Next project is the LED upgrade of my Penthouse light. It's a florescent fixture that comes standard with the Penthouse top from Sportsmobile. You can see the light in this picture on the far end of the top before it was installed.

The light runs off the starting battery and other owners had reported concerns with the amount of power the fixture used. Because of this I was always leery of using it too much when boondocking. Decided to convert the light over to LED since I could get the same amount of light for less power.
A lot of other people have done a similar upgrade. I could have done it more simply but thought this was would be a good chance to mess around with some different led's.
I want the replacement leds to produce at least the same amount of light as the original florescent fixture so need some way to measure the light output. A photoresistor should work. A photoresistor is a type of semiconductor whose resistance decreases with increasing light intensity.

Hook a voltmeter across the photoresistor leads and place it on the resistance setting. The voltmeter reading won't give me an absolute value for what the light output is but I can use it as a relative value to compare florescent to led.

Stand the fixture on edge on my workbench and center it on this "line" where two planks meet.

18 inches from the light is the photoresistor centered on the same line.

The measurement from the photoresistor is expressed in ohms.
The brighter the light the lower the value.
Here's what I'll be using as my power source. Two 6 volt golf cart batteries connected in series which then supplies 12 volts.

I'll be replacing the florescent light with this type of led. It's 75 white leds mounted on a ridged board that runs off 12 volts. It comes with a plastic tray that the board slides into for mounting. I bought it at
DealExtreme for $10.28 which included shipping.

The good thing with buying from DealExtreme is the items are low priced. The bad thing is you're buying directly from China and it can take weeks for your items to arrive.
Here's the back of the led board. The 75 leds are actually 25 groups of 3 leds. Each group of leds has it's own resistor so the board could be separated between each grouping and wired directly to a 12 volts source. Since the board is 20 inches long it will have to be cut. Mark the board between two led groupings close to the middle.

Using a hacksaw carefully cut across the board.

After separating the board into two pieces, one piece will still have the power wires attached to it while the other piece will need wires soldered on.

Cut the mounting tray same length as the led board.

For testing use double sided tape to stick the two led boards to the back of the fixture. Flip the fixture back and forth to measure the light from each type of light.

The florescent light gives a reading of 3.2 ohms and the led is 2.8. The led reading is less than the florescent light's reading so the led is outputting more light.

Here's where I almost made a big mistake. The photoresistor is measuring the light directly in front of the fixture which is where the leds have an advantage since they are directional while the florescent light is more scattered. I also forgot to test with the diffuser in place.
Test again with the diffuser.


With the diffuser in front the florescent reading is 4.1 while the led is 4.0.
I also measured the amount of power both lights were using. The florescent is using 1.2 amps and the led 0.45 amps. Since the led is using less than half the amps as the florescent and recognizing the bias the photoresistor measurement has towards the leds I decided to add one more 10 inch segment of leds to the fixture.

The light output with the diffuser is 3.4 and power usage is now 0.69 amps.

continued -