Repairing the Room Lamp Control Module aka Ignition Key Cylinder Light Relay (1 Viewer)

[Timpanogos Slim]

Edit 8/6/2013: SOLUTION FOUND. It's the little glass component on the far side of the brass tabs. Cut it off. Or replace it. See if i care.

I have that "ignition cylinder light never turns all the way off" issue.

We've seen the threads - here and here. The part number is 85965-60010 and it costs over a hundred bucks new and $50 used.

Or you can just remove the bulb. Or, even easier, remove the module. On my truck it is behind the passenger side kick panel, right under where the diff lock ecu would be if i had lockers.

But those options are the easy way out.

It turns out to be a fairly simple electronic circuit:

This probably cost a buck to manufacture in 1990.

I see two transistors, two capacitors, four diodes, and half a dozen resistors.

The only components that are at all likely to fail are the transistors and maybe the electrolytic cap. And it's not the electrolytic cap (already replaced it).

The contacts were dirty, but a judicious cleaning with 1000 grit sandpaper and then deoxit d5 didn't make a difference.

The taller transistor is a 2sb647c and is easily sourced on ebay.

The shorter one is 2sb637c and a bit harder to find, though there is an NTE equivalent available from Fry's for less than $4 shipped:

http://www.frys.com/product/1975880?source=googleps&gclid=CI202b3v5LgCFQnhQgodzRoA6g

But i don't like NTE's business model so i consider them a vendor of last resort. Ironically, their business model involves cornering the vendor of last resort market.

So I'll be frank: I want my ignition switch lamp to switch off properly, and I'm willing to debug this issue to do it. And unwilling to buy another module.

if one of the black diodes is cooked, the one in the middle is not connected to anything (wire harness is only 3 pins - unused output?) so i might try swapping them around. I haven't identified them yet but i would not be surprised if they are plain old rectifiers and i could just drop in 1n4001 diodes which i have a-plenty.

Perhaps i should pull out the unused diode and see if i can identify it.

I will probably order a baggie of the tall transistor from a chinese ebay vendor tonight. Though, a real man would just find a modern part with the same hfe curves.

 

[890man]

I sure wish I knew what the hell your talking about. I can pick my way around schematics but I don't know how all this stuff works as a whole.

 

[Timpanogos Slim]

Oh, I don't understand it extremely well either. But if I trace out the circuit, well, it's basically a timer.

I pulled zener diode and the transistors to test them separately, tested the black diodes by lifting one leg.

The black ones are effectively a 1n4001. But they all work.

Unsure on the zener. Can't identify it definitively. Marked 3383 which might mean 8.3v.

The tall transistor tests out fine. The shorter one tests in the right hfe range but makes my multimeter display the battery icon, which supposedly means something is wrong.

The smaller transistor seems to interchange with bc556 which is european/american thus wired inversely.

I may actually have a big bag of bc556. Too late to check.

The symptom seems to be that when closing the door the ignition light dims and stays dim instead of turning off. So current is leaking. So I suspect zener or the transistor that makes my dmm act funny.

This could all be replaced with a 555 timer but what would be the fun in that?

 

[landcrshr]

Oh, I don't understand it extremely well either. But if I trace out the circuit, well, it's basically a timer.

Draw-out the schematic and post it for us.

Unsure on the zener. Can't identify it definitively. Marked 3383 which might mean 8.3v.

The numbers usually don't translate that way.

And don't worry too much about the the hfe curve, they're just used for switching so no Q-point to set up.

My source for electronic components: www.mcmelectronics.com

 

[Timpanogos Slim]

Draw-out the schematic and post it for us.

The numbers usually don't translate that way.

And don't worry too much about the the hfe curve, they're just used for switching so no Q-point to set up.

As for the zener, I presume it is a zener because my nifty component tester w/ automatic identification is pretty sure it is a 96 ohm resistor. I presume this means that the test voltage is below where it starts zenering.

I can find no record of a 1n3383 or 2s3383 or etc. I should put it back in the truck and measure voltage across it.

I suppose I could shove in some other pnp to92 parts I have.

Will trace it out.

 

[landcrshr]

I suppose I could shove in some other pnp to92 parts I have.

Right, except the taller transistor isn't exactly a TO-92, it'll have a higher power rating (more like around 1W iirc).

That one is likely to be the current source for the lamp and also a likely source of the leakage.

Good luck!

 

[Timpanogos Slim]

Right, to92h or whatever they want to call it.

I may have some to220 pnp parts. I know i have some to220 pnp darlingtons, but that's sledgehammer overkill.

 

[landcrshr]

Darlington prolly not a good idea, too high-gain, may never shut off. Then you'd just have the same problem!

 

[Timpanogos Slim]

Either that or the fade-to-black would occur in just a fraction of a second.

But I'll see what I've got.

 

[landcrshr]

One way of trouble-shooting this would be to de-solder the base of the output transistor (presumably the tall one) and see if it still conducts. If so, that is where the leakage is. If not (and the lamp stays off) then the problem lies in the timer circuit.

Of course you could replace all the semiconductors in that time.

 

[Timpanogos Slim]

Yeah, and i've just realized i should have a cache of mje253 or bd140 somewhere. Should work a treat.

 

[Timpanogos Slim]

Hey, it turns out there is a simplified schematic in the wiring diagrams.

