Higher Voltage: Fool Your Alternator?

[4pyros]

MrAl;
I am still trying to catch up on some of the old threads and found the one that started this one.
Many years ago I ran across someone that isolated his alternator and installed diodes in series with the ground to increase the perceived output voltage.
It was an old trick to try and get more power out of car audio amps.
Its a vague memory now, I never tried it myself and don't know if it would work.
Its just a thought.

 

[MikeMl]

...Many years ago I ran across someone that isolated his alternator and installed diodes in series with the ground to increase the perceived output voltage.

If you do that, the diodes would have to be rated at 60A+. If you place a diode in the sense wire going to the VR, then you can do it with a 1A diode.

 

[MrAl]

Hi guys,

Well unfortunately i dont think either of these ideas will work. I'll explain what a pain this is...


First, the ground diodes:
Breaking the connection between alternator and ground and inserting a diode in series means that the alternator puts out 14v even when the battery is only 13.3v (0.7v diode drop for example). This means any resistance in the ground lead makes the charging current go down instead of up. I dont know if that is the case with all alternators, but it is with mine unfortunately.
Second and maybe most important, the mounting of the alternator is with a large bolt on the bottom which bolts right to the engine block, and the top bolt is smaller but bolts to a steel arm that bolts to the engine block. Now the top mount might be replaceable with a fiberglass rod, but there's no way to replace the bottom mount without some serious steel reworking. It would just be too hard to do.

Next, the 'sense' wire diode:
This could work but i have a feeling it wont because i think my alternator gets it's sense lead signal from INSIDE the alternator, as strange as that sounds. In other words, this alternator does not have an external sense lead. I believe there are three diodes inside that provides the sense voltage, and because they are inside that means there is no outside sense connection.
I think of it like this (and i could be wrong):
Imagine we have a sense lead from battery plus to the alternator along with the 'normal' heavy gauge wire that charges the battery. The voltage regulator sees the sense voltage and adjusts the charge voltage/current as needed. All is well and good and very good control over the battery charging this way. But now imagine that the sense lead, running through the engine compartment and maybe the ignition switch, develops a higher than normal resistance, what happens then? What happens then is the battery over charges.
Now maybe some manufacturers dont mind this potential problem, or maybe they have a build in fail safe mechanism. I dont know for sure, but i only care right now what mine has. I think mine gets its sense voltage from inside the alternator so it's going to be hard to rig this thing. What makes me more sure of this is that i've seen other mod's done where they go right inside the alternator to modify the small diode trio. Adding a diode in series with THAT seems to up the voltage, but then it goes up by the full diode voltage which i am also not sure i want to see happen. But i think that would be better overall because i can also install a switch to short out the diode when i dont need the monthly super charge, and that would be really nice.

Maybe what i have to do is try Mike's idea and at least see if it works or changes anything. If the voltage goes down by a little that's ok at least i would then know for sure. That would tell me if the small diode trio is just to run the excitation winding and not really the sense voltage, or it does in fact do both. Maybe also the alternator might use the diode trio output AND a sense wire. Maybe the diode trio voltage is the over charge protection feature so that if there is too much extra resistance in the sense lead the diodes take over, but i still dont know for sure because there is no schematic and impossible to see the internal voltage regulator guts because it is potted.

Any more ideas would also be appreciated.

For example, of your two or more cars or other transportation, what is the 'normal' charge voltage of the battery?
Im not looking for an estimate here, but an actual measurement with the engine running. It should be somewhere between 13.5 and 14.5 but i would need to know which, 13.6, 13.7, 13.8, 13.9, 14.0, 14.1, 14.2, 14.3, 14.4, 14.5, 14.6, 14.7, 14.8, 14.9, 15.0, etc. One of those should fit the actual measurement.
If you could measure one or more of your batteries with the engine running that would be great as that would help too with this problem. Thanks in advance. The measurement(s) should be done after the engine has run for at least a few minutes. and of course with the engine still running.

 

[KeepItSimpleStupid]

I did that about a month ago and got 13.5 V. 13.8, I believe is the target. That's what radios used to be spec'd at. I have a SM for an amp for a 2000 GM vehicle. Says 11-16 V. The 16 is there because of the alternator going Nova on you. The SM says the "test voltage" is 13.2 V.

It's rainy, cold and dark right now, so I can't check another vehicle. Some cars, way back when, had a tab that you ground to force the alternator to generate full output. Doing so, was scarey.

