Ooops!! Anyone know about short circuit caused by an illuminated switch?

ANYONE SEEN THIS PROBLEM WITH DIODES USED AS LIGHTS? IS IT A COMMON PROBLEM?

I suspect that the rad fan doesn't always come on when it should so I installed a bypass switch to turn it on manually if she got a little hot.

Worked great last summer going thru' the mountains but now it's turned cooler it's no longer needed. So I replaced the switch with an illuminated one (common make - made in UK) as I was replacing the radiator anyway.

Luckily, I warmed the car up to test the rad for leaks. I had my back turned when the idle speed pulled down and I heard a "Crack!" I turned to see smoke rising from the harness! That's one way to find the fusible links!

I put the meter on it - nothing made sense! Pulled the batttery ground - still nothing made sense. Finally replaced the fusable link with 6" of 20 GA wire (Don't ever do that) and powered the fan direct from the battery. 6" OF FLAMES 3" HIGH! (fusable links have non-flamable insulation - normal auto wire is SURPRISINGLY FLAMMABLE!!)

Well, found the problem but still took a while to figure out that it was the switch I had put in. It has a ground wire on it to power the light - this is hard-wired to the power-out side of the switch EVEN WHEN THE SWITCH IS TURNED OFF!!

If the light was a bulb this would be OK but since it's an LED (Diode) it appears to directly ground that line if power is fed into it from the other end (which happens when the temperature switch switches the fan on)

THIS IS VERY DANGEROUS!
1) The parts counter guy figures auto wire is almost as good as a fusible link
2) My Haynes manual (CRAP!) implies that wire will do!
3) There is no way of knowing that the switch will act that way - even if you checked continuity you would not know what amperage the diode would allow.

Anybody replaced your fusible links with wire? You won't believe how good it burns!

Anybody seen this before? Is it a common problem with diodes used as lights?

  KenF