Automotive 6 Volt Generator Transistor Voltage Regulator

[Danwvw]

Yup, As near as I could tell testing the mechanical voltage regulator the field bias resistor for a mechanical reg should be about 2 or 3 ohms but I bet original mechanical reg resistors have a high negative temperature coefficient probably they go up to 10 ohms or more.
The 10 watt resistors I have shown on top of the generator in the mock-up are 10 ohms and 7 ohms.
I could use the 10 ohm one to start with but if just 1 amp was needed at the Armature (ie.. D+) it may not be possible with these generators as they tend to not want to start charging.
Another way to start them charging is to go ahead and connect D+ to B+. (ie.. the Battery).

 

[Danwvw]

Wow! Passed the first testing:

 

[Danwvw]

Putting it on the Scope, it does have an Oscillation at pin 7 the I comparator output when the current limit is at threshold.
A .1 mf cap to ground seems to remedy it, however it introduced a bit of hysteresis into the voltage comparator. I suppose it's time for a test where the generators output voltage is simulated to see how well it would regulated. I don't have a variable DC power supply!

 

[rjenkinsgb]

I'd not worry too much about that, as the output transistor looks to be working inverted the way you have the test circuit.
That means positive feedback rather than negative.

 

[Danwvw]

I kind of liked it that way (inverted) as I was able to omit the 2n2222 transistors etc.. It makes enough voltage too but I have it set to about 4 volts with about 7.5Volts powering the LM311's with the voltage divider there on the gate. I did testing and found anything over 3.7 volts kept the MOSFET cooI.

Built the prototype up and tried to test it on a real 6 volt Generator yesterday But I can't get this Un-Know 6 volt Generator to charge unless I power the field with my 6 volt battery charger. (DF to Neg and Ground to Pos.) It should work with D+ Positive and Ground negative and the DF to the 10 ohm but it doesn't load that way. ?????? It's like someone may have wired the field windings wrong. I think they should be in series between D+ and DF with a polarity such that it would motor counterclockwise. It motors clockwise! These VW generators are meant to charge running clockwise.

 

[rjenkinsgb]

It may just need repolarizing.

From your diagrams, the field and armature should be in parallel if you connect DF to D-
It should then run the same direction regardless of which way you put the charger across D+ and D-

Let it run for a short while with the wires connected the correct direction. Whichever way it runs, that's the way it needs to turn to generate - once the speed is high enough, the back EMF will increase to the point the current flow reverses (or it would if connected to a battery.

It may need quite a lot of speed to self-energise. The output polarity is dependent on any residual magnetic field in the metalwork.
If that is wrong, try connecting the field to 12V for a few seconds, that should leave a bit of residual magnetism.

Conventional dynamos with field common at the ground end are given a start-up bias from the ignition warning lamp, but I don't see any similarly simple approach with that odd setup.

 

[Danwvw]

Yes, I did those things I think, But it could just be the 10 ohm field resistor is too high!

The 10 ohm broke off so now I am testing the prototype on a different 6 volt gen in the car with a 7 ohm field resistor.

It won't start charging by itself however when I jumped out the barrier diode it kicked in and was charging just a little. I could remove the jumper then and it would keep charging and did respond to the Voltage Adjustment however even turning the voltage adjustment all the way would not get it over about 6..5 volts or so. Then after it got warm it quit holding in probably the I "current" limit is not temperature stable enough may need adjustment, letting it cool down now.

I should have put LED diodes on the LM311 outputs so I could see what was going on but didn't want to load them.
I will do another test with 3 ohms on the field to ground and see what that does.

 

[Danwvw]

Just tried it again after it cooled off. Still with 7 ohm field resistor. This time at a fast idle about 1500 RPM on the motor more at the generator, it was staying charging after jumping the barrier diode to make it start charging. The 500 ohm Voltage adjustment seems to adjust the output voltage measured before the barrier diode from 6.4 volts 6.8 volts.

 

[Danwvw]

OK with my 3 ohm Field resistors actually three 1 ohm resistors in series. It will start charging if revved a little but over charges revved up however adjusting the current limit (1K 10 turn pot) will stop the over charging. The current adjustment seems to be the one in control with it I can adjust the generators output voltage all the way up to 9 volts revved with the headlights on. That's before the barrier diode it's about 8.2 volts at B+ after the barrier diode. The 10 turn current limit pot has about 1 turn in the range where it actually adjust it.
This design does work it's just spongee! Video

 

[rjenkinsgb]

I wonder if it is worth adding an eg. 5V relay plus resistor, to short out the 10 ohm resistor until the dynamo produces some reasonable voltage?

