Maybe I'll take the alternator off of my '88 since I know it's good and throw it into that one...

 

That would verify if it is the alternator or something in the wiring at least.

Am I talking to myself?

Pull fuse 12.

I think I see what you are getting at Naru. Current from the alternator is backflowing through a shorted diode in the cluster and providing the ignition circuit enough power to still work, even with the warning lamp may be in series with the circuit. Is that basically your thought?

Edited April 9, 201213 yr by Cougar

I think I see what you are getting at Naru. Current from alternator is backflowing through a shorted diode in the cluster and providing the ignition circuit enough power to still work, even with the warning lamp in series with the circuit. Is that basically your thought?

 

Can`t be anything else.

 

Pulling the fuse may or may not work,The contradictory and incomplete Subaru Factory wiring diagrams are not clear on the exact location of the fuse.It may be on the "wrong" side of the bulb.

 

Unplugging the round connector(s) from the gauge cluster WILL work.

Opening the light gauge charge light wire between the alt. and the dash will work too.

If motor now stops ,diode is shorted.

Edited April 8, 201213 yr by naru

Can`t be anything else.

 

Pulling the fuse may or may not work,The contradictory and incomplete Subaru Factory wiring diagrams are not clear on the exact location of the fuse.It may be on the "wrong" side of the bulb.

 

Unplugging the round connector(s) from the gauge cluster WILL work.

Opening the light gauge charge light wire between the alt. and the dash will work too.

If motor now stops ,diode is shorted.

 

Awesome, somehow I missed that... I'll give it a shot tomorrow! I do again thank you for the help, I was totally lost on what it could be... I'll pull that fuse and report back soon! (and no, you weren't talking to yourself that's why I was asking for values on the diode... because if it's bad obviously I'd like to replace it... but still gotta check if thats the case first.) As Cougar stated, and you confirmed, it's all that would really make sense.

 

I think I was writing my response when you posted your's Naru... I didn't even see it till now.

Edited April 9, 201213 yr by kabarakh
additional info

You should be able to replace the diode with something like a common 1N4002 diode.

Naru, thank you! You were correct. The fuse was indeed on the wrong side, so pulling it didn't help, but I followed Cougar's suggestion on the diode value (thank you Cougar!) and wired it in line right after the alternator and now the car shuts off just like it's supposed to:drunk:

 

Appreciate all the input and help! Thank you everyone!

You're welcome for the help. This was a pretty good one. Naru called it well.

As much as I would like to take the glory,I was not the first to point out that the alternator was backfeeding the coil.

 

Since removing charge light fuse#12(it is between the alt. and the charge light after all) had no effect I`m having serious doubts about that part of the theory.

 

I think the coil was being backfed somewhere else.

Might be able to find it by unplugging fuses etc. but who cares?

 

Elegant solution in the end.

After going back and reading the first few posts I see it was Gloyale that first called the bad diode. So hats off to Gloyale on that good call.

 

The alternater does provide a backfeed voltage on the lamp lead when it is working. That is what turns off the warning light when things are working correctly. I can't remember how things are tied together exactly but if I remember correctly I think fuse 12 protects the circuitry tied to the ignition switch so the fuse is tied between the switch and the rest of the circuitry. I may be wrong about that though. It has been a while since I have looked at that area and that old of car.

As much as I would like to take the glory,I was not the first to point out that the alternator was backfeeding the coil.

 

Since removing charge light fuse#12(it is between the alt. and the charge light after all) had no effect I`m having serious doubts about that part of the theory.

 

I think the coil was being backfed somewhere else.

Might be able to find it by unplugging fuses etc. but who cares?

 

Elegant solution in the end.

 

Just to follow up... I think you were in fact right Naru... My fuse box lost its label in the depths of time, so I just pulled out the fuses one by one. Nothing killed it (although right after that I found the diagram and checked again on the charge fuse specifically) so I think you're original assessment was right.

 

But as you said, who cares an elegant and cheap solution ($1.39 for three diodes at radioshack.)

 

My hats off to you as well Gloyale! This was a fun one.