Restoration of 1952 MG TD 2

Author: Bob McCluskey
First posted: 1 Sept 2000
Last amended: Dec 2015
Please email Bob McCluskey
Car No TD/11935
Engine No XPAG/TD2/12333
Body Type 22381
Body No 11301/78948

PLECTRICAL Long ago, in the pre-restoration period, when this car was used daily for general domestic duties, it was never completely reliable: the lights were always a bit erratic and it always seemed to crank very slowly, which made it hard to start. One day as I was trying to coax it into life, I noticed a wisp of smoke coming from the carburettor linkage; and when I touched it, it was hot enough to make me wish I hadn't. Evidently the electrons used this linkage as a shortcut on their way back to the battery after trying to persuade the starter motor to live up to its name. I can't guess how they got there, or where they went afterwards, but I didn't think they should be anywhere even close. To keep them away I ran a heavy duty wire directly from the battery earth terminal to the starter motor fixing bolts. Thereafter it cranked well, and that prompted the second fix, which was to run an earth wire from each light right back to the battery earth, avoiding earth returns through corroded steel or painted panels. It looked messy (of course, by then the whole car looked messy, anyway), but after that I had no more trouble from the electrics, and if there were any bits which still relied on return through the chassis they were obviously well-earthed or non-critical, because they didn't worry me.

original tail-light fitting with one double-filament globe
This is the original tail-light fitting, showing clearly enough why I didn't like earth returns through bodywork or chassis.
front sidelight with two-filament globe
And this was the front sidelight fitting. To enable flashers, I'd soldered wires directly to the terminals on the new two-filament globe. The spring contact for the original single-filament globe was needed to hold the new globe in place, so I'd insulated it with a piece of electrical tape. It did the job, but it was a bugger to change the globe.
Also I'd bodgied up flashing turning indicators, which are something of an embarrassment to me now as I think about them. I'd replaced the single filament sidelight at the front by a double filament globe, without changing the globe-holder (what I'd done was to insulate the contact provided for the single filament, and solder wires directly onto the terminals on the globe itself - how innovative was that?). The rear turning lights were a bit more conventional, using the stop light filament: when you were turning, the brake light filament on that side would flash, and if you braked at the same time, the brake light on the opposite side would come on and stay on while you braked, but the turning side would still flash. Many cars of that period worked that way, for example late TDs, TFs and MGAs, as well as Holdens and many others. But to my mind it was a confusing system and I was surprised it was legal.

So when I came to rewire the nicely painted restoration, my first thought was to make sure that the earth returns didn't depend on the breakdown of the paint film or absence of corrosion: everything would be earthed to either the battery earth terminal or the starter motor fixing bolt - and those two points would be connected together with the thickest wire I could find. And the second thought was to find a way to make turning indicators work properly, with the turning light distinct from the brake light, which meant three filaments in the rear lights - Ivan's mate John had shown me his lights working, so although at that time I hadn't seen his solution I knew there must be room for two globes inside the rear light.

No-one seemed to have any other constructive advice, except to recommend buying a loom from a professional vintage cableform manufacturer. By way of practice I'd already rewired the Lotus, and the lights and horns and other ancillary equipment now work reliably, without intermittent faults. I'd taken out several feet of redundant wiring, and cut out all the nasty black snap connectors, replacing them with neat, modern terminal blocks with brass connectors. But it took much longer than it should have done. The cost of wire, of terminations, of time and (especially) of anguish, really didn't seem to me to make the effort worthwhile, and more than that it didn't look good - even though I'd threaded the wires through heatshrink plastic it was hard to keep them all straight and all the right length, and where there were branches in the harness they just looked scruffy. A bought loom would fit properly, and it would come in a nicely finished professional-looking cotton covered harness, just like the original. So I followed advice, and ordered a loom from a supplier here in Australia - name and contact supplied on request.