See item #4, here: http://www.hartin.com/alternator.htm I repaired these in my day. They had an internal replaceable regulator. See http://store.alternatorparts.com/regulator-6.aspx
Some, I think now are controlled by the ECM or whatever module it is.

One terminal of the rotating field was grounded to change the output. The other goes to Bat.

So, you HAD the BAT terminal, or screw that went direct to the battery. The case ground that went to the engine block. A terminal for the idiot light and a terminal for switched 12 V that goes to the non-grounded side of the rotor.

That said, it may be 'likely" to gain access to the "other" field terminal.

How, might mean taking it apart, and bringing the wire out safely. Probably easier said than done. You had to insert something like a paper clip when putting the brushes back to push them far into the brush holder. Then pull it out once assembled.

Back in 1968, the earliest car I worked on when I was a kid was easily rebuildable. You could press out the top-hat diodes and solder. The regulator was external. Not old enough in the 6V car days to care except what I read in old car electronics books. 1974 GM was easy, 1982 Toyota was not.

The alternatorparts link shows some set points. reminder, it's not at the battery terminal. 100-200 mV could be dropped across the BAT wire.

 

[MrAl]

Hi KISS,

That's very interesting, that the voltage settings are all over the place. I see 13.2v up to 14.8v, so i wonder how that can be possible.
Does this mean that the car manu's dont know how to set the VR properly to begin with?

It's also interesting that you quote 13.5v on your one alternator. I've never seen mine that low, ever.
I am pretty sure that my last alternator put out 14.2v or close to that, but this one puts out 14.0v tops and might go down to 13.8 after a while of driving. That's what worried me from the start.

The voltage setting is interesting because when the voltage stays around 14.0v on mine the battery does not charge properly, but then again i am sort of experimenting a little now with this older battery. I find that when i charge it with a manual charger at a higher voltage, the internal resistance comes down and so it takes a charge better in the car. But as days go by, the resting voltage falls (over the same period) lower and lower. Right now the estimated charge is around 50 percent, while several days ago with the same rest period it would have been around 60 percent, and before that 70 percent, etc. At one time it was 100 percent, and if i take it back inside to charge it i can get it back to 100 percent which will drop to maybe 80 percent over 2 to 4 weeks. That's what else is interesting because i am using an old battery that at one time had been discharged all the way down to 0.00 volts.

I guess the new battery would hold the charge better, but before i put it in i wanted to check with some other people about the charge voltage. I would like to have a good idea if 14.0v is a good charge voltage or not. Most batteries are the same, so i cant see how the VR's can be so different.

When the weather clears up if you can get that secondary reading that would be nice too as we can compare the three. I appreciated this and i wont forget that you did it.

 

[KeepItSimpleStupid]

The voltage at the battery "should" be "all over the place" with respect to the "regulator set points". The reason is simple. You have 30 to 120 Amp alternators or so depending on the model. "Load" will change the output because you don't do sensing at the battery terminals.

A "better place" to do "regulation checks" is the alternator terminal heading to the battery on the alternator. There it will be influenced much less by the amount of accessories turned on.

I tend to write down in the service manual, or I used to, the drop from the alternator to the battery with the heater, AC and high beam on. It was my way of measuring the alternator output (a reference) for troubleshooting. I now have an AC/DC clamp meter.

 

[MikeMl]

Actually, the VR in most modern automobiles (not Mr Al's ) senses the main electrical bus voltage, which is in the power distribution box where the big, main fuses are....

The main alternator output B lead is a #4-6 wire that goes directly to the + lead of the battery. There is a separate #8 wire that feeds the power distribution box. There is a small IR drop between the battery + pole and where the VR samples the bus voltage caused by the total electrical load (ECU, Dash, Fuel Pump, Lights, AC) . The VR sense voltage is not reduced by the IR drop along the B wire or along the box feed wire because those voltage drops are not inside the feedback loop to the VR. The VR holds the main bus voltage within mV, but the battery terminal voltage might be slightly higher by a few tens of mV.

Kiss is confused about the VR setting to achieve a fully charged battery. I have measured hundreds of cars, trucks, tractors, boats, and airplanes, and the VR setting has always been between 14V and about 14.7V with the engine turning 1200rpm or above. When Calcium was added to batteries about two decades ago, the setting was bumped up by about 300mV from what was common prior to ~1990.

