# Discrete negative voltage regulator oscillating under load

I'm trying to save a few volts voltage drop and use a discrete voltage regulator, having a feedback doesn't really matter because the current draw is very low.

In the diagram below, if I remove C2 then under load the output starts to oscillate at ~50 Hz and to around 100 Hz at higher currents.

simulate this circuit – Schematic created using CircuitLab

It's not related to the gain of transistor. I tried a TIP127 which is a Darlington with minimum gain of 1000.

Why is the output oscillating under load without C2?

It's not oscillating

What you are seeing is ripple from C1 causing excessive variation of the current through D1 and varying its voltage.

That voltage then directly affects the output voltage.

You could make the filtering of the voltage to the zener more effective by splitting R1 into two resistors of 1k each and putting C2 at that point to ground.

A filter capacitor directly across the zener is not very effective because of the low dynamic resistance of the zener - it acts like a resistor of only a few ohms so needs a very large capacitor to get much filtering.

A current source (as you mention in the comments) is also a way of improving the design but much more complex.

How much ripple do you see at the output? How much ripple at C1?

Post a scope shot if possible.

• Ripple is ~20mVPP and C1 ripple is ~80mVPP at ~10mA, higher current is anything higher than 10mA (like 100mA), How do I filter that ripple without adding a big capacitor at the base of the transistor? Feb 1, 2021 at 20:38
• I added a constant current source for D1 and that fixed the problem, thank you. Feb 1, 2021 at 20:50
• C2 in this circuit forms a gyrator which reduces the capacitance needed and helps filter the output. I would change C1 to several hundred microfarads. that will lower the input impedance and also add high frequency bypassing. You could try swapping C1 and C2 and see what happens.
– Gil
Feb 1, 2021 at 21:28