The supplier's website offered only two possibilities for TDs - an early one, with dash mounted dipswitch combined with the horn push, and a later one with floor mounted dipswitch, and turning flashers through a relay. Well, mine had a dash-mounted combined dipswitch/horn-push, and no flashers, so that's what I asked for. With some misgivings I accepted the snap connectors offered - those black bullets that everyone who's ever tried to track down an intermittent electrical fault must hate more than whining back axles - on the grounds of originality and the suppliers assurance that they only used copperbased connectors so there wouldn't be a problem with corrosion. I did specify a separate earth wire from each light going right back to the battery, rather than just going back through the body and chassis, and for provision for turning indicator lights by way of an extra wire to each side light and rear light. What came was a professional-looking harness with an instruction sheet specifying early TD using RF95 regulator, so I opened the package with enthusiasm and a light heart.

Well, it probably wouldn't be fair to say that my experience with this was on a par with the American door timber: but there were more problems and disappointments than I thought were strictly necessary, and when I first posted this section I was angry and intemperate. Much of the piss and vinegar has dissipated now, so I have rewritten it, omitting many of the picky problems which annoyed me so much.

Most of the problems came from the loom itself. Firstly the wires had been individually numbered with little slips of sticky paper wrapped around the wires. But in Sydney's warm climate the glue on the little slips didn't seem to stick well to the coated wires, so even when I opened the package many of the numbers - which in any case didn't agree with the numbers given in the manual - had already come off their wires, and unfortunately the connection instructions which came with the cableform refered only to the numbers. All the same, because they were colour coded, and (subject to the exception noted below) the colours agreed with the manual, it was still possible to work out what ought to go where; but it would have been helpful if the numbers hadn't come off, and if they had been the same as the manual, and if the instructions had refered also to the colour coding.

All the wires were about 6 inches too long - better than being too short, but why not make them right? (In about 1963-64, when I had my MG TB, I was working as a circuit design engineer on the prototype of the world's first supersonic swing-wing fighter plane, the English TSR2. The wiring harnesses for these prototypes were made in the wiring shop below our design offices. There I learned that because of the heat caused by friction from the speed of the airframe through the atmosphere, the length of the fuselage was expected to increase by several inches in flight, and the wiring harness was deliberately made longer than the length of the airframe at rest. However MGs are distinctly subsonic, and the car hasn't changed for more than fiftyfive years: how hard could it be to get it right?)

And although I had specified - and paid for - earthing wires from each light all the way back to the battery terminal, its not what I got. The front lights weren't wired as I'd expected, but close enough. But the rear lights and brake lights weren't even close! As well as the (thin) wire for the flashers, and standard cotton covered wires for driving lights and brake lights, each tail light had three black earth wires emerging from the cableform, one with the texture of the original wires and two smooth black plastic covered wires. Well, I thought, one earth wire for each filament. But I found that the two black wires didn't go anywhere: it was really only one wire looped into the harness and back out again, and the same on the other side. And what's more, because the harness was several inches longer than necessary and would have to be trimmed to length, these looped in wires would be cut out from the harness. They had no functionality at all! And checking the continuity of the real earth wires, I found that the earth wires for the two tail lights and the number plate lights were connected through snap connectors to a single wire which was nicely stripped, and crimped and soldered to a 5/16" lug ready to be bolted to the chassis using one of the bolts used to mount the rear damper. However the rear damper bolts are 3/8", and to the best of my knowledge have always been 3/8". The lug was too small to drill out to the right size, and I had to cut it off and crimp and solder the correct size lug: not a major problem, but symptomatic. And instead of a wire for each light going all the way back to the battery earth point, as I had specified and paid for, it seemed that I had one wire serving all three lights, which returned through the chassis, exactly what I had said I wanted to avoid. Responding to my comments, my supplier replied: "if you have only one wire coming out of the terminal that bolts to the rear shocker, it would appear we felt the positive earthing of the tail lights and extra earthing for the indicators was sufficient to meet your needs.... I do believe the earthing for the tail lights will be satisfactory provided you have a clean connection at the earthing point". So in spite of everything, the earth return for the rear lights was to be via the chassis, and would rely on there being no paint and no corrosion at critical connection points. I put it to him that his second-guessing me (and my specification) in making his own judgement of what would be sufficient to meet my needs was disappointing.

after the regulator the loom runs down the firewall
Anyway,as I began to trim the wires going to the regulator, I found the real problem.

