Reversing Lights in a dummy car?

Are the lights meant to turn on only when in reverse?

A tether is the easiest way.  Failing that, we need to know if it's command or conventional.

There are some cheap remote control devices that could be pressed into service to send the light status to the dummy car,

Take one of these C:US:3160" target="_blank">4 Channel Wireless RF Remote units, rig a relay to push the button on the transmitter, and use the receiver in the trailing cars to operate the lights and other functions as well.

Yes, rear markers off and headlights on when the tail car becomes the front of the train (power car pushing from the rear, its headlights off and markers on - the board handles that)

Of course with the rear car as the tail car, red markers on, headlight off.

Right now I just have red lights connected to the in car lighting so they are on regardless.

Yes, rear markers off and headlights on when the tail car becomes the front of the train (power car pushing from the rear, its headlights off and markers on - the board handles that)

Of course with the rear car as the tail car, red markers on, headlight off.

Right now I just have red lights connected to the in car lighting so they are on regardless.

Actually, I think the lowest "out of pocket" approach is the tether, and it's also the simplest by far to implement.  Options #2 and #3 involve some custom circuitry.  The Minitronics 50-004-01 4 pin tether would seem to be the cheapest and simplest option at $13.95.  Also, it's just a simple wiring exercise to get it working.  The tether is so small that it's easily hidden under the couplers, so it's a very attractive option as well.

Well, he's running a subway so presumably this might be, say, 4 total cars (or more?).  That would be 3 inter-connects at $13.95 each or over $40 just for the wires/connectors.  Of course sourcing your own connectors and wires could save some coin.

There would still be some minor component level work - adding resistors or what not.  That is, I don't know how MTH does subway powered-unit lighting but if the headlights use bulbs and he uses LEDs in the tail-unit, then some fussing is still required.

The "problem" with a tether approach is it's like that potato-chip jingle from several decades ago - no one can eat just one.  That is, the first thing I'd be thinking if I installed tethers is how to add one more wire to put the interior lights of the trailing cars under powered unit control.  So when shutting down the engine, or turning off the Interior Lights under DCS control, all the lights in all the cars turn off.  Then, once I got that solved, I'd be thinking...

Yes, tethers would get expen$ive.  My R-17 is an eight car set (2 power cars at the head end) so that's seven sets of connectors and the R-142a set is another four.

Besides, these are not kept on a layout all the time, so each would have to be connected and disconnected for every use.

Before anyone asks why I don't just use the second powered car at the back end of the train, it's because I sometimes run less that eight cars and my rear car has a detailed cab end (safety chains, q-car H2A coupler, etc).

I wonder if you could use a small motor as a generator to power the lights, or at least let a circuit know which way the car is heading.

The power car in the R-17 set uses bulbs, the R-142a set uses LEDs.

For now I'll probably just wire 18v bulbs into the car light bar as I mostly use these with DCS so will have good track voltage.

Well, the bulbs vs. LED's for MTH are not an issue, a resistor solves the problem of driving the LED.

I agree that multiple cars does complicate the issue, I was thinking two cars.

I think for a multi-car solution, I'd be back to my little remote hacked up to do the job.  Detecting the direction of movement would involve wheel sensors and more mechanics, so I'd personally tend to lean the other way.

Of course, for a simple reversing light, only two wires are required, and you can "roll your own" perfectly acceptable tethers for peanuts from snappable headers.  Here's a sample, the mating connector is made with the snappable machine socket strips.  This brings the cost down to pennies a connector.  Obviously, for the tethers, you'd use all black heatshrink for the construction.

I forgot about the friction based switch, that would be a simple thing to do!  You could use a latching relay and a simple leaf spring so that when it goes in one direction it contacts and flips the relay.  Then it would stay there until you went in the other direction long enough to flip it the other way.

Can you identify a specific part/switch that would fit the available space?  I think the issue is what's readily available.  Once the directional sensor is identified, the "electronics" to then drive the correct lights is the easy part (in my opinion). 

I only mention the quadrature encoder because it is a readily available mechanism that fits the available space.  It's overkill as it provides speed information as well as direction.  All that's needed is the direction.

Truthfully Stan, I was thinking of a home-rolled solution for this.  I'd probably try to do it with a micro-switch with a long tab.

