Yet Another Relay Board Project?

If you use NC/NO relays, you can chain as many relays as you want in as the diagram illustrates. Shows 3 wire and 4 wire example of 2 lamp dwarf signal. First relay in chain sends AC down the chain, and each following relay passes along the AC. Any relay being energized open the chain of AC, and sends GND down the rest of the chain. Dwarf Signal wired so red lamp is connected to same AC as starts chain, and then has its other lead (internal for 3 wire, externally for 4 wire) connected to the green lamp. Green lamp has its other lead connected to ground. When AC comes down the chain, green lamp is lit. If a relay is energized, ground comes down remaining chain, and red lamp is lit.

@MED posted:

If you use NC/NO relays, you can chain as many relays as you want in as the diagram illustrates. Shows 3 wire and 4 wire example of 2 lamp dwarf signal. First relay in chain sends AC down the chain, and each following relay passes along the AC. Any relay being energized open the chain of AC, and sends GND down the rest of the chain. Dwarf Signal wired so red lamp is connected to same AC as starts chain, and then has its other lead (internal for 3 wire, externally for 4 wire) connected to the green lamp. Green lamp has its other lead connected to ground. When AC comes down the chain, green lamp is lit. If a relay is energized, ground comes down remaining chain, and red lamp is lit.

That's a very ingenious setup MED; looks like it should work great. So if you have 2 or 3 insulated rail sections say along a section of track, the same 2 light signal(s) switch on and off as a train makes its way through the sections, right? And it would not matter which direction it was moving I guess. Or you could have one section on each converging track at a turnout say, and the signal would operate no matter which section was being activated. That has some interesting possibilities.

Rod

The only requirement is that the relay that starts the chain is the only one whose contact is connected directly to AC, and that the AC connected to the relay that starts the chain is the same AC as goes to the signal(s), all will work. The AC at relay and at the signal has to be the same so that the one lamp in the signal that will have AC applied to both terminals does not "glow dimly". Relays within the chain can be wherever you want. No directional at all. Any one relay in the chain being energized will change which lamp in the signal is lit.

You can also create a relay chain where the first relay has ground to its NC contact, and AC to its NO contact - all the rest of the relays in the chain have AC on their NO contacts - this is just the reverse of my drawing. Thus, when no relay is energized, ground is passed through the chain. Any relay being energized sends AC down the rest of the chain. This reversal may be needed based upon your particular application.

Perhaps this link to the earlier thread works?!

To this day, Lionel documentation shows need for an EXTERNAL relay when two or more 153IR occupancy detectors drive a 3-wire signal or 3/5-wire crossing gate.  In the linked thread, member Dtrainmaster proposed a method to eliminate this external relay.  But it required modifying the circuit board inside the 153IR to isolate the "common" terminal of the relay.  Unfortunately, as I recall Dtrainmaster removed his diagrams for various reasons - one of which was Lionel made multiple versions of the 153IR that would have required differing modifications making it quite the exercise. This is because the 153IR internally connects the "common" terminal of the relay to a power wire.  In effect, this means ALL the "common" terminals on multiple 153IRs are tied together.

It appears the current proposal by MED is similar in requiring modification (cutting circuit board traces) of the 153IR to isolate/access the common terminal of the internal relay.  Of course if we're just talking generic SPDT relays like Rod's new board, then the common terminal if free to be wired as proposed.

Stan, thanks for posting a working link to that previous topic; I seem to have problems with linking for some reason. Probably because I do it so seldom. It seems MED's drawing can be taken to mean the relays in the 153IR's, or maybe using external relays triggered by insulated rails or whatever. Maybe MED will see this and comment.

There was no way I wanted to make internal changes to the 153IR, and create a bunch of one-of's. Thus my relay board creation at the top of this topic. You could also use this relay board as shown in MED's drawing, wherever you wanted on the layout, triggered in whatever way you wish. In truth I knew nothing of grj's relay board that he posted, up a few posts. I must have missed that whole topic somehow. It seems like it would offer all the same or better functionality, and is also DCS compatible. My creation has no DCS choke onboard because it was not really intended to be installed in engines or rolling stock, though you certainly could. In this case you would want a 22uH choke in the trigger power circuit, no big deal. What I wanted was a small all in one board that could run on any typical layout trigger voltage, AC or DC, and had a voltage regulator onboard, so no other buck converter or extra components are required. If there's a future revision, it will likely get an smd onboard indicator led, so you know when it's triggered. We shall see. Right now it's doing what was intended and there will be a few more hooked up.

