend,

So the solution to the instability was because of the configuration of the integrator and also the capacitor which was needed for the ADJ pin.

I reduced the R4 value from 1K to 470 ohm and then placed C1 between ADJ pin and non-inverting input of the op-amp. that solved the oscillation problem with and without load.

For the second problem I added a buffer stage to the reference voltage, so Q1 only have to pull down the reference voltage which will draw way less current.


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.