How long should I drive the coil of a latching relay?

Question

I have inherited a project that uses HFD2/005-S-L1-D single coil latching relays. The datasheet is here.

It lists 4 parameters which may be what I want:

• Operate time: 4.5ms
• Release time: 3.5ms
• Set time(latching): 4.5ms
• Reset time(latching): 4.5ms

I believe the 'Operate' and 'Release' times are for the non-latching versions of this relay. That leaves the 4.5ms 'Set' and 'Reset' time.

I have it on decent authority(one of my predecessors actually did Science!) the debounce time of the contact is in the neighborhood of 4.5ms. This project has a bank of relays that needs to be switched in groups, due to the current limits of the microcontroller.

The debounce time can't be rushed, of course, but I'm hoping to save some time by only waiting the full debounce time at the end.

What I'm currently doing:

switch some relays
wait 4.5ms
switch some relays
wait 4.5ms
switch some relays
wait 4.5ms
return

What I hope I can do:

switch some relays
wait .5ms
switch some relays
wait .5ms
switch some relays
wait .5ms

wait 4ms
return

I have plenty of development units I can abuse, so I'm definitely going to run some tests to see what I can get away with, but I'd feel a lot better with even a little bit of Theory to support my Practice.

Any insight would be appreciated.

Answer 1

You must maintain the current through the coil for a minimum pulse width to guarantee the set/reset will reliably occur over the life of the relay, over temperature, over variations from relay to relay (with that mfr expect a lot) and voltage variations-- so (given that your limitation is the maximum number of relay coils energized at one time) I don't see any advantage to not switching the maximum number you can, waiting the minimum time with current flowing, then shut it off and proceed to the next lot.

I'm not sure your 'Hongfa' relay is actually fully specified. If you look at a similar type of Japanese relay (the Japanese usually have superb documentation) you will find that the set/release times are specified at 3ms but the minimum set-reset pulse width is specified as 10ms. I would actually give it a bit more than the minimum just to be safe. You might want to try to contact the manufacturer for a number, sometimes they will actually respond.

No doubt if you test it you will find the typical minimum pulse width and switching time under nominal conditions is much less, but I think it would be a mistake to depend upon that.

Then, after the final set/release begins you wait the operate time + debounce time for the contacts to be reliably seated, which is probably similar to the minimum pulse width if the relays are similar to the Omron part.