Cougar

You can place a little wedge in between the accelerator cable or just pull on the cable if want. You may be able to just read the voltage without setting up the idle and not have very much of a difference. I just wanted to find out what the voltage is at a normal engine speed. If the voltage at the 'L' lead is still around 7 volts and the output voltage is above 13 volts with the engine running you are done. Replace the alternator.

Well done. You won't have to look at those contacts ever again I would think.

Here is a link to a book that may be of interest to you. I have this book and it is pretty good. http://half.ebay.com/cat/buy/prod.cgi?cpid=1173023&domain_id=1856&meta_id=1

Akira, I went back through the posts again to try and see what happened. It does seem originally that the alternator was not charging and the light was on. The voltage output was less than 12 volts. In post 50 you stated that there was about 7 volts at the 10 amp fuse with the engine running. Is this still true? If so, we need to check to see what is causing the voltage drop there. Check the voltage at the 'L' lead while the engine is still running. If the volage is still around 7 volts there then it appears the alternator is at fault. I know you have tried others with the same results but it has to be the trouble. In the beginning things were a little different somehow. Something changed on us while testing. If the problem is in the alternator it may have happened when your dad was probing around and the alternator worked briefly. Possibly some of the exciter diodes failed. That would explain some of this but it wouldn't explain why you have the proper voltage now, if that is the case.

The fact that you now have 13.6 volts input makes me think the alternator is working for some reason but the warning light is on. Before, the voltage was about 11.2 volts with the engine running, correct? If the alternator is working but the warning light is still on then there is a few things things that could be wrong. First, lets measure the alternator output to verify what it is putting out the proper voltage with the engine running around 2,000 rpm. We want to have around 14.5 volts there normally. Verify that the warning light is still on. If the warning light is still on then either, the diode has an excessive forward bias voltage, there is a connection problem after the diode, or the alternator is bad. To find out where the problem is, measure the voltage across just the diode. If the voltage is more than .7 volts then that is a problem. If that is ok, then measure the voltage between the 10 amp fuse and the 'L' lead of the alternator. There should be very little voltage drop, no more than .2 volts. If that is ok then measure the voltage at the 'L' lead of the alterntor to ground. It should be 12 volts but it isn't. Do all these tests with the engine running around 1,500 to 2,000 rpm. According to your tests there is a 6.6 volt drop on the line after the light. This is too much. Let me know what the voltages are.

Are the voltages you gave with the engine running or just with the ignition on? I assume that the light is on while you read the voltages. If you didn't make the measurements with the engine running let me know if the light is on while the engine is running and tell me what the voltage is on the connection between the light and the diode. If the light is on with the engine running check to see if the light goes out if you pull off the 'L' lead off the alternator. If it does go out then you may need to replace the alternator.

This is beginning to sound like a conspirasy.

There are a number of books already available to learn about using a meter and troubleshooting technics. Half.com has some good prices on used and new books.

You may find that using a digital meter, for resistance tests especially, will give you better results. They are designed so they will not exceed the the threshold voltage of a diode unless in the diode mode. Using a analog meter may give a false reading in certain circuits. Your meter should have a 20,000 ohms/volt sensitivity for voltage. If you have a cheap 1,000 opv meter that can cause errors also in high impedence circuits. A decent DVM can be purchased for around $30 dollars if you need one. I like Fluke models myself but they cost more.

Most shops should be able to this.

I'm not sure if this will work for your model or not. You could try turning the ignition on and off three times with the drivers door open. I think that is what I needed to do on my '01 Outback.

Thanks for the info Mwatt. I have a '02 model year. The problem is intermittant and I need to find out what the dealer will do on this, as I still have a warranty. I need to see if they will just replace the module even though it works most of the time.

I agree with what he says. The price seems real fair to me. You don't want to have to go back in there to change the belts. I would use genuine Subaru replacement belts also.

Unless your coolant is due for a flush I would hold off on that. Instead, Iwould have a block check done to see if exhaust gases are getting into the coolant. Is is a cheap test to do and you will know what is happening. If the test is negative then you should have no problem. Your timing belts may be due for changing.

I think my Town & Country van is having this same problem.

You could try swapping the ignitors to see if that changes things on the spark. I think shops can test the exhaust system by pressuring it and checking the back pressure. Welcome to the forum BTW.

I think you should be able to do this without much trouble. It may take you about half the day to do, I'm not really sure. Make sure you understand how to do the alignment of the timing marks if you do this. You don't want to have to redo the job because of a misaligned belt. You may want to replace the water pump also while your in there.

To John in KY: Your problem may be due to a bad fusible link or perhaps the ignition relay. I would check out the ignition switch area carefully. If you don't have a wiring diagram you may need one to fix this problem. If the CAS were the problem it would seem there would be a code showing that. If that is not working correctly it would cause this kind of problem.

What kind of tests are you doing, and what kind of meter are you using?

Your welcome for the help!

Ok, thanks for the clarification. Your last statement says that the 15 amp fuse has a black wire tied to it. This wire is the side of the fuse that should should supply power to the fuse, correct? If so, you need to see why power is not getting to that wire. The voltages you are seeing on the WR wire right now are not really significant. We need to have 12 volts powering that line. The arrowhead side of the diode needs to have a positive bias on it for the diode to conduct.

Those lights are in the field circuit of the alternator. It would seem your new unit has failed. If a new alternator solves this be sure to check the output lead of the alternator to the positive battery post and engine ground to the battery ground. Make sure there is no voltage drop between the points. This will cause the alternator to overwork. It may be a good idea to have the battery load tested also.

If you are measuring the diode in the standard resistance mode then I don't think that is a valid test. You need to be in the diode test mode. If you want to prove the diode is good then place a 12 volt source on the anode side of the diode and see if the alternator works. If it does, then the diode is ok. If the diode was the problem then you should see 12 volts on the anode side. If I remember correctly, the brown wire coming from the alternator ties to the 10 amp fuse, through the fuse the wire turns black and that then ties to the light and the diode. Is that correct? Also, what color wire ties to the anode of the diode? It is that wire that needs to be tracked for a problem of what is blocking the voltage.

Yes, a voltmeter is what you want to measure with.

I was only suggesting the jumpers at various ground points. Like between the engine block and the negative battery post. For the alternator output connection it is best to measure the voltage between the output connection and the positive battery post with a load on the system, like the lights and blower on. If the connection is good then the voltage drop should be less than .1 or .2 volts.