I'm making a circuit to alternate its input, depending on the voltage that's feeding the system. I used a voltage comparator with a transistor and a relay to do this:


The simulation works pretty well, but when doing it in real life, I have already burned 2 op-amps. I noticed the resistance between pins 3 and 4 got pretty low (from like, 1k to 20 ohms after the op-amps were burned).

I was trying to understand what the problem was, and by making some changes in the simulation (image below), I noticed that pin 7 is somehow with the same voltage of the BAT1. Like in the image below, the simulation starts with the relay contact already changed. I don't know if it is the problem, but it might have something to do with it. I can't understand what's happening.

When I tried to change the pin 7 connection to BAT1, it seems like it works. But my idea was to preserve the voltage from V1 (it would be a photovoltaic panel to feed the load, if there is no power, then the load wouldn't be fed).


Thanks in advance.

  • 2
    \$\begingroup\$ A small comment on the schematic. In general it is well laid out. The exception is Q1 which is drawn upside-down. The emitter should be on the ground side so that voltage decreases from the top of the schematic to the bottom. It would also avoid two wire-crossings. Putting Q1 right directly under the relay would make it even better. \$\endgroup\$ – Transistor Aug 30 '19 at 6:06
  • \$\begingroup\$ Thanks, I'll take care of it next time. \$\endgroup\$ – João Victor Aug 30 '19 at 13:12

The problem is the comparator has to have a supply voltage for all cases. So, V1 as well as Vbatt have to be connected to pin 7. You should use ORing diodes for this.

The reason that the comparator gets damaged is due to the internal clamping diodes. If there is no supply voltage connected to pin 7 and Vbatt is connected to pin 2, Vbat will burn these clamping diode and damage the comparator.
These clamping diodes are the reason you measure about Vbat at pin 7 in simulation. In simulation, components however don't burn/damage.

Speaking of using ORing diodes: why not replace the complete schematic and use 2 ORing diodes?

I don't have access to the circuit editor: do check this answer for an example of ORing diodes: https://electronics.stackexchange.com/a/453741/200815

| improve this answer | |
  • \$\begingroup\$ Just out of curiosity, are you saying that anytime you connect voltages to the input pins when the op amp has no supply voltage, it will get damaged? \$\endgroup\$ – David777 Aug 30 '19 at 7:02
  • \$\begingroup\$ Thanks for the reply Huisman. I live in Brazil and apparently there's no such thing as ORing diodes here, not that i know of. I did some research and all i could find was in english/spanish. This projects needs to be completed quickly, so i was thinking... Can I do it with normal diodes? like the 1n4007 for example. If I cannot do it, i think i will just conect the 7th pin to the BAT1 and remove it's connection with V1 . \$\endgroup\$ – João Victor Aug 30 '19 at 13:32
  • \$\begingroup\$ ORing is the circuit topography, any diode will work. Schottky diodes have a smaller Vf and therefore less power loss. If that's even too much, ideal diodes or other active power switching ICs would work. \$\endgroup\$ – Cristobol Polychronopolis Aug 30 '19 at 13:49

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.