According to the diagram in my manual, the live side of the horns should be connected to the A2 terminal on the fusebox, and the other end of the fuse - the A1 terminal - should be connected directly to the battery. So I trimmed and connected the horn wires to A2 (noticing but not being too concerned that this A2 was on the fusebox, whereas the instructions were to connect them to A2 on the regulator, and at the same time noticing that the colour supplied was maroon, whereas the manual said brown/green), and then I looked for the supply wire. I expected to find a brown wire emerging from the cableform, ready to trim and connect to the fuseholder at A1, and I expected the other end to emerge somewhere on the other side of the car, ready to bolt onto the lug on the starter switch.

Well, there was no such wire.

As I puzzled over this I realised that the electrics on my car seemed to be a compromise somewhere between the very early TD (which had the RF95 regulator with 9 terminals and built in fusebox, combined ignition/lights switch, combined dashboard-mounted hornpush/dipper switch, and no turning indicator), and a later one for the TD/TF/MGA (which had a different regulator - possibly RB106/1 - with 5 terminals on the regulator box and separate fusebox, and ignition switch, horn push, light switch and dipper switch which were all separate, and a relay box for the turning lights). I had the later RB106/1(?) 5-terminal regulator like the later system, but the other features - combined ignition/light switch, and combined dash-mounted horn/dipper switch - were as for the early TD and earlier TC. (To my surprise, I realised that the lights weren't fused in either system, but in both cases the current did go through the regulator to regulate the current from the dynamo.) The cableform which had been supplied was for the RF95 system, as the wiring instructions said, where the horn circuit is shared with the lighting circuits. So of course there wasn't provision for a wire from A1 terminal to the battery, because in the RF95 system the A1 terminal on the regulator goes to the A terminal on the ignition switch. So not everything fitted easily.

Most of the MGs I saw at concourses were exactly like mine, but a few had the 9-terminal regulator. And I gradually realised that there weren't just two systems for TDs, but three: the early system, using the same wiring system as the TC, still using the RF95 regulator and simply replacing the weak TC horn with the more powerful windtone horns (but wiring them the same way), an intermediate system like mine using the RB106 regulator and wiring the horns directly from the battery, not through the ammeter and regulator, and the later system using RB106 regulator, and having flashers and separate dipswitch and hornpush, etc. In line with MG's policy of minimising costs, they seem to have saved drawing office costs by giving the same drawing number to the schematics for the first two systems (of course at that time quality assurance and document control hadn't taken control of our lives in the way we now take for granted). The only difference I could find was the new regulator and the wiring for the horns, and perhaps they thought such a small change didn't merit a new drawing number. In the regulator the fuses were separated out: the A3/A4 fuse (for stop light, wipers and petrol warning light) was never really part of the regulator circuit anyway, but the A1 connection on the A1/A2 fuse (for horns) was now separate from the A1 connection on the regulator and was connected directly to the battery. Originally the TC horn had been wired through the fuse to the A1 terminal, through the series winding in the regulator to the A terminal, then through the ammeter to the battery. So when the horn was used, it would draw current through the ammeter and would pull in the series contacts in the regulator to heat the dynamo up a bit. Well, that sounded OK to me, so why would it have been changed? The only explanation I could think of was that the pusillanimous TC horn was changed because no-one could hear it, and that was because it drew no power. In the early TDs it was simply replaced with the much louder windtone horns, originally using the same wiring plan. But the windtone horns were louder because they drew much more current: maybe the engineers found that the regulator couldn't handle the extra power, and maybe the ammeter couldn't handle the much higher current, possibly leading to premature failure. Maybe they thought it would be better to simply draw more current from the battery than to try to compensate through the charging circuit: after all, you'd never use the horn enough to flatten the battery, except in Sydney, when you'd deserve all you got.