Rod

Since @Rod Stewart you said you had a new relay board, I assumed you were not speaking at the IR153, which has the pwr in and C of relay connected together. My idea only works if you have C, NC, and NO independent of any other circuitry.

@Rod Stewart posted:

...

There was no way I wanted to make internal changes to the 153IR, and create a bunch of one-of's. Thus my relay board creation at the top of this topic

...

Yup.  I remember thinking back then in the linked thread that the chances of someone going in and cutting PCB traces and similar surgical procedures was about 0%...especially when adding a single external relay for maybe $10-15 was a perfectly functional solution.  That is, if you're already paying $45.99 (MSRP) per 153IR and have two (or more) then it is what it is.

Now since I challenge anyone to show me that they are a bigger cheapskate than me, I was thinking about this exact scenario.  You have two (or more) 153IRs...or MTH ITADs...that internally tie the internal relay common to one of the power inputs.  You need an external relay to control a 3-wire signal or crossing gate.  You offer something about $5.  That's a new price-point.  But it's all about defining the problem narrow enough to simplify it even more.

As you point out, you don't need the DCS choke.  Yeah, the choke is a whopping 10 cent part but it takes real-estate and you pay for the boards by the square inch.

By the same line of reasoning, since we are talking 153IR detectors, they come with the "DELAY" adjustment which eliminates relay chatter typically associated with insulated-rail triggering.  In other words, if using an EXTERNAL relay with multiple 153IR detectors you DON'T NEED the capacitor to demote relay chatter...since there is NO relay chatter from a 153IR!

I also argue that if you are using multiple 153IR's, you are probably using AC Accessory Power and not trying to do things "on the cheap" like using track power.  In other words, there's a super high likelihood that you are using 14V AC to power the 153IR.  So this demotes the need for a voltage regulator IC...even if they are only a dime a piece (plus PCB real-estate property tax!).

That said, akin to that old TV game show "Name that Tune" or whatever it was called where people claim they can name a tune with only the first few notes...how about an AC compatible relay with only 2 parts:

This is eBay Asia pricing so you can actually do better from other Asia suppliers if willing to live with the shipping delay.

I'm not up on a bare-board PCB pricing but I'm wondering if, at quantity 10, we can get it down to $1 per AC-relay all-in?!   Obviously, $1 is an arbitrary goal.  But it makes for good discussion fodder.

I see that you guys are still at it.
Rod, nice job on the board.
Interesting idea using an AC relay, Stan. That takes me back to my first 2-color signal problem, trying to use those pressure activated 153C Contactors. Very difficult to get those to activate with a light-weight PW caboose and then fully spring back to NC after a heavy engine. That led me to trying an AC relay, then only having to adjust for one track contactor position. (I couldn't use the insulated-rail method back then because I didn't have the special Super-O track pieces).

If Lionel had only let that 153IR relay run free, with an untethered common, or even with a jumper option, we wouldn't be here now.
Dave

@MED posted:

Since @Rod Stewart you said you had a new relay board, I assumed you were not speaking at the IR153, which has the pwr in and C of relay connected together. My idea only works if you have C, NC, and NO independent of any other circuitry.

That's kind of how I interpreted your multi-relay drawing MED. This relay board could work really well for your application, and they are inexpensive.

Rod

Stan2004 said:

"By the same line of reasoning, since we are talking 153IR detectors, they come with the "DELAY" adjustment which eliminates relay chatter typically associated with insulated-rail triggering.  In other words, if using an EXTERNAL relay with multiple 153IR detectors you DON'T NEED the capacitor to demote relay chatter...since there is NO relay chatter from a 153IR!

I also argue that if you are using multiple 153IR's, you are probably using AC Accessory Power and not trying to do things "on the cheap" like using track power.  In other words, there's a super high likelihood that you are using 14V AC to power the 153IR.  So this demotes the need for a voltage regulator IC...even if they are only a dime a piece (plus PCB real-estate property tax!)."

Stan good point on the built-in anti-chatter nature of the 153IR. Besides using this board with the 153IR, I was also envisioning use with insulated rail sections and who knows what. So I figured besides smoothing my half wave ripple, the 100 uf cap should also mitigate chatter, a good feature for a universal application module. Pretty inexpensive too. Actually the most expensive part is the relay @ about 75 cents each from Digikey for quantity 10.

