I needed to make a decent light on the cheap, which I started in a different thread.

I used one of those old Eveready flashlights (2 d cell type with the white switch, very common in the US) that I had lying around. I replaced the light bulb with a 5D cell (5V-7.5V) 5.5 watts, 4500K 150 lumens 920ma) bulb. I drilled 2 holes in the bottom of the flashlight and ran wires into it and soldered the pos directly to the bulb and wrapped and soldered the neg wire to the copper connecter that is part of the switch. I ran the wires to the back of the bike and put an 8 AA cell holder and put 6 2500mah 1.2V nimh cells(18 watt hours) in it with 2 dummy cells (just shorted the 2 cell spots with aluminum foil). The cell holder has pretty stiff springs, so I think the pack itself isn't deficient, and I'm pretty sure it's good enough for <1 amp

(Also, I moved the pack to back of the bike to free up the cage, I bought a small dog collar at the dollar store and am using it to hold the pack in place right above the rear brakes, can't even really see it, looks much better now)

Last night it was very dark and raining and I had brought along a spare set of batteries in case the ones I had died. With the ground wet and the rain and the poor lighting, I was having a really hard time seeing the road conditions ahead of me (I have it tilted so the main light is about 25 feet in front of me). Wanting to go a bit faster, took one of the dummy cells out and put another cell in it and it really, really helped and got me home that much faster.

Just out of curiosity today I put a multimeter across the cell pack with the light off and read 9V, pretty typical for 7 NIMH cells. But when I turned the light on, the volts dropped to 7.5 which is 1.07V ( down from 1.285)per cell, that's a pretty big drop. The cells are in a good state of charge and aren't old and are a decent brand. The 920ma draw is only .368C and NIMH cells should be able to sustain at least .5C or even 1C or greater. The only thing I can think of is that either the batteries are damaged, or the switch is making a poor connection to the bulb housing and causing too much resistance in the circuit (I guess I can check that). If so I can just solder the negative lead to the bulb and wire in a simple switch, or is this drop normal on such a low draw?

I know that nicads have much lower internal resistance, so I was thinking of replacing the battery pack with C sized 1.8ah ni-cad cells. This would give me less run time, but I could make up for the difference by putting 2 6 cell batteries in parallel, but I can't do that until my rack gets delivered. Since they are made of very light plastic, the weight is no big deal. Would you anticipate a similar voltage drop on a nicad pack?

This setup is working for me, but I don't want to damage the cells. Should I go and use the Nicads? According to the data sheet both the C and D cells are made up of a sub-C 1.8AH cell (they are the yellow Eveready D and C cells, they are identical)

Any suggestions?