What is your source for this information? I've never heard of this and have read a number of books on solar, and lived with solar electric systems nearly my whole life, I know that LED's put out a small amount of power while in the sun (you can easily try this with a volt meter and any LED), but I've never heard of a normal light bulb being able to do this, so unless you have something to back this up, I'm reasonably confident that this would not work.

MD2000: I often use Polyurethane to glue stuff, it has more tensile strangthe then silicon caulk and doesn't seem to brake down in sunlight like silicon caulk does.

"normal" lightbulbs can't do it, you need two electrodes, and a lot of headlamps with a common ground, you need to get a burned out one so the circuit is open one side. It's a thermal effect more so than a photic effect. It's related to thermionic emission.

Actually, I'm looking forward to getting my hands on some reject or surplus OLED panels to play with that LED effect. Though I also wonder if there's a way of wiring up Thin Film Transisior LCD displays to get something out of them.

Edit: Oh here we go, best explanation I can find at the moment...
https://en.wikipedia.org/wiki/Thermionic_converter

I have heard of this before, but the efficiencies you are quoting are with an optimized system, a headlight has a few serious issues.

1.Headlights front glass acts like a diffuser to homogenize and direct the beam, which would make the sharp focus necessary to heat the single filament impossible.
Assuming that somehow was not an issue, the pointing accuracy of the reflector to heat only one filament, and not the other would be pretty tight.

2. The two filaments are thermally connected at the base, and are in a vacuum, so the unheated one would soon reach nearly the same temperature as the hot one reducing the temperature difference and output.
3. Sunlight delivers about 1 kw / sq meter, the small amount of sunlight that a headlight would intercept would be a fraction of this.
4. Headlights are heavy and expensive, and would be pretty lumpy glued to a car roof :eek:

Sorry, I think I will stick to solar cells.;)

10 amps, 125watts is way too much.... My 97 honda D16y8 at 2k rpm uses exactly 37.5 watts (3 amps @ 12.5v) during the day with nothing on but the motor.

Yeah, I wasn't really seriously considering headlight assemblies, just the bulbs. They are only likely thermally tied at the ground plane. So when the main beam has blown, you use the ungrounded side for your cold side and the lowbeam sheilded inside the internal reflector as your hot side. I'd figure you'd see a max of about 12W out of a headlamp bulb behind a letter (approx A4 size) fresnel lens. Probably would get about 3 or 4 W out of a headlamp assembly if the focus landed in the right place.

Anyway was more just pointing it out as a curiosity.

measuring amp draw on your car
 

How did you measure that amp draw on your car

https://www.harborfreight.com/cpi/cta...emnumber=95652
https://www.harborfreight.com/cpi/cta...emnumber=95683

Neither of those lists DC amperage measurement in its description, but I know one does it, and I'm pretty sure both do. I have both. Too bad I can't find them.

The solar panels are a great idea and remember that the battery can keep you running for hours so that when you are parked the solar panels will continue to charge to make up for driving on shady roads. You can also charge the battery from the grid if needed to keep it fully charged later in the day so that it will always be fully charged overnight. If it needs charging then it will charge from the grid but if it doesn't it will not use much electric keeping it charged.

You might consider removing the alternator comletely if the belts can be run to keep the water pump happy so you can get rid of the extra drag of the belt.

I was just thinking that the Lithium Iron cells I recently purchased would be perfect for this application because they maintain a slightly higher output voltage during their discharge than a lead acid battery (13.4 volts) and hold it really well until almost dead and don't suffer from discharge cycling as much as lead does (3-4k cycles is typical) as well as are not bothered by sitting partically discharged. The 4 cells needed only weight 13.5lbs for a 40ah battery too.

Lithium Iron cells
 

Where did you purchase your batteries also, where did you get your info???

A guy in California was working with someone that purchased more then they needed for the projects they are working on so I purchased 18 LFP40AHA cells from him last month. The pack of 4 I actually put on a UPS scale 0-50lbs and the way Lithiums work is they age faster when fully charged and cycle life is not affected by letting them sit at a less than full charge state unlike Lead acid batteries which really loose capacity when left discharged. I have them sitting on my desk in front of me right now with 17 connected in series and when I measure 4 in a row I get 13.4 volts. I also have done a few discharge cycles on a couple of cells and find that they hold their voltage pretty steady for most of their capacity if you don't draw too many amps out of them.