Latching Toggle Relay with powerfail remembrance

[AL13N]

Trying to make an easy and cheap latching relay to be triggered by a pulse (soft touch).

I came up with the schematic below (not an electronics engineer):

The basis is the latching relay at the bottom: This is the one that is the actual relay, it's latching so that there is no power usage when it's in either state, and after a power failure, the state is still the same.

The 3rd relay is a latching one to preselect the other state of the lowest relay.

The first relay is non-latching in order to have a non-pressed state (which the switch does not have)

The 2nd relay is non-latching in order to keep on until the switching is completed.

I put on a capacitor in the hopes of disconnecting the first relay after some time, so that it if you press and hold the switch long enough that you don't fry the latching relays.

I wasn't sure if some diodes were required or not in order to keep the current correctly flowing...

Please give me some assistance in this, thx.

 

[Colin]

You just need a latching relay, a switch, and some bits.

 

[AL13N]

You just need a latching relay, a switch, and some bits.

what do you mean, a switch? like, a programmable switch, or something?

the latching relays i found were not pulse-driven, just either reverse polarity, or dual coil...

 

[AnalogKid]

The text in your image is so small I can't see what the system power is.

Here is a single-coil impulse relay that uses a mechanical interlock to hold the contacts in either state with zero power applied (after the pulse). Many coil and contact options.

ak

 

[crutschow]

the latching relays i found were not pulse-driven, just either reverse polarity, or dual coil.

All latching relays can be pulse driven and will remember there last state when the coil power is removed.
That's why they are called latching relays.

 

[Colin]

Latching relays with memory have a little brain inside that remembers the last position and when you turn them on again, the little brain wakes up and looks at the position and says: "Yes, that's correct."
That's why you pay a fortune for them.

 

[rjenkinsgb]

Latching relays with memory have a little brain inside that remembers the last position and when you turn them on again, the little brain wakes up and looks at the position and says: "Yes, that's correct."
That's why you pay a fortune for them.

They are generally far simpler than that - just a magnet that sticks to one or another armature (or vice versa) and can have it's position changed by attraction or repulsion from a coil or two coils. The contacts stay in position without any power needed due to the magnet.

The original is likely the "Carpenter relay" used with teletype systems etc., with newer ones using a similar principle but scaled down.
https://www.flickr.com/photos/russell_w_b/5075821618

 

[AL13N]

I guess it wasn't clear from the drawing, it's for 12V (or around that) ; i was looking for ratchet relay; but i can only find those for 230V and around that.

I would love to have these "single pulse" ratchet latching relays; but i haven't found them.

that's why i was trying this scheme; one switch, no current flows if the switch isn't pressed; and after power fails, the last state is still there (due to the latching relay). considering i can buy these 4 relays in components for less than 10€ ;

plus i don't want to make it to complicated, like adding memory or something...

If someone has a better idea to accomplish these 4 requirements, that would be nice:
* low voltage (~12V or 24V max)
* single pulse from the same switch turns it on and off
* no current flows when the switch is not pressed (or not long thereafter)
* when power fails, the last state is still there (preferably mechanically)

 

[AnalogKid]

I guess it wasn't clear from the drawing, it's for 12V (or around that)

Looks like 12 Vdc to me:

ak

 

[rjenkinsgb]

I can't find any pulse relays with low voltage coils, or at reasonable prices anyway.
(There are such as selectors and motorised rotary switches, at crazy prices).

The Tycho one AnalogKid suggests does appear to be suitable.


Other than that, the simplest circuit I can come up with uses two, dual coil bistable relays plus a changeover button, or a normally open button plus a simple relay with a CO contact.

The basis of that is a "master-slave" bistable. Each relay has the two coils connected to the NO and NC of a contact on the other relay, with common of that contact fed from the button.

The other side of the coils go to NO and NC on the same relay (as the coils) with common to 0V.

When power is applied, the "working" relay will take the same state as the other relay, shutting off it's own coil power as soon as it changes over.

If you duplicate that, cross-connected between two relays but with the control contact NO/NC reversed on the one fed from the button NO, the effect should be that the first relay changes state when the button is pressed and the second relay changes to the same state when the button is released.

The only power consumption is as each relay changes.

One of the relays needs to be a four pole type to control the external circuit, as two changeovers are needed in each to provide the basic function.

 

[AnalogKid]

The Tycho one AnalogKid suggests does appear to be suitable.

Perhaps if you tell us what about it is not suitable, we can suggest other alternatives.

ak

 

[AL13N]

Perhaps if you tell us what about it is not suitable, we can suggest other alternatives.

ak

actually, he says it is suitable.

I'm just thinking of the price and availability, but that does seem easier...

It's a bit big, i was kind of hoping to use it on a test board, but it's not really a pcb component...

