I replaced the power transistor on the board in my bike light with a higher rated, larger one. It handles high/low/off to the light, which is a BB500 driving two 3W TV1Js. There's some background info here, but it's not really necessary.
The new transistor is a Radio Shack TIP31 NPN transistor in a TO-220 case. It works great as a replacement, handling both high and low modes without excessive heat (it gets noticeably lukewarm after 1-2 minutes, and definitely warm in 5 minutes).
However, I was metering it in operation, and I managed to short L+ and L- through the multimeter in current mode. The batteries were around 4.6V under high-mode load (around 861mA to BB, 1380mA from batteries) before the mistake. The path was battery, transistor (collector-emitter), multimeter. I realized something was wrong when the emitters turned off, so the short was in place for up to a second. Not completely understanding, I left the light running for 15 more seconds, then I noticed a funny smell. It took me a few more seconds to identify the transistor as the source and to burn myself on it; then I cut power.
The multimeter and the transistor seem to be okay. The multimeter measures up to 4A, and its fuse didn't blow. The transistor, once allowed to cool down, performs exactly as it used to. Its relevant max ratings are Ic = 3A and dissipation = 20W.
Should I worry about the transistor and just replace it now? Once I finish building the light, it will be very difficult to replace anything. The heatsink will be epoxyed into place in the headlight body, and the transistor's back will be AA'd to back of the heatsink. In normal use, the transistor shouldn't get much hotter than warm, if the heatsink is as effective as hoped.
The new transistor is a Radio Shack TIP31 NPN transistor in a TO-220 case. It works great as a replacement, handling both high and low modes without excessive heat (it gets noticeably lukewarm after 1-2 minutes, and definitely warm in 5 minutes).
However, I was metering it in operation, and I managed to short L+ and L- through the multimeter in current mode. The batteries were around 4.6V under high-mode load (around 861mA to BB, 1380mA from batteries) before the mistake. The path was battery, transistor (collector-emitter), multimeter. I realized something was wrong when the emitters turned off, so the short was in place for up to a second. Not completely understanding, I left the light running for 15 more seconds, then I noticed a funny smell. It took me a few more seconds to identify the transistor as the source and to burn myself on it; then I cut power.
The multimeter and the transistor seem to be okay. The multimeter measures up to 4A, and its fuse didn't blow. The transistor, once allowed to cool down, performs exactly as it used to. Its relevant max ratings are Ic = 3A and dissipation = 20W.
Should I worry about the transistor and just replace it now? Once I finish building the light, it will be very difficult to replace anything. The heatsink will be epoxyed into place in the headlight body, and the transistor's back will be AA'd to back of the heatsink. In normal use, the transistor shouldn't get much hotter than warm, if the heatsink is as effective as hoped.
