Relays (DC): 99.9% Less Power & Latching Option




About: Exploring the cosmos one synapse fire and one mouse click at a time.

Relay switching is a fundamental element of electrical control systems. Dating back to at least 1833, early electromagnetic relays were developed for telegraphy systems. Prior to the invention of vacuum tubes, and later semiconductors, relays were used as amplifiers. That is, when converting low power signals into higher power signals, or when remote load switching was beneficial or necessary, relays were the state-of-the-art option. Telegraph stations were linked by miles of copper wire. Electrical resistance in those conductors limited the distance the signal could be communicated. Relays allowed the signal to be amplified or "repeated" along the way. This is because wherever a relay was connected, another power source could be injected, boosting the signal enough to send it further down the line.

Electromagnetic relay switching may no longer be state-of-the-art technology, however, it is still widely used in industrial control, and where true galvanic isolated switching is desired or required. Solid-state relays, the second of the two primary categories of relay switch, have some advantages over electromagnetic relays. SSR's can be more compact, more power efficient, cycled faster, and they have no moving parts.

The purpose of this article, is to show a simple method to increase the power efficiency, and functionality, of standard DC actuated electromagnetic relay switches.

Go To Build Instructions

Step 1: The 3 Common Electromagnetic Relay Types

1. Standard Non-Latching (monostable):

  • Single coil of magnet wire surrounding a core of low magnetic permeability (only magnetized when coil is energized).
  • Switch armature held in it's stable state (not pulled in) by a spring.
  • Requires a DC voltage to be applied to the coil, in either polarity, to pull in the switch armature.
  • Requires a continuous current to temporarily magnetize the pole piece on the armature and hold this state.
  • More current is required to pull the armature in than is required to hold it in.

Uses: General purpose.

2. Latching (bistable):

Single Coil Type:

  • Single coil of magnet wire surrounding a semi-magnetically permeable core (remains lightly magnetized).
  • Switch armature held in unlatched state (not pulled in) by a spring.
  • Requires only a short pulse of DC power to be applied to the coil, in one polarity, to pull in and magnetically latch the switch armature in this state.
  • Requires only a short reverse polarity pulse to be applied to the coil to unlatch.

Dual Coil Type:

  • Two coils of magnet wire surrounding a semi-magnetically permeable core (remains lightly magnetized).
  • Switch armature held in unlatched state (not pulled in) by a spring.
  • Requires only a short pulse of DC power to be applied to one coil, in one polarity, to pull in and magnetically latch the switch armature in this state
  • Requires only a short pulse of DC power to be applied to the second coil, in one polarity, to unlatch.

Uses: Outside of industrial control, mostly used for RF and audio signal switching.

3. Reed Type:

  • Single coil of magnet wire surrounding a core of low magnetic permeability (only magnetized when coil is energized).
  • Closely spaced spring metal contacts hermetically sealed in a glass tube (reed).
  • Reed is positioned close to the coil.
  • Contacts are held in the stable state by their spring tension.
  • Requires a DC voltage to be applied to the coil, in either polarity, to pull the contacts opened or closed.
  • Requires a continuous current to magnetically hold the contacts in the non-stable state.

Uses: Almost exclusively used for small signal switching.

Step 2: Pros & Cons of the 3 Types

1. Standard Non-Latching (monostable):


  • Usually the most readily available.
  • Almost always the lowest priced option.
  • Versatile and reliable.
  • No driver circuitry required.


  • Not power efficient when conventionally driven.
  • Produce heat when energized for long duration's.
  • Noisy when switching.

2. Latching (bistable):


  • Power efficient, sometimes more so than SSR's.
  • Once actuated, hold either state even when no power is present.


  • Less readily available than standard relays.
  • Almost always priced higher than standard relays.
  • Usually fewer switch configuration options compared to standard relays.
  • Require driver circuitry.

3. Reed:


  • Usually the most compact of the 3 types.


  • More specialized, less available, fewer options.

Step 3: Squeeze That Juice Like a Miser

A conventional way to reduce the holding current of a standard relay, is to connect the coil through a series resistor with a large value electrolytic capacitor paralleled with the resistor. Most non-latching relays only need about 2/3 (or less) of the actuation current to hold state.

When power is applied, a surge of current sufficient to actuate the relay, flows through the coil as the capacitor charges.

Once the capacitor is charged, a holding current is limited by and supplied through the paralleled resistor.

Step 4: Maximize Your Miserly Mischief

With a carefully placed neodymium magnet and a little circuitry, the holding current of a DC relay can be reduced by greater than 99.9%. This technique will work with many... mini relays, but not all.

Before Proceeding:

  • Soldering skills and basic electronics knowledge are needed to complete this project.
  • Circuit supply voltage range is 5VDC to 24VDC.
  • Due to voltage drops and reactances in the circuit, choose a relay with a coil voltage rating approximately 25% lower than the supply of your circuit. For example, with a 12V supply: 12V - 25% = 9V.