So this seemed to explain the anomalies. It accounted for why the wires to the horn were the early colours of purple and purple/black, not the later colours of brown/green and brown/black as I had expected. It accounted for why the fitting instructions directed the horns to be connected to A2 on the regulator, while of course with the separate fusebox there was no A2 on the regulator, and it accounted for why the brown wire I was expecting to connect the A1 terminal to the battery was absent.

Anyway, absent the brown wire, I thought I could make a neat and effective solution by wiring A1 on the fusebox to A3 on the fusebox, which is connected directly to IG on the ignition switch. I reasoned that when the ignition was on, it would be exactly as though it were connected to A1 on the regulator, so it would work exactly the same as the early version except that the horn would only work when the ignition was on (and my very clear memory is that this car did work that way in the past - but perhaps my wiring had been changed to discourage small boys). It did occur to me that if current was the problem, maybe the ignition switch wouldn't be able to handle the increase; but I thought the horn would only be used when the ignition switch was well and truly on: it's not as though the horn current would be turned on and off by the ignition switch itself, in which case one might well expect sparks and premature failure.

But on reflection, I thought that after all I would run a separate wire directly to the battery, external to the harness. Because I only had coils of blue or orange wire, I thought I'd have to put up with it being the wrong colour; but conveniently, there was an extra brown wire (one of the few where the number hadn't come off), separately enclosed in a sealed plastic bag. It was the right colour, it was neatly terminated with a lug at one end, the right size to bolt onto the starter switch, and it was exactly the right length to terminate at the fusebox, and clearly it was meant for just that purpose. The wiring instructions said this extra wire should go from the A1 terminal on the regulator to the starter switch (which of course is connected directly to the battery). This must have been wrong because the wire already on the A1 terminal was brown/blue and went to the A terminal on the lighting switch. So why was the extra wire supplied? And why were the fitting instructions wrong? It seems to me that if the manufaturer thought it worth supplying, he must have suspected the possibility that it might be needed, and that would suggest that he contemplated the possibility that a terminal might exist somewhere which could be connected directly to the battery, and that not all the confusion was mine.

Now it's true that my purchase order didn't specify clearly what I needed in respect of regulator. Some of the fault is mine, because all the information I have now was available to me then. At that time - mea culpa - I hadn't thought through the subtleties of MG's electrics and Lucas's regulators during the evolution of the car, and knew nothing of RF95s or RB106s. But I do believe my supplier could and should have known and foreseen and clarified the issue and should have made it clear that there were more than two systems.

Now I don't know whether this supplier has embraced ISO9000; but if it has, their auditors ought to have a close look at contract review and records: just conceivably, fifty years ago, when quality management systems were still only used by defence procurement and by telcos, you could understand MG reissuing a drawing without changing the number, or not being sure about what they had supplied, but I would have hoped for a bit more traceability these days.

John's solution with two globes
John's solution, showing two globes in one tail light lens. You may find it hard to believe that John has been as scrupulous as possible about the restriction on washing cars during this water crisis.
My solutionMy solution
My solution, clearly inspired by John's
Turning Lights John kindly took his car apart so I could see his solution to the turning lights problem, and inspired by him I made up my own fittings. The globeholders themselves were standard items from an auto accessories supplier, but were more cheaply made with a much lower build standard than the originals (of which just enough had survived for me to be able to make the comparison). They came with some unnecesary bits which I discarded. I made the baseplate and mounting brackets from 1/16" brass and the globeholders soldered in, so I expect no problem with corrosion or continuity. I don't know how John organised his filaments; in mine, the top globe is a single filament for the stop light, and the bottom globe has two filaments, of which the top filament is for the driving/parking light, and the bottom filament is for the turning flasher. The earth wire is soldered to the bracket. On the back of the light, connections to the main loom are made with modern brass spade terminals, covered with heat-shrink tubing. If the lights need to be removed at any time, it will be necessary to cut the tubing off in order to separate the spade fittings.

new sidelight fitting with one double-filament globe front view of the new fitting
rear of the taillight showing heatshrink over spade terminals
And finally, inspired by the success of the rear lights, I made up similar fittings for the front sidelights in order to be able fit two filaments with a degree of respectability. In this case the baseplate was 1/8" brass, and drilled and tapped for the mounting bolts. The earth connection was made by way of a lug underneath the front mounting bolt. As with the rear lights, connections to the main loom are made with modern brass spade terminals, and covered with heat-shrink plastic tube; if the lights ever have to be removed (for example, for panel-beating), it will be necessary to cut off the tubing in order to separate the fittings.

