QUESTION ABOUT LATCHING RELAYS

@JohnActon posted:

From what I understand there are three different types of latching relays 1; Twin Coil where coil A latches the relay and coil B unlatches it.  2: Single Coil type A, where the polarity of the power on pins 1 and 2 is reversed to unlatch. 3: Single Coil type B, where one pulse latches the relay and a second pulse (without reversing the polarity) unlatches the relay.  It is this third type that I am looking for and I don't know the proper name of this type latching relay to start looking for one.  Can anyone help ?          Thanks,      j

An Impulse relay may be what you're thinking about.  Subsequent trigger signals (of the same polarity) change the relay state.

https://www.mouser.com/datashe...57/105A_755-7090.pdf

If you're into low voltage stuff, say 5 vdc, you can find latching relay modules with timers that provide delay on/off on ebay. Look close but most of the relay contacts are rated for 10 amps at 250 volts so they work well if you're switching track power.

How about some specifications?  Coil voltage, contact current rating, contact configuration, etc.

3~5V COIL  DPDT 1~2A contacts. It would fall into the category of a signal relay. SteveH, said it.  Impulse relay.   Ever have one of those brain freeze moments.  I must have been searching for the term for three hours last night.   Thanks gentlemen !        j

Here's 16 possibilities in stock at Digikey with a 4.5 to 5.0 volt coil and 2A contacts.  I didn't know if you wanted surface mount or thru-hole, so I left both.

Single Coil Latching Relays at Digikey

@JohnActon posted:

3~5V COIL  DPDT 1~2A contacts. It would fall into the category of a signal relay. SteveH, said it.  Impulse relay.   Ever have one of those brain freeze moments.  I must have been searching for the term for three hours last night.   Thanks gentlemen !        j

So is this Mission Accomplished?  I'd be interested in seeing a reasonably priced Impulse Relay that has a 3-5V DC coil.

One aspect of latching relays not yet mentioned is if they require power to maintain the latched position.  In other words is the memory mechanical or electronic (as a 1-bit memory).  There may be a benefit to having the relay always turn OFF when you turn off your layout...so that it is not ON when you next power up the layout.

Anyway, if you found what you need by way of an Impulse Relay, then great.  Otherwise try search terms like "flip-flop relay" or "start-stop relay".   I've used the following widely available (Amazon, eBay, etc.) module though I don't understand enough about your application to say if it's suitable.  For example, there was an application to have a flip-flop relay alternate a turnout each time the train tripped an insulated-rail trigger. This module was "fooled" by dirty track - easily solved but demonstrates the difference between the real world and the back of an envelope.

So far all single coil relays require reversal of the polarity to change the state. Some of them are not bi-stable. I think something like Stan proposed may work however it is larger than I would ideally want and needs three conductors. Perhaps even four if you have to run to and from a remote button. I would like to keep it two conductors on the input side. I'm thinking I may be able to cobble something together with a flipflop. It's looking like no one makes a true impulse relay in such a small size. Or I haven't found it yet.          j

If you're thinking about an active circuit with power all the time, there are tons of ways to do the job.

@gunrunnerjohn posted:

If you're thinking about an active circuit with power all the time, there are tons of ways to do the job.

  No permanent power.  I'm trying to use a pulse with two conductors to start a random event. One pulse starts it one ends it. Come to think of it, a Lionel two position e-unit may do the job.  Could be a bit power hungry for my pulse to operate. I have several will give one a try and then brag about it.  If it works.                       j

If you only want 2 conductors on the input side, why not use a single-coil latching relay - the ones where you reverse polarity to flip latch.  Then drive the two conductors with a DPDT switch that is "cross-wired" to send one or the other polarity. 

I want to be able to control this with the Cab1 flipping a dpdt switch is not an option.  Thanks for the suggestion never the less.          j

If you're controlling it with a CAB1, what is receiving the signals?  Is this for accessories, rolling stock, or locomotives?

Automation Direct has an Alternate Action relay, ARX-2C-120A-TL
in either 1 pole SPDT or 2 pole DPDT. Click on Specifications for info and
wiring diagrams. Unfortunately, it is a 120vAC
coil. No specs on pulse  length shown. It is really a solenoid that operates
a push-ON, push-OFF switch. It is typically used to alternate the use of
pumps so they both get the same wear.

