IF I can find enough cores 

(see wanted ad here (  http://theamcforum.com/forum/dead-r2am4-style-regulators_topic91124.html" rel="nofollow - http://theamcforum.com/forum/dead-r2am4-style-regulators_topic91124.html ))

I will gut the original regulator and put a new regulator circuit inside and re-pot it. 

I am bread-boarding a couple of circuits now to test the basics for operation, stability, and so on. 

I will then move to producing the actual PC board that will fit into the housing and run further testing and if that works out, will pot the circuit and test further. 

This will HOPEFULLY mean I will be able to reproduce the regulators - however, I will need cores - regulators that I can salvage the housing and pigtail from. 

Using a semi-NOS alternator for the testing to ensure nothing is skewed by marginal alternator parts.
(semi-NOS as the pulley shows use where I've used it for testing, LOL - can't call it true NOS now, can I?)

Once that smoke escapes, you can't put it back..........and the device stops working. 

That's been proven time after time - from computers to cars to stereos to your Christmas lights - once the smoke gets out, it can no longer function.

Whoever invents a method to replace the smoke, he'd be wealthy.

I need to go over my Motorola technical data to see exactly what sort of ramp they had figured in for what temp ranges. I have all that data from Motorola's documentation and literature......

I got another bread board and am putting together another design with temperature compensation and a different transistor arrangement using darlington pair........... 

I'll try it out, test each types as much as possible and figure out which does best.
The boards are restrictive in size, about 2.5" x 1.675" so mounting the large transistor may have to be done with stand-offs to get everything to fit well. 
The third-party replacements used much smaller transistors and very small heat sinks so my build may be over-kill but Motorola did use some hefty components, too.
I dug out my old PC board bits - I wonder if the etch solution is still good after three decades? I know the old "resist pen" is all dried up HAHA. I pretty much got my goody out of it years ago anyway.

Motorola up into 1970 used TO-3
The third party TO-220 used a small heat-sink riveted to the transistor, it was free-standing.

Motorola didn't mount the TO-3 transistor to the case or housing. It was totally isolated electrically.

They used thermal-conducting potting materials. They conducted heat to the case but not current. 

I've demonstrated how quickly they respond to heat by warming the up with a hair dryer while watching the voltage drop on a running alternator on my test stand. I was curious as to how close to the Motorola specs. They have a "ramp" specifying the voltage change at various temperatures and I was experimenting on live regulators with heat. 

I may change transistor case styles as it's actually far more than needed - I can easily handle the power dissipation and amperage with a smaller transistor and the TO-220 style. 

My second version DOES use a darlington pair in a TO-220 case, am building a mock-up tonight - maybe.... my wife needs a quilt mounted on the machine tonight (she's already been paid for it, she was commissioned to make it special for a guy)

Is being able to tweak or adjust voltage important at all for any reason, to anyone?

Our cars were never equipped with adjustable regulators so it just seemed that once things were set at the correct voltage, they'd be sealed, even if they originally had a trimmer on the board. 

With the holidays and health things I've not made a lot of progress the last few days but hope to get back at it shortly. 

The Motorola is out in my shop and it's too cold to go back out there just for a photo now....

Also shown - the first version I made with the larger transistor - same case style Motorola used, and the breadboard I'm working on with the second version of my circuit.

(I hope to have a working prototype in a proper case released soon)

This is the version I already tested and have the board designed for........

it works fine.

This is the third-party or replacement regulator (upper left corner of pic) and the new circuit I'm working on that does not use a Zener diode for "sensing" or regulation.

(Then Lucky became interested in what I was doing and wanted to carry some of the parts away, so I stopped.)

I'll have to replenish my PC board supplies because this is the last single layer copper one side board I had and my pens were pretty well dried up. I have enough solution for a few more, perhaps but will likely have someone else make the boards if I end up doing enough of these. 
It takes a lot of time to make the board, drill the holes, etc. 

Doug

Anyway, I had to shift some things around from my original board layout when I determined that it wasn't easy to get this transistor, or even one with the same specs, in a different case style.

