[TimClifford]

A fun and fantastic two weeks of driving my 1994 SAAB 900 Turbo Convertible, and I've hit my first real problem. The battery died on me.

I got a jump from a friendly passerby, then half way home the battery died on me while driving. Damn. Sounds like the alternator.

Ok, I dissect the problem further, and it looks like the alternator is giving charge, just not enough (about half a volt). The battery light on the instrument panel doesn't turn on, and a local mechanic shows me how to get the alternator up to an even 14 volts if I just connect a test light from power to a connector on the back of the alternator. Now the instrument panel bulb, known to affect charging when it burns out, is the most likely suspect, and I vaguely remember thinking "Why is the battery light so much dimmer than the others?"

Which would mean I need to buy a 70 cent replacement part and live happily ever after. But then I go and swap out the charge warning light bulb with another, less important bulb (check engine, hah), and lo and behold that doesn't make a lick of difference. Hell, the battery bulb doesn't look burnt out, and I can't even get the leads for the battery indicator bulb to show a voltage (but to be honest it is a damn awkward angle).

If I could get the alternator to manually turn on (the friendly neighborhood mechanic wasn't terribly descriptive in his methods), then I'd at least be in the clear to drive the thing while I figure out the underlying problem, but so far no luck on that front. Is this info pointing to some other likely culprit? What is going on here?

 

[SPGreg]

I would check the voltage regulator on the alternator. It's a couple of screws. You can see if the brushes are worn down. They're spring loaded but when they get short there's insufficient pressure (& obviously insufficient length) to maintain charging.

 

[scantar]

x2 on checking the voltage regulator. Also check the connections on the back of the alternator as well as grounds from the battery and the alternator.

If it was just the bulb, you should be able to excite the alternator by revving the engine above 2000 rpm.

 

[crwchf01]

The voltage regulator is part of the brush assembly and can be removed from the alternator without removal of the alternator. Check for more than 5mm of exposed contact brushes. If not replace the assembly. Brushes themselves are available, but for the cost, it is just as easy to replace brushes and regulator.

 

[TimClifford]

Never would have guessed the voltage regulator. Rev'ing the engine over 2k seems to affect the voltage, but I've gotten conflicting readings over the past day between the lady holding my half broken multimeter at home and the mechanic who towed me using his battery/multimeter combo on the side of the road (ouch and double ouch).

Are there any good visual diagrams of the alternator in question? I can sort of guess what the voltage regulator is, but I haven't been able to tell B+ from D+ from Adam with any of the diagrams I've found online or in Bentley.

 

[scantar]

On the back of the alternator there are 2 screws...can't remember if they're Phillips or not. They were on my 318i (same alternator) Undo those and the whole regulator will slip off into your waiting hands.

 

[Geoff Weeks]

Forget the chargeing system light, about all it is good for is to tell you the belt broke, any other defect will not likely turn on the light.
Absent a good chargeing system tester (and the knowedge on how to use it) Jim M posted a good test.
Start the car, and check the voltage at the alternator and battery, they should agree.
Next turn on all lights, blower A/C etc and run the car at 2800 rpm, the voltage should be close to the voltage measured before and both should be around 14 volts, 13.8 minimum. If not then look at the wireing and make sure BOTH the output and ground wires are makeing good contact. The ground wire on the Saab system carrys the same current as the output wire, so needs to be as heavy and in as good condition.
I just picked up a Snap-on MT1552 Vat tester, after being without for some time. There are many ways to test a chargeing system but few have the tools. An occilascope or a good VOM that can pick-up small a/c componant in a DC line, or one of the many porfessional grade testers like the MT1552. Most DIY's don't have access to these and Jim's test is the best I've heard of for the DIY. It still doesn't pinpoint the problem, but does tell if things are working. If it fails the test, then I would take the car to someone who has the proper test equipment and knows how to use it (not a given just because they own or have the equipment) and pinpoint the problem. So much better then thowing parts at something and hopeing you got it fixed.

 

[dmgb5]

Are there any good visual diagrams of the alternator in question? I can sort of guess what the voltage regulator is, but I haven't been able to tell B+ from D+ from Adam with any of the diagrams I've found online or in Bentley.