Roger

GRJohn, R2LCO8.  Roger, It's far to large, not to mention the inverter needed to power it. I guess I will have to come clean. I am wiring up the Williams PRR L1 Mikado that I have been working on far too long.  I've had an idea for some time of labored smoke. It must be easy to turn on and off and we have a front coupler function which is not used on most of our steam locos. I did an experiment on another loco using the coupler pulse to start a timer. However you could not turn the smoke down at the precise time though the timer strategy works ok for blowing the whistle on old RS-1 & 2 boards . With an impulse relay you hit front coupler once to turn it on and once to turn it off.  I already wire idle smoke into some of the locos that I convert to TMCC however my efforts with labored smoke have not panned out though I think if I can find or make a impulse relay small enough or if as I mentioned a couple posts ago a two position Lionel eunit will work if the coupler pulse is long and strong enough.  There are enough variations on the flipflop that even a small common single coil DPDT  relay can be made to work.  It would not even have to be a latching relay as the flipflop can take care of that. Though by the time you build that circuit you might as well let the flipflop switch an scr or triac   However it sure would be easy if small signal relays come in an impulse version. You would not need all the extra circuitry.  Also I don't see why this same relay/ coupler pulse could not be used to blow the whistle on old RS-1 sound boards. I have several of those old RS-1 steam boards and would like to install them in the grandkids trains. BTW I did some rpm measurements on the motor in the MTH smoke unit that some will find interesting. Will start another thread for that.                            j

While I stopped for dinner I was browsing eBay and ran across something that I am sure will work. The only thing I don't like about it is the size of the connectors. They can be removed and the wires can be soldered directly to the board.  This was the last one the seller I bought it from had though I doubt he is the only seller selling these.           j

Another solution I found which will turn any relay into a one button push on push off. latching relay. Search ebay for "Bistable flip-flop latch switch circuit module button trigger power off m.19" These things are tiny and would match up well with tiny signal relays.  You could drive an optocoupler off your front coupler pulse to act as the button to trigger the switch.  Imagine the neat stuff you can do with the unused front coupler function on your steamers.   There are more than a few variations on this flipflop on ebay and a little searching will turn them up. One even switches up to 5A  and is only slightly larger than this one. Both sell in the $3.60 range.                         j

I built a small board that allowed me to control some lighting using the coupler outputs.  This was for a BEEP that didn't have couplers.

I also have a board design that does automatic two-stage smoke, it turns it down when you're stopped and full power when you're moving.  It also has contacts to control the cab light and Rule-17 LED Lighting.

@JohnActon posted:

  No permanent power.  I'm trying to use a pulse with two conductors to start a random event. One pulse starts it one ends it.

@JohnActon posted:

... I think something like Stan proposed may work however it is larger than I would ideally want and needs three conductors. Perhaps even four if you have to run to and from a remote button. I would like to keep it two conductors on the input side.

Perhaps semantics, but unless there is a "traditional" mechanical latching relay, you need "permanent power" for the module to remember which bi-stable position it is in after the trigger pulse vanishes.  But if the application does not require the electrical isolation of a traditional relay (2 wires control input, 2 wire switched out) then the options indeed expand.

@gunrunnerjohn posted:

I built a small board that allowed me to control some lighting using the coupler outputs.  This was for a BEEP that didn't have couplers.

What are the characteristic of the Coupler Output?  AC/DC, voltage, timing, etc.?

Is that a 12V DC dual-coil relay?  And how did you choose 4.7uF NP caps?

Clever John.  If you ever want to add couplers to your beep you could use your motion detector to control the cab lights. I think you like to just try things whether you ever use the idea again you know it works. I get side tracked all the time trying out things that may or may not work.  One of the uses I have found for those cheap regulator boards that you warned us all about is to put a bridge between the reg board and the motor and use it as a motion detector. The motor never has more than 15vdc to it so it does not cook the little regulator.  I drive a small relay with the board and use that to get idle smoke. If I want an action to happen in forward or reverse I just use diodes coming off the motor instead of the bridge.  Using these reg boards and relays to control loco functions I can cut the tethers down to five wires two for electrical pickup two for the motor and one for chuff. I guess the drawbar with some jumpers soldered to it could eliminate the chassis ground cutting the tether down to four wires  I will post some photos when I get through all the Easter obligations.  Happy Easter everyone.          j

@stan2004 posted:

What are the characteristic of the Coupler Output?  AC/DC, voltage, timing, etc.?

Is that a 12V DC dual-coil relay?  And how did you choose 4.7uF NP caps?