If a flooded-cell lead-acid battery is charged only to ~13.8V or less, it never reaches full charge, it sulphates, and has a short life. One that is recharged to 14.2+V, especially when driven daily, seems to last three to five years. Even though 14.5V sounds too high, I think that this setting is compromise based on the fact that most vehicles are driven an hour or less in the morning, and again in the evening, and this setting provides quick recharge, keeps the battery at near full charge, provides a modicum of equalization, etc.

Another confusion factor is that standard automotive VRs are temperature compensated. They have a temperature sensor (usually a NTC thermistor). They reduce their voltage automatically as the engine compartment comes up to operating temperature. For example, my 2004 GMC pickup measured at the battery is at 14.7V after a cold start with engine reved. After been driven for 30min, with the engine compartment hot, the battery voltage is about 14.4V.

I have been fighting the temperature compensation issue with my Cessna 182. In early models, the battery, alternator, and external VR were all mounted in the engine compartment. Later in the model run, the battery was moved aft into the fuselage aft of the baggage compartment to improve the CG related handling of the aircraft. This leaves the VR subject to engine compartment temperatures, while the battery is freezing in an unheated part of the fuselage. It is one thing if the battery and VR are co-located and subject to the same temperatures; it is another where the VR is sensing the engine compartment temperature, but the battery is elsewhere.

This causes chronic undercharging of the battery, especially in winter months, because the VR reduces the charge voltage during flight to 13.9V or less, and the battery never recharges properly. The solution is to trick the VR....

 

[KeepItSimpleStupid]

This https://www.w8ji.com/battery_and_charging_system.htm guy may have some useful info.

 

[MrAl]

The voltage at the battery "should" be "all over the place" with respect to the "regulator set points". The reason is simple. You have 30 to 120 Amp alternators or so depending on the model. "Load" will change the output because you don't do sensing at the battery terminals.

A "better place" to do "regulation checks" is the alternator terminal heading to the battery on the alternator. There it will be influenced much less by the amount of accessories turned on.

I tend to write down in the service manual, or I used to, the drop from the alternator to the battery with the heater, AC and high beam on. It was my way of measuring the alternator output (a reference) for troubleshooting. I now have an AC/DC clamp meter.

Hi again,

It should not matter what the alternator can put out or what loads are on, but i would find it hard to believe anyway that they would test some regulators with headlights and high power audio system turned all the way up,and some without anything on. Hence the voltage spread seems to wide to me.
I did check at the alternator heavy positive terminal, and only see less than 0.1v drop when the battery is charged up. What i am saying here is the drop is not that important unless there are accessories turned on, and only then if they draw significant current. That's because once the battery comes up near the target charge voltage, the current naturally cuts back, so the voltage drop must reduce as well. For example as i am sure you know, 0.1 ohm with 10 amps drops 1.0 full volt, while with 1 amp it only drops 0.1 volt, quite a difference.

So the reading i really need is the one where there is nothing much turned on. If the regulator senses the battery voltage though, then it may matter less.

Thanks for the link too, it looks like that guy has the meter right on the battery which is where i would like to see the readings taken.

 

[MrAl]

Hi Mike,

Actually, the VR in most modern automobiles (not Mr Al's ) senses the main electrical bus voltage, which is in the power distribution box where the big, main fuses are....

Do you know how they run the wire to that box, is it through the ignition switch or directly wired or maybe a relay ?

The main alternator output B lead is a #4-6 wire that goes directly to the + lead of the battery. There is a separate #8 wire that feeds the power distribution box. There is a small IR drop between the battery + pole and where the VR samples the bus voltage caused by the total electrical load (ECU, Dash, Fuel Pump, Lights, AC) . The VR sense voltage is not reduced by the IR drop along the B wire or along the box feed wire because those voltage drops are not inside the feedback loop to the VR. The VR holds the main bus voltage within mV, but the battery terminal voltage might be slightly higher by a few tens of mV.

Kiss is confused about the VR setting to achieve a fully charged battery. I have measured hundreds of cars, trucks, tractors, boats, and airplanes, and the VR setting has always been between 14V and about 14.7V with the engine turning 1200rpm or above. When Calcium was added to batteries about two decades ago, the setting was bumped up by about 300mV from what was common prior to ~1990.

If a flooded-cell lead-acid battery is charged only to ~13.8V or less, it never reaches full charge, it sulphates, and has a short life. One that is recharged to 14.2+V, especially when driven daily, seems to last three to five years. Even though 14.5V sounds too high, I think that this setting is compromise based on the fact that most vehicles are driven an hour or less in the morning, and again in the evening, and this setting provides quick recharge, keeps the battery at near full charge, provides a modicum of equalization, etc.