This what the voltage regulator looks like...

B+ is the terminal with the red thick gauge wire running from the alternator to your starter and battery and is your high amperage alt output.

D+ is the terminal with the small green/white wire going to it. This is the low amperage output on the alternator and goes to the warning light in the dash.

Also both B+ and D+ should have the same voltage output.

 

[TimClifford]

I was posting from my phone earlier (stranded and bored while the battery charged enough to get me home), so I may not have been particularly clear on a couple of points. The charge indicator light (47E on Bentley) isn't getting power, but ground is connected to the green/white wire on the alternator (D+ connector, exciter wire).

If I short the 12V connected (B+) to the exciter wire (D+), then it'll kick the alternator on and everything changes. Before, rev'ing the engine would increase voltage over the battery. After, rev'ing the engine will lower the voltage over the battery. Before, idle voltage on the battery was 12.7ish. After, it's 13.XX. Clearly the alternator is doing something when I hit the exciter wire, but I don't know what.

I just spent the last hour or so testing connections, if I'm reading the wiring diagrams correctly, then the break in the circuit is somewhere between fuse 7 and the instrument panel.

I just tracked down a ratchet wrench with a philips head on it, so I might go back out and take a look at the voltage regulator before it gets dark. I also need to get a reasonable DIY test of the belt tension, as it squeaks, but I don't want to break any teeth over what might just need some oil.

More ideas are welcome! I bought a c900 because I wanted to get a little dirty and solve problems, and this has certainly been an interesting one so far.

 

[Geoff Weeks]

. The charge indicator light (47E on Bentley) isn't getting power, but ground is connected to the green/white wire on the alternator (D+ connector, exciter wire).



far.

What? The ground is connected to the green and white? IF the D+ terminal is grounded, it will not charge and will most likely burn up some diodes inside.
Jumping B+ to D+ while it is grounded will cause all kinds of problems.
SO what is connected to what?
Belt tension can be easly assesed by pushing on the alternator fan with your thumb, if it will turn the belt is loose, glazed or damaged.
Picture would help.

 

[TimClifford]

I'm going off of this schematic:
http://saab.znet.lv/02_Tech_Info/Saab 900 M93 S.pdf

Again I'm being unclear- the chain goes like this:

Battery + -> Ignition Relay -> Fuse 7 -> Instrument Panel -> Charging Light Indicator -> D+ on Alternator -> Ground

The negative side of the charging light indicator socket had a clear connection to D+ on the alternator (less than 1 ohm resistance), but the positive side of the socket, the side that should go to the battery+ through a chain of stuff, wasn't connected to the fuse. Somewhere between fuse 7 and the charging light indicator socket is a break in the connection, which, apparently, is a ton of wires stuck way up behind the dashboard. Fun Fun.

 

[dmgb5]

I was posting from my phone earlier (stranded and bored while the battery charged enough to get me home), so I may not have been particularly clear on a couple of points. The charge indicator light (47E on Bentley) isn't getting power, but ground is connected to the green/white wire on the alternator (D+ connector, exciter wire).

If I short the 12V connected (B+) to the exciter wire (D+), then it'll kick the alternator on and everything changes. Before, rev'ing the engine would increase voltage over the battery. After, rev'ing the engine will lower the voltage over the battery. Before, idle voltage on the battery was 12.7ish. After, it's 13.XX. Clearly the alternator is doing something when I hit the exciter wire, but I don't know what.

I just spent the last hour or so testing connections, if I'm reading the wiring diagrams correctly, then the break in the circuit is somewhere between fuse 7 and the instrument panel.

I just tracked down a ratchet wrench with a philips head on it, so I might go back out and take a look at the voltage regulator before it gets dark. I also need to get a reasonable DIY test of the belt tension, as it squeaks, but I don't want to break any teeth over what might just need some oil.

More ideas are welcome! I bought a c900 because I wanted to get a little dirty and solve problems, and this has certainly been an interesting one so far.

