Two quick questions - Relays and Capacitors

Check out @gunrunnerjohn 's stuff at Hennings Trains. He has a lighting module that will work for your passenger cars along with an insulated rail relay. Not sure if it will work as you need but he's on the forum all the time and should see this reply.

https://hennings-trains.shopli...cal/john-will-assoc/

The signal issue sounds like you probably need a custom solution, I don't see any of the "canned" solutions doing what you're describing.

OTOH, the flickering lights in passenger cars I have a handle on.

LED Lighting Regulator

• Requires a small fraction of the power of conventional lighting
• Provides adjustable lighting intensity
• Storage capacitor on module provides flicker-free operation
• MTH DCS compatible, will not degrade DCS track signal
• Simple connections to power and LEDs

@Cheap&NothingWasted posted:

1.  I am attempting to retrofit two-color track occupancy signals into a section of layout that is already completed.  It has three tracks running parallel that I want to signal as one block (they merge at one end into one track and two tracks at the other).  I know having an insulated rail would be the easiest, but the wiring is not setup to allow it.  Best I can do is about a foot of insulated rail at each of the 'exit' points from the block (4 total).  ....

Trying to understand your block description.  If the 3 tracks merge to 1 on one end and 2 on other end, don't you then have only 3 (not 4) 'exit' points?

So if I understand your idea.  A train entering the block from EITHER direction hits ANY of the insulated-rail triggers and sets a relay (or whatever) to signal RED.  When the train exits the block on the other side, it hits the insulated-rail trigger and resets the relay to signal GREEN.

I'm not aware of an industry-standard name for this behavior that would make searching easier, but a "bistable" relay might work.

Or you might see the term "toggle mode" to describe a relay behavior where, for example, pressing button opens garage door - then pressing same button closes it.  As shown in above example, there is a bunch of electronics surrounding the relay itself (the blue cube).  That is, relay modules of this ilk use a standard relay...and surround it with DC-powered circuitry to effect the alternating on-off function.  Using a more sophisticated/expensive relay (like a latching one) can make the supporting DC-powered circuitry simpler/cheaper.

Problem is I don't think you will find such a relay or relay module for 18V AC operation...or if so it will seriously lighten your wallet.

But if you want to pursue this tack, 12V DC seems to be the most common voltage used for these bi-stable/toggling relay modules.  Hence, there is some additional homework to be done.

Separately,

1. Suppose you turn on the "system" and a train is entirely within the block.  As described, the signal would probably go "green"...and when it leaves the block it would toggle the relay and now the signal is red.  Hmm!

2. Suppose a train enters the block and signal turns red.  But the train stops on the insulated section and reverses and backs out of the block.  The signal is still red.  Hmm!

3. Suppose you have an unusually long consist that spans the entire block.  Train enters block and signal turns red.  Train reaches exit trigger but entry trigger is still active.  So when train exits block the relay only received one trigger and signal is still red.  Hmm!

And so on.  I appreciate the keep-it-simple (or chips-fall-where-they-may) approach but it can be useful to identify the must-haves vs. nice-to-haves vs. don't cares... OGR being a discussion forum after all!

@Cheap&NothingWasted posted:

...

2.  I'm looking for recommendations on capacitors to stop my interior car lights from flickering.  Some have LEDs with on-board rectifiers and resistors and some are plain bulbs using whatever track voltage is (~18VAC usually).  I don't know much about capacitors, so I have no clue what specs to look for.  I plan on buying these in bulk, 50 or 100 at once.

The two key specs for a capacitor are capacitance and the DC working voltage.  Capacitance is how much energy it can store and for this application would be measured in microfarads typically abbreviated uF.  The voltage spec is the maximum voltage that can be stored.  There's all kinds of EE math and equations but to get your arms around the application - if for no other reason than to get a cost estimate - for your LED cars you're looking for something like 470uF of capacitance and 35V DC working voltage.  You can go higher on either spec.  There is a sequence of common values.  For example 330uF or 680uF might be the next lower/higher values you'd find.  And 25V and 50V DC next lower/higher voltage ratings. So you might be able to tack-in just a capacitor to your LED cars and effect some level of flicker reduction.   You say your LED cars have a rectifier in them.  The rectifiers convert AC voltage to DC voltage.  Capacitors only "store" DC energy and can/will react violently if AC track voltage is directly applied to them.

In round numbers, LEDs as used in O-gauge lighting applications are ~10 times more efficient than bulbs.  So for ballparking the situation, you need 10 times the capacitance so maybe 4,700uF.  The voltage spec is the same...so 35V (or more).  The physical size of capacitors with these specs are often a show-stopper but assume space is not an issue.  Since your bulbs are directly connected to AC track voltage, you must splice in circuitry such as rectifiers to convert the AC to DC.  A rectifier will be less than 5 cents and will be physically small relative to the size of the capacitor.  Because of the higher power involved with bulbs, there are reasons why you may want to use 4 rectifiers (instead of 1) but these are details.

