Automotive 6 Volt Generator Transistor Voltage Regulator

[shortbus=]

Don't have a design, sorry. But from my experience back in my early days the most usual faliure in a voltage regulator was the points/contacts burning up. The solenoids themselves and settings would be good but the contacts no longer would carry the load. So starting with a good regulator and putting mosfets after the contacts would take the load off of them since they would only be switching a very light load.

 

[Danwvw]

Yes, perhaps place a MOSFET across the DF to Ground contacts but it would need electronics to control it. Actually it could be inplace of the contacts and the contacts used to control it without electronics maybe?

As side note unrelated!
This has me looking at the schematic its for a 12 volt generator voltage regulator and I am a bit puzzled by it as I thought the diode was a field freewheel diode, It's apparently some kind of TVS Diode that protects against overload by increasing resistance. Does anyone know what the diode is? (Note! The 6 volt mechanical regulator does not have the diode it's just connected there i think.)

**broken link removed**


The diode is actually wired in series with the second coil on the field relay. I drew this trying to understand the operation better:

 

[Danwvw]

I can get these on e-bay out of New York. MTP75N03HDL

These came today: FDP025N06.

 

[rjenkinsgb]

I'd guess it's to route all the current through the right-hand coil if the dynamo voltage is less than the battery voltage, so it the disconnect circuit is more sensitive.

The current winding on the left hand one is only needed at high current to prevent overload, while the right hand one needs to be sensitive to small variation.

 

[Danwvw]

The 6 volt mechanical voltage regulator did not have a diode and it was wired just a little different.
https://www.thesamba.com/vw/archives/manuals/voltage_regulator_diagram_closeup.jpg So in theory thinking positive to negative current flow. Current from the generator D+ goes through the current windings in series first then through the voltage windings that are configured in parallel. Looking at the close-up link the fields are opposite on the relays however the current and the voltage windings are all going the same direction within a coil. This means that the current and voltage coils work together to increase magnetism . The 6 volt mechanical regulator uses about 2 or 3 ohms for the resistor which is enough to cause overcharge by itself when the engine is revved. What happens to limit the current is the field relay starts to make the normally open DF to D+ connection which shorts out the field which limits the current.

 

[shortbus=]

The 6 volt mechanical voltage regulator did not have a diode and it was wired just a little different.

Probably due to the 6V having positive ground.

 

[Danwvw]

Don't have a design, sorry. But from my experience back in my early days the most usual faliure in a voltage regulator was the points/contacts burning up. The solenoids themselves and settings would be good but the contacts no longer would carry the load. So starting with a good regulator and putting mosfets after the contacts would take the load off of them since they would only be switching a very light load.

If two MOSFETs were configured like the (common, Normally Open, Normally Closed) field relay on the mechanical regulator even with feedback control to prevent gating them both at the same time there is a risk of blowing both MOSFETs as together they provide a direct short to ground .
However the idea is interesting to me as it more fully emulates the Mechanical Design.
Example:

 

[Danwvw]

Probably due to the 6V having positive ground.

The 6 volt VW gen's are Neg Gnd. those small generators were very basic, simple, cheap and trouble unless everything in the electrical system was right. Actually I have a new respect for them now after trying to understand there exact operation. Not sure when they started using the diode, probably after the 12 volt generator.

 

[rjenkinsgb]

I think the other contact just provides "flywheel" effect, to allow the field current to circulate until it dies off.
The flywheel diode does the same thing so no need with the electronics one.

 

[Danwvw]

The way mechanicals operate is the field relay only pulls in enough to open the NC connection at normal RPM's. They start hitting the Normally Open if revved up high. At least that is what I have seen when I removed the cover and watched them.

Here is a Schematic Draft for the next prototype it has the pull-up resistor. I think I will put all the resistors and capacitors on one board then the IC's, MOSFET and diodes on a second circuit board with terminals for the gnd, generator, shunt, output diode connections etc...

 

[rjenkinsgb]

Looks good, except I'd suggest the 0.1uF cap on pin 2 should be the other end of the 1K, so it takes noise off the reference divider but does not short the output of the other 311.

Possibly also add a low value resistor, 100R ? in series with the FET gate connection.
The 10K pulldown is redundant now it has a pull-up instead.

Looking at it and thinking about what will happen as it starts up, there is a possibility that the voltage control could see it as overvoltage, until the dynamo output exceeds the 6.2V zener level and the reference stabilises.

I'd swap the 680R feeding pin 3 for a 1.8V zener; that guarantees it will stay lower than pin 2 when it should and will also give fractionally better control.

 

[Danwvw]

Looking at it and thinking about what will happen as it starts up, there is a possibility that the voltage control could see it as overvoltage, until the dynamo output exceeds the 6.2V zener level and the reference stabilises.
I'd swap the 680R feeding pin 3 for a 1.8V zener; that guarantees it will stay lower than pin 2 when it should and will also give fractionally better control.