If you have the Bentley diagramme in front of you...you can see that D+ wire which is the goes to the warning light(47E) is diode protected. Which means current can only flow one way. So when car/alt is not running the green/white wire acts as earth wire to the warning lamp with 12v coming from the blue wire which goes to ignition switch. The current that flows through the lamp acts as the exciter for the alternator creating the magnetic field needed for the alternator to start up. Once the alternator is running D+ (and B+) will show voltage and the warning lamp in the dash goes out because it no longer has an earth.

 

[Geoff Weeks]

Ok, I see what you are trying to say. The D+ is not a ground connection but acctually goes to the + feed to the voltage regulator, when the alternator is running D+ will equal output voltage. When stopped, the voltage is zero. When a lamp is placed in series with the battery the current will flow from the ign switch thru the lamp to the voltage regulator via D+ connection, thru the field to ground. Once the stator voltage rises the voltage will be the same on both lamp terminals and the light will be out.
If your lamp is inop, the alternator will still turn on if engine speed exceeds about 3000 rpm. Once it is on, it doesn't turn off when the speed falls below 3000.
It doesn't sound like you are getting anywhere close to 14 volts, which would indicate a problem.

 

[Geoff Weeks]

If you have the Bentley diagramme in front of you...you can see that D+ wire which is the goes to the warning light(47E) is diode protected. Which means current can only flow one way. So when car/alt is not running the green/white wire acts as earth wire to the warning lamp with 12v coming from the blue wire which goes to ignition switch. The current that flows through the lamp acts as the exciter for the alternator creating the magnetic field needed for the alternator to start up. Once the alternator is running D+ (and B+) will show voltage and the warning lamp in the dash goes out because it no longer has an earth.

I beg to differ, the D+ wire allows current to flow both ways (or at least mine does) when you shut off the engine, the charge light comes on (because now the d+ or stator has voltage as long as the engine is still turning) while the ign switch is off, and the lamp in series with the ign loads now has power on the D+ side and grounded thru the ign loads on the other. This is one way to make sure your chargeing system light is working, to watch it when you turn off the key. The light fades as the engine comes to rest (or as the stator voltage drops to zero as the field fades).
I don't know what aBentley shows as I don't have one, but the link the O/Per had shows no diode, and my car definately doesn't have one.

 

[dmgb5]

Sorry, that should read that the warning lamp is diode protected so that current flows only one way through it.

 

[Geoff Weeks]

Sorry, that should read that the warning lamp is diode protected so that current flows only one way through it.

No it is not, as I have stated, current will flow in one direction when the ign is on and the engine off, the oppsite direction when the ign is off and the engine is still turning.

 

[dmgb5]

No it is not, as I have stated, current will flow in one direction when the ign is on and the engine off, the oppsite direction when the ign is off and the engine is still turning.

Are you feeling particularly argumentative right now? How do you achieve ignition off and engine still running? My warning lamp does not come on when I turn off the engine/car. Bentley clearly shows diode on battery warning lamp circuit for both the alternator/charging system and combined instruments wiring diagrammes covering MY85-93/94. I know the Bentley is fallible. I have checked the service manuals Misu posted but there is no wiring diagramme for alt/charging system.

 

[Geoff Weeks]

Are you feeling particularly argumentative right now? How do you achieve ignition off and engine still running? My warning lamp does not come on when I turn off the engine/car. Bentley clearly shows diode on battery warning lamp circuit for both the alternator/charging system and combined instruments wiring diagrammes covering MY85-93/94. I know the Bentley is fallible. I have checked the service manuals Misu posted but there is no wiring diagramme for alt/charging system.

If you turn off the ign, the engine doesn't stop instainiously, it revolves for a few seconds. During this time, the stator voltage is above the ign circuit voltage and current flows from the stator to the ign circuit thru the lamp, lighting it.
This is how all charge system lights work, and the one in my Saab 900 is not different then any others. Go out a try it on your Saab. with the engine idleing turn the key off while watching the light.
I've checked the factory wireing diagarms in my factory manual and while they are very hard to read, to best of my ability I can find no diode. The fact that the light operates the the way it does with the key off, engine turning confirms there is no diode. Unless a change was made in later models (mine is an '85) for some unknown reason, I would say once again the Bentley is wrong.
A diode, wired in parralle with the bulb would result in a redundant turn on signal but would prevent the light from working, but would not prevent the light comeing on when the key is shut off (as the engine winds down). Putting one in series with the lamp would prevent it comeing on when the key is turned off but would serve no other funtion.

