Latching Relay That Follows +12VDC Trigger and Uses No Power Unless Switching

I have an application where I have a +12VDC circuit that I need to use as a trigger for a latching relay.

So, when +12VDC is present, the relay would SET. When +12VDC is not present, the relay would RESET.

However, I need to have no power consumption when the system is quiescent (i.e. power draw during switching is fine, but there can't be a power draw during non-switching times).

I've tried a few things, including wiring the latching relay coil through its own output contacts (so that the latching side only received a ground when the relay was in the unlatched state), however this doesn't provide enough time to latch the relay.

Any ideas would be greatly appreciated.

P.S. I'm using an Omron MY2K DC12, but happy to change to something different.

Thanks.

• Have you tried a bi-stable relay? Feb 2, 2020 at 19:21
• Yes - that's the Omron mentioned above. Not sure how to trigger this to un-latch on the loss of +12V trigger, and not sure how to stop power draw after the latch or un-latch event. The coil still pulls ~ 100mA of current. Feb 2, 2020 at 19:23
• "I need to have no power consumption when the system is quiescent" - No power consumption? A more more reasonable spec would have a maximum quiescent consumption. What is the maximum acceptable current draw when +12V is applied? Feb 3, 2020 at 6:50
• Their are two relays with the same part number !!! The OMRON MYK DC12 is NOT a latching relay. But this version is ! BUT some pages say 'latching' but the linked datasheets are non latching. You'll need to ensure that what you get IS a latching version. Feb 3, 2020 at 10:20
• Hi Russell. The version mentioned is the MY2K (not MYK), which is a bi-stable latching relay. Hope that helps. Feb 3, 2020 at 13:04

4 Answers

You need a capacitor large enough to hold enough energy to reset the relay when the +12 disappears. Minimum pulse width is stated as 30ms. Assuming a 1V drop would be okay and allowing for say 45ms time, you'd need a capacitor C = $$\\frac{t \cdot I}{\Delta V}\$$ or 0.092A * 0.045s/1V = 4,000uF. So maybe 4700uF/16V.

Then you need a control circuit to detect the power loss and pulse the set and reset coils. It may be possible to do an adequate job in your case with a few passive parts and a 2N6028 but if you need reliable operation in more pathological situations such as slow brown-out of the 12V line you might want to consider more sophisticated circuitry or even a microcontroller.

• OK - that's definitely helpful. I typically work in the industrial control / automation world, and I would normally just use a PLC and be done with it. Trying to do this on an automotive platform with limited energy capacity is a whole different beast. I'm not worried about voltage stability issues...the component feeding the +12V is actually a relay (? MOSFET) designed to drop power when the system voltage goes below 11.9V. Feb 2, 2020 at 20:37
• You could use an SCR instead of the PUT, just reverse the polarity of everything else. Feb 3, 2020 at 2:37

The circuit below does what you want. It

• Provides an ~= 100 ms latching pulse when ignition is turned on

• And similar delatching pulse when ignition is turned off.

At all other times the relay coils draw no current.

This is a "cut down" version of my circuit from this answer to "Power Latching Relay off/on in response to 12V ignition and 1 second ground pulse".

If there is a single ignition feed input then an overall power supply capacitor - charged via a diode (not shown) needs to provide enough energy to hold the IC on and to provide relay delatching energy. If an unswitched 12V feed is available then the circuit can be powered from that. After the turn-off action current drain is around 1 microamp.

I have left an additional a 1 second input pulse option to show how the relay may if desired follow a shorter pulse. The two inputs are essentially identical.

A "better" solution may be to use an Arduino or similar as more flexibility is possible if the spec "evolves", as happens. But, this should do the specified task.

Inverters are a package of hex Schmitt trigger inverters.
MUST be Schmitt trigger. (CD40106, 74c14, ...)
There are two spare inverters which can be used in various ways if desired.

D1 accepts 12V positive on ignition
D2 (optional input) accepts a 12V positive going alarm pulse.

R1C1 and R3C3 plus following inverters provide 1 second delays.
A negative going edge pulls the cap low and the resistor then charges it high.
When the alarm pulse activates C1R1 the output pulse lasts the lesser of the alarm pulse or C1Ra timeout. As the set/reset relay only needs a 10 ms pulse 'this is not a problem".

R1C1 & R3C3 both provide ~~= 100 ms time constant but values vary by a factor of 10 as examples of how different values can be used.

