After reading many accounts of how this circuit works, I felt compelled to investigate further as I did not understand how two positives would cause a light to operate. They won't. One must be somewhat negative to complete the circuit. Internet searches turned up the same basic explanation, still was not buying it. I think it was being oversimplified.
This is how I think the alt light works:
When the key is on and engine not running, there is 12 volts + at the alt light power side coming from the fused side of fuse #9. The other side goes to a junction on the relay board with D+. With the key to off and a meter connected between D+ and ground at the relay board, the reading is 12 ohms. As soon as the key is turned to on, the reading jumps to 12.5 meg ohms and the light comes on. If the wire for D+ to the alt is removed, the reading stays the same and the light stays on. Removing the VR caused the reading to jump to infinity and the light goes out.
For the light to work, there has to be power to one side of the light and some resistance reading to ground for the other. Looking at the wiring diagram, one can follow the blue wire to the junction at the relay board at D+, then up thru the VR to a set of relay contacts, then down thru a ( resistor ?, not sure) and then down to the DF connection and on to the rotor where the current will produce a magnetic field. After the rotor, it goes to ground. When the alt spins enough RPM's, a voltage is produced and fed back to the VR, causing the relay to open and removes the ground path for the alt light.
I could be entirely wrong here, but this is what I see and my readings more or less confirm it. If you see an error in my thinking, please post and let me know.
Thanks,
Tom

I recall you need a light or resistor to "excite" the alternator into producing power particularly at low engine speeds. As you stated, once the power is flowing the light drops out???

Light comes on when the alternator out voltage on the red lead is lower than the battery voltage. . reverse polarity flow. . .

Red white is switched battery 12.5 or so. . that is what lights the bulb at idle. . the bulb is connected to potential lower than 12 .5 volts thru the diodes in the alternator. . with the alternator turning, the voltage out of the alternator on the Blue wire to the regulator reaches the same voltage as the red [B+] out to the battery so there is no differentiation in voltage so no flow anymore thru the light and the light dims. both side now being equal voltage. .
.
and yes, the alternator needs voltage to excite the fields to begin the charging cycle.
spinning an alternator without hooking up a batter will not produce voltage. . generators did it by
having magnets in the outside fields. .

Really simplified version. .left out the technical stuff. It's only confusing

Your findings are correct.

The GEN light has 2 functions:

1) Indicate when the main charging path is not functioning correctly. There are 2 charging paths (see diagram). The one on the left powers the voltage regulator. The path on the right is the main path and powers the vehicle's electrical circuits and charges the battery. The GEN light is connected between these 2 paths. If the main path isn't providing enough voltage, there will be a difference of voltage across the GEN light and it will light up.

2) Provide the minute current necessary (when the engine is first started) to energize the magnet in the armature (the part of the alternator that turns). This is called bootstrapping since the alternator pulls its voltage up by its own generated voltage.

#2 above is why if the GEN light is burned out or missing the alternator may fail to start up. It is possible for the alternator to start up since the armature magnet may have some residual magnetism when the engine is off.

A couple of minor misconceptions in the OP:

The light does not need voltage on one side and ground on the other to light. It needs more voltage on one side than on the other--and it doesn't really care which side is higher and which is lower.

The thing at the alternator is not a resistor, it is a diode. It keeps the current from flowing "backwards" through it except under some specific circumstances.

Toolguy and Spoke both have the functions correct. When you turn the key on with the engine not running, the red/white wire gets +12V. Since the alternator isn't spinning, the wire coming from it has about 0V. Current flows from +12V to 0V through the bulb, and the bulb lights up.

When the engine is running, the alternator produces voltage. So the blue wire from the alt will be at about +12V, as will the red/white wire. No current will flow, so no light. Note that you can get resistance on some connections that drops some of the voltage on one side or the other of those, which will make the light glow faintly; and there are other failure modes that will make the light shine in different ways.

And yes, the light provides resistance to help "bootstrap" the alternator so that it starts charging.

--DD

Wow! I'm still a little dazed but that was some "on point" info. I'm glad you guys filtered "toolguy's" simplified version. Thanks a million.

Originally key on, engine not lit, the battery feeds current throgh light tobthe D+ which is simply a path through the windings to ground because theregulator has D+ linked directly to Df. This current magnetizes the Df windings and when the spinning starts D+ becomes 13 volts really soon. That balances the battery and the light goes off. If there is a failure in the diodes charging the battery, the battery will go low, D+ will stay high and now the light will come on with electrons flowing in the opposite direction. If the original lamp wattage is too low, below 1.2 watts, it is possible that you wont bootstrap Df with enough current.

Every one else covered the basics very well above.