Another confusion factor is that standard automotive VRs are temperature compensated. They have a temperature sensor (usually a NTC thermistor). They reduce their voltage automatically as the engine compartment comes up to operating temperature. For example, my 2004 GMC pickup measured at the battery is at 14.7V after a cold start with engine reved. After been driven for 30min, with the engine compartment hot, the battery voltage is about 14.4V.

I have been fighting the temperature compensation issue with my Cessna 182. In early models, the battery, alternator, and external VR were all mounted in the engine compartment. Later in the model run, the battery was moved aft into the fuselage aft of the baggage compartment to improve the CG related handling of the aircraft. This leaves the VR subject to engine compartment temperatures, while the battery is freezing in an unheated part of the fuselage. It is one thing if the battery and VR are co-located and subject to the same temperatures; it is another where the VR is sensing the engine compartment temperature, but the battery is elsewhere.

This causes chronic undercharging of the battery, especially in winter months, because the VR reduces the charge voltage during flight to 13.9V or less, and the battery never recharges properly. The solution is to trick the VR....

That's what i suspected about the 13.8 or 14.0v charging voltage, at the battery terminals. That dosent seem like enough.
What else is i often use the car for very short drives, like under 2 miles, which means the battery doesnt get much time to charge anyway. So over time the voltage drops lower and lower as the car is used over about a two month period. You can tell the series resistance goes up quite a bit too, because even a small load will cause the battery voltage to drop a little while when it is fully charged it barely changes at all with the same small load.

 

[KeepItSimpleStupid]

That storm is still around. The reading I did a few months ago was idle, steady state, recently started.

The 100 mV drop under load used to be a benchmark for me. I'd use that and the voltage as the "first cut" for charging issues. The next would be AC ripple and then I'd scope ripple to figure out if diodes are blown. But now, it's alternator, battery and rarely connections.

On the few alternators, looked at:
a) Worn brushes
b) bad bearings
c) Bad regulator or diodes usually caused by loose battery connections (The paint shop or regular shop didn't tighten the battery terminals)
d) A totally broken wire on the rotor - (the car just died)
e) Loose belt

and one really hard one to find: A corroded bulk-head connector in a 1973 GM vehicle.

I didn't have a clamp-on AC/DC ammeter until last year, but I could usually borrow one. That wire was an uncalibrated shunt and a benchmark.

 

[debe]

Here was my approach to your problem in winter & short trips. It is a Bosch altenator that had a built in reg. I was lucky enough to find a brush holder that fited where the orig reg was fitted. I used an external reg that had a wire loop that if it was cut raised the voltage, I cut the wire loop & fitted a switch, which is in one position for winter & a higher voltage. Or other position that was suitable for summer 14.2V. There is a circuit included for an adjustable regulator. Was also having problems with the exhaust manifold so close to the built in reg with failures due to heat.

 

[MikeMl]

The Bosch system that Debe posted is like Mr. Al's Hundai, and follows the original Motorola Patent. Note that it is not sensing the battery voltage directly; only very indirectly. The premise is that the voltage at D1, which comes from the diode trio, bears some resemblance to the battery voltage. That is both the source of the current that excites the rotor and is the voltage that is compared to the Zener voltage to control the rotor current. Note that the Bosch system uses an idiot lamp which backfeeds a bit of current backwards from the battery via the ignition switch to bootstrap the alternator.

Even the Prestolite (Ford) alternator and VR in my 1967 Cessna is not built this way; the VR has a true "remote sense" wire that comes from the main electrical bus. It doesn't have the extra diode trio inside the alternator or the idiot lamp in the dash, either. It has a separate wire that switches the VR on and off (the rotor current also flows along this wire).

 

[MrAl]

Hi Mike and debe,

debe:
Thanks for the nice schematic and technical paper. That looks like it is similar to mine, or at least close, as the three small diodes go to the VR as i suspected. Take a look at the drawing i made below.

Mike:
Ok then that must be similar to mine. You can see that the three diodes go to the VR which is what makes this difficult.
I consider loading the D+ line to make the internal diodes drop a little more voltage, but i feel that wont be enough because after all i can not load them too much or else they will blow out (along with the regular load current for the coil).
Take a look at the drawing below and see what you think. The proposed solution is shown as that single blue diode.
The idea that the regulator will see 0.5 or so less volts then it sees now, and so will try to push more current through coil #4. That should make coils #1 through #3 put out more current which should result in more voltage at B+.
Hopefully the extra diode drop wont have any secondary effects.
It's a shame the regulator circuit is potted or else this would be cake to fix.
Thinking about using a 1N5400 series diode. Not sure if the diode can be soldered into place, but i suspect it should be ok.