And you are bang on; I use 14vac for most accessory power, supplied by a built up PS using a Hammond multi-tap transformer. I also have a 16vdc PS for powering the old 022 switches, and a 12vdc PS for some accessories and building lighting and suchlike. Nevertheless I wanted something that could work with a wide range of trigger voltages, without the need to change or add components. That's how I got to this point. All good. And I like your ideas and suggestions.

Rod

If you don't mind waiting a little, you can do a more universal relay board with a 10A relay, the relays are less than 60 cents each from AliExpress.  Just make sure to beef up the traces to the relay contacts, 10 amps requires about .2" wide traces using 1oz copper.

@gunrunnerjohn posted:

If you don't mind waiting a little, you can do a more universal relay board with a 10A relay, the relays are less than 60 cents each from AliExpress.  Just make sure to beef up the traces to the relay contacts, 10 amps requires about .2" wide traces using 1oz copper.

That may be the Mark II version John. In reality though there's no sense re-inventing the wheel; if I need a 10 amp version I will just buy your modules through Hennings.

Rod

@Rod Stewart posted:

In a recent OGR topic there was a great deal of discussion about using Lionel 153IR sensors in parallel, to control a 3 wire 3 color track block signal; titled "Wiring Dual 153IR's to operate one signal light. Can it be done without relays?". ...

Pardon the broken record, but if we narrow the objective to the multiple 153IR (or ITAD) problem, the Lionel 2-component solution (bridge rectifier + 12V DC relay) can be the makings of "another relay board project"!  Lionel recommends a Radio Shack 15 Amp relay which is arguably overkill for a signal or crossing gate!

Above shows what I'm calling the "Dollar Tree" version using the 2 components I identified in an earlier post.  This "AC relay" module is only 3 Amps (250V AC)  but that ought to be more than adequate for signals and crossing gates!  AND. The relay "common" terminal is isolated so can be used with the clever cascading configuration described by MED and Dtrainmaster!

When I read and observe all the amazing circuits that you experts come up with, I wonder why there is always the nagging issue about providing DC to make things work. Each project seems to require, as a separate piece of circuitry, the engineering and building of a DC supply.

I suggest that when a person gets to the point of having a sophisticated layout that can use these circuits, why the obvious (to me) next step isn't recommended. Why not create a layout-wide DC bus (or busses) that is/are referenced to the return circuit of all the AC sources, so that wherever the DC is required, all the hobbyist needs to do is splice onto the DC bus, anywhere on the layout.   We see lots of folks running constant AC busses for accessories and signaling. It seems that as a "policy", we ought to think about providing DC, too.

A simple filtered rectifier circuit, of adequate amperage for all the expected applications, can easily be built for a couple dollars. It can be powered by a separate AC supply, or get its input from the output of an existing transformer.

@stan2004 posted:

Above shows what I'm calling the "Dollar Tree" version using the 2 components I identified in an earlier post.  This "AC relay" module is only 3 Amps (250V AC)  but that ought to be more than adequate for signals and crossing gates!  AND. The relay "common" terminal is isolated so can be used with the clever cascading configuration described by MED and Dtrainmaster!

Don't you potentially get relay chatter with no filtering?  Some relays are too quick to react and you hear the contacts chattering.

Gunrunnerjohn--is this the same problem that your INSULATED TRACK SIGNAL DRIVER is designed to solve?

Don Merz

The addition of say a 100uf filter cap would damp out any relay chatter, and it could likely fit just between the two terminal blocks on your minimalist perf board Stan. But then we've added 50 % more components and probably 25 cents in cost. And assuming a 12vac supply, the coil voltage will be up around 16 vdc, which might be a bit much? What's the max voltage rating for that coil?

Rod

@Don Merz 070317 posted:

Gunrunnerjohn--is this the same problem that your INSULATED TRACK SIGNAL DRIVER is designed to solve?

Don Merz

Yep, and I added DCS signal protection and some extra simple wiring options. Also, the cap and the optional expansion cap allow you to have a delay before the relay drops out.

There's no magic in my board, just a solution to a problem that crops up a lot.

@gunrunnerjohn posted:

Don't you potentially get relay chatter with no filtering?  Some relays are too quick to react and you hear the contacts chattering.