There are (at least) two considerations specific to your application:

1) A capacitor is a "dumb" energy storage component.  Unlike a battery which maintains a relatively constant voltage as it provides power, a capacitor predictably loses voltage from the moment it starts to provide back-up power.  And on the "charging" side, when storing energy into the capacitor, the voltage might ramp up to more than you'd expect...for example for 18V AC track voltage, the DC capacitor might reach ~25V DC.  How LEDs vs. bulbs react to what can be wildly swinging voltages begets more equations, more do-the-math, and so on.  Can be boring to the non-enthusiast.  Most flicker-reduction circuits employ some type of regulation circuitry so the voltage presented to the lights is more constant (like what a battery would provide).  It's not that these circuits are particularly expensive, but it's a far cry from simply tacking in just a capacitor.

2) If your car will be used in a MTH-DCS command environment, you need to add a so-called DCS-choke between the track pickup and the capacitor since capacitors in lighted cars are known to degrade the DCS control signal.  If this applies to you there have been many OGR threads on this.

---

As an aside, the issue/topic of installing capacitors in incandescent cars comes up every so often.  In 100% of the cases, LED conversion is the consensus opinion/suggestion.  I am simply addressing your capacitor question at face-value.  If for whatever reason you want to proceed with some scheme for flicker reduction for incandescent bulbs, I will work with you in the spirit of a high school science project.

FWIW, my S gauge LED illuminated passenger cars each use one 470mfd, 35V capacitor. They run in a Legacy environment and provide between 1/2 and one second of ride-through until the LED's dim and go out. These are full 80'scale cars so in S scale they are 16" long. You may want more than 470mfd for O gauge cars depending upon how many LED's are used.

@stan2004 posted:

Trying to understand your block description.  If the 3 tracks merge to 1 on one end and 2 on other end, don't you then have only 3 (not 4) 'exit' points?

A stub-end industry siding comes off of one track.  I am not sure if I want to make an 'exit' track there, but I figured I'd mention it in case that made a difference.

@stan2004 posted:

I'm not aware of an industry-standard name for this behavior that would make searching easier, but a "bistable" relay might work.

Or you might see the term "toggle mode" to describe a relay behavior where, for example, pressing button opens garage door - then pressing same button closes it.

Yeah, that might work, it depends on how long it will tolerate being triggered.  For $4 it's definitely worth a try, and if I fry it, oh well.  Thank you for the suggestion.  And now to think of it, most modern electronics now work this way.  No more solid SPST on/off switch, they have a soft-switch you have to hold down a few seconds to turn on or off.

@stan2004 posted:

Problem is I don't think you will find such a relay or relay module for 18V AC operation...or if so it will seriously lighten your wallet.

But if you want to pursue this tack, 12V DC seems to be the most common voltage used for these bi-stable/toggling relay modules.  Hence, there is some additional homework to be done.

I can use the 24VAC relay purchased when I thought I could make insulated rails for the whole block to have the AC pulses trigger the DC Relay.  Connecting two relays together is still easier than programming chips and such.

@stan2004 posted:

1. Suppose you turn on the "system" and a train is entirely within the block.  As described, the signal would probably go "green"...and when it leaves the block it would toggle the relay and now the signal is red.  Hmm!

2. Suppose a train enters the block and signal turns red.  But the train stops on the insulated section and reverses and backs out of the block.  The signal is still red.  Hmm!

3. Suppose you have an unusually long consist that spans the entire block.  Train enters block and signal turns red.  Train reaches exit trigger but entry trigger is still active.  So when train exits block the relay only received one trigger and signal is still red.  Hmm!

For 1 and 2, I plan on wiring a momentary contact switch into the circuit so it can be set correctly should that occur.  For 3, it's a mill layout and the main track through the block is about 10 feet long, so any train longer than 10 feet is probably just a tail-chaser and wouldn't be paying attention to signals anyways Thank you for your help in this, I will let you know how it ends up.

@stan2004 posted:

Capacitors only "store" DC energy and can/will react violently if AC track voltage is directly applied to them.

Yes, my experience with capacitors and AC power, both back in high school shop class and recently repairing my TMCC J 611, was that mixing the two is an appropriate Independence Day activity .

@stan2004 posted:

The two key specs for a capacitor are capacitance and the DC working voltage...Because of the higher power involved with bulbs, there are reasons why you may want to use 4 rectifiers (instead of 1) but these are details.