 

[landcrshr]

Looks like an uncoventional design to me, but then I've seen these simplified diagrams be pretty far from the actual design.

Set up like that, either of the transistors or the zener could be the culprit (as well as the timer circuit).

BTW: looks like your '94 dome circuit differs from my '97.

 

[Timpanogos Slim]

Looks like an uncoventional design to me, but then I've seen these simplified diagrams be pretty far from the actual design.

Set up like that, either of the transistors or the zener could be the culprit (as well as the timer circuit).

BTW: looks like your '94 dome circuit differs from my '97.

I'll see how close it is to reality later tonight.

Not at all surprising that a 97 differs from a 94. I even have seen a few different versions of where the module is supposed to be.

The circuit is truncated a bit, what you don't see off to the left is the vanity and 'personal' lights up front.

 

[landcrshr]

If you're like me and want to get to the bottom, I would start with removing the small transistor from the circuit. That would isolate the timer from the switching.

 

[Timpanogos Slim]

Right now the transistors and zener are out. So leaving out the small one makes it easy enough.

 

[blkprj80]

How is the ground for that module? I can see where a bad ground could cause the room light to be on real dim.

The schematic doesn't show a connection to GND. That has to be there or it just won't work electrically, in my opinion.

 

[Timpanogos Slim]

How is the ground for that module? I can see where a bad ground could cause the room light to be on real dim.

The schematic doesn't show a connection to GND. That has to be there or it just won't work electrically, in my opinion.

Yes it does.

Through the bulb inside the ring around the ignition switch.

No real progress tonight. I spent the evening looking for something in the volkswagen, which turned into cleaning the volkswagen. And then it was dark.

But i did find a power PNP in the junk box that will maybe work. 2sa900. TO-126b package and Ic of 1 amp, so, twice as big as the original big transistor.

 

[blkprj80]

I looked up the wiring diagram for mine ('96) and there is no ground wire. That seemed odd until I noticed, as you had already clearly stated, it is a very simple circuit.

If you depin pin 4, I presume it turns off completely, right?

My money might be on the big aluminum electrolytic. If you desolder it and fire up the board, the light should go from on to off without the fade out. If it's completely out of the circuit, you can rule it out entirely.

What's the little cap do?

Do you have a picture of the solder side, or better yet, a reverse engineered schematic? Sketch it up and take a photo if you can.

I may never use this information, but you can bet I'll find another circuit that is very similar, maybe even on my rig! This should be an easy fix.

 

[Timpanogos Slim]

Right, it builds up charge in the capacitor when pin 4 is grounded, and it releases it through the bulb in the ring until it's supply runs out - which should run out shortly after pin 4 is lifted.

At any rate, it is the zener diode.

The component they added to protect the circuit from damage has in fact caused it to fail.

I tried landcrshr's idea and tried it with all components except the smaller transistor installed. This resulted in no bright light around the ring when the driver's door was opened, but did result in constant low glow.

To determine if it was the taller transistor or not, I swapped out the tall transistor for the 2sa900 i found in my junk pile. This resulted in the same low glow.

I added back the small transistor in it's position and I was right back where we started.

So i lifted one end of the zener and voila - it's behaving as it should.

My plan is to put in a string of four 1n749a which i guess would protect it from spikes above 17.3v. I coulda sworn i had some 11v zeners but i don't see them. I found these and some 1n5255b and those won't do me any good.

If you have this problem and you like living on the edge, clip either side of the glass component between the brass tabs and the edge of the board. Your module has been surviving without it's services for some time now, so it's apparent you don't need it. Worst case scenario, the electrolytic capacitor goes "BANG!", the passenger side kick panel smells of phenol, and the light turns off immediately after closing the door. You are already living with this possibility.

If you prefer to return it to original design function, I have no idea what that might be but 16v sounds like a reasonable high limit for a nominally 12v device to me. Go to mouser.com and search for a 1n4745. Which ever part comes up with that search that strikes your fancy, given that it has wires coming out the ends, will work just fine. look at the picture to be sure. If all you see is a little rectangle, that's not the one. This is a 16v 1W zener. It'll be bigger than the glass one, which was probably half or quarter watt. Make sure the stripe is oriented toward the side with the capacitors and transistors when you install it.

If you wish to alter the length of time that the light stays on after shutting the door, replace the electrolytic cap with a larger or smaller value (for longer or shorter time). Use a 25v electrolytic just because it won't be more expensive and it will be erring on the side of safety (since it is the least voltage tolerant component on the board -- seriously the transistors are rated for 100v). Be sure to observe polarity when installing the new cap. The stripe goes towards the brass tabs. The stock value is 100uf but it has probably gotten tired with age -- a brand new 100uf will return it to original factory length. 47uf would be about half, 220uf would be a bit over twice as long. i imagine if you use an LED in the ignition ring, it will be lit a lot longer. With an LED, you might try a 4.7uf cap for short duration and fade-out. DO NOT use a tantalum capacitor, they are little bombs. And again, remember to buy a through-hole part, not an smd part.

I'll take a picture of the solder side before i reassemble the enclosure and put it back in the truck for good. It'll be messy from how frequently I've been reworking it and i don't care to clean it. It just turns a lamp on and off, we're not playing tchaikovsky here.