Too much juice for Vans dimmer?

[McFly]

I finished installing an old school AT150 transponder in my panel yesterday. This is the last electrical component installation and the last hole in the panel is now filled. When I was testing the panel lighting last night, the entire panel went dark after a minute or so. The problem is repeatable (I can recycle the panel light after a minute). I suspect that I am taxing Vans dimmer beyond its limit and it is overheating but cannot say for sure.

My setup is 10 Vans gauges, 2 EL strips, 1 ICOM radio and 1 Narco AT150 transponder (this was the last piece installed and suspect it is drawing a lot of juice). Sorry, no current draw numbers.

Is my theory about the dimmer on target or should I look elsewhere? Assuming the problem is with the dimmer, would installing a larger heat sink help? Any suggestions? Thanks.

[vlittle]

If possible, you need to measure the total load on the dimmer bus at various voltage outputs. Given what you said, I suspect that the regulator on the Van's dimmer is going into thermal shutdown.

I have a smaller load than you do on my dimmer, and I calculated the junction temperature to be too high. I replaced the heat sink with a larger one, and now I'm confident that temperatures will be in spec.

For reference, my current load at 10.25 Volts output was 0.9 Amps. With a 14.1 volt input, this gives 3.465 Watts of power dissipation.

However, this is not the worst case. The worst case is at 4.8 Volts output, 0.55 Amps load, giving 5.115 Watts dissipation.

5 Watts is a lot, and needs a bigger heatsink than Van's supplies.

If you are technically inclined:

ThetaJC (TO-220 case) 3 degrees C per Watt
ThetaCA (heatsink) (whatever you choose depending on the part number)
DeltaTJ = Power(Watts) * [ThetaJC + ThetaCA]
TJ (Junction Temperature) = DeltaTJ + TA (Ambient Temperature)

With a 40C ambient air temperature (under the panel), a heatsink of 14C/Watt, I get 126C as maximum junction temperature of the regulator. (I'm comfortable with 125C, so this is close enough).


Vern Little

[McFly]

Vern, thanks for your help and confirming my suspicion.

I believe I have found an elegant solution that is going to work even though I used bubba engineering, which is more my style.

I scavenged a couple of CPU processor heat sinks. The sink I used is billet aluminum and looks like a bed of nails. I milled the base (with a belt sander) until it was 3/32 thick. I removed a few ?nails? on top and milled that area flat so that the sink could slip under the dimmer component that generates the heat. I then trimmed the sink to fit the circuit card, drilled it and attached it. I guess I increased the area of the heat sink by more than a factor of 100 but who knows.

I tested it in my 95+ degree garage with zero air movement (as registered by my personal sweat-o-meter). I used various dimming settings, each time letting the sink reach its temperature equilibrium. It worked great; the sink did get pretty warm but not quite hot. I?m going to call it good.