If the current control can regulate it, then the voltage control should be able to? The current part controls the feedback to the voltage section, so the 311 is working, it appears to just be the voltage dividers feeding that which are out of range at present?

Re. the rather touchy control, that's why I suggested a fixed resistor plus a lower value preset. It means much more of the preset range is in the wanted region.

Looking promising though, for a first prototype it's going well!

 

[Danwvw]

Thanks "rjenkinsghb" for your continuing interest, Your being most helpful. I am not sure I understood presets totally, Those would be the fixed resistors on the LM311's Right? Here is the Updated Schematic. Measured the ohms of the pots and it looks like I had the voltage all the way up trying to get more, I will try backing it down etc...

Yeah, the 3 ohms Field Resistor will not even hold it in at Idle. I need to do a bit more testing and see if I can go even lower. Relay, That's going to need a tough relay!
I want to try to put the scope on it and see what the gate is doing.

Schematic with the resistor values: E's pot set to 3 ohms, I's pot set to 385 ohms (Note! the shunt is that coil of copper wire, the .005 ohms is from the charts on copper wire that size and length.)

 

[Danwvw]

Did more testing. The 3 ohm field resistor is too low as it's overriding the regulator. It's going up over 9 volts even when the regulator is dropped out. But here are some photos of the Gate voltage on the scope when the regulator is kicking in.




2 volts per cm @ 5ms

 

[Danwvw]

One thing, kind of an Ace in the hole is the Generator Charge Wire, It's Key Switched 6 volts through a light bulb. I just tried connecting it to UnRegulated 8 volts supply, It helps a lot getting things started, may be able to go back to 10 ohms or whatever works best to not overcharge on the Field Resistor and still get it started. It would need a lower resistance lamp or perhaps we could chance some things in the circuit to accommodate using this switched voltage source.

 

[Danwvw]

Ok it dialed in! Both Voltage and Current adjustments are controlling!

Installed the 7Ohm Field Resistor, Connected the Gen Lamp normally it goes to D+ to the U-8 supply Output.. It still needs help starting to charge but no longer drops at idle and it will even start charging by itself after running and hot. Am able to control it nicely from about 6.5 volts to 9 volts at B+. With B+ set to 7 volts with the headlights on fast idle it will build to 7.1 or so revved however without the headlights on it goes to 7.4 volts revved. Gen lamp works correctly only it's dim when on at all.
See the gate waveform on the Scope here: 7Ohm

 

[rjenkinsgb]

OK;

How about moving the IC & reference supplies (the 47 ohms & zeners section) to ignition switched power? That should lock the FET "ON" until the engine starts and the dynamo output builds up?

That should allow a high value resistor to work, as the field will be at maximum anyway, until the regulator kicks in.

[Or add small schottky recs so power is taken from either D+ or ignition?]

 

[Danwvw]

Yes, I have kind of thought about that.
To keep the Electronic Voltage Regulator Design here compatible with stock wiring we only have :
4 connections to the cars wiring,: B+, Load, Gen Lite wire and ground. To the generator D+, DF, and Ground.
The Gen lite wire is switched 6 volts through a light bulb and that has the job of lighting the Gen lamp when the generator is not charging, It does not have enough power to power the generator field (It's normally already connected at D+). It would have to have some kind of relay to get power from B+.
Connected to the LM311 power supply It helps though, it is keeping the gen light from coming on at idle even though the MOSFET drops out and it quit glowing dim when revved as it did before this way it's just not bright enough.
I think it's time to look at components.
What about a PTC Thermistor a Posistor for the Field Resistor? And A better Barrier Diode to help with Gen light glow? Getting over a volt drop across the Barrier Diode when revved and loaded. PTGL5SAR1R0M1B51A0. This Thermistor may not be high enough wattage though?

Posistor

 

[rjenkinsgb]

I can't think of any simple component combination that would boost start-up without also messing up operation while running.
If you put a PTC in place of the resistor, it will be "reset" any time the FET locks on due to undervoltage.
What is the field coil resistance?

The lamp glow could be reduced by adding a diode in series.

Once it's running, a large cap or supercap across the 311 supply could possibly keep that stable and maintain control of the FET when the engine revs drop too low for normal dynamo output?