  • Soldering Iron & Solder
  • Multi-Meter
  • Hot Glue & Gun

The housing of the relay shown was cut open for demonstration purposes, but shouldn't be necessary in practice. In most cases, placing the magnet directly above the switch armature on the top of the plastic housing as shown will work. It may be necessary to try magnets of different size and strength. As well as trying thin plastic or cardboard shims underneath the magnet, to space it further away from the armature.

Magnet Placement:

  1. Place the magnet as shown in Figure 1.
  2. Connect one multi-meter probe to the common switch terminal. Connect the other probe to the normally open switch terminal. Set the multi-meter to the maximum ohms scale, or continuity test mode.
  3. Apply power to the coil (in either polarity).
  4. If the meter shows all zeros, or the buzzer sounds, remove power from the relay coil.
  5. If the meter continues to indicate the switch is holding closed, secure the magnet in this position with tape or hot glue.
  6. If not, reverse polarity to the coil. If the switch closes, remove power, if it holds, secure the magnet in place.
  7. If the relay doesn't latch in either polarity, try adjusting the magnet position, or placing a thin non-magnetic shim underneath the magnet.
  8. Continue to repeat these steps until the switch actuates, and remains latched when power is removed. Then fix the magnet in place.
  9. Mark the power supply polarity that works on the relay body.
  10. Reverse polarity to the coil again and confirm that the relay unlatches. If not, further adjustment may be required.

Adding the magnet in this way, causes the relay to function as a latching relay. The following circuit, allows the relay to behave like a standard non-latching relay, but with over a 99.9% reduction in holding current.

Building the Circuit:

Solder the components together as shown in the sequence of photos at the top of this section. Be sure to match the circuit polarity to the polarity previously marked on the relay body. If the orientation is inverted to whats shown in the photos, the magnet can be flipped over.

Circuit Functionality:

When power is applied, the relay coil is briefly energized as capacitor C1 charges. Simultaneously, the base of Q2 is reverse biased because it is tied to the supply ground. This holds Q2 and Q1 OFF, and also allows supply current to pass through R1.

After C1 charges, virtually no current flows through the relay coil, accept for a tiny capacitor leakage current, typically dropping to less than 1μA. The relay holds state because the neodymium magnet has latched it. As long as the circuit is connected to the supply, the base of Q2 is held reverse biased by the current passing through R1. This is how to calculate that current: If the supply is 12V, then 12/2,000,000Ω (R1 2MΩ) = .000006 A or 6μA.

When the supply is disconnected, the base of Q2 is forward biased through R1 by the charge stored in capacitor C1. This turns ON Q2 and Q1, allowing the charge stored in C1 to be rapidly discharged through the relay coil reverse polarity, which unlatches it. D1 prevents Q2 from becoming reverse biased by the discharging of C1.

Step 5: Latching Option

With a few additional components, the circuit can be configured to allow selecting between non-latching and latching modes.

Additional Components:

Q4 is the latching switch, and Q3 is the unlatching switch. The B connection only needs to be pulsed to ground momentarily for the relay to latch. The A connection only needs to be pulsed to the supply positive momentarily for the relay to unlatch. When S1 and S2 are closed, the circuit is in non-latching mode. When S1 and S2 are open the circuit is in latching mode. In latching mode, a micro-controller on a shared supply can be used to control this circuit, by providing the momentary pulses to the A and B connections.

Electronics Tips & Tricks Challenge

Second Prize in the
Electronics Tips & Tricks Challenge



    • Plastics Contest

      Plastics Contest
    • Make it Glow Contest 2018

      Make it Glow Contest 2018
    • Optics Contest

      Optics Contest

    26 Discussions


    7 days ago

    This is basically the make up of a tattoo machine, 1 or 2 coils. I tried using a little solenoid switch and building it into a tattoo gun/machine. This is without any other components. It worked as the switch was intended to i suppose, the armature bar was pulled down just enough to break the circuit then reconnected. I need for it come down roughly a 1/4 in. What do i need to do? Add another component, resister,, etc...??? Or wrap the coil with many more turns to strengthen the magnet???

    1 reply

    Reply 5 days ago

    I don't have any suggestions to offer about making a tattooing gun.


    Question 13 days ago

    Do you have suggested values for the capacitor and resistor in the very first circuit?

    4 more answers

    Answer 13 days ago

    Here's an example of how to do a rough calculation for the resistor value. For a 12V relay with a coil resistance of 280 ohms, 12/280 = .043. So the coil will draw about 43mA. 2/3 of which is about 30mA. 12/.03 = 400, so the nearest standard resistor value is 390 ohms. A 1/4W resistor should be okay, but if it gets hot use one rated at 1/2W. A 1000 microfarad capacitor should work for most applications. Make sure to select a capacitor with a voltage rating slightly above your supply. Ultimately you will probably need to try different values until you find what works for your application.