One statement in the OP's explanation jumped out at me though:


You cannot measure the resistance of a powered circuit.
So trying to explain the difference in resistance to ground before & after energizing the circuit is not measureable with a DMM in resistance mode.

stugray,
OMG, you are so right! I don't know what I was thinking yesterday. I do know that you can't take resistance readings with power applied. Must have forgot yesterday.
Dave_Darling,
The light must have a path to ground, even if there is a resistance between the light and ground. I measured 12 ohms initially.
I was talking about the two rectangular "boxes" in the voltage regulator. They don't look like diodes. One is all blacked in, the other is left uncolored or clear. I am thinking they are resistors.
toolguy,
The light doesn't get connected to a potential less than 12.5 volts thru the diodes, it's connected thru the voltage regulator and then thru the field to ground.
I still think the relay in the voltage regulator has the major part in keeping the light off when there is voltage being produced by the alternator. With the normally closed contacts open, there is no path for the light negative side to get to ground. I guess it is like the chicken or the egg. If no voltage is present, light lights, if there is voltage present, the relay operates opening the path to ground, but the light would not light anyway as the two voltages would be the same. Although, looking at the regulation side of the circuit, if the relay did not close and remove the ground, it would be almost like the initial "bootstrapping" was in place and the alternator would not vary it 's output to keep up with demands, so the relay does have an important part in this circuit.
I guess I am just being too technical, but when I see something I don't fully understand, I reach out to the guys who may have the answer for me. Often times I have found out information that has helped me in other projects.
Thanks for the reply,
Tom

Could you expand on this? My alt light just came on (and I changed the alt for a tested good one about 6 weeks ago); fan belt still tight; no obvious observed problems so I removed the alt and took it to the free FLAPS for testing and it tested good. Great good alt but the alt light is lit. Need things to check. Now that the alt is out, I'll check the alt harness and the 14 pin harness on the relay board. Want to drive to a local event this Sat so I'd like this resolved. TIA.

Check the voltage regulator.

There is a test where you unplug the VR and connect two pins on the harness side and start the car and measure the ALT output voltage.
I dont remember which pins, but I have it written down at home (a serach on here might turn it up).

Anyway, I had a problem where I was barely getting above 13 Volts with the car running.
I did the test, and the ALT was putting out 16 Volts with the VR bypassed.
I put in a different (still used) VR and then I was getting 14.5 Volts running.

So that check can tell you if the problem is with the Alt or the VR.

If some of your alternator's diodes are failing the light will flicker. If there is resistance in the circuit, the light can come on dimly. There is some other failure mode that makes the light dim at idle, and get brighter as the RPMs increase. (Failed VR? I forget!) And there may be others as well.

I used to be able to find a chart with different light behaviors and possible causes, but I can't find it now...

--DD

The alt I removed this morning and tested good has been reinstalled. Just have to connect the starter connection and reinstall the heating ducts and air guide. Funny, I have a VR sitting out on the tool table so I might as well try that too. Break for lunch ... BLT's!

Rats! All put back together and Alt light still on; changed VR, still on ... no power to oil temp gauge and step on brakes, alt light gets brighter, tach indicates less rpm than it sounds like. Going to check brake light switch wiring circuit ...

So, after looking at the schematic, I decided to check out fuse #9 with it's circuit to the tach, brake light switch, fuel gauge, etc. Fuse was blown. Replace fuse, all is good (except to figure out why the fuse blew, but I'm beat and no more today).

Moral of the story: If your alt warning light goes on, check #9 fuse before you R & R the alternator.

So, there is no joy in Mudville, the alt light is still on. I installed a new solid state VR and the light is still on. I think i need to test/measure voltages at the alt warning light; the blue wire and the red/white wires as it seems the explanations have the +12 V from those 2 wires controlling the light operation.

Alternator testing good may not completely test all the functions of the alternator. Not sure the test would find a damaged diode on the charging or VR sides.

Here's a little description of how the diodes work in the alternator. The alternator is a 3-phase alternator with 3 separate windings spaced 120 degrees from each other.

With full diode bridge rectification, the correct alternator output across the VR would look something like this. The black line is actual voltage since the VR has some capacitance to filter the ripple voltage from the alternator.

Click to view attachment

If you lose one phase of the alternator because of a open circuited diode, the voltage would droop between the operating phases and the GEN light would light a bit.

Click to view attachment

If 2 phases are lost, the droop in the voltage is more pronounced and the GEN light is brighter.

Click to view attachment

So with burnt diodes, the GEN light is on at idle, rev the engine and the GEN light dims. This is because there is less time between active phases of the black trace, the voltage across the VR and one side of the GEN light is almost 14V like across the battery on the other side of the GEN light.

Click to view attachment

Tested at Advance Auto parts. I really don't want to R & R the alt again. When I started and revved the engine, the light did not dim. I need to get out and measure voltages on either side of the alt light. I think I can get the DMM probe in the plugs. First try did not get 12 V but about 3 V. Had to stop for other things ... So on either side of the alt light there should be 12 V? That's what I read from the above info.