The turning light switch came from a Holden Commodore. It's not exactly the right shape - I would have preferred a straight switch, rather than the cranked switch which all modern cars seem to have - but it had the advantage of being almost as cheap as possible. It has an extra set of contacts for a headlight dip-switch and flasher, which I might connect up at some time in the future, or I might use for something else.

repaired hornpush
Dipswitch and Hornpush Well as I said earlier, this car had the combined dipswitch and hornpush, identical with TCs and early TDs. Thinking first about the hornpush, it was a dreadful combination of incompatible metals and poor design. The live side of the horns is connected to the battery via the A1/A2 fuse; the horn is sounded by earthing the other side via the hornpush button. In essence, the hornpush consists of a piece of brass strip riveted (with a steel rivet) to a baseplate and a brass terminal lug (with a steel lugscrew) to connect to the horns, and a separate flexible strip connected to earth, again with a steel rivet, brass terminal lug, and steel lugscrew. The flexible strip is connected to earth, and is kept in its rest position by a strip of spring steel. When you push the horn, the flexible strip makes contact with and earths the fixed strip. Well, Ok, you might think.

The first problem comes from the earth system: the flexible strip is earthed via a thin annular steel strip attached to the lug. This strip goes around the edge of the hornpush, and is drilled for the bolts which hold the hornpush/dipswitch assembly together. The horn is earthed through the contact of the assembly bolts with the edge of this steel strip as they penetrate it, so the integrity of the earth contact depends on the bolt making a contact with the edge of the metal annulus. In the first picture you can see the annular steel strip, and in the last picture you can see the bolts (the bolts and the holes in the annulus have been highlighted). If the bolts make a contact with the edge of the mounting holes they may make an earth, and the horns might work; but I think you can see that it's likely to be less than completely reliable. Any corrosion, any wear, no earth, no horn. Any significant resistance will limit the current drawn by the horns, which in turn will limit their effectiveness, and instead of a healthy bellow, they sound more like a cow in labour. The bolts make contact with the back of the dash panel, which is itself earthed: yet more opportunity for poor earth connections. Now all this may have been OK for the TCs with their pusillanimous horns which draw next to no current, but I really don't think it works for the more powerful windtones - and thinking about that, it may explain why the hornpush and dipswitch were redesigned for later TDs and TFs.

Also in the first picture you can see the brass terminal and the connection to the annular strip; what you can't see is the other brass terminal, because of the second problem: when I tried to undo the rusted-in steel grubscrew, the whole thing came out of the bakelite substrate, and the same thing happened when I tried to undo the one you can see.

In my case the steel annulus was corroded, so it couldn't make a contact anyway. I could have bought a new switch, but it would have had exactly the same design fault the thought of which is just offensive. So I set about to fix the existing switch.

The first imperative was to make a decent earth. Since by now both brass lugs were missing, leaving 3mm holes through the substrate, I replaced them with 3mm ss nuts and bolts, and soldered wires, one for earth and the other for the active hornpush, directly onto the brass contact strips below. I used a cyanoacrilate glue (superglue) as a threadlock. So now the hornpush at least would work reliably and effectively.

The dipswitch had similar problems. One of the terminals was severely corroded. There was a small crack in the bakelite. I drilled out the corroded terminal, and replaced it with a stainless steel bolt; I had to file the nut to fit into the body of the switch, and to make a good contact with the tumbler. I soldered the wire into a loop and fitted it under the nut. That meant that the wire did not exit the switch through the original hole, and I had to drill a new hole in the bakelite. The nut was fixed using cyanoacrilate glue, which was allowed to wick down through the threads by capillary action; the small crack was fixed the same way.

A quick coat of paint, and modern terminations, and its better than new.