 

[Danwvw]

Ok, I have never played with the PTC thermistor but it looks like the right one would do the job nicely, the mosfet may not reset it as it's pulsing quick and the thermistor is only going to reset if it cools down below a certain temp and the current thru it drops below minimum I would think they are slow to reset? And I see what your saying.
I am thinking the field Resistance is about 1 ohm Testing with the mechanical reg seemed like the voltage DF to ground was running between 5 volts and .8 volts I should check that against the electronic design prototype.
Did some more testing with the same circuit with the 7 ohm field resistor. It really controls better than I ever expected. The voltage is working and holding like a rock while adjustable from 6 to 9 volts. The soggy behaviour is the barrier diode.
I probably need to put a higher value in for the shunt. I have just set the current adjust as close as possible to where it starts to drop the voltage. (Voltage Pot set to 171 ohms) (current pot 397 ohms) is just right running the 7 ohm Field Resistor to get out 7 volts to the battery with the high beams on.
On the scope with High beams on and the Right Turn Signal going I am seeing Noise within the Gate on Pulse, Pretty High Frequency It's the aftermarket flasher relay in my turn signals that is the source of the noise I think? Probably need some more little .1 uf caps in the regulator circuit somewhere.

Here is this info. again:

Field voltage is 2.4 Volts with lights & reved. and 1.7 volts no lights reved. So minimum field current is 2.4/1.3=1.85 Amps or 1.7/1.3=1.31 Amps.
So what this means to me is that adding 120 watts load at say 7 volts about 17 amps to the Generator Output changes the field current about 0.54 amps.

 

[rjenkinsgb]

Thanks for that, I do remember seeing it now...

As you say, if the minimum field is over 1A then seven ohms should be fine.

I'd not worry about any apparent oscillation, it should be switching rapidly while in regulation; the FET voltage may vary drastically but the current through the field and its flywheel diode will only be varying a fraction.

You can add caps to the reference side of the voltage divider and across the inputs of the current 311, but I'd not put anything anywhere else within the feedback loop, on any switching type signals.

A large poly cap of some sort between D+ and D- would be OK to help reduce spikes there.

If the current shunt is really 0.01 ohm, it should be OK? Even half that should still give enough voltage for reasonable control.

If you want to make adjustment easier, change the preset for a fixed resistor around 50 ohms lower than the working value plus a 100 ohm preset in series; that will make the adjustment five times less sensitive, if you presently have a 500 ohm preset.

 

[Danwvw]

Thanks "rjenkinsgb" for your insight, I am amazed at the handle you have on electronics, it's great to chat with you on this or anything. I have enjoyed prototyping this little 6 volt generator voltage regulator with you, everything has gone so smoothly unlike other electronic Ideas I have had.
I have worked for Hewlett Packard years ago building Oscilloscopes but I was working as an Electro Mechanic at the time maintaining production equipment did a lot of High Vacuum stuff, Built some controllers and got know everyone and all the Jobs they did building CRT's. I was always a bit of a pain for the bosses as I would design my own electronics for machines rather than get something already made from off the shelf. Then I got into computers doing systems management for HP but that job was really the Building Systems Electronic Tech Job.
After leaving HP I got an Electronic Engineering Job doing High End Home Theater Sound equipment, mostly testing, troubleshooting and repair for the design engineers. That all is in the past. So now it's been a lot of fun for me to play with this Voltage Regulator,

I will post a final schematic with resistor values that will fix the voltage range so it can't burn out the generator etc... I want to build another one before changing 680 / 1.2K ohm input divider resistor values as thing are just too tight with the prototype..

Will post a final schematic for anyone that wants to make there own when I figure out the Field Start Circuit design whatever it turns out to be. I did order some 2.5 ohm 2322 661 54711 PTC Thermistors to try on the field.

Still need to source a better barrier diode, A TO-247 package with an electrically isolated heat sink contact would be nice, something that did not have much change in forward voltage drop at around 25 Amps.
As always will try your suggestions on input values. Testing is kind of a wild thing to do, the header exhaust on my little 36 horse VW is noisy and the heat from the engine is blowing on me as I rev the engine and try to take readings without getting the voltmeter leads caught in the generator belt. Fun!
Wish you could see the Scope waveforms it's producing, difficult to catch on camera. The amplitude on the gate builds as the engine revs until the generator output voltage becomes stable and the gate pulse amplitude reaches about 4.4 volts. Its ON width varies from just a spike to a over square wave depending on load and RPM. It's still drops out at Idle however it comes back in without having to over rev the engine like it needs the first time stating. Part of the starting charging problem is this generator, it does this too on the mechanical voltage regulator. The generator seems to lose its residual magnetism, If left more than 3 days.