The Southern Michigan Railroad Society bought two cabooses from Conrail in 1985. These are used for passenger operation. These were wired for 12V electrical power, and were supplied with circa-1970 Edison batteries. Last year I cleaned one up with a vinegar soaked rag, and charged it.

It worked.

At first, the battery could be easily exhausted. Then the interior lights, desk lamp and marker were replaced with LED bulbs. This is fairly well concealed behind the original lenses. The rear marker is brighter and draws 0.08 amps (1 watt). The interior lights no longer flicker when lights or the marker switches on and off.

The caboose parks far from a power plug. Running extension cords was a major chore. So this was attached to the irons which once held the roof walk.



The strapping material "springs" in place and is held only by one bolt in the corner. Above it, a sheet of plexiglass (lower right corner) is bolted across the irons to protect it from branch hits. The wires were routed properly later.

Below decks, it connects through a top maker's least expensive charge controller to the battery. This prevents overcharge of a lead-acid; it doesn't get near overcharge for an Edison. It also uses PWM to turn overvoltage into more useful current. Aside from the addition of this module, the original battery wiring panel is unchanged.

The panel is rated at 20W (aimed straight at the sun and not covered with dirty plexiglass), so I expect 5-10 watts out of it. The panel was sized for the available space, though I later learned a 30W panel would have fit. Measure twice... sigh! There's a similar site at the other end of the caboose.

In the month since it was installed, there have been several revenue trips, much use, and no plugging in chargers. Nonetheless it managed a night run quite successfully, burning a rather bright LED headlight.

Since operations are generally once a week, this serves as a completely automatic way keep the battery topped up and ready. The solar panel is rated at 20W if aimed directly at the sun and not covered by dirty scratched up plexiglas. I expect 5-10 watts ordinarily, given that it has a whole week to top up in most cases. During consecutive days of operation, the crew would probably need to cut the headlight when not near crossings, or top up with a battery charger. This is December, so the light will only get better.

I'm rather pleased with the results so far :)

[ amazon links pay me commish http://cmp.ly/5 ]

Very interesting way of integrating modern technology with operations. May I assume the solar panel is waterproof? The only question I have is whether the plexiglass affects the wavelengths of the light coming in. You'd have to run tests both with and without the protective plastic, but right now, if it works, so much the better!

What were your costs associated with doing this? Where did you purchase the equipment from? How did you set up the LED bulbs in the various housings? We have been talking about the same issues regarding lighiting equipment for display purposes.

John Deck

> waterproof?
You bet. They live outside. This one more than most, the builder was a real fan of epoxy (or hopefully, urethane "epoxy" as it is UV resistant). Many solar panels have a little junction box with terminals inside. Not this one, it just had 6" wires coming out of a potted box.


Good question, certain bulbs were devilishly hard to find. I hunted down LEDs that would fit the sockets. I did not alter any housing at all.

The 20 watt solar panel was $63. (the 30 watt would have been $90.) The Controller was $28.

Red marker: about $5 to $9, they are commodities.
Interior lights: four at about $20 each.
They are coming down. Everything is coming down all the time.

Total project cost around $200 excluding the headlight.


DETAILS ON BUYING MARKER LIGHTS:
The flashing marker light used an 1156 automotive base (1-lamp, typically parking light). This is a very common size of LED sold at all truck stops and many auto parts. Here was my first try, note how it lets you pick the angle, very useful. It was too big and struck the reflector bezel, so I went down to a smaller bulb and it was plenty bright (on axis).

Color LEDs are 20+ times more efficient in this category because of 3 things.
a) makes only light and not heat (4-6x more efficient)
b) makes only the right color, so the filter isn't wasting most of it (2-4x)
c) aims and focuses where you want, reducing reflector losses (1-4x depending on your reflector/lens).

DETAILS ON BUYING INTERIOR LIGHTS:
You know most bulbs in your house use a standard screw-in base called an Edison E26 also called E27. Our caboose lighting uses the same base, except it's 12VDC. Don't mix them up with your other bulbs and label zealously. This goofy voltage/base combo is commonly used in RV lighting, and that's a good search term.

They are a challenge to find in LED and CFL, and even then, it pays to be picky. With LED and CFL bulbs, the issue is color "temperature" (which is really hue, not temperature). Think of those annoying blue headlights some cars have. The unnatural bluish colors are called "cool white" and have higher K-numbers like 5100K. Traditional incandescent colors are called "warm white" and have lower K numbers <= 3300K.

The old incandescent bulbs go in all directions. So do CFL's. Most LED products are directional, centering on the axis of the bulb. The ceiling lights preferred side-light, but the reflectors and lenses made good use of the LED light. That was not so in the desk lamps, and I needed to use CFL there. I'm not thrilled about it, they look ridiculous, and are fragile.

LED's I have found: My favorite is this, $20. which is warm white so it passes for traditional. There's also this very blue one (6500k), $22. And here is a $13 one available in cold or warm white.

CFLs I have found: here , and one other whose source I don't recall.

eBay is a treasure trove of these products, possibly worth a look if you don't mind dodgy sellers. I just saw an all-direction LED there which would solve the desk lamp problem. Remember they have to be 12V.

Summary:
Test any of these when they arrive. Users report a lot of DOA issues (and first-month issues on the CFLs). I expect them to be reliable after that.

