Any way to do an interlock with a latching relay?

[julian1]

The simplest way I can think of to do an interlock, is to break the coil circuit of a normally-open relay. The end-user must complete the circuit by making a connection to enable output.

The issue, with latching relay, is that it behaves like a flip/flop and can assume any state at power on.

But latching relays are attractive since they only consume current when switching, and there is less potential for magnetic and electrostatic issues.

[Kleinstein]

If the circuit is build right, one can make it in a way, that the relay state does not change at power on. So interlocks handled in software between latching relays should persist even after a power cycle.
In some cases one can read back the current state of the latching relays.

[julian1]

One strategy might be -

- place a diode in series with the latching coil - with polarity such that any signal power/glitch/surge can only turn the relay off.
- apply power with a timer at startup. To cover a change in relay state after last power-off, due to knocks/transport etc

- to enable the interlock, and allow digital signals to control the relay - just short the diode.

Edit . scratch that. As Kleinstein suggests, I think sensing the current relay state, using a pole of a double-pole Form-C, is a better starting point for working out the required transitions.

[julian1]

I think the following gets closer.
Using a nor flip-flop to represent a latching relay.  The current relay state is sensed (as Q ) which would correspond to a spare DT pole.

- If interlock-sense is not enabled (pullup R1 is left hi), and relay is sensed on, then turn off.
- If interlock-sense is enabled (R1 shorted to gnd), then drive relay according to ctrl signal and current relay state
- As additional measure, interlock has 2 extra pins to short diode D1. D1 prevents any glitch/failure from turning on the relay if interlock is not enabled.

It's a bit fiddly, maybe there is a simpler way?.