I just remembered – and I don’t think you are going to like me for the remembrance – but when I was willy-nilly tossing stuff out of my shop back in the early-to late-90s, in that timeframe, I tossed away a lot of things that I just thought would never have any value.
Four or five of those voltage regulators for your car were amongst them… I am glad that Niagara is a long way away from me, now.
Yessir Rob. I’m hoping that’s what it is. After a few tests in situ I was getting a steady reading of only a few volts at the regulator at 3000 rpm instead of the 16-16.5 specified. Corroded points according to the diagnostic. Took the unit out to check and clean up the points. Zero resistance across the cutout points closed, so good there, but across the regulator points the same resistance reading either open or closed, 60 ohms or so. Cleaned up the points and now no resistance with the points closed so maybe found the issue. I’ll reinstall and resume the exercise tomorrow. Good hint. I’ll smear a little dielectric grease on the rivet connections. Thanks again!
Geez, Wiggs, you probably threw away a bunch of eBay treasures. 
… you have noidea: another was a 5-gallon bucket FULL of new 36 hp VW pushrods (integrated lifter), still in cosmoline.
I’ve kept stuff for decades “just in case” and never did anything with them so tossed them out. A week or two later I’m kicking myself when they would have been useful. Karma.
Success!
Regulator reinstalled, engine fired up and the ignition warning light went out. Pulled the positive battery lead with the engine idling and it wanted to die so increased revs to around 1500, pulled the lead and she kept running strong.
That was fun.
Thanks to all for the advice - Robert for the education on digital v. analogue multimeters that steered me away from the generator to the regulator, Clive for the Lucas Tech course providing a clear understanding of regulator function, @Erica_Moss for the troubleshooting guide and Rob who correctly called the points as the culprit, seconded by Tom, along with all the other expert advice. A team effort.
(Sounds like I’m giving an acceptance speech at an awards ceremony, but what a valuable resource this board is, eh?).
Now waiting for Welsh to get back to me on the compatibility of their XK120 water pump.
I did precisely that with a very rare, and made-from-unobtanium set of Boge front shocks for my BMW Isetta: I had no idea what they went to, but I kept them for years, because they were kind of cute.
Literally about a month before I got my Isetta, I junked them. I’ve been kicking myself ever since then, and I don’t even own that little pile of junk anymore!
I know you’ve said you found the XK120 cramped for space, but an Isetta?
Roomier than a 120!
However, the rate of acceleration… was like a T bill, gathering interest (apologies to P. Egan)!
Well folks, that was short lived. Took the 120 out for another spirited run. Ran well. As much fun as the last time. Parked it in the driveway for awhile. Restarted it and the ignition warning light’s on again. Checked continuity and resistance at the points and still good. Obviously still missing something.
The points have been known to stick closed.
Yes. In this case, the cutout relay IMO.
Yes, it’s the cutout relay (right again, Robert) but it’s still a puzzle.
On a fresh startup the ignition warning light is on. Battery not charging and no power is getting to the ignition system. Pull the positive battery lead and the engine dies. Checked the cutout points and they are open when they should be closed.
With the engine running and the ignition warning light illuminated, I pressed down on the cutout armature (red arrow) to make contact with the core (blue arrow) thus manually closing the points (green arrow) which stay closed when finger pressure on the armature is released, so manually closing the points has caused the core to be energised.
the ignition warning light goes out, the battery is being charged and the ignition system is energised normally. Pull the battery lead and the engine keeps running.
Shut off the engine and the ignition warning light goes out. Restart the engine and everything is functioning as normal.
Engine now off, pressing down on the cutout armature once again causes the points to close and stay closed, thus energising the core, and the ignition warning light illuminates and stays on, until such time as the positive battery terminal is pulled, which de-energises the cutout core and opens the points, which remain open when the battery terminal is reconnected.
Restart the car and everything’s working as it should … but for how long?
Intermittent grounding problem?
(my brain hurts)
Some type of interference with that cut off arm: a problem with the armature, or an intermittent open in the windings, or in the pivot points?
Yeah, guess I’ll check all that out tomorrow. I’ll also take Rob’s suggestion to solder up the rivets in back, in particular the rivet securing the ground terminal.
My new water pump is on its way from Welsh. Getting closer to the end of the fettling stage
You keep thinking that, old boy… 
Hey, an old man can hope, young fella.
Sort of. There are two coils on these things, sorta like on a starter solenoid. That’s because with the magnetic circuit open (relay not “pulled in” it takes a lot of current (pull in current" to generate enough magnetism to move the armature. But once it’s pulled in, it takes much less current to keep it in (“holding current”). Making up symbols for convenience, the pull-in current Ip (through the thicker coil winding with resistance Rthick) is given by Ip = (Vgen-Vbattery)/Rthick. So to pull in the relay, the generator has to be producing a voltage significantly higher than the battery’s–that’s the whole point, namely to allow current flow only from gen to bat, not the other way. Once the relay is pulled in, the gen and battery voltages will either be the same (voltage reg relay closed–charging) or different (voltage regulator relay open; it buzzes between these states). To prevent the cutout relay from cutting out again, there’s a smaller coil (like in the starter solenoid) that’s wired between ground and the generator output (armature). This current is sufficient to keep the relay closed but not to close it. When the generator stops, this hold voltage disappears (as does the pull-in voltage), and the contacts open.
Or so I think (been a while). Manually closing the points and having it stick closed suggests the hold winding works, but the pull-in doesn’t. Since it works sometimes I think the pull-in winding must have continuity. Maybe the adjustment screw can be be used to move the points a little closer together. It should also help to rev the engine to try and help it pull in. That’s enough; I probably don’t understand exactly what’s going on.
Ah. So we’ve (well, you’ve) ostensibly narrowed down the (intermittent) problem to the heavier series coil. Getting closer …
Hi Nickolas,
I can think of a couple of causes. First is that the cut-out contacts are still a bit dirty, so that they only give good electrical contact when you are first giving good physical contact by pressing them together.
If you are happy that the contacts are clean, measure the resistance across the shunt coil. That is from D to E terminals (disconnect one of the external connections - else you will be measuring the generator resistance). If it is open circuit, then the shunt coil is blown or there is an internal break, such as at a riveted connection. The only resistance spec I can find is for a later regulator at 9 ohms.
If there is a short in the shunt winding, then its resistance will be lower and it will have reduced pull. When running at speed, try holding a screwdriver near the core to see if it is attracted - i.e. there is some pull from the winding.
If the shunt winding is good then I would suspect poor connection to it, or the contact adjustment needs changing. If the poor connections were external to the regulator then I would expect the regulator circuit or the charging circuit to misbehave, which you indicate is not the case.
No harm in checking the external connections. On the earth/ground side there is generator to engine, engine to body, regulator to body. On the D side check at generator end and at regulator end.
Hope you find the problem.
Regards,
Clive.
Some lights are coming on. Is that smaller coil always live?
With all the help here I’m interested in finding it too. It’s an enjoyable puzzle.
