Power Supply DC Input Simple Over Voltage Protection

[Jacques St-Pierre]

Hello,

For a long time, I have been using a simple zener (or TVS) paralleled to the input combine with a simple PTC Resettable Fuses in series with it to protect the input of a small DC-DC power supply. That DC-DC is use to provide a 5Vdc output at less than one amp. The input is 8 to 14Vdc and the goal was to prevent reverse polarity input or voltage over 15Vdc to destroy the circuit. It works, but most if the time at the sacrifice of the zener, that blowup short with over voltage. So we have to replace the zener, but the rest of the circuit is safe.

Does someone have any simple idea to prevent the zener from shorting each time?

Bye
Jacques

 

[prestonee]

Have you looked into using a crowbar circuit? It uses a Zener as well though and sounds similar to what you mentioned. but you might verify that.
-Pb

- - - Updated - - -

The crow bar works by shorting the output to ground via SCR, this will prevent output circuitry from seeing over voltage and at the same time cause large current on the input triggering your resettable fuse. This configuration I believe saves the zener, and puts the current sinking burden on the scr (scrs are made for this use).
-Pb

 

[Jacques St-Pierre]

Yes it may one way, but to what I saw, a crow bar will use a large power resistor to limit inrush current to the SCR and prevent it from blowing up. Sadly, my design do not have the space to use a large resitor maybe a small one will do it. But on the same line of idea, what do you think of adding of simply adding some resitor in series with the zener to limit inrush the time the PTC react?

 

[prestonee]

The resistor will affect your overvoltage threshold I believe. I would need to see a drawing I think.
-Pb

 

[Jacques St-Pierre]

Here a print screen of the supply input.

 

[prestonee]

before the zener kicks in it will see ground on the node between it and the proposed resistor, so the vin will be across it. once vin goes too high, the zener will kick in, it will then immediately see the node between it and the resistor shoot up like a voltage divider, this will cause the zener to kick back off. so you would need to verify the fuse would trigger before the zener disengages.
-Pb

 

[BradtheRad]

In case this might suit your purpose, here is a shunt regulator.



The simulation shows the power supply ranging between 15 and 20 V. You would need to size the transistor to withstand whatever is the anticipated maximum.

The 1 ohm resistor represents some amount of resistance in your power supply.

The diode is to prevent backward current if the supply were to be hooked up in reverse polarity.

 

[prestonee]

Basically you are still building a crowbar but using the bjt instead of an scr. the affect is a more gently current sinking rather then the scr turning on hard like a switch, wouldnt this cause a draw back of the bjt not responding fast enough to a sudden voltage spike?
I admit power electronics is not my field so i could be wrong in my conclusion.
-Pb

 

[Jacques St-Pierre]

I think the input enter some sort of oscillation; when the voltage will be high enough to turn the zener ON, the BJT will turn on and the PTC will open, so the voltage at the BJT will fall down and the current drain will stop flowing. But now with no current flowing, the PTC will reset, and the cycle will probably repeat endlessly until the input voltage will drop under the maximum.

What about replacing the BJT by the SCR?

The result may be the same, but with a more brut force.

 

[prestonee]

the idea is he needs the fuse to trip before the 'reset' occurs

 

[picgak]

Hi Jacques St-Pierre
use TL431(variable zener) along with a Transistor insted of ZENER it can handle upto 36v

 

[Jacques St-Pierre]

I may be wrong, but I think, any crowbar method may end up oscillating because of the presence of the PTC.
I will test it and let you know of the result.