That too was my first thought.  That is, Lionel's 2-component "AC relay" has 15A relay which suggests a relatively hefty relay-coil which could store more holding energy in its magnetic field.  Their documentation does not provide any discussion or analysis of chatter...just this diagram:

Not to make the eyes glaze over, nor to hijack Rod's thread but here's the Dollar Tree AC-relay with 14V AC applied:

When the full-wave rectified voltage drops to zero (red trace), the relay coil current (orange trace) does NOT collapse to zero.  Lots of tedious mathematics to analyze but one way to look at it is the coil stores energy and when the voltage disappears the coil releases that stored energy to keep the relay closed.  As the scope shows, the coil current only drops to ~5 mA when the coil voltage is 0V.  This 5 mA is more than adequate to keep the relay closed until the voltage returns just a few milliseconds later.  As shown in an earlier post, this particular HK4100 relay only need about 2 mA of coil current to keep it closed!  Lather, rinse, repeat.

I set up the scope to calculate and display the real-time relay coil power in Watts (white trace)...since, as they say, "heat kills."  What's coincidental to the point of seeming contrived is with 14V AC applied, the average coil power was exactly 0.2 Watts...which is the nominal coil power for this relay!

Interesting, the relays I've used have chattered on full-wave AC without a filter.  That's why I went with a regulator and filter on my relay module, just figured it was the path of least resistance.

@Rod Stewart posted:

The addition of say a 100uf filter cap would damp out any relay chatter, and it could likely fit just between the two terminal blocks on your minimalist perf board Stan. But then we've added 50 % more components and probably 25 cents in cost. And assuming a 12vac supply, the coil voltage will be up around 16 vdc, which might be a bit much? What's the max voltage rating for that coil?

To be clear, I think we are dealing with 2 types of relay chatter in the context of O-gauge train operation with AC voltages.

1) 60 Hz chatter where many of these small relays are mechanically "fast enough" to rapidly move its contacts thereby buzzing 60 times per second (if half-wave rectified with a single diode like your design) or even 120 times per second (if full-wave rectified with a bridge rectifier).  A 100uF capacitor would certainly provide an adequate reservoir of stored energy to remove the line frequency chatter.  Some techno-geeky back-of-envelope calculation that I don't know the answer to as I type this so let's see if I make a fool of myself:

While the nominal coil power of this relay is 0.2 Watts, we know (once closed) it will stay closed if you provide as little as about 0.002 Watts (1.2V x 2mA) which is pretty remarkable when you think about it!  We'd need that power level for, at most, 1 60 Hz line-cycle for half-wave operation and possibly even a chopped-sine transformer with pulse-like AC.  Let's just call it 20 millisec.  0.002 Watts for 20 millisec = 40 microJoules.  Bear with me.  A 100uF capacitor charged up to 16V DC stores about 10 milliJoules of energy...or 250 times more than required to deal with line-frequency chatter!  I was doing the math as I typed so maybe I'm off a bit but this sounds about right.

2) The other relay chatter is from intermittent wheel-to-track contact in insulated-rail triggering.  You need much more capacitor energy to mask this type of chatter.  Let's just say you want to mask drop-outs of 100 millisec or 1/10th second which seems a ballpark duration for flickering signal bulbs I've seen.  So that's about 50 times longer than the line-frequency chatter but a 100uF still stores adequate reserve power for that task too!

Again, the Dollar Tree version (NO capacitor) has no hope of guarding against the 2nd type of relay chatter!  This version is narrowly focused on the dual/multiple 153IR signaling application where you don't have the insulated-rail wheel chatter.

Separately, you ask about the coil voltage if a capacitor were placed in the Dollar Tree version.  If it charges up to 16V, that would indeed push the 15.6V spec'd max voltage.  So, yup, that would be a case where a resistor or whatever to limit coil power would be a good idea.

No secret to what my relay module has on board, it is designed to eliminate both types of relay chatter.  It's a few more parts, but it has considerably flexibility.  The resistor was just included to minimize the power dissipation of the linear regulator.  It isn't technically needed, but with 18V powering the relay module, and the relay energized for long periods of time, the regulator got up to around 80C-85C, so I thought I'd split the load a bit and dissipate a little power in the relay.  It dropped the high temperature to around 70C-75C, mission accomplished.  I could probably make it 100 ohms and help the cause a bit more...

Click on graphic to expand

You can use @gunrunnerjohn's relay module with multiple IR153s. Requirement is that all the IR153s have to be powered by the same AC source - this is because we connect all the Normally Open (NO) outputs of the IR153s together, and there will be no electrical problems if all are from the same source. Since the states of the NO output is open circuit or AC, we have to wire to the relay module a bit differently than if you were connecting the relay module to an Isolated rail.

Simply connect the GND that sevices the IR153s to the Sense Rail input of the relay module. Connect the wire from the IR153s NOs to the Center Rail input of the relay module.