Solution Finder

Creation of a Negative Rail from One Positive Output

Posted by: scs2016 on 10/26/2021

Hello all.  I have since prototyped a circuit that allowed me to create a negative rail from one single, positive output.  The transformer was designed to produce only one positive output.  By rectifying the negative portion of the output switching waveform and filtering, then zener regulating, a low current (10 mA max) negative rail can be produced that is stable over load.  This allows the addition of op amps now that have dual rail supplies, etc....

I am looking for any feedback or thoughts on this approach.  One thing I did notice is that without the 22 ohm resistor, the supply will not start up.  Is this due to initial current spikes into the filter capacitor cause the controller to shut down?  If so why wouldn't this occur on the positive output as well?

The rectified unregulated voltage on the filter capacitor is around -22 VDC.  Ripple is within 10 mV under load (10 mA).





Submitted by PI-Jodec on 10/28/2021

Hi Shawn,

Thanks for using Powi products. I think this is what happened to your circuit. When the primary switch is on, some of the energy stored in the magnetizing inductance is transferred to the negative rail of your circuit when the 22ohms resistor is present. Removing the 22ohm resistor during startup will draw a large amount of current since you've shorted the output (Capacitor is shorted initially), this might draw all the energy from the primary so when the switch turns off, there is no more energy left to transfer to the secondary side of your circuit which causes your supply to fail at startup.

We can verify this by monitoring the control pin of the IC, If there is no external feedback/supply current fed into the CONTROL pin by the end of 17ms, the capacitor on the CONTROL pin discharges to 4.8V and enters protection mode.


PI- Jodec