The coupler outputs are full wave AC track power, the impulse is roughly 1/4 second.  The caps were actually something I had on hand and the right "small" size, I just wanted to kill any chatter.  It is indeed a 12V dual-coil relay.

@JohnActon posted:

Clever John.  If you ever want to add couplers to your beep you could use your motion detector to control the cab lights. I think you like to just try things whether you ever use the idea again you know it works. I get side tracked all the time trying out things that may or may not work.  One of the uses I have found for those cheap regulator boards that you warned us all about is to put a bridge between the reg board and the motor and use it as a motion detector. The motor never has more than 15vdc to it so it does not cook the little regulator.  I drive a small relay with the board and use that to get idle smoke. If I want an action to happen in forward or reverse I just use diodes coming off the motor instead of the bridge.  Using these reg boards and relays to control loco functions I can cut the tethers down to five wires two for electrical pickup two for the motor and one for chuff. I guess the drawbar with some jumpers soldered to it could eliminate the chassis ground cutting the tether down to four wires  I will post some photos when I get through all the Easter obligations.  Happy Easter everyone.          j

I looked at adding couplers to the BEEP, but it seemed like a tall order with the odd coupler mount, so I decided that I didn't need that option.

@gunrunnerjohn posted:

The coupler outputs are full wave AC track power, the impulse is roughly 1/4 second.  The caps were actually something I had on hand and the right "small" size, I just wanted to kill any chatter.  It is indeed a 12V dual-coil relay.

Got it.  Yeah, I feel like I'm playing a game of Whack-a-mole.  Initially it seemed like the answer was a 3-5V DC Impulse Relay - what I'd think would be unusual, hard-to-find, or expensive in today's day-and-age.

But apparently, the requirement is not a relay at all, and activated by an AC pulse from the Coupler output of a TMCC "receiver".  And if an electronic non-relay module is indeed capable of adjusting the power to a smoke unit, then the need for a mechanically latching mechanism is not needed at all.  That is, when power is removed from the engine it is OK for the smoke power to return to a default level.  The ability to use electronic latching using a digital flip-flop chip or whatever vastly expands options and cuts cost.

So the discussion has morphed into how to use an engine's unused coupler output(s) to provide additional features.

I asked about Coupler timing because I've used this technique in MTH PS2/3 engines.  It's just that the MTH Coupler outputs are less than 0.1 seconds in length.  They have gobs of power (being able to drive several Amps into the solenoid coil) but is a short pulse.  As I highlighted in my previous post, the bi-stable electronic module proposed as a solution requires a trigger pulse that is 0.2-2 second long.  Of course if you have gobs of Watts available, an MTH PS2/3 coupler pulse can be trivially extended to be at least 0.2 seconds.  But no problem if the Lionel coupler pulses are 1/4 sec.

In the case of MTH PS2/3 engines, an unused Coupler output can even be used in Conventional mode... by pressing sequences of the transformer Bell and Whistle buttons which trigger the Front and Rear couplers.  The coupler pulses in the PS2/3 engine boards are the same whether triggered by DCS command control or by conventional.  And there is so much "power" even in the short Coupler pulse.  I think the last example I posted on OGR about using the Coupler pulse was to power an RF module to send a wireless control burst to a receiver in the tail car of a consist.  So you could fire the Coupler once to turn on a feature in the caboose or tail car...then fire the Coupler again to turn off a feature in the caboose or tail car.

Stan, a perfect illustration of defining the specific problem and environment before trying to solve it!

So here is a method by which you can wire a standard single coil DPDT relay so it will latch. I don't know if it would be applicable to your project, but I used this circuit a couple of years back to make an offshore 10 amp DC power supply kind of fool proof. (Most of them have a nasty quirk whereby if you unwittingly connect the load up first, then power it up, it will go poof, and its basically junk.)

This addition requires you to push a PB to activate the relay and connect the load; and push another PB to disconnect the load. If you follow the circuit you can see how it works. It can be used with any garden variety DPDT relay, or 3 pole/4pole, as long as you have one spare set of contacts. The idea was presented to me by forum member rtr12. And I must say the PS has worked just fine ever since. One nice feature is that if you have even a short power blip the relay will unlatch, and even if the power comes on almost immediately, the load remains disconnected until you hit the connect PB again.

Thought it might add another option for you that doesn't require purchase of a special impulse relay.