Well, I think that you may have found the problem why it's not starting charging without help! Thanks!

 

[rjenkinsgb]

I hope so, it should be able to self-start..

 

[Danwvw]

Yes, I did some testing today,
First I got a couple of lithium ion batteries that together make 8 volts and I wired them through 5.6 ohms to the Un-Regulated supply. and it took a bit of revving cold but it did start then once started it was ok the gen light still comes on at low idle till it warms up but. it starts charging without revving after that.

Second I removed the 8 volts from the un-regulated and connected through 5.6 ohms + 7 ohms to the reference supply, It pulls the voltage down to the 6.2 volts reference however it would not start charging on it's own with just the reference supply powered by batteries.

Third connected both supplies to the 8 volts with the 5.6 ohms on the batteries plus 7 ohms on the reference supply only and it worked beautifully.

Fourth removed the 8 volt batteries and made a jumper to the ignition coil voltage where running voltage is around 6.8 volts used a 150 ohm resistor (which doesn't help much) on the reference supply while tying it directly to the unregulated supply and this too starts charging beautifully!

 

[rjenkinsgb]

Have you changed that resistor for a zener?
Three diodes in series should do temporarily, if you don't have a 1.8V one.

And, are you now using the pull-up version of the gate drive or still with the prototype that has a pull-down resistor? I think that's what will make a massive difference as it no longer needs the 311 to be powered to start charging.

 

[Danwvw]

No to both, Still playing with the original prototype and design, Perfecting the original design using the coil voltage. Ordered some 2 volt zeners couldn't find any 1.8's from my e-bay supplier.
I will try the 2 volt zener and the pull up on the new prototype that won't use the coil voltage.
I spent all morning dialing in the best start charging behaviour using the coil wire power through a .3 volt diode directly to the unregulated and a 50 ohm now from the same diode to the reference supply.
Did some testing.
Given 120 watts for both Hi beam headlamps:
Given 14" of .047" dia. copper wire for the shunt:
Headlights off shunt drops .123 volts,
Headlights on Hi shunt drops .345 volts.
Measured 5.746 volts to headlights after Hi Lo Switch and fuses ,
Watts/Volts = Current so 120watts/5.746volts = 20.88 amps total for both headlights,
Shunt Ohms = change in volts across the shunt .345-.123 = .222 /20.88=0.01063 ohm.
The current potentiometer is adjusting as it should it can be adjusted to take precedence over the voltage adjustment with the headlights on or when the battery is low.

Oct. 10, 2019 Updated Original Design prototype 6 volt Neg Ground Automotive Generator Voltage Regulator:

 

[rjenkinsgb]

OK, I'd not be too concerned about the initial start up function until the prototype is updated.

 

[Danwvw]

Yeah, it's starting beautifully though. The diode to the coil was most helpful then the 50 ohm resistor on the reference makes it start charging better too and once operational it's not dropping out. I suppose that it should be designed to run on 3 volts ideally.

Here is the next plan! Because the Unregulated is slightly lower than D+ voltage even with 2 volt Zeners, Voltage and Current inputs are only going to be 1.7 volts below supply voltage. Using a STPS4015 .3 volt diode from D+ to Unregulated will help and also adding another shunt.

 

[rjenkinsgb]

Hi Dan,
The input to the voltage reg half needs the diode, as you have it drawn.

The current reg will not work with one diode like that as both inputs need to be proportional to the varying D+ voltage.
You would have to replace the 680R feeding the other input with a 2V zener as well.

To get the adjustment back, add a single diode and 1K from the zener to ground and put the pot across the diode.
That gives pin 3 a range of 0 to -.6V relative to the output side of the shunt, with much more precise control range.

Or just stay with all resistors in both legs of that; I can't figure out for sure at the moment what the voltages will do during start-up.

Adding yet another 2V zener in series with pin 7 may help if the modified circuit has a problem with the current sense part starting, it will prevent the voltage reg being clamped until the supply increases somewhat.

 

[Danwvw]

Ok Thanks I was wondering about that. Still need some parts to finish the New Prototype:

Yesterday I played with the Current Adjustment and gained a better understanding of how it's working on the original prototype. . Kind of a problem is the 14"s of 16 Gauge Copper Wire shunt gets Hot when the headlights are on, Voltage Drop across it goes up to .8 Volts because it's resistance changes so much with the added 20 amps load. But still the Current can be adjusted to provide higher output voltages than needed. Also added a el-cheapo digital Volt Meter on the Dash to keep an eye on voltage while driving.

New Prototype has separate boards for components and a Heat Sink for the STPS40L15CW Output Diode: