Resurrecting this thread for progress. We had thought replacing the burned-out diode between the IGN term on the upper alt to the FLD term on the Twinput solved the charging issues. But over the weekend, prolonged idling caused the front battery to cook, H2S spewing and a hot battery.
So after consulting the book of smarts (the TM-20) and paying attention to the wiring diagram F-10 in the Appendix had something kick off in my gray mush. And then consulting TB 43-0001-39-8 on the Duvac wiring modification, this vehicle had not been properly modified. So I set off on the journey into problem resolution.
Sure enough, the TB had not been applied. And to make matters worse, the line connecting the Twinput terminal A+ to the small terminal of the engine wiring harness block HAD MELTED!! Looked like it had been "repaired" with some strange blue wire (main is orange). And the crimp terminal was scorched, melted in two.
This meant the Twinput did not sense battery 1 voltage, hence the boiled battery. Ach zo! This was repaired IAW the TB, making the wiring to the TM-20's Appendix F-10. But still, voltage was way too high on the rear, And low on the front. Still the problem remained. So we pulled the Duvac and Twinput, cleaning 25 years of off-road dust and making sure connections were clean. The Zener diode pair (12-0-24 volt, anode ground) had burned out, another thing to worry about. The 12 volt side is a a 1N4744a which is a 15 volt Zener, the burned out one is on the 24 volt side so it should be a 1N4751a.
So I run diagnostics IAW troubleshooting section, pulling the 20 amp fuse for the engine and noting the voltage change. As expected, the change was not within the TM steps. Thoughts of pulling the volt meter relay danced in my head along with new regulators. But with two regulators, ONE must be working so I swapped. When pulling the regulators off, the IGN terminals were highly corroded. Seems Leece-Neville had used lightly plated steel fasteners, 25 years meant corrosion was certain. Slapped the now cleaner regulators on the alternators and SUCCESS!!!!
Voltage on the dash voltmeter is right in the middle green, right in the middle of the white hash. Now when tested with a DVM, and cheap ones at that, the meter is reading the PEAK voltage, the Duvac's switching SCR means there is plenty of "overshoot", much like water hammer in a plumbing circuit as a silicon controlled rectifier like the Duvac is like a fast acting gate valve. The noise is suppressed by the Zener diode pair on the Duvac.
Now I need to see if I have any of these filter diodes...and am assuming the value for the high side, if the 12 volt is a 1N4744a, the 24 volt must be double that which is a 1N4451a.