So the alt terminal D+ found in track 76 in the schematic should be 12 V. The terminals T 3a, T 3b are the plug for the alt harness on the relay board and T 14/2 connects there too; I should be able to measure at the plug...

Remove the gauge bucket with the GEN light. Start the car. Measure the voltage on both sides of the GEN light to ground.

One side should be the battery voltage about 13-14V. The other side should be the same but will likely be lower. If the one side is lower, likely there is an issue with the alternator (or wiring). I threw in the (or wiring) because a wire issue could cause just about any odd behavior.

When did the GEN light start lighting up? Did you do any work to the car before?

BTW, what is the battery voltage when the car is idling?

Last time I saw a bad vr I replaced it with the solid state unit. It would not work. Replaced it with a new oem hella and all was well.

I have a bunch of VR's laying around so I could try them.

Spoke, the Alt light lit up ~2 weeks ago. Blew #9 fuse and when I replaced the fuse, it seemed all was OK (see posts above that are about this). I have been working on the car but not on the electrical. I've been trying to get a CIS system to work. Only related things are a 009 dizzy with Pertronix. I've looked for shorts, I have the gauge cluster out and will measure on both sides of the alt light. I even took the brake light bulbs out and Dremel wire wheeled the surfaces wondering if oxidation was the culprit. Thanks for the ideas and I'd appreciate any explanations and thoughts. Probably won't get to do any work 'til Sunday. Going for PA peaches at Brown's in Loganville, PA tomorrow (SO quality time).

If that's the voltage across the battery posts (with a decent voltmeter--not the stock one in the center console!!) then your battery is not charging, and is about 3/4 dead. Re-charge it and keep troubleshooting the charging system, because it ain't workin!

--DD

So instead of starting a new thread, I'm going to necro this one and camp onto it because it's so useful.

Racecar: no need for a GEN light, I have a full LCD dash with voltage display and warnings. I removed the dash, relay plate (carbs) and mounted the voltage regulator directly to the sheet metal (easier in/out of drivetrain). Of course, the alternator won't start charging until I connect a test light between the battery and touch it to the red VR wire terminal.

Anyone have a rough idea what the resistance is on the dash bulbs to I can wire a resistor to the red terminal (don't have one handy)? Or, is any resistance at all required; can I just go with a fused jumper wire on it?

Thanks!

I would take a guess at about 200 ohm should do. Being 14V maximum across this resistor, the resistor wattage should be greater than 14V*14V/200 = 1W. A 2 watt resistor should do.

The functioning of the regulator is a bit sneaky and not necessarily obvious (though honestly you can say this of a lot of pieces of heavily cost-optimised technology).

Click to view attachment

Things to know:

- The alternator will only produce an output when current is already flowing through the field coil (marked G). This is the coil that spins when the alternator input shaft is turned by the engine.
- There are a couple of capacitors (close vertical lines) in the regulator drawing; I've covered some of their functions here but almost certainly not all of them.
- I only marked two of the resistors in the drawing, there's another that's blacked in - being blacked in almost certainly means something, but it's not clear what.
- Current takes time to start up / slow down, especially when it's flowing through coils like the field or stator coils in the alternator.
- It takes more current flowing through the coil to turn a relay on than it does to keep it on (so the 'turn off' current is lower than the 'turn on' current).
- Like most folks I'll use positive-to-negative when describing current flow; don't be upset that electrons actually move the the other way. 8)

When the engine is off, and immediately after it's started, current flows from the battery, through fuse #9, through the alternator warning lamp (K2), through the D+ terminal on the regulator, through the relay in the regulator and out the DF terminal, to the DF terminal on the alternator, then out the alternator ground back to the battery (current always flows in circles). This provides the necessary field coil current to bootstrap the alternator.

Once the engine has started the voltage across the stator coils begins to rise. Current will start to flow out the D+ terminal once the voltage rises above a certain point (tough to calculate due to the way the field coil behaves, but probably a few volts), and then out the B+ terminal once it rises above the current battery voltage.

While current is flowing out both terminals the voltage on each terminal will be nearly the same (since in both cases the current's coming from the stator coils). B+ will never fall below battery voltage; D+ will never go (much) higher than battery voltage / B+, but may be quite a bit lower at times.

As B+ is connected to the battery, and thus to the other side of the warning lamp (K2) this means there will be no current flow through the lamp and it will go out once D+ reaches battery voltage (no voltage difference -> no current flow).

The voltage on B+ and D+ will continue to rise (assuming the alternator is spinning fast enough) as the battery charges.

As the voltage on the D+ terminal continues to rise, the current flowing through the regulator relay coil and resistor R1 (circled in red) also rises.
Eventually this current is sufficient to energize the relay and it switches, disconnecting the field coil (and in fact shorting it to D- to ensure that no leakage current continues to energize it). The upper of the two capacitors in the diagram helps prevent damage to the relay contacts during this switching operation.