Panel Lights Well the panel lights are a real bugger, as well as being expensive. I used four of them, partly from symmetry, and partly because I wasn't going to have a panel light dimmer switch, or a foglight switch, but I did want turning light warning lights. rear of instrument panel There is a resistive winding in series with the globe, made of very fine (ie easily damaged) resistance wire. Because of the construction of the light, the panel dash hole can be only marginally bigger than the diameter of the series winding, meaning extreme care must be used in fitting them. The fitting is held in place from the back with a spring and circlip arrangement, which means you have to use some sort of spring compressor to fit them. I used a 3/8" W open-ended spanner. The globe is held in position by the plastic lens itself, by bearing onto the front of the globe, so I suppose the purpose of the series winding is to partly to reduce the brightness of the globe, and partly to stop the globe from melting the plastic. The winding's resistance is approximately 85 ohms, and the resistance of the globe itself is approx 9 ohm. That means that without the series winding the globe would dissipate about 16 watts; with the winding it is approx 0.15 watts. But the globe itself is marked 12V, 2.2W, so it is a puzzle. The tolerances on the windings are so wide that some of the globes glow brightly, whereas some of them are hard to see even in the darkened garage. It must be possible nowadays to get very low wattage bulbs, so doing away with the need for the series winding, and it really does seem to me to be taking originality to silly lengths to copy such a difficult and expensive method. And now I can tell you something else about these panel lights: they are physically very fragile, so that while I was undoing the grub screw on the brand-new fitting (which was very tightly screwed into place) for the earth-side connection, I sheared off the whole of the back of the light, connections and all. All very unsatisfactory. Too late to change now! This sort of thing accounts for why MGs lost market share to Triumph, because all the little fiddly fittings which had been inherited from early T-types (and even earlier!) were all expensive and unreliable, and all the little brackets and adjustments had to be chromed, and added cost but no value, whereas Triumph started again right from the beginning and took advantage of modern methods and materials and built a cheap, reliable, attractive, stream-lined fast car.

The instrument lights worked fine for the speedo and tacho, but the little bracket for the light for the oilpressure gauge and ammeter seemed wanting: there was nothing to keep the light in, and given the problem with light pollution in the Sydney sky I thought it important to do my bit. So I made up a little fitting to hold the globe and fit snugly around the two instruments. The only problem is that the light is a little bright, so I might have another go. This picture shows the fitting, and also shows how modern connectors will allow the dashpanel to be removed simply and quickly.

windscreen wiper motor Windscreen Wiper The windscreen wiper is described as a three-pole wound armature running in a two pole field. I connected mine up to a power supply, and found that, like so much else, it had seized up during the long interval since it was last used to wipe windscreens. Unfortunately electric motors take a large current when they are stalled, and can heat up quite dramatically. I'm not sure now whether one of the armature windings burnt out before or just after it got hot enough to melt the grease and allow it to start turning. When it did turn, a loose wire from the burnt-out pole struck the field pole, followed by a shower of sparks with a smell of burning insulation. It was really quite spectacular for a very short time, before the armature wiring unwound itself and the wire got jammed between the armature and the field laminations. At that stage I thought it wise to disconnect the power and unjam the armature, although it was probably too late to save the motor. But in fact I found that only one winding had failed, the other two were still continuous, and just out of curiosity I took the gearbox apart and tried just the motor. And I found to my amazement that it worked. So then I took the whole motor apart, cleaned and regreased the gears and replaced some of the dodgy insulation, cleaned the commutator and brushes, and tried again. So now it is working quite well as a two-pole wound armature in a two pole field - better at any rate than it ever did when it was in daily service as a three-pole wound armature.

