Forum

[Adrian]

Sounds like you're using an Electrex World CDI stator kit, Stinkwheel! Only 50W output for ligting on some models...



If you want one switch to control AC and CD side by side, perhaps a double-pole toggle switch will do the job?



https://www.ebay.co.uk/itm/Toggle-Switc ... Swh9hZ72RR



A.

[stinkwheel]

Aye, one of those but as far as I can see, it's the only setup (other than a later Indian thunderbird system) that will do both ignition and charging on the end of the crank. This one claims to do 65W. I'm using 1970's C90 switchgear on it (for appearance, bare alloy and uncluttered). There's no way I'm taking all the load through the switches though, I reckon two relays and a diode will do it but it's very much at the back of a fag packet stage for now. Here's what I have for now, might give OP some ideas too?

[Aethelric]

Although I've had a fair bit of experience on other makes, this is my first Enfield, and I don't know anyone else with one. I assumed the crap lights were just a characteristic. But I thought I'd take some measurements. I got 7.5V at the headlamp when reving quite hard. No, I thought, it should be better than that.

An investigation showed that the amber lead from the alternator was connected to chassis. Nothing fundamentally wrong with that and it would explain why the return for the pilot light was connected to the amber lead on the main bulb. This seems to hav been covered in other posts

BUT I also found the stator winding had a ground connection somewhere in the middle of the coils. Either/Or is OK but not both. Checked the stator - It was fitted backwards so that the bare soldered joints were facing the crankcase and one of then was shorting to the inner chaincase. So a coil was shorted out.

Now I just have to refit it the right way round it making sure the stator/rotor gap is OK. Spec says 10 thou nominal, 6 thou minimal. A google search reveals aluminium beer cans are 4 thou thick. Two thicknesses should do it.

Time for a beer. .

[Aethelric]

Stinkwheel, I can see the point of the relay for the headlight, but I'm not convinced you need one for the sidelights etc. If you used LED bulbs for these, they may well take less current than the relay.

If you are going to use two relays, maybe use a spco relay for the main and dip and avoid sending the bulb current through the dip switch?

[Adrian]

I would want to keep the amber AC return separate from anything on the rest of the system. The only other bulb powered off the AC would be the main beam warning light (as fitted at the factory, at least), everything else comes off the DC side.



The Thunderbird/Machismo (and very early 350 Electra) CDI alternator should still have a reasonable out put, but is not an easy swap as it uses a different mainshaft on the crank - the keyway is in a totally different position for the trigger arm, onto which the un-keyed alternator rotor slots. Part number 502042/A or possibly 500214F. Not showing on our hosts' parts search, but they do have some unlisted stuff.



A.

[stinkwheel]

Aethelric, pleased to hear you've found something. I was thinking that I find my AC headlight more than adequate and that the dip at low revs is pretty much only when I'm stopped anyway. There again, I'm aware that I have pretty dgood eyesight and other people are fussier than me about lighting.



Adrian. Yes, you're right, the switchover relay is for the back of the next "fag packet" (I design looms on a series of 4x2" cards which I put onto a big diagram once I have it worked out, saves getting overloaded with information. It's a design technique my higher technology teacher showed me). And yes, I probably don't need a relay on the side lights. There again, I also intend to use multicore wire and hide the relays/battery/fuses etc in a tea caddy air filter so there will be one cable to/from the switchgear then a seperate one going back up to the lights so it may as well have a "clean" supply. It can afford to be very fine gauge cable if it's just running relays.

[Aethelric]

I did a test on the AC regulator. It is NOT a pair of back to back zener diodes.

It is a single SCR set to trigger at about 17V in one direction only! This was not what I expected, but on reflection:-

(If you are not electrically minded this may seem gobbledygook)

At low revs and voltages it will do nothing. this will be up to 17V peak or 12Vrms

As the revs and voltage increase slightly above 12V the SCR will trigger thus removing some of one side of the sine wave. I suspect there is a capacitor in there to give a bit of phase delay.

With careful choice of phase delay and trigger voltage it will limit the RMS voltage to around 12V.

But the waveform will be a strange shape when it is regulating, and a digital meter or even a moving coil meter will not give an accurate reading of the rms voltage.
It is a simple but very elegant design, giving a regulator which intrinsically has very low power dissipation with minimal components.

[stinkwheel]

It's certainly an order of magnitude better than any other direct lighting system I've seen on a motorcycle. Most of them rely on the combination of bulbs fitted drawing enough power to prevent them going over-voltage and popping at maximum revs. Even down to using ballast resistors when on sidelights only. You used to get what I referred to as cascade bulb failures. If one bulb went, the rest would follow one after the other as they were progressively fed more and more power.

[Aethelric]

The alternator short was probably not the exposed solder wires. Someone had fitted shake proof lock washers under the nuts - wrong size - and one was spreading and reaching a winding and piercing the insulation. Lock washers removed, lock-tight added.

Good grief - I have a really bright headlight.

Now to test it I need to go for a ride in the dark - on a moonlit night in mid-summer - without disturbing the neighbours.



Now maybe I don't need the original mod I suggested, but I do think that the regulator may be better located closer to the headlights.
T
The problem is measuring the voltage. A DMM will not give an accurate reading on a clipped waveform. Maybe the regulator sets the voltage slightly high to allow for volt-drop in the wires? If so, moving it would cause blown bulbs!!





The Generalised Iceberg Theorem:- 90% of EVERYTHING is beneath the surface.

[dave p]

I've recently improved the headlamp of my 2008 traditional Bullet by a factor of erm, quite a lot.
One "Neolite" H/lamp unit from our hosts and one Phillips Xtreme Vision H4 60/55 watt bulb. The difference is amazing, no qualms about night riding now.