# SCR crowbar circuit gate voltage?

I'm designing a simple overvoltage / reverse voltage protection circuit for a 24 V battery system, with plans to upgrade to a 48 V lithium setup in the future.

I selected the parts to trigger on a voltage slightly above 60 V. I haven't used SCRs before in a circuit, but from my understanding it will begin to amplify current similar to a BJT when the threshold voltage on the gate is exceeded.

Note: [Battery+ In] is the battery connection, [Battery+ Out] is the load side

simulate this circuit – Schematic created using CircuitLab

My concern is that the leakage current through D1 will place the gate voltage at at least 1.1V due to the forward drop of the 1N4007, and that leakage current will amplify and cause power loss through the SCR. Most SCRs I have been looking at have a threshold of 0.8-1.1 V, would choosing an SCR with a threshold voltage of 1.3-1.5 V be enough to counteract this leakage?

I'm not sure what parameters are important in the SCR. The PTC can have a hold current of 3 A, with a trip current of 7.5 A. The datasheet says conducting around 15 A will trigger it the quickest, so the SCR will need to be able to handle at least that. The +BATT input will be fed into 2 wide input range DC-DC converters to power other devices, each with filters on the inputs.

If there are any other problems with this circuit I'm glad to receive any feedback!

• A P channel FET should have its source connected to the incoming (higher) voltage that you want to switch on/off. Mar 29, 2021 at 20:57
• The reason I have the FET oriented that way is for reverse voltage protection, otherwise the body diode could conduct if the battery polarity was swapped. It should conduct briefly to bias the source side to put it into saturation, but I could be wrong. I don't often use P FETS Mar 29, 2021 at 21:15

My concern is that the leakage current through D1 will place the gate voltage at at least 1.1V due to the forward drop of the 1N4007

The leakage current of a 1N4759 zener diode is about 5 μA: -

So, this current will be insignificant for voltages of 48 volts and probably will be until the voltage is in the low 50s and this will probably be sufficient for crowbar protection.

that leakage current will amplify and cause power loss through the SCR

Miniscule power losses I would anticipate. It certainly won't cause the SCR to trigger until it rises to several mA typically so, check the SCR data sheet and find out what the typical trigger current is for that device.

Also note that the sqm40p10 PEMOS, like most MOSFETs, include an intrinsic body diode:

Wired the way you currently have it, a voltage spike could go straight through, via the body diode, to the load.

• If I swap the orientation I believe it adds a path for a reversed battery to conduct through the body diode instead, which is more of a concern to me. I have inductor / capacitor filters on the 2 switching regulators that this goes to, so I'm assuming they should handle small spikes. I haven't really considered large spikes due to the size of the batteries attached, is there an easy solution you would propose? Mar 29, 2021 at 21:18
• Typically in such a scenario, two PEMOS devices are placed back-to-back and both are switched at the same time. That way, power cannot flow either way through them while their gates are "off." Mar 29, 2021 at 22:03
• Note that if BatteryIn was, say, 12V and BatteryOut were 24V, connecting BatteryOut would try to push 24V into the 12V battery. That must never happen. Mar 29, 2021 at 22:05