This is really good info that I doubt I would have found easily, if at all, so thank you for that.  I've put diodes in some bulb cars to half-rectify the input 18VAC down to 9VDC (to prevent overheating), so if I wire the cap in after the diode, would it still work?  Also, most of my bulb cars left are cabeese, work cars, etc - the kind where you can't see much of the interior, unlike a passenger car.  So cap size is definitely less of an issue.

@stan2004 posted:

Most flicker-reduction circuits employ some type of regulation circuitry so the voltage presented to the lights is more constant (like what a battery would provide).  It's not that these circuits are particularly expensive, but it's a far cry from simply tacking in just a capacitor.

That sounds like a good reason to rein in the project and just install one of GRJ's modules in the worst offenders.  Our 2-rail to 3-rail conversions have sometimes left a car with just a single pickup roller, those flicker almost constantly unless on newly-cleaned track.

@stan2004 posted:

2) If your car will be used in a MTH-DCS command environment, you need to add a so-called DCS-choke between the track pickup and the capacitor since capacitors in lighted cars are known to degrade the DCS control signal.  If this applies to you there have been many OGR threads on this.

And we are converting over solely to DCS (with TMCC hooked into it).  Thanks for enlightening me about another quirk of DCS.  We already have 24VAC bulbs hanging off the ends of several sidings on the layout to improve the track signal (it does work!).

@stan2004 posted:

As an aside, the issue/topic of installing capacitors in incandescent cars comes up every so often.  In 100% of the cases, LED conversion is the consensus opinion/suggestion.

Yeah but I can't just simply pull bulbs and put in LEDs + rectifier + cap.  If I'm opening up my C&NW MTH Challenger coaches, then I'm wiring in bicolor white/blue LEDs with a switch so I can have prototypical day/night lighting.  See how a simple conversion can snowball?

Thanks for your help.  I'll work on my worst flickerers and let you know what I did.

@AmFlyer posted:

FWIW, my S gauge LED illuminated passenger cars each use one 470mfd, 35V capacitor. They run in a Legacy environment and provide between 1/2 and one second of ride-through until the LED's dim and go out. These are full 80'scale cars so in S scale they are 16" long. You may want more than 470mfd for O gauge cars depending upon how many LED's are used.

That agrees with what Stan2004 said.  All my cars are 70' or 72' in O-scale, so that should be adequate for my needs too.

@Cheap&NothingWasted posted:

A stub-end industry siding comes off of one track.  I am not sure if I want to make an 'exit' track there, but I figured I'd mention it in case that made a difference.

Yeah, that might work, it depends on how long it will tolerate being triggered.  For $4 it's definitely worth a try, and if I fry it, oh well.  Thank you for the suggestion.  And now to think of it, most modern electronics now work this way.  No more solid SPST on/off switch, they have a soft-switch you have to hold down a few seconds to turn on or off.

..

This is really good info that I doubt I would have found easily, if at all, so thank you for that.  I've put diodes in some bulb cars to half-rectify the input 18VAC down to 9VDC (to prevent overheating), so if I wire the cap in after the diode, would it still work? Also, most of my bulb cars left are cabeese, work cars, etc - the kind where you can't see much of the interior, unlike a passenger car.  So cap size is definitely less of an issue...

If you're still working on this.  The answer depends on what you mean by "still work."  The half-rectify method provides DC bulb power on 1/2 of the alternating AC voltage.  By adding in a capacitor after the diode, the capacitor will charge up during 1/2 of the AC cycle...and then discharge on the other 1/2 cycle to fill-in-the-blank so to speak.  As you increase the capacitor value, the bulb will get brighter and brighter.  There is some tedious math but one could calculate the capacitor value at which the brightness would be the same as with 18V AC directly applied to the bulb...something probably in the 1000's of microFarads.

A coincident effect is you gain some flicker protection.  So as you increase the capacitor size, the bulb gets brighter and you get additional flicker reduction.  My guess though is by the time you increase the capacitor to a large enough value to get acceptable flicker reduction, the bulb will be too bright.

Hence, if you want both a dimmer bulb AND flicker protection, you need more electronics than JUST wiring a cap in after the diode.

@stan2004 posted:

I bought two of those.  One didn't work at all, the other only switches when power is on.  I am sure they could have fit the bill, but the ones I received didn't even have the same part number on the PCB as in the pictures.  At least I got my money back without having to send them back.

Bought and bench tested this last night, works exactly the way I needed it too.  More expensive, but likely more robust too.  Plus I can run two different voltages on the output if necessary.

And I've not put any more thought into the car lighting problem, except to buy a dozen bayonet-base LEDs to replace Lionel 1445 bulbs in various cars.