Rod

Thanks Rod This is going to be inside a locomotive and depends on electrical pulses to trigger so a push button solution will not work.        j

@stan2004 posted:

Perhaps semantics, but unless there is a "traditional" mechanical latching relay, you need "permanent power" for the module to remember which bi-stable position it is in after the trigger pulse vanishes.  But if the application does not require the electrical isolation of a traditional relay (2 wires control input, 2 wire switched out) then the options indeed expand.

Stan, of course you are right about the need for permanent power. I was referring to the pulse which by definition goes away.    There will be permanent power to Vi- and Vi+ from the track through a bridge. They also mention a pulse duration between .2 and 2 seconds.   My initial thought was I could use the pulse from the coupler through an opto coupler to use as a trigger . But that only works if the relay actually can be triggered with a .2 sec pulse.   If not my next idea is to use a capacitor to store enough energy, from the coupler pulse, to keep the led in the opto coupler on long enough to reliably trigger it.  Once I have that ironed out the plan is to install a second resistor in the smoke unit using track power through a bridge and then the relay.  Adding this should give the loco three smoke levels as I already have idle smoke wired up.  Your thoughts are surely appreciated.                               j

FWIW.  The DZ1008 relay, is a latching relay. 

@Mike CT posted:

FWIW.  The DZ1008 relay, is a latching relay.

Really?  Here's the schematic of the DZ-1008, I don't see how this is a latching relay?

Now that we've pinned down the specifications, why not use something simple like this TE Connectivity 6-1462037-6 $4 single coil latching relay with some simple steering logic to toggle the polarity and switch it between states?  It seems it should solve the problem and is certainly fast enough to react to the TMCC coupler pulse.

@JohnActon posted:

...

(1) There will be permanent power to Vi- and Vi+ from the track through a bridge. They also mention a pulse duration between .2 and 2 seconds.

(2) My initial thought was I could use the pulse from the coupler through an opto coupler to use as a trigger .

(3) But that only works if the relay actually can be triggered with a .2 sec pulse.

(4) If not my next idea is to use a capacitor to store enough energy, from the coupler pulse, to keep the led in the opto coupler on long enough to reliably trigger it.

...                          j

So I don my peanut-gallery hat:

(1) If you are generating DC power using a bridge rectifier (full-wave rectification), the DC- output of the bridge is no longer the same as AC ground.  I know nothing about Lionel engine electronics but it seems there have been issues.  For example, is the smoke unit or whatever you are controlling galvanically/electrical bolted to the chassis presumably picking up "ground" from the chassis?  If so, using a bridge rectifier will create an additional smoke unit.

(2) The module you show has what appears to be a type 817 opto coupler on it just to left of FET.  It's unclear what it is isolating in the module and the documentation I saw on Amazon is skimpy.  So you may already have an opto coupler.  Or maybe it was a typo and you meant "thru THE opto coupler" rather than "thru AN opto coupler."

(3) Relay?  Are you referring to this solid-state module as a relay or are you saying you are using this module simply as a bi-stable flip-flop latch that then drives a generic non-latching relay which you must add?  I'd think if you are adding a relay (perhaps to gain load isolation from chassis ground per (1) above) you'd be better off with a DIY latching relay circuit.

(4) Since, as GRJ points out, the coupler pulse is AC, it might be prudent to add a diode to keep options open to use a polarized capacitor when energy storage is involved.  The other weird thing which you will presumably soon find out is what the heck they mean by "0.2-2 sec low pulse trigger".  Taken at face value, "low" trigger tells me the trigger pin should be connected to "ground" or 0V to trigger the module. The other trigger pin is called "Trigger Signal Input Ground".   I suppose this means you can "short" the two trigger pins (T and G) to effect a "low" trigger pulse.  Anyway, another detail to work out but would affect the implementation of a capacitor to extend a duration of a "high" trigger voltage!

@stan2004 posted:

(1) If you are generating DC power using a bridge rectifier (full-wave rectification), the DC- output of the bridge is no longer the same as AC ground.  I know nothing about Lionel engine electronics but it seems there have been issues.  For example, is the smoke unit or whatever you are controlling galvanically/electrical bolted to the chassis presumably picking up "ground" from the chassis?  If so, using a bridge rectifier will create an additional smoke unit.

Big NO-NO as a rule!  Most of the stuff I do will forgo a bridge rectifier for that very reason.  I use a full-wave bridge in my LED lighting module because I don't need a common ground.  However, the Super-Chuffer, YLB, etc. all use a single diode to retain the common ground as that's necessary for TMCC and early modular Legacy compatibility.