Precision 5V0 clamp circuit ?

[chipeater]

Hello,

I would like to protect the output of a LDO from transient.
But I would like to protect it with something else than a TVS to clamp at a specific voltage like 5V.
Do you know a way to do this?

Thank you

[Benta]

Well, there are several ways to deal with transients.
One is to dissipate it (as heat) which a TVS will do.
Aother is to deflect it to somewhere harmless. The easiest way of doing that is a reverse diode over the LDO, sending the transient to the input storage capacitor.
But "precision"? No way. Why?

[Zero999]

The output and whatever is connected to it, will be able to handle slight over voltage, for short lengths of time.

The TL431 can be used as a precision clamp, but it's much slower than a TVS and probably won't be able to handle a much energy. I suppose it can be added to a 5.6V zener diode, to handle longer pulses.

[Doctorandus_P]

Why no TVS? They are made for this task and they are available in lots of sizes (peak dissipation) and voltages. Why would you want more precision? Having a precision device (<100mV ? tolerance) for the same speed as a TVS would be difficult the achieve..

[David Hess]

A fast clamp comes down to holding the base of a PNP bipolar transistor at about 4.4 volts, connecting the emitter to the 5 volt output, and the collector to ground.  Then if the 5 volt supply increases, the extra current gets diverted through the emitter to ground.

[Kim Christensen]

For very brief transients, a passive lowpass filter (Or just a capacitor) on the output of the LDO will work. The ESR of the lowpass filter will degrade the effective output regulation a bit, but maybe that's not an issue for your application. Otherwise a "reverse diode over the LDO" like Benta mentioned is usually good enough and should be added for longer "transients" anyway.

[dietert1]

One could use a large capacitor precharged to 4.7 V and a schottky diode to clamp transients. If the circuit is also meant to be a DC clamp the cap needs a strong enough shunt regulator.

Regards, Dieter

[bdunham7]

I would like to protect the output of a LDO from transient.

Exactly what sort of transient are you looking to protect against?  Where would it come from and what would it look like?  If you are protecting whatever is powered by the LDO (as opposed to the LDO itself) how much voltage are you willing to tolerate and for how long?

[srb1954]

Hello,

I would like to protect the output of a LDO from transient.
But I would like to protect it with something else than a TVS to clamp at a specific voltage like 5V.
Do you know a way to do this?

Thank you

Use an overvoltage crowbar sensing circuit like the MC3423/3523. The MC3423 has a relatively accurate sensing voltage and is quite fast acting. It is originally intended for triggering an SCR to clamp an overvoltage but could also be used to turn on a large transistor or MOSFET to temporarily divert the surge energy away from your LDO output.

Unfortunately the MC3423 is now obsolete but you can still pick them up on eBay. There may similar chips available from other manufacturers.

[golden_labels]

chipeater, could you draw that situation to use an post it as an attachment to the post? Are you protecting downstream from transients from LDO regulator⁽¹⁾ or the regulator’s output from transients caused by downstream? Because for the latter you can use a diode across the regulator: power sources can usually accept overloads, which would evaporate anything you can put on the output side. Why is precision needed? Could you quantify that precision?

⁽¹⁾ What would be their cause, though? The very purpose of a regulator is to make the voltage constant. Transients are present, but greatly attenuated or extremely short.

[Doctorandus_P]

This has the smell of an:

https://en.wikipedia.org/wiki/XY_problem