 

[Geoff Weeks]

BTW I am not trying to argue for argueing sake. Electrical diagnoisis more then any other requires knowing how the circuit works, and how it doesn't so you can eliminate possibilitys based on how the electrons flow, and what will happen when componants fail.
For example, if you replaced the lamp in the chargeing system, with a solid connection the alternator would backfeed the ign circuit and key would not shut off any componant that is common with the charge light feed.
In another post I explained why a sudden loss of fuel injection would not cause the tach to dive to zero, thus cutting the diagnois time by focusing on items that would cause that to happen. While both a loss of fuel and a loss of ignition will cause a stumble, only a loss of ign feed will show on the tach. Without the info that the tach goes to zero, there would be much more to check, and you would be correct that fuel would be one of them.

 

[dmgb5]

If you turn off the ign, the engine doesn't stop instainiously, it revolves for a few seconds. During this time, the stator voltage is above the ign circuit voltage and current flows from the stator to the ign circuit thru the lamp, lighting it.
This is how all charge system lights work, and the one in my Saab 900 is not different then any others. Go out a try it on your Saab. with the engine idleing turn the key off while watching the light.
I've checked the factory wireing diagarms in my factory manual and while they are very hard to read, to best of my ability I can find no diode. The fact that the light operates the the way it does with the key off, engine turning confirms there is no diode. Unless a change was made in later models (mine is an '85) for some unknown reason, I would say once again the Bentley is wrong.
A diode, wired in parralle with the bulb would result in a redundant turn on signal but would prevent the light from working, but would not prevent the light comeing on when the key is shut off (as the engine winds down). Putting one in series with the lamp would prevent it comeing on when the key is turned off but would serve no other funtion.

BTW I am not trying to argue for argueing sake. Electrical diagnoisis more then any other requires knowing how the circuit works, and how it doesn't so you can eliminate possibilitys based on how the electrons flow, and what will happen when componants fail.
For example, if you replaced the lamp in the chargeing system, with a solid connection the alternator would backfeed the ign circuit and key would not shut off any componant that is common with the charge light feed.
In another post I explained why a sudden loss of fuel injection would not cause the tach to dive to zero, thus cutting the diagnois time by focusing on items that would cause that to happen. While both a loss of fuel and a loss of ignition will cause a stumble, only a loss of ign feed will show on the tach. Without the info that the tach goes to zero, there would be much more to check, and you would be correct that fuel would be one of them.

Okay, Geoff my turn. I will try to present my case in a civil manner as I don't want this to turn into another B202 alternator thread...as we are already getting off topic. I understand you have great deal of knowledge when it comes to alternators and their repair. I imagine you must work with them in some sort of professional manner or deal with them on a regular basis.
While I don't work with electrics in a professional manner on a regular basis I do have firm grasp of electrics, its laws, and principles especially in an automotive application. I understand what your saying and I 100% disagree with you.

My warning lamp does not illuminate while turning off my T16 as you suggest. I did go out and checked for you. I also checked the missus's A4 estate. Nope, same as the T16.

I'm not sure where the confusion is when I say that the warning lamp is diode protected. I don't understand why you think there is no diode there when there clearly is. The diode is in series with the warning lamp circuit and what wrote still holds true...

The D+ wire which goes to the warning light(47E) is diode protected. Which means current can only flow one way. So when car/alt is not running the green/white wire acts as the earth wire for the warning lamp, with the positive 12v coming from the ignition switch through the blue wire. The current that flows through the lamp from the ignition switch acts as the exciter for the alternator creating the magnetic field needed for the alternator to start up. Once the alternator is running D+ (and B+) will show voltage output and the warning lamp in the dash goes out because it no longer has an earth to complete the circuit.

The warning lamp will only illuminated again when the alternator is no longer producing a charge/current at the D+ terminal and provides a path to earth.


This is what is presented in the Bentley wiring diagramme...