Dynamo I had promised myself to install a modern alternator instead of the dynamo. I can't remember now why I broke that promise, but I had the dynamo rebuilt by an auto-electrician, who also showed me how to condition the field winding to ensure that the remanent magnetism in the soft iron core of the field magnet had the correct polarity. The cost of reconditioning was considerably more that a secondhand alternator would have been. Initially I fitted a "no maintenance, no retensioning" pulley belt from a Japanese supplier. No retensioning means no stretch at all, and that meant not being able to wind the belt onto the dynamo, as we all used to do, because it has to stretch a little to go over the flange. So I ended up having to remove the dynamo completely, thread the belt over all three pulleys, and then bolt the dynamo up again. But worse than that, the section was so small that there wasn't enough friction with the pulley wheel, and the belt slipped with an irritating whining noise. So I went to my MG supplier and bought the correct belt and there has been no more trouble.

voltage regulator Voltage Regulator Well having gone with the dynamo, there seemed little choice but to use the voltage regulator, even though it must be the least understood part of the car. I think that part of the confusion comes from the circuit diagram, which doesn't show clearly that one pair of contacts is normally open and one pair normally closed, or which are operated by the series coils and which by the shunts. In fact the circuit schematic makes it look as though the Normally Open contacts are controlled by the series windings. But they are not.

As far as I can tell, it works like this:

The battery is connected through the ammeter to terminal A on the regulator (A for ammeter, obviously), and thence through the series winding to terminal A1 on the regulator which in turn is connected directly to terminal A on the ignition switch (this lettering system makes me think that an earlier iteration of the regulator might have had only four terminals, A1 and the series winding between A1 and A being omitted, and the battery connected to terminal A on the ignition switch, and through the ammeter to terminal A on the regulator). Anyway, when the ignition is turned on, current flows through the ignition warning light to terminal D (D for dynamo), and thence though the output winding of the dynamo to ground. The ignition light glows. At the same time, current flows through the regulator frame, through the normally closed contacts to the field winding F.

circuit diagramWhen the dynamo starts to spin its output voltage begins to rise. As it begins to rise, it is fed back to the field windings through the normally closed contacts. At the same time, the voltage on both sides of the ignition light equilibrate, and the light begins to dim, and finally goes out. Current begins to flow through the shunt windings, and when the dynamo is spinning fast enough, this current is enough to pull in the normally open contacts. Now current can flow from the output (terminal D), through the frame of the regulator, through the (now closed) normally open contacts, through the series winding, and through the ammeter (which therefore shows a charging current) to the battery.

When this charging current through the series winding is high enough, it pulls in the armature and opens the normally closed contacts which have been allowing current to the field windings. The field current is now limited by the 100ohm resistor in parallel with the (now open) normally closed contacts, and the output voltage falls. The current through the series winding is reduced, and is insufficient to hold the armature. The normally closed contacts close again, and current is restored to the field windings. Output current is regulated by the on/off cycle of the field current. The duty cycle depends on the regulator settings, but try to resist the temptation to adjust them.

When extra current is drawn by the sidelights or headlights, the extra current flows through the series winding between terminal A and terminal A1. I believe this series winding is wound in opposition to the main series winding, and therefore acts to keep the normally closed contacts closed longer, allowing the dynamo to keep charging at higher revs.

Although seemingly complicated, in fact there is little to go wrong with it. Except for this. The soft iron cores of the series and shunt relays also have remanent magnetism. This can be enough to slow down the release of the normally open contacts. If this happens when you turn off the ignition, then current - large amounts of current - can flow from the battery, through the ammeter, through the series coils, through the (now closed) normally open contacts, through the shunt coils - which therefore hold the normally open contacts in - and through the output windings of the dynamo to ground. There is very little restance in any of these windings, and therefore very little to limit the current. You can tell if this happens because the ignition light glows brightly even though the ignition is off, and because the ammeter needle wraps itself around the end stop. If this happens, you should start the engine again, open the bonnet and remove the regulator cover, and manually open the contacts while leaning over to turn off the engine. And I suggest you do this quickly, before there is time to do damage, like setting fire to the car or burning out the armature.

Contact Breaker and Coil. Initially I used the original coil and condenser, although I did put in new the contact breaker points. It worked OK, but eventually, because of a problem with overheating, I replaced it with a new Bosch coil. I suppose I could have dealt with this subject in the chapter on overheating, because that's where it arose. And because the coil is undoubtedly electrical, I could have dealt with it here. But because of the confusion surrounding coils in a positiive earth system, I chose to give it a section all of its own.

send me an email

Top of Page