Removal of the field coil current causes the stator coil to stop generating (this isn't instantaneous; the field coil current falls, the magnetic field starts to collapse, the stator induced current falls...).

As the stator voltage falls the current out the B+ terminal falls, reducing charge current to the battery / rest of the vehicle.

At the same time the voltage on the D+ terminal and thus the current through the relay coil / R1 also falls; some current flows via K2 but this is insufficient to keep the relay energized. This causes the relay to switch back, returning the alternator / regulator to a state similar to the bootstrap state. At this point, there are several sources of current for the field coil; the bootstrap path through K2, the residual charge in the lower of the two capacitors, and the the stator via D+ and R2; one or more of these (probably all three contribute) cause the field coil current to rise and the cycle repeats.

The on/off process repeats very quickly; it can be several hundred times a second. During the off phase some current may flow through K2, but the average current will be low enough that the lamp won't light. The duty cycle (ratio of on time to off time) will vary based on the load on the system; more load will cause the B+ / D+ voltage to rise more slowly, delaying the time until the relay switches. With less load, it will rise more quickly. The time the charge circuit spends off will (tend to) be more constant, as it's related to the value of the lower capacitor and the characteristics of the relay coil, which don't vary with load.

HTH, and apologies for any errors or confusion this might cause.

That's some beautiful stuff right there (along with the prior posts).

So lemme ask two unrelated alternator questions.

- After posting, I took the racer out for a spin to check the drivetrain install (it's a caged racer but actually has plates and insurance). I was just cruising 5-10 minutes locally so didn't "excite" the alternator (not worried about battery level) but 5 minutes into the drive I glanced down and the dash was showing 13.5V.

How did this happen, and is it possible that this is a non-problem for a higher-revving race car?

- In the street car, I had a glowing GEN light that would be dim at idle and get brighter with RPM. I did some voltage checks and found that the B+ battery voltage was ~13.5V so it was charging fine, but the D+ voltage was full alternator output, from ~14.5V at idle to ~17+V at 3000 RPM. I drove it like that for a few months until I got motivated to replaced the alternator (the only thing the D+ was going to was the GEN light); a new voltage regulator did not resolve it, but a replacement alternator did.

Why is that? What failed in the alternator for that to happen?

Your description just begs more questions, sadly. Did the alternator self-excite? (residual magnetic field, current leakage into the field winding?). Is your battery just a monster (or are you running a 16V setup?), do you trust the dash's voltage reading?



If D+ is significantly higher than B+, then the high side of the main rectifier or something else between the stator winding and the battery (B+ terminal hardware, battery strap not tightly connected, etc.) is probably hosed.

Not sure about what you're doing with D+ only going to the GEN light... that's not how it's supposed to work. 8)

"Self-excite" is the only thing I can think of. But here's the rub: prior to my removing the replay plate and re-wiring everything, the VR was sitting on the relay plate, as normal. However, the stock dash was already removed and a Stack 8130 was stuffed in there, with no wires going to it for GEN/ALT light. I removed the Stack last year to install a Race Technologies DASH.

And yet...the alternator was charging. I *presumed* (but never verified) that the previous owner had stuff a resistor or bulb someplace in there, but was never curious enough to find it. However, today we know for a fact that since I removed the relay plate it's not connected.

And yet...?

And here's another rub. Some years ago I removed my VR from the replay plate on the street car; I described it in this thread here. At the time, I was not aware that the common knowledge was that the alternator would not "self-excite" without a GEN bulb..."and yet"...when I did that the voltmeter was showing charging. I only added that jumper wire in order to get a GEN light indication in case of failure.

So...I do wonder if it's actually needed.

Battery in the street car is a Miata AGM, and battery in the race car is a very small Deka ETX14 battery.


Ok, you're correct. To clarify, the GEN light is the only D+ connection to the car. So I drove it like that, not worrying that it was going to fry any equipment in the car (the VR didn't seem to mind). I still have that alternator in a box in the garage, I'm likely to send it to an electrical repair shop to keep as a spare. But it wold be nice if it's a simple bench fix (while I'm handy with electtrical systems basics, I'm a mechanical guy, not an electrical guy).

It is possible that the armature has some residual magnetism in it at rest such that without the current provided by the GEN light there is enough magnetism to get the alternator started. Once started, the GEN light is not in use thus the alternator is providing all the energy needed to drive the armature.



Sounds like a diode failed in the D+ circuit but it doesn't make sense to me. The D+ voltage drives the VR and armature and the VR keeps D+ in regulation around 14V. The VR has no idea what the battery voltage is doing. The VR only senses D+ to D- and keeps that voltage to 14V. In a fully functional alternator, the battery voltage will also be 14V.