First ever project looking for efficiency improvements and guidance

Lots of really good points, above. But I’ve noticed a couple more things that I didn’t read being spelled out (though I could have missed them while scanning through this thread).

  1. It looks to me like the D2 flyback diode across the coil of K1 is backwards. You want that diode reverse biased in normal operation or the relay coil will only see the breakdown voltage of the diode (about 0.7V). With the diode reverse biased (anode to ground, cathode to the +V signal) then it will act as an open circuit during normal operation, but when the relay coil is de-energized the back EMF pulse will be shorted out by the diode and not get injected into the rest of the circuit.
  2. You say that the signals that you want to switch by the relay are 230Vac. But according to the datasheet that @ForrestErickson posted the maximum switching voltage for the contacts in that relay is 125Vac. This is the wrong relay for switching 230Vac.

Another thing to consider with relays is how the flyback diode affects the release time of the relay contacts, and as a consequence of that the life expectancy of the relay contacts.

One way to improve the situation is to use a driver transistor with maximum voltage much higher than the supply voltage, and then put a zener diode or transient suppression diode of 2x supply voltage (or more, depending on what the transistor can take) parallell to the relay coil. Either use a dual polarity transient suppressor or put i in series with the flyback diode to increase the voltage in the conducting direction.
Note that the transistor will see the sum of supply voltage + transient suppression voltage + additional margin, so choose the transistor at least 1,5x to 2x (supply+transient).

Another method is to use an Avalanche rated MOSfet. Quite a lot of MOSfets have ratings for avalanche current in their datasheet.

Also, regular BJT’s do not suddenly explode when their voltage rating is exceeded, but they will also go into avalanche mode. (And effectively act as “zener” diodes) at a voltage which is typically some 20% to 50% above the voltage stated in the datasheet. I do not know if (some) BJT’s are explicity tested and rated for avalanche conditions though.

But the main thing here remains:
230Vac on a PCB is a harsh learning environment, and if OP’s experience is not enough to catch blatant errors in a simple relay circuit, then working with 230Vac on a PCB is a very bad combination.

1 Like

Diode definitely backwards for spikes

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.