Edit: Whoa, more than $40 a piece ... quite expensive... maybe i could use a smaller cheaper version of it...

 

[AL13N]

I can't find any pulse relays with low voltage coils, or at reasonable prices anyway.
(There are such as selectors and motorised rotary switches, at crazy prices).

The Tycho one AnalogKid suggests does appear to be suitable.


Other than that, the simplest circuit I can come up with uses two, dual coil bistable relays plus a changeover button, or a normally open button plus a simple relay with a CO contact.

The basis of that is a "master-slave" bistable. Each relay has the two coils connected to the NO and NC of a contact on the other relay, with common of that contact fed from the button.

The other side of the coils go to NO and NC on the same relay (as the coils) with common to 0V.

When power is applied, the "working" relay will take the same state as the other relay, shutting off it's own coil power as soon as it changes over.

If you duplicate that, cross-connected between two relays but with the control contact NO/NC reversed on the one fed from the button NO, the effect should be that the first relay changes state when the button is pressed and the second relay changes to the same state when the button is released.

The only power consumption is as each relay changes.

One of the relays needs to be a four pole type to control the external circuit, as two changeovers are needed in each to provide the basic function.

I kind of started with that system: but i didn't think to connect the other side through their own contact towards the 0V. Instead i got sidetracked into the datasheet which said that the coils couldn't be powered too long. the only thing that's missing is now to make sure the NO switch is powered for long enough to do the switching.

I'm gonna try to change my diagram into what you said.

 

[eTech]

I guess it wasn't clear from the drawing, it's for 12V (or around that) ; i was looking for ratchet relay; but i can only find those for 230V and around that.

I would love to have these "single pulse" ratchet latching relays; but i haven't found them.

that's why i was trying this scheme; one switch, no current flows if the switch isn't pressed; and after power fails, the last state is still there (due to the latching relay). considering i can buy these 4 relays in components for less than 10€ ;

plus i don't want to make it to complicated, like adding memory or something...

If someone has a better idea to accomplish these 4 requirements, that would be nice:
* low voltage (~12V or 24V max)
* single pulse from the same switch turns it on and off
* no current flows when the switch is not pressed (or not long thereafter)
* when power fails, the last state is still there (preferably mechanically)

Hi

What is the load voltage/current requirements? In other words, what is this circuit used for?

Seems there is a much easier way to do this than with all those relays.
if possible, are you open to a digital electronic solution?

eT

 

[AnalogKid]

Assuming the reason for the latching and unlatching is to connect or disconnect power or signal circuits, what are the signals or power levels being switched?

ak

 

[AL13N]

Thx for all the responses:

• I was looking into 4PDT dual coil latching, but didn't find anything suitable :-(
• What i'm switching is just a variable 1-10V DC input for a dimmer
• I'm not adverse to a digital solution; but i'm trying to cut down on price and complication level; since this is only a temporary solution; as i'll later (in a couple of years) i'll be making a complete digital solution with hundreds of sensors to switch 30 of these dimmers but also other outputs
• i've bought some big test boards, so i don't even have to solder because it's only for a couple of years as a temporary solution
• the normal open switch is connected through a 10m to 30m twisted pair (UTP cat5), so i'll need to account for some voltage drop
• i haven't decided on power supply yet, but it must be lower than 24V

That should give some background info.

since i can't find a multipole DT latching relay, and a 12V DPDT is less than 2€ ; i'm trying to keep things cheap here. though i'm trying to cut down on power leak, since this i'll be 24x7 (and 25 times)

I'm gonna try to find a way to do a version with DPDT's only

I'm looking at EC2-12TNU and V23079B1203B301 and i'm seeing they already have 1kOhm resistance, and even have enough with 9V to switch, so it seems like i won't even have to use some resistors in series with it.

After this, i'll need to find a nice cheap 12VDC power supply than can handle 25times that kind of schematic, so i guess somewhere above ~15W

 

[AL13N]

new schematic with 4 relays: de latching relays have ~1kOhm and the simple SPDT has 360Ohm ; i don't think i need to account for voltage drop due to the cable for the open switch (twisted pair cat5e : 10 to 40 meter), since all of them seem to work at 9V too...

would this work?

 

[crutschow]

Certainly a CMOS flip-flop with a coin-cell backup would be much simpler and cheaper.
A lithium coil-cell would provide back-up for years before needing replacement.

 

[eTech]

HI

You could do something like this and use a single coil latch relay.
Just make adjustments for whatever supply voltage your using.
Might need to modify or remove POR though.

eT

 

[AL13N]

Certainly a CMOS flip-flop with a coin-cell backup would be much simpler and cheaper.
A lithium coil-cell would provide back-up for years before needing replacement.

would this really be cheaper than $6.07 ? (which is the price for the 4 relays?)