Some of these lights are multi-voltage, e.g. 12-24 volts. Can this work outside that voltage range? Good chance. They are using electronics internal to the bulb to "chop" the input voltage into the voltage they want internally. The chopper doesn't care unless you exceed its capacitors' breakdown voltage, or force it to draw too much current because voltage is too low.

[ amazon links pay me commish http://cmp.ly/5 ]

This is a very cool use of current (no pun intended) solar technology. Many of us with RVs and travel trailers have added solar power to keep batteries charged when "dry camping" (no hook-ups).

What's really neat (IMHO) about this particular use though is that there is historical precedence. A number of railroads experimented with solar power on cabooses years ago. Two that stand out in my mind are some of the Southern Railway bay window cars, which used solar power strictly for powering the marker lamps on yard and transfer cabooses, many of which did not have axle-driven alternators.

http://www.srha.net/public/x201update13p2.htm

Another car that made use of solar technology was Western Pacific 452, which was rebuilt by WP in 1979 with many experimental technologies, including solar panels that provided a slow charge to the batteries to keep them topped off whether the car was in motion or not. You can see the panels on the roof of the car in this shot.

http://www.wprrhs.org/theheadlight_erra ... 52_002.jpg

AFAIK, Conrail never used solar on its cabooses but at least the use of solar on cabooses in general is not without some historical precedence. Very cool!

There is also precedent for this idea. The Southern used solar panels on quite a range of their cabooses (cabese?)

Resto on these has been attempted with quite good results.
http://www.srha.net/public/x201update13p2.htm

I came across this item, which has a very interesting tie-in (and note the lead photo, featuring an 0-6-0T):

http://www.guardian.co.uk/environment/2 ... lar-panels

Official site; too bad some people in Pennsylvania seem unable to have a similar vision for some of the anthracite facilities that remain there:

http://www.museumwales.ac.uk/en/bigpit/

There has been a discussion of solar and other renewable power sources for an electric railway here, too:

viewtopic.php?f=1&t=29372

viewtopic.php?f=1&t=29799

Have fun.

The primary motivation in this caboose (OP) was to preserve the Edison battery. The only "green" was green extension cords, specifically, not having to run them! Because historically nobody ever bothered, which is why the battery had been neglected in the first place. Anyway everyone presumed the battery was long-dead, since conventional wisdom is batteries last 5-10 years.

Now with the solar panel, the battery gets used, and displayed. People get another think about that conventional wisdom. Which has an impact on electric cars, homepower and other 21st century energy management strategies that make a lot more sense with a more durable battery. So I guess it is kinda green!

I'm happy to give an update after a year of actual use.

First, the system works exactly as intended. No one had touched the system since December, they went to use it in May, and it ... just worked. The battery stays "topped up" and is not overcharged and does not use water excessively. (the controller is intended for lead-acid batteries and will tend to undercharge a nickel-iron.)

Concerns about undercharging and sizing were tested this cloudy October with three 3-day weekends in a row, Thurs-Sat-Sun, Thurs-Sat-Sun, Fri-Sat-Sun. All using the big headlight continuously the whole trip, for backing moves, so 3 hours a day. I expected the final back-to-back 3 days to finally exhaust it, but it didn't.

The headlight was originally dual 18 watt lights with 30 degree beam spread. I upgraded the headlight to a single 42W LED headlight with 8 degree spread. The narrow spread means the light is much brighter head-on, and crews call it brighter than the locomotive headlamp. Sitting in the yard, the old lights clearly illuminated 1000 feet or so. The new light throws more like 1/2 mile. Having flagged crossings with the train coming at me, the light is bright and eye-catching. You can't not notice it.



One other thing about this LED light, notice its operating voltage is 10-30 volts. That's because it uses a switching power supply to supply the LEDs constant current, and allow a wide range of input. A side-effect is that it works like a "Joule Thief", drawing more current as input voltage sags. If our battery sagged, the light would compensate. (this picture was on the last run of the last weekend, the worst case scenario.) In this type of device, minimum voltage is defined by maximum current. Maximum voltage is defined by insulation inside the device. That tends to have a lot of "headroom", so it'll probably work fine on 32V or even 64v. If you're running it off a resistor ladder, you only have to get "iin the ballpark" and the light handles the rest.

Edison battereies were used almost universally on railroad passenger cars. They were serviced by shop workers. I am told that every five years or so they were taken apart, all the grids were thoroughly cleaned, then they were reassembled with freash electrolyte and were good for another five years of service. The Edison batteries were reported to to last for the life of the car.

I find it funney how on a shove move the FRA only requires a "white light" allowing any shop to make one out of common parts. But for a FRED you need to have an FRA approved unit. Also considerimg that during daylight all you need for a EOT is a red flag this dosen't seem to make much sence.

P.S. LEDs do get hot and if you don't take this into consideration your LED won't last long.

Depends on the LED. If you are buying a pre built LED assembly, that is not your worry. If you are building your own assembly then it matters.

Single color LEDs can run cool. If they are not, it is because they are being overdriven to make more light from a smaller device, reduce current and that should fix it. White LEDs make blue internally and rely on "phosphors" to map their blue light into the full spectrum. Phosphors by nature run hot. Fluorescents do the same but with UV. Feel a fluorescent bulb.

You can make a white LED by combining several color LEDs to look white. E.g. Get an RGB LED light or strip and light up all 3. But the color iss notchy and eon't give good rendition.