# Getting a higher voltage from TL431 in a crowbar circuit

I want to protect a circuit from overvoltage using crowbar circuit, the input voltage shouldn't get above 50 V.

• TL431 output is 36 V max, can a zener diode in series like schematic below protect the TL431? (is the circuit below correct)

• Should I add the triac gate trigger voltage to the set voltage?

simulate this circuit – Schematic created using CircuitLab

BT139 datasheet

TL431 datasheet

• Do you really want to short out the supply abruptly or simply block it with current limiting softly such as a Zener and PTC . There is thermal protection already. Is this a real need for device failure protection? Commented Nov 16, 2020 at 18:25
• Yes a PTC is the general abbreviation for the brandname Polyfuse which may be used with any voltage limiting clamp by choosing the holding current threshold. They are n't intended to be precise parts so a temperature rise to 85'C is expected to increase the series resistance sharply to limit current determined by the voltage clamp. The power dissipation must be considered in all parts and your crowbar must be used if VI=P is too high to be reasonable. then the the PTC limits current and a power cycle to reset the SCR below holding current threshold. Commented Nov 16, 2020 at 20:00
• For a DC power driver only an SCR is needed which has a negative resistance and thus drops voltage to ~1V + IfRs and if the OV is removed and then current drops , it cools resistance drops a couple orders of magnitude to the rated R value and it operates normal but all maximum values must be observed and not exceeded Commented Nov 16, 2020 at 21:08
• Of course, you could always choose a better transistor with higher voltage rating, and then you don't want a transient to false trigger a crowbar, so you might add a 1ms RC filter Commented Nov 16, 2020 at 21:12
• @TonyStewartSunnyskyguyEE75 Thanks for all the explanations. Commented Nov 16, 2020 at 22:23

You want to 'set' the TL431 to your desired nominal crowbar voltage. Read the TL431 datasheet and look at the block diagram to understand why.

You also likely will want to reduce R3 considerably. The TL431 needs 1mA, so maybe a couple hundred ohms. Otherwise current will flow thorough the gate unnecessarily before it is supposed to trigger. Probably not enough to trigger it, even under corner conditions, but there's no real advantage.

Agree with Andy, set up a free simulator such as LTspice (you may have to add the models for the parts) and test this yourself, then ask questions if it does not make sense.

• "You want to 'set' the TL431 to your desired nominal crowbar voltage. Read the TL431 datasheet and look at the block diagram to understand why." I don't get it... Commented Nov 16, 2020 at 19:29
• What relationship can you divine between the cathode voltage and the op-amp inputs? Commented Nov 16, 2020 at 19:30
• Correct. So it is comparing the divider voltage directly to the reference of 2.495V nominal. Commented Nov 16, 2020 at 19:36
• Yes, to both and as I said about R3. Commented Nov 16, 2020 at 19:38
• The gate trigger voltage appears at the cathode. How does that influence the REF input? Not at all . Commented Nov 16, 2020 at 19:45

Just to add to the previous answers, a TRIAC usually needs a current pulse to fully turn on. You could add a PNP transistor at the output of the zener diode, which discharges a previously charged capacitor to the gate of the TRIAC. Here is a small simulation:

The cyan color depicts the current through the TRIAC.

• I couldn't find a single source mentioning this in crowbar circuits, thank you. Commented Nov 16, 2020 at 19:54