I am open to ideas, certainly worth checking out. The current through the field is less than 10 amps. One critical rating here is the "on ohms" Prototypes 2 & 3 use the Motorola HDT Mosfet which is only 8m Ohms of resistance when turned on. In (prototype 1) I used the IRL540 which had 77m ohms of resistance and it worked ok but the kick in rpm is higher with it.
Just calculated the (+-)15 volt supply loads above But didn't add in gate current. Any idea what that would be at -15 Volt? The leakage current is only 100ua and the gate ohms I think is just for High Frequencies. Not sure. The resistors and the gate diode in the "Prototype 4 Schematic" above use the JFET UJ3N065025K3S limit gate current.
I read somewhere that P channel JFET's are not developed for high current applications. The 6 Volt Generator Regulator circuit needs the load to be on the Drain side. So I used the N-Channel MOSFET.
The JFet gate junction conducts when the voltage is positive. In that DS you linked, they show ~ 2,4Ohm.
For negative voltages, the gate current should just be the charge and discharge current of the gate capacitance.
That capacitance is always problematic when trying to switch quickly.
If the comparator current capacity is not enough, the Fet will switch with a linear slope. That slope will cause heat in the Fet.
I am guessing that you know all this already.
I am not really much help.
BTW...
I found this thread because I have a 1951 Mercedes Benz L3500 truck where the Bosch regulator is shot. I wanted to replace the insides with something else but keep the vintage look.
Great! Yes be a bit of work to do it using the Jfet but worth a try. I priced the +-15 volt supplies on e-bay and one can be had for 5 or 6 dollars but as for the JFET above I think it's more like a $15 part. The VW 6 Volt Aftermarket regulator is something like $39 for a mechanical if your generator is a 6 volt neg ground one. Delco others may be available. Now you can buy off the shelf 6 volt electronic regs for vws for 61 lbs https://www.paruzzi.com/uk/volkswagen/electricity/dynamo-and-alternator/1950/ and a coulple of vw websites are selling them in the us for about $80. If your generator is like 100 amps or something the thing that takes up the space is the output schottky diode heat sink. I have a 3" heat sink out of an old mac computer that's on my prototype 2 that gets pretty warm with the headlights going.
Bosch Voltage Regulator, 12 Volt Generator Models, 113 903 803E allows generator to charge the battery when it needs it, but not when it's fully charged! Overcharging will overheat and destroy your battery! New regulators prevent this risk!
vwparts.aircooled.net
It uses 3 output diodes mounted to the housing and potted. Not something you could tear apart and use. Perhaps just come up with a 12 volt Design yourself. The LM311P E is what sets voltage. Just raise the reference voltage from 6.2 to 9.Volts or something Really. though the other input resistor values could be about the same maybe double them. The. LM311P I circuit could probably just use double resistor values. If using a (+-15) Volt 1Watt converter supply though current needs to be limited to below 33ma total per rail. The thing that concerns me about the "On State" JFET design though is reaction time. I think it might over shoot 14Volts at startup. Also if it looses power supply power it will go to full un-regulated output on the Dynamo.
I will be happy to follow along and comment if you want to build your own 12 volt design.
You have done a great job on this. i will probably not design anything at this stage because I have so many other things to worry about.
Will the JFet design overshoot because the Recom module startup is at too high a voltage? At a glance, It appears have an undervoltage lockout on the diagram for the module.
I think your design can work with 12V quite easily with just a few mods - as you already said.\
With all these complications, one has to really appreciate the design of the original mechanical regulators!
I also wonder if one could repair them by replacing the contacts with something salvaged from a relay.
I am not picking up on what your saying "Recom module startup is at too high a voltage" and what are you referring to when you say "undervoltage lockout" But with the (Normally Closed to ground) the JFET has it will need to react very fast to not overshoot output voltage on startup. If the (+-)15 volt supply is slow to turn on, over voltage at startup will be a problem.
Yes it gave me a whole new appreciation for mechanical regulator technology.
Repair? Maybe with the right parts. On the VW mechanical I have gotten them going again by filing the contacts and cleaning and if needed setting the voltage by bending relay wiper or ground contacts a little. Usually though that is risky they can fail and burn out the generator.
Hi, your schematic is more advanced than the spice model one. It does not have current limiting As for your other query, I meant that from memory, the Recom modules have a minimum voltage before they start. As for the undervoltage lockout, I was mistaken. I now see the transistor and Zener near C1 are actually a voltage limit/regulator for ~5V.
The circuit will not be able regulate / turn off the JFet before the Recom module has started up and during that time, there will, as you said, be full output to the battery.
In practice, I am not sure if it would be a problem as it would be very short. one would have to see how it works.
But, in case of component failure, it will also overcharge the battery something terrible. It will likely fail ON instead of OFF.
One more notable difference between your design and the "Bosch" one, is that they have the diode D1 at the D+ side and the whole regulator circuit on the B+ side.
That will take the diode drop out of the equation and it allows the circuit to be powered before starting the engine.
Do you think that helps with eliminating the original problem we are trying to solve of it not starting the regulation?
My guess is that it will (if the battery is not too flat) but on the flip side, it will always sit there drawing some power from the car battery.
The work around I use is to run it off of ignition ON voltage keep D+ D1 to B+ circuit separate. I power the ic's from two sources.
1st source (Ignition SW)
2nd source is (D+)
Aha but then why is there a problem with it not starting the charge initially? I was under the impression that it was because there was no power initially to turn on the Fet.
Or was that problem resolved with the power from the ignition?
Yes that is right. This schematic above I have running in my beetle and it gets its power initially from the ignition. Just a. extra wire. I had wanted to make a design that did not need an connection to battery power. Your Idea of the JFet is the best Idea yet. Just have to figure out how to make it work right.
One way to make it work might be to use a tiny rechargeable 12 Volt battery inside the Regulator. It might hold. it's charge with very low current IC's or something. Such a regulator might need a button on it to press if its battery is dead to make it start working.
In hindsight, the easiest solution for a drop in replacement, no extra wire, will be to just move the diode from the output to the D+ side.
Like in that spice diagram.
Then the circuit will have initial power (in fact permanent).
It may need tweaking of the power consumption in the design a bit.
Just looked up the no load supply current to the LM311P it's (.4ma output Low) and (7.5ma Output high) the negative is (-5ma) so it's probably capable of operating in the off mode at .4ma plus whatever else you want to power off of it like the quiescent current of the LM193. (1 ma) Bosch uses a 5 volt regulator to run everything which uses 6ma quiescent current but it would draw that plus whatever load is on it's output. Looks like 7K + 13K that's say 12.6v/20K=.63ma some maybe the total for the Bosch Voltage Regulator is only 6ma + .5ma + .63ma + 1ma = 8.13ma .
This probably explains why I had to charge the 12 battery in my VW Bus every 3 weeks or so when I ran the bosch solid state 12 volt regulator.