Connect with us

Two-coil latching relay arrangement

Discussion in 'General Electronics Discussion' started by TomHoward, Jun 22, 2013.

Scroll to continue with content
  1. TomHoward


    Jun 22, 2013
    Afternoon all

    I am looking to make a sort of two-coil latching relay arrangement which hopefully will hold its state without power. I can find impulse relays, which will change state based on input to a single coil and hold without power (pulse on, pulse off sort of thing.)

    What I have is two 240V rails with a common neutral, which are energised for short amounts of time (seconds not milliseconds), and I'd like to be able to tell which one was energised last via closed contact.

    I only need one closed contact based on one rail (A) being energised, and reliably cleared when the other rail (B) is energised, but unaffected if the same rail (A) is energised for a second time before B.
  2. KrisBlueNZ

    KrisBlueNZ Sadly passed away in 2015

    Nov 28, 2011
    Hi Tom!

    Welcome to the electronics point forums.

    Thanks for giving a thorough problem description up front. That saves a lot of time.

    If your "impulse" relays are latching relays (and they sound like the same thing), you can get them with single coils (bipolar drive), dual coils with common connection, and dual isolated coils. A latching relay sounds like exactly the thing you need.

    You mention two 240V rails with a common Neutral, so I assume you're talking about AC. In this case, you can use any of those three types of latching relays; the only difference being the circuitry that drives the coil from the 240VAC voltages. For simplicity and cleanness I would use one with two separate coils.

    Unfortunately it seems that latching relays are not made with 230VAC coils, so you need a way to reduce the voltage and convert it to DC: either transformers, or a capacitor dropper circuit. For a transformer circuit, the requirements are minimal and a web search will show you how it's done. But transformers are a bit bulky.

    For a capacitor dropper circuit, I would use a latching relay whose coil current requirements are as low as possible. This would mean a high coil voltage. A search on Digikey shows only one latching relay that's in stock and has 48VDC coils: so it's probably safer to use a relay with 24VDC coils.

    Digikey has 10 suitable types of dual coil latching relays with 24VDC coils in through-hole style (pins for mounting onto printed circuit boards or stripboard). They have coil currents ranging from 6.3 mA to 15 mA. That range is well-suited to a capacitor dropper power circuit.

    Here are all the suitable relays I found on Digikey, in order of increasing price. Any of these would be suitable. Most of them are a standard size but there are a few that I've marked that have different form factors.

    Omron G6SK-2DC24 DC24/Z2664-ND/369275 USD 4.75 coils 12.5 mA
    Panasonic TQ2-L2-24V USD 4.92 coils 12.5 mA wide, low height
    Panasonic TX2-L2-24V USD 5.12 coils 8.3 mA
    Panasonic TXS2-L2-24V USD 5.66 coils 6.3 mA
    Panasonic DS1E-ML2-DC24V USD 6.30 coils 15 mA
    Panasonic DS2E-SL2-DC24V USD 6.37 coils 7.5 mA
    Omron G6EK-134P-ST-US-DC24 USD 7.41 coils 8.3 mA
    Omron G6AK-234P-ST-US-DC24 USD 8.05 coils 7.5 mA low height
    Omron G6AK-274P-ST-US-DC24 USD 8.05 coils 7.5 mA low height
    Panasonic DS1E-SL2-DC24V USD 8.58 coils 7.5 mA

    Here's a schematic and simulation of a suitable circuit. The relay coil is replaced by a resistor with a value from 3810 ohms (to simulate a 6.3 mA coil) down to 1600 ohms (to simulate a 15 mA coil).

    You will need two of these circuits, one for each mains supply. Each circuit drives one coil of your latching relay.


    C1 is the main component in the circuit. Its capacitive reactance allows it to drop the 230VAC voltage down to a much lower voltage without actually dissipating any significant power as heat. This is definitely White Man's Magic, not quite up there with quantum physics and string theory, but still Pretty Darn Cool (TM).

    C1 must be a specific grade of capacitor, for reasons of reliability and safety (fire risk mainly). Here's a suitable part:

    R1 is a fusible 100 ohm resistor such as and is designed to fail quickly if C1 fails for some reason.

    R2 is a "bleeder" resistor that is intended to discharge C2 when the circuit is disconnected, for safety.

    D1~4 should be 1A silicon rectifier diodes such as the 1N400x series. These are polarised components; the plate (not the triangle) is called the cathode, and corresponds to the stripe on the component body. These are available with voltage ratings from 50V (1N4001) to 1000V (1N4007). Although they are connected to the mains, in this application they never see more than about 30V across them, so you can use any member of that family. Here is a 1N4007 made by Fairchild:

    Alternatively you can replace D1~4 with a 1A bridge rectifier such as a DF04M to save a bit of board space:

    D5 and D6 are shown as two separate diodes, but that's only because the circuit simulation software I'm using (LTSpice) doesn't have the exact part I want. They should actually be replaced with a single diode, a 1N5361B zener diode, which has a forward voltage of 27V and a power rating of 5W. It will not dissipate 5W in the circuit, but I chose a big grunty zener in case of failure of C1. The cost difference compared to a smaller zener diode is minimal.

    C2 is a 10 uF, 50V electrolytic capacitor such as It's a polarised component; the top connection in the diagram (the hollow plate) is positive. This corresponds to the long wire on the electrolytic; the polarity is also marked on the body of the electrolytic itself.

    The relay coil is represented by a resistor on the schematic.

    The RAVG and CAVG components are not part of the circuit; they are used in the simulation so I can see the average (mean) coil voltage.

    The combination of C1, D1~4, the zener diode and C2 produce a DC voltage that is limited to about 27V. The traces at the top of the image show the coil voltage (the red trace) and the average voltage (green trace). The horizontal axis represents time, in seconds.

    The mains supply is turned on at 0.2 seconds, and you can see that the coil voltage rises rapidly. It stays between about 23V and 28V because of the action of C2, a smoothing or reservoir capacitor, which holds the voltage up during the times when the mains voltage is not changing quickly and cannot provide much current through C1. The average voltage is about 27V.

    The mains voltage turns off at 1.0 seconds and you can see that the coil voltage drops quickly. Even with the 6.3 mA relay coil, which causes the slowest voltage drop, the voltage to the relay coil will drop below its specified minimum voltage in less than 0.1 seconds.

    I chose component values very conservatively for this circuit. The mains voltage, though nominally 230V (or perhaps 240V where you live) has a certain tolerance. I like to make sure that a mains-powered circuit will work properly from an input voltage as low as 210 VAC. Also, C1 will slowly lose its capacitance over time; this type of capacitor (metallised film) is designed to do this when a mains voltage surge occurs. As the capacitance reduces, the current available to the circuit drops as well. The relay coils are rated for 24V DC but this is only a nominal rating; they are specified with a maximum voltage at least 20% higher than this. Finally, these values are chosen so that you can use any of the relays I listed. They have coil currents between 6.3 mA and 15 mA which is quite a significant range. This may be useful in future if you have to replace the relay and the original one you used is no longer available.

    The latching relay has two independent coils; each one must be driven by a circuit like this. I would keep the two circuits fully isolated from each other all the way to the relay.


    All of the circuitry in this design, apart from the relay contacts, is either live or half-live and is an electric shock risk. Never touch any part of the circuit while it's connected to the mains. Once you have constructed and tested it, make sure it is mounted in such a way that it cannot come into contact with anything conductive. Ideally, full encapsulation is the best option. The only connections to it should be the mains inputs (insulated wires) and the relay contact connections.

    If you're not confident constructing mains-voltage circuitry, get someone who can advise you and check your work. If you can't, don't use this design, and if you do, and something goes wrong, don't try to make me responsible. I cannot be there to check that you've done everything properly and safely, so I cannot be responsible for anything that may go wrong.

    Attached Files:

    Last edited: Jun 23, 2013
  3. TomHoward


    Jun 22, 2013
    Thanks Kris

    It's UK mains supply voltage so technically 230V + 10%.

    As you say I can find DC latching relays with dual coils - either with a common connection or reverse polarity - but I'm really struggling with 230/240VAC,

    After some more research I've found Omron MK2KP-UA-AC240, which is dual isolated coils, and Finder 13.11 - which is a common dual 240VAC relay designed for Set & Reset in emergency call systems - the type you find in sheltered/warden housing or in disabled toilets.

    However the Omron is over £100 and I haven't got a price on the Finder yet. However for the cost of the components and a small 24V relay it seems sensible that I could construct your circuit.
  4. KrisBlueNZ

    KrisBlueNZ Sadly passed away in 2015

    Nov 28, 2011
    Right. I think that circuit is your best option.
  5. duke37


    Jan 9, 2011
    Hi Kris,

    Would the circuit work without the bridge rectifier (D1 to D4) and a diode between D6 and C2. This would keep the 'ground' line at neutral and so safer. C2 may need to be a bit bigger but relays are inductive and can stand some ripple.

  6. KrisBlueNZ

    KrisBlueNZ Sadly passed away in 2015

    Nov 28, 2011
    Yes Duke, that's a discussion we had on an old thread:

    I prefer the bridge design because both halves of the mains cycle are used, so C1 can be half the size of the other option. This saves more money and space than you lose with the extra three diodes. With the arrangement you suggested, about half the current that's available from C1 is wasted when the zener conducts in its forward direction.

    As for isolation, my understanding is that both Phase and Neutral should be considered live, and must be fully isolated, so the advantage is not clear. That would apply especially if the mains plug is reversible, which I think used to be standard in America.

    Yes, C2 would need to be increased as well, but it's C1 that's the main reason I prefer that design.

    To the OP, please feel free to follow the discussion on the thread I linked to. The design he has is also valid and may have a safety advantage over my bridge-rectifier-based design.
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day