# Calculating resistor values for STEF01 overvoltage protection

EDIT : I was making a mistake and there isn't an issue in the data sheet (see answer below). Keeping here for reference to others.

The STEF01 is an electronic fuse by ST with features such as soft start, overcurrent protection, undervoltage protection, overvoltage protection: The overvoltage protection is configured by setting values for R1 and R2, but I can't understand how this is done correctly.

Page 9 of the data sheet has the following equation: (resistor values in k Ohms)

e.g. for a VClamp of 48V: R2 = 100k, R1 ~= 18128k However this doesn't match the values denoted in the graph: (the equation doesn't seem to match the graph).

Notes:

The circuit is something like this: Where RH, RL are R1, R2 respectively, and the divided voltage is compared to 1V.

Page 8 states that the internal divider current is fixed to 10 uA. This is a bit strange and might be the key to understanding this.

## 1 Answer

It looks like it matches the graph to me. It's a log-linear graph, and the pink line almost reaches the 20,000K line, which fits the 18,130K calculated.

The 10uA they are referring to is the current at the point of switching, not under any other conditions.

The circuit looks like this: simulate this circuit – Schematic created using CircuitLab

We know the current through R_B is 10uA at the point of switching. We know the current through R2 is 1V/R2 at the point of switching.

The voltage across both R1 and R_A is Vout - 1V.

So the current through R1||R_A must be (Vout - 1V)/(R1||R_A) = 0.01mA + 1V/R2

We know R_A is 2.7M so 1/R1 = 1/(R1||R_A) - 1/2700K

.. and the datasheet equation falls out easily.

• Yes - you are right. I was doing something wrong in my calculations elsewhere, and then posted the right calculation in the post above. Should I delete this question since it's not really useful to others? – Elliot Woods Feb 24 at 5:35
• I’ve explained the circuit in a way the datasheet fails to, so it might be of benefit to others. – Spehro Pefhany Feb 24 at 5:37