• both comments say "100 mS pulse" on the schematics. is that intentional, considering the resistors are different? Feb 3, 2020 at 8:25
• @Maple Good spotting. That was modified from my circuit here where I DID purposefully use two different Rs and Cs to provide a 1 second pulse. I changed the C'2 and missed the Rs here. The pulse length is not very critical . The relay spec is confused (see my comment below OP's question) but 15 mS pulse MAY be enough. If so the 100k/100nF may work OK, but, I'll change the upper 100 nf TO 1 Uf. Feb 3, 2020 at 10:24
• @RussellMcMahon Thank you - this is excellent! I did pull the specs for the relay and it lists the following (same for set and reset): Time: DC 15 ms max; Minimum Pulse Width: DC 30 ms. Also, there is an unswitched 12V feed available. I'm not quite sure I understand how this would be used though. Feb 3, 2020 at 13:12
• 1. Set the pulses to not exceed max ALLOWED on time. Be aware that spec sheets MAY also contain typical and "maximum worst case REQUIRED" on times The latter is shorter than max-on-allowed. 2. If there is a 12V continuous feed available then running the IC and relay coils from that removes the need for a "holdup" capacitor (that I and others have mentioned) to provide energy for the system after ignition is removed but while the release coil needs powering. | Drain from an always on supply is zero for the relay except when switching and around a microamp for the IC .... Feb 4, 2020 at 0:11
• ... (IC could be depowered once relay is off. ) Feb 4, 2020 at 0:11

I had the same idea as Spehro Pefhany use a capacitor to store energy and a thyristor to dump it through the relay

simulate this circuit – Schematic created using CircuitLab

I came up with this, but I'm a bit worried that it might trigger from a supply ripple and short-circuit the supply.

but this isn't going to work with the Omron MK12K because this relay seems to use a mechanical latching mechanism and the relay coils probably have the same action with reverse polarity.

simulate this circuit

• Latching relays almost invariable use two coild - one to latch and another to delatch. Some may have a single coild with bipolar drive for on/off. The one used here is confused by the datasheets not agreeing with the detail pages but it is almost certainly a 2 coil relay. Feb 4, 2020 at 0:13
• Single coil versions are available, and I also have some with a centre-tapped coil. Feb 4, 2020 at 2:19
• Yes. And/but: One needs to define how a relay works when providing a circuit to drive it. The originally posited relay has two coils - a set-on and reset-off. | A single coil latching relay may be pulse on, same polarity pulse off (eg toggle on pulse) or pulse on, opposite polarity pulse off. | A centre tapped relay is PROBABLY the same as a two coil relay (eg centre +ve, ground relevant end of could to set-on or reset-off. But whatever is being driven needs to be defined for the circuit to make sense. | Your circuit looks useful for a single coil opposite polarity set/reset system. Feb 5, 2020 at 2:17
• To make it worse the specified relay has two sets of data - one states two coil latching and the other is single coil non-latching - bith with same part number. | Here is the 2 coil. BUT This does not mention latching or two coils. latching - or so it explicitly says. | Feb 5, 2020 at 2:20
• Here is a MY2k datasheet that shows two coils, as he specified. Other pages with the same model name show a single coild non latching :-). infinite-electronic.hk/datasheet/f2-MY2N-CR-AC220-240.pdf Your circuit could be used with a two coil version by using a diode in series with each coil so you charge C1 by one coil and discharge it via the other. Or you could change the SCR discharge path - that may even elimnate the short-supply-on-glitch risk Feb 5, 2020 at 2:27

This may depend on the model of the relay. The relay I use (W30L) latches reliably while removing the latching power with its own contacts. All three instances I tested latch no problem even under reduced voltage (tested under 3V, and it is officially a 5 V relay). The manufacturer confirmed that the relay is capable of such an action, even if not mentioned in the official data sheet. They also have the 12 V model. It is a power relay but not very big. The coil of the 5 V version draws 200 mA.

Relay contacts are springs, and the anchor moves some distance before the closed contacts open. Depending on the construction, this may be sufficient to make the latching process irreversible without any capacitor assisting.

simulate this circuit – Schematic created using CircuitLab

You can simply use the manual switch on the top of this relay to reset it. You need the second power source or maybe supercapacitor to do this electrically (opposite polarity or another coil if present can move the anchor back). A bistable relay needs power for latching, you cannot achieve this by removing the voltage.