So I shifted some things around on the board to open up a space and decided to do it similar to how Motorola did it - multi-levels. They mounted the switching transistor to a "shelf" in the housing and had the main board directly below that in the main body. That is why some ORIGINAL regulators have a housing that is not solid under the tag - that's where the transistor is, on one level, while the board is more "around back". They had to have an open face to screw the transistor to the "shelf" - which aided in heat transfer, but made it more complicated to assemble the regulator guts into the housing. Once the transistor was in place and secured, then they potted it and the material flowed around the "shelf" to encase the transistor and board, the transistor one one side a bit below the tag, the board in back, on the other side.
I will mount the transistor on stand-offs above the blank area of the board (at least for this design)
Depending on the original regulators I can get to salvage the housing and harness from I can also mount the transistor exactly like the originals - assuming I can get my hands on a few of that design.
The transistor case itself in the case of a TO-3 is the collector and is electrically "hot" so must be electrically insulated - otherwise the secret ingredient of any electric device, the smoke, will escape.
I note that they did like I'll do - test each circuit and tweak it as needed to get the regulated voltage desired by soldering resistors and other parts on the back side of the board to "adjust". The third party replacements ground out the carbon resistors where Motorola placed resistors on the back side of the board to "adjust" the voltage before it was potted.

My other design has a trimmer - I may end up doing something similar since even with modern parts of tighter tolerances, you get enough variation it will bug me.

This is what I used - it's pretty nice but it doesn't have all of the "case types" for transistors, and it's not simple to add something like the TO-3 (well, maybe it is and I just haven't figured it out!!)
It's been handy and helped me when I moved things around as it keeps connections and so on. 

I tried to add that TO-3 but it looked as if I had to draw it out myself and I just didn't feel in the mood to get the datasheet out and copy in all the specs and draw. I'm no artist, that's for sure......... Mom was and my little brother is, I guess that skipped me.

My diode and transistor stash started out with a box of military surplus stuff from the early 1960s. I think some of them even have date codes roughly 1964 or so. I added a few up through my teen years but then got into other things and that interest waned........... until the last few years.
I have a few early timer chips (you know, the 555) and other ICs I bought back in the 1970s and some early metal film resistors (before the 2% and better tolerances became common)

TO-3
TO-18
TO-39
TO-52

Found TO-220 under "package" - I suppose because it could be transistor, SCR and other things so used that for my other circuit.

Thanks.

There are likely others - things I have from way back that aren't in this, but those aren't really important right now.

Thanks.

Anyway, here's what I hope to test tomorrow if I get time - all parts on board basically how they would be in the end save for the transistor which I hope to do like Motorola did, and my board will lend itself very well to their design or layout.

But for now it's mounted on the PC board using metal stand-offs to hold it above some other components and make things fit, and hold it solid. 
I found myself getting so picky that I was actually lining up the resistors so the tolerance band was always on the right end with the board leads facing me......... geesh, what does it matter?
And the transistor had to sit perfectly straight, a line through the screws perfectly parallel with the edge of the board. Good grief, once it's done who will ever see it - just get it done!!
Ha, the other two transistors are mounted so they are the exact same height off the board and in a perfect line. I guess this could be a show regulator for AMO?
Naw I'd be docked for a scratch on one of the resistor stripes. 

Will do again later in a warmer shop and let it run a whole lot longer.

So far the only part to get really warm to the touch was the resistor that feeds the field excitation before it starts charging (engine off, key on)
There is no flow through this while it's running, only with key on and engine off, but that's the same wattage as used by Motorola and the third party replacements. 

The board would easily fit an original Motorola shell.......... more testing, and will test with it potted in a shell, but so far success with version or circuit type 1.
(On to type 2 soon)

Switch on, alternator not turning (same as key on, engine not running) the light is on as it should be and the field is being excited.

Alternator spinning at about 3400 RPM, charging, light completely out (no glow).
Charging and regulating.