 

[MikeMl]

A Schottky diode would raise the voltage by about 0.3V; a Si diode would raise the voltage by about 0.7V. The Max current that would flow through this diode is about 3A. The extra diode-trio for the field inside the alternator is made out of ~5A diodes...

It really isn't that hard to completely bypass the internal regulator, and replace it with an external one, either homebrew or a store-bought one like the Prestolite one I put in my boat.

I have built regulators for homebuilt aircraft from scratch. The aircraft homebuilders like to use a modern lightweight automotive alternator. Because of in-flight fire danger, it would be stupid to use a Motorola-Bosch-Hundai-style alternator in an aircraft because if the regulator transistor shorts, the alternator runs away (produces full current output, remember it is a 60A current source with a compliance of >100V), and there is no way to shut it off.

I have discarded the built-in regulator from several Japanese alternators, exposing the internal connections, and remoted the new, homebuilt regulator. If you want, I can dig out the schematic for the homemade, external VR. When installed in the aircraft, there is a 60A pull-able breaker between the battery and the main alternator output, and there is a separate pull-able 5A field breaker that runs the VR, allowing complete isolation of the alternator if it goes bad in flight. You cannot pull-over to the side of the road

Here is one you can purchase.

**broken link removed**has a lot of alternator/VR parts.

 

[debe]

This is the reverse engineered adj reg which is quite easy to construct & work quite well. I rev engineered it as they were no longer available & I needed another one.

 

[MrAl]

Hello again debe and Mike,

debe:
Very nice photos, that explains a lot. I'll have to look that over and think about this.

Mike:
The diode physical size seems to have a lot to do with the current rating, and the diodes i've seen that were 3 amp were much bigger than similar ones that are rated only 1 amp. With that in mind, the diode trio in my alternator has three diodes with body length about one half the length of a 1N4004 series diode and about the same diameter, so they are smaller than a 1N4004 diode. That makes me think they are only rated for 1 amp, but i could be wrong of course as i can not see a number on them.
One other interesting thing though is that the driven winding measures 3 ohms with nothing else connected (shaft removed from the alternator and measured directly on the slip rings).
The six larger size diodes used for the main three phase rectification are surface mount and quite large rectangular shaped.

I thought about completely rebuilding the voltage regulator section but im not sure i want to go that route because of the environment it has to run in, under the hood, and im not sure i want to run wires from the alternator all the way into the passenger compartment as that would be about a four foot run.
Granted it would be really nice to be able to adjust it for whatever i wanted it to be.

When you said 'bypass' the regulator, did you mean remove the old one completely and replace it, or just use a permanent fooling regulator that just electrically bypasses the old regulator?
I ask because i cant remove the regulator because the housing also holds the brushes, but i could possibly see using a parallel circuit to the regulator that drives the low end of the coil, and that would effectively take over.

One more little question, do you know if the coil in these things is driven in a linear way or bang bang just on or jsut off? The linear way would push a very variable current through the coil, which might range from 100ma to 3 amps (or whatever the top end is) while the bang bang would just turn it on and off.

Oh yeah one more thing, the old alternator looks like it has the slip rings all carbonized. I wonder if that is why it stopped working. It worked intermittently for a while before it stopped altogether.

 

[KeepItSimpleStupid]

Some measurements:
Temperature measured with Fluke IR thermometer.
Voltage measured with Fluke 77.

2000 GM vehicle; Just started; alternator temp: 25 Fahrenheit ; Voltage at battery: 14.3 V idle; 14.3 V Revved

2000 Toyota Vehicle;
Just started; Alternator temperature: 26 F 14.3 Idle; 14.2 Fast idle
Idle for 20 min; Alternator temp 125 F; 14.13 V; 14.13 V revved.

 

[Mosaic]

Did

View attachment 89313 View attachment 89314 View attachment 89315 View attachment 89316 This is the reverse engineered adj reg which is quite easy to construct & work quite well. I rev engineered it as they were no longer available & I needed another one.

Did you make that PCB with a chisel?

 

[debe]

That side of the board was covered in resin, & yes it had to be chiselled off to reveal the traks.