MTH subway trucks are derived from the PCC trolley truck block. So they're really just slightly slimmed-down locomotive trucks. 

---PCJ

Well then, I'd say some close-up shots of the dummy-car subway trucks in question are needed to continue down this path.

Another market for MTH to use the PS-2/3 technology of the Coors Beer Set with a small PS-2 board placed in the trail engine and commanded by the lead engine.  When in reverse the engine would tell the trail unit to turn on reverse light and turn off markers.  G

The mouse sensor idea sounds good.  Plenty of male mice around (the ones with the balls).  I'm no good at the electronics side though.  I can build something if it's laid out for me, but have no idea how to design it.

The mechanical friction lever would probably also work.  My thoughts on the shortcomings of that would be the reliability of the friction surface.

In HO, Marklin has a sleeve on a cab coach axle that is a loose friction fit. A wiper then touches contacts biased in the direction of travel. Reverse direction, and the wiper moves with the direction of the axle's rotation to touch the other contact changing the lights from red to white, or vice-versa.

Thanks Mike - a good idea, but as you can see the MTH trucks are pretty much a solid slug of metal.

-};>~

-};>~.  Well that's being polite.  My first thought was, Oh xxxx.

OK, is it practical for you to take a photo of the guts of a trucks?  Or is that a solid block too?  On the motorized diesel trucks I've worked on, the "lid" of the solid block comes off to expose the guts but who knows what surprises lay ahead...

I think I'm back to the remote control transmitter/receiver.

The only thing visible in the top of the dummy truck is the slot where a motor worm would go. Otherwise, it's a solid block where the axles go through--power is carried to the other axle via spur gears on the side of the block.

Another suggestion is to use an inexpensive electronic reverse unit to control the lights. I believe this idea was used on a set of Lionel LIRR M-7 cars.

---PCJ

I'd have to pretty much strip the car down to do so.  The trucks fasten to the floor from the inside, under the interior casting.  I don't think there's anything to see, just a flat metal top with a hole in it and a mount point.  The axles are threaded through the block with the wheels pressed on afterwards - no way to remove them.

With that solid block with no access to the axle(s) some of the other alternatives start looking pretty good!

But to follow this trail to a conclusion, dead-end or otherwise, what would you think of mounting a few tiny optical components in the yellow box shown below.  They would shoot an optical beam at some white stickers you'd put on the inside face of the wheel as shown.  I took the liberty of marking up one of your photos showing the area in question.  Mine is from a diesel powered-truck but it looks like there's a similar amount of wheel "overhang" beyond the solid block for a reflective optical sensor to shoot a beam at the wheel and get a reflection.

This would be the same concept as the tach sensor and the striped flywheels in MTH engines except the sensors would operate in quadrature-mode to sense direction. A few wires would have to find their way to inside the chassis. The optical components would only be a couple dollars and you'd still need a few more dollars of components inside the chassis. 

I'm presuming the subway is PS/2, so it has to respond to DCS commands.

Does this mean all the gears are in there for the dummy trucks?  In other words could one install just a shaft (no motor) with the correct worm gear and you would now have a shaft that spins when the wheels turn?  That would be even better than accessing an axle since the gear ratio would mean the worm shaft would spin faster than the wheel making direction sensing virtually instantaneous.

WRT the reversing unit, wouldn't this be for conventional only? Or is there something special about Lionel M-7 sets that it can talk to separate electronic reverse units like the MTH Coors set that GGG mentions can magically talk to the trailing cars?

I don't believe there are any gears in the dummy truck. The center of the axle may be exposed in the casting where the worm would normally go. I only have a power truck exposed (it was swapped out from an R-21 set that was randomly jamming), so it's only speculation about what's actually exposed.

A friend of mine runs two power cars in a train, so that took care of the lights as the two were MU'ed in DCS, despite running at opposite corners of the train.

I had overlooked the part about doing this in a command environment when mentioning the Lionel M7's. For MTH subways in a command environment, I'd just tap the headlight/tailight of the lead car, and equip the intermediate cars with pass-through tethers running headlight power to the tail lights and vice-versa. Its not like one does switching with subway trains, and the real ones have cables hanging beneath their drawbars too.  