    Reply 12 days ago

    You can actually do a better calculation by looking at the relay's data sheet. They all specify an an operate voltage and release voltage (sometimes current is given instead) so if you look at the data-sheet for OSA-SS-205DM3,000 for example, which has a 5v, 47 ohm coil, you see the operate voltage is actually 3.75v and the release voltage is 0.25v. So your resistor only needs to allow the coil to be over 0.25v! I'd go for 1v to be safe, so assuming you have a 5v supply you get 1/47=0.0213, 5-1=4, 4/0.0213=187.79, so nearest value is 180 ohms. The resistor power is given by V^2/R, V is still very close to 4, so 4*4/180 gives 0.09W, so only a tiny sized resistor is needed . To get the capacitor (without doing really heavy maths) you need to know the pull in time of the relay, which you get from a graph on the data-sheet. At full power (532mW) this is 7mS, but since the pull in current is going to be falling, allow a bit more, the coil power at 3.75v is 3.75x3.75/47=300mW, which gives 9mS. You can judge how much the capacitor can charge by going by the operate voltage - don't go below this value. Use an on-line calculator to determine what capacitor will charge to 5-3.75=1.25v in 9mS through a 47 ohm resistor (the coil), to save doing heavy maths: and you get 665uF. Nearest higher standard value is 680uF.
    Of course none of this takes into account the inductance of the coil but that's not given. I assume 9mS is too long a time for it to be important.
    Whew! I never did that before - hope it's right!
    Thanks once again for a really useful instructable, I would never have found out the reason for discrepancy in relay coil voltage and operate voltage without you prompting me to do this!


    Reply 12 days ago

    I agree completely that it's always best to refer to data sheets (when possible), and do precise calculations to determine the optimal component values. I'm sure your more exacting explanation will be helpful to many. Thanks for taking time to put it together.


    Tip 12 days ago

    The world-at-large consideration of a relay of course is its
    real world application. One could have a relay just to hear the sharp
    "click" when it operates, but usually there are wires connected to
    the contacts that do something, and what that "something" is what
    defines the relay.

    Diagrammed below is the EXISTENTIAL relay. The coil circuit
    is routed through a normally-open contact so that the contacts never operate.
    Can't get any more existential than that!

    Should one force the contacts closed by hand then the relay
    will stay energized as long as voltage is supplied, thus reinforcing the
    existential quality.

    Existential Relay2.bmp
    1 reply

    Reply 12 days ago

    Is that a twinge of sarcasm I detect? If the purpose of this article were to demonstrate what is being switched, rather than the method of switching, said load would certainly have been the focus, and included in the diagrams. I'll leave judgement about the "real world" validity of what I've shared up to each individual creative minded reader, rather than the "world-at-large". Thanks for sharing your existential "tip" though.


    13 days ago

    Useful and educational instructable. Now I know why a relay station is called a relay station!
    I don't think I will go so far as using the whole circuit with the magnet, but I will definitely use the trick with the capacitor in future!
    As an aside, you missed out an important pro of reed relays, which is they can be made to stand much higher voltages than normal types of relay. Also you can "roll your own" by getting a separate reed switch and adding a coil to it. You could probably also incorporate a holding magnet by doing it that way.
    Voted :)

    3 replies

    Reply 13 days ago

    Your interest, and vote are appreciated. I assume you're referring to reed contacts being sealed in glass, so not exposed to air, therefore less likely to arc when switching higher voltages?


    Reply 12 days ago

    Ah, yes, that is indeed an important trait of reed relays that I will include in the next edit of the article, after the judging for the contest concludes. Thanks for pointing out one of the things that I didn't think of mentioning.


    13 days ago

    Wow, this is great, do you think that this will extend the lifetime of those relays?
    Thanks for sharing

    1 reply

    Reply 13 days ago

    The coil will certainly be less stressed because it wont heat up. The mechanical wear may be reduced slightly, because the permanent magnetic field may reduce the friction at the switch armature pivot point. The switch contacts will be susceptible to the same deterioration though.


    14 days ago

    Realy nice trick, I will defenatly use this in a next project

    Thanks, you got my vote.

    1 reply

    Reply 13 days ago

    I appreciate your comment, and your vote. If you use this technique, please come back and share how it works out for you.


    Question 15 days ago

    Great writeup clearly explained, thanks!! <-two thumbs up
    Are there IC drivers which do the equivalent?

    2 more answers

    Answer 14 days ago

    I used the IR IPS6041 high side driver for a 4 wire to 3 wire trailer light converter. It has thermal and overcurrent protection and other features. It takes almost no driver current when off or on and switches up to 24 volts DC. They and similar drivers are used in vehicles and are more reliable than relays.


    Answer 15 days ago

    I'm glad you found my article useful. I'm not sure if there are IC drivers on the market that offer similar characteristics. I doubt there are any with the same exact functionality and ultra low holding current spec.