I have the following schematic, which would be used as a over-voltage and reverse polarity protection:


simulate this circuit – Schematic created using CircuitLab

Both D1 and D2 are rated at 15V+-5%, which should theoretically produce maximum breakdown voltage of 15.75V. Lab power supply is used as a voltage source. If I'm using jumper wire instead of fuse F1, then PSU just goes into current limiting mode, but crowbar part works as expected, but not at the desired voltage of 15V. I've tried it several times, both with and without load R3 (I'm using a heater as a load). The results are the same:

  1. Voltage at output drops to zero when input exceeds 16.9V. If PSU is set to 16.9V, same voltage is at output. When I add 0.1V more, crowbar fires and output drops to zero.
  2. I've measured voltage drop across D1, it exceeds 15V. If input is 16.9, then D1 voltage is about 16.3V, way too high compared to spec.
  3. I'm using second DMM to measure D1 voltage. When input is getting closer to 16.9V, sometimes when I touch D1 with probes, output voltage drops to zero, but crowbar part does't fire (PSU shows 16.9V at the output).

Questions are: how to explain what I see, and how to tune crowbar to fire at exactly 15V? I think I could use 13V Zener, but I don't like acting without clear understanding of what's going on.

  • 3
    \$\begingroup\$ Have you looked at the SCR trigger voltage between the gate and the cathode needed to fire the SCR? \$\endgroup\$
    – G36
    Oct 24 '19 at 15:02
  • \$\begingroup\$ Also, consider of using the TL431 instead of a Zener diode. See the example electro-tech-online.com/articles/… \$\endgroup\$
    – G36
    Oct 24 '19 at 15:10
  • \$\begingroup\$ the 100 ohm resistor may be way too low. Also, the Zener does NOT fire; there is no sharp turnon in a Zener. \$\endgroup\$ Oct 24 '19 at 15:21
  • \$\begingroup\$ @G36 SCR is VS-25TTS08S-M, datasheet says Vt is 1.25V, if I'm reading it correctly. Still have no idea how to explain what I see \$\endgroup\$
    – Zhenek
    Oct 24 '19 at 15:56
  • \$\begingroup\$ @analogsystemsrf how can I calculate proper value? \$\endgroup\$
    – Zhenek
    Oct 24 '19 at 15:58

The rated zener voltage for a BZX384 appears to be at 5mA, while you are using the device with 7mA through the resistor plus whatever current in needed to trigger the SCR.

Try increasing the resistance of R1, this should lower the trip voltage of your crowbar circuit.

  • \$\begingroup\$ How did you calculated 7 mA? \$\endgroup\$
    – Zhenek
    Oct 24 '19 at 16:18
  • \$\begingroup\$ I assumed that you needed 0.7V to trigger the SCR, and that voltage is developed across a 100 ohm resistor. It would have been helpful if you had provided a link to the SCR datasheet. If it needs 1.25V to trigger then you need to choose a zener with a lower voltage if you want to crowbar near 15V. It's impossible to trigger "exactly" at 15V so you must live with some error. \$\endgroup\$ Oct 24 '19 at 18:34
  • \$\begingroup\$ oh, now I see. Thanks! But I'm still a little bit confused. Datasheet says that triggering gate voltage is 2V (max). But many other sources state, that gate-cathode behaves much like diode, i.e. voltage drop is 0.7V \$\endgroup\$
    – Zhenek
    Oct 24 '19 at 18:40
  • 1
    \$\begingroup\$ @Zhenek Both are true, typical and max are rarely the same. You'll expect to see any voltage between those, with different devices, times of the day, temperatures, etc etc. Use a TLV431 as your voltage detector to trigger an SCR/Triac properly. \$\endgroup\$
    – Neil_UK
    Oct 24 '19 at 18:48
  • \$\begingroup\$ Welcome to the real world. Your SCR will trigger with some voltage less than 2V on the gate. Maybe 0.7V, maybe less, maybe more, and with variation due to temperature. Design accordingly. \$\endgroup\$ Oct 24 '19 at 18:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.