The only question mark is whether the PS2/3 headlight/ taillight power supplies will tolerate a couple of additional LED's running off them. What say our resident PS2/3 experts?

---PCJ

There are no gears in the dummy truck and the axles are not accessible.  The idea of mounting a sensor on the end of the truck would probably work.  I still like the idea of the mouse sensor if it would work.

Stan, Lionel has a Small unit that receives TMCC commands and works in conventional also.  The original version was an LCRX.  It had directional lights, couplers and output for RailSounds.  Otherwise your correct, getting conventional to work is easy, but command changes won't work with a conventional reverse unit.  How about an accelerometer? G

Don't laugh George, the accelerometer isn't a bad idea, and I suspect they're pretty cheap.  I see some suitable ones on Digikey for around $11, like the Analog Devices ADXL78.

Going with the optical sensor, why not just read the ties of the track as the stripes?  Or would the contrast not be enough for the sensor to work?

I guess that depends on the track, how it's ballasted, etc.

This is a very interesting idea.  As discussed earlier, the current technology in computer mice uses "optical" technology rather than the rubber ball with rotating wheels.  They illuminate the mousepad or surface with a light (usually a red LED) and have a camera-like sensor that essentially takes pictures of the illuminated surface hundreds of times per second.  It's sophisticated pattern recognition software albeit integrated into a chip to analyze images looking for changes which reveals both speed and direction of motion.  All for less than $10 retail!  Pretty amazing stuff.

So, yes, I like the concept if that's where you were going but I think it would take some heavy-duty experimenting with lenses/optics to "focus" the camera on the rail ties...and then figure how to extract the direction info from the electronics.  I think it would might take all one's spare time for years to implement this from the ground-up.  And even then, the lights would disagree with the powered-unit until the consist starts moving.

OTOH if you meant the track ties are proxies for the black/white or on/off stripes on an MTH flywheel, then picking up on GRJ's comment, it too would take some experimentation to establish contrast levels for different track schemes and oddball cases such as how to ignore weird sensor data when going over, say, switches/turnouts.

I like the accelerometer, it would be reasonably stable and not require environmental considerations.

George, the mercury switch with a latching relay would probably be an interesting solution.   You'd want something that stays on after the acceleration ceases.  Of course, when you slowed down, the lights would reverse.

I think I'm back to the little RF transmitter triggered from the backup light in the locomotive.  It would not get confused.

Yes momentum of the engine would factor in.  I am just throwing ideas out until something sticks  G

I have opened up a couple of mice (the computer type) and I noticed that the scroll wheel sensor is small, compact, and easy to remove.  It would also be easy to mount with a small wheel running against the tread of the rail wheel.  I have NO idea how to power or use the output from this unit to do the direction sensing.  Does anyone here have enough knowledge to design the needed circuit?  I can build it if I have a schematic.

Alternately, can anyone point me to a page that explains how these sensors work and how to use them so I can muddle through the design process?

Here's a page that explains how optical encoders work.  I doubt you're going to want to design this in a circuit, a uP is clearly the way to do this task.

http://www.electronics-tutorials.ws/io/io_2.html

If you can extract the "sensor" part of the mouse and reliably couple it to the car's wheel via a rubber tire (or whatever), then I think the electronics side is within reach.

Since your persistence is admirable (many ideas here fade to black for whatever reason), I cobbled together this simple circuit.  It's somewhat half-baked using obsolete parts but it's what I had in my parts stash.  It demonstrates the idea of using an LED shooting a beam to 2 phototransistors.  The mechanical mouse probably integrates the LED and phototransistors into a single package but the circuitry to drive it is the same.  I hot-glued the parts to the truck of a diesel so it's kind of a mess but there's a 4-wire cable back to the circuit.  I only hooked up 1 Red LED and 1 white LED, a +5V supply needs to come from somewhere, and there are other short-cuts which make this more a proof-of-concept than a ready-to-build schematic.

A 10 sec video shows the circuit in action.  Note that this method operates at arbitrarily slow speeds.  Some of the "inertial" detection schemes including accelerometers or other momentum-based sensors can have problems with slow speeds if there's not enough "action" to trigger the detector.