Alternator - The Build

From the last post, I've got pictures of the Kubota alternator all dismantled.  Next comes the modification pieces.

I've made an adapter ring plate to which flat bars will be welded to hold the case of the dynamo to the engine block.  This has the challenge of interference with the side of the case.  The cap screws that I will use will have to be ground down somewhat, and the side of the dynamo case ground down to accommodate the screws.  This should hold the case firmly to the engine block.

The next big item is machining the crankshaft adapter.  This will be a copy of the back of the armature, which will allow the magnet of the alternator to bolt directly to the crankshaft.  Here's a picture of it just started out of a piece of roundstock.  I'll need to get a parting tool for the lathe before I finish the part.

Here are the plans for the finished part.

Alternator - Proof of Concept

I'm fed up with the original Sunbeam S7 Lucas dynamo, and the MC45L dynamo that I bought. Neither one produces enough electricity in a foolproof fashion.
I don't want to have to worry about if there's enough battery to get me home anymore, and I want to run the lights all the time. I also don't want an after-market alternator conversion that does not look like the original.

The plan is to create an alternator conversion that produces 6 volts, is fully encased in the original housing, and works reliably.

I've seen others who have used a Kubota tractor permanent magnet alternator to do the job. Its only 3-1/2" in diameter and shorter than the MC45L dynamo. That means it should fit inside the casing. The trick is to mount it up.



It comes as a flange mount unit for belt drive, with its own bearings.

[edit]  I actually did use the bearings after all to keep the stator assembly centered in the magnet.  The magnetic forces are VERY strong radially, and will pull the stator into the magnet if there is the slightest mis-alignment or looseness.  A spacer piece was machined to hold two of the bearings into the stator and then the bearings were placed on the center bolt with their own machined spacers.

The permanent magnet rotor can attach directly to the S7 engine crankshaft, and the stator can mount to the engine block, or dynamo casing. All that is critical is that it runs centered and true. Should not be a problem for a guy with a lathe in his workshop.

Here's the sample fit up.  The center bolt presses out of the rotor cup easily.  This will allow me to make a bolt (left-handed Whitworth of course) to fit in there and mount to the crankshaft.  The question is whether 'tis better to bolt the stator to the engine block, or to bolt the stator to the outside of the dynamo, and put the rotor cup "upside down" against the crank.  I'll have to make an adapter to provide the seal to the front of the engine regardless.  The stator to the front makes it easier for the wires to come out to the terminal block on the dynamo casing.

The final part is the regulator / rectifier from a scooter.  I've bench-tested the rectifier / regulator up to 6 amps load.  It works fine with 14 vac input, and does not overheat.

Here's a VERY ROUGH sketch of the assembly.  I keep changing the design, so my dimensioned sketch is pretty bad.  I will require a minor modification to the dynamo case.  The cap screws will interfere with the side of case, so the case will need to be ground away in two small spots not visible from the outside.  The cap screws can be modified slightly to reduce the amount of grinding to the case.

MC45L Dynamo arrives

Just bought the proper dynamo for my S7 Deluxe.  The one that was on it was from a the earlier S7 model.  This one should have more current capability, since it has updated field coils, and four brushes instead of just two.  I knew it wouldn't include the bolts, so I probably will have to get those elsewhere... or use the ones that are on the bike.  It looks like its in fairly good shape for its age.

The commutator is a bit rough, but shows no signs of arcing.  It should be 100% functional.  The brush wires need routed, and the connector needs fastened down too, but it looks like its all there.  I may have to dress up the seal surface on the back of the rotor if I decide to use it. There's some pitting.  Its kinda a shame that the cover has been bead-blasted.  I would have preferred that it still had the original patina.  That's not to say that I couldn't GIVE it some patina.  Hmmm.

Hopefully, this all works and I can sell the older Lucas MC45O dynamo that's on the bike now to someone who needs it to make their original bike proper. (sold it!)

Decreasing Electrical Load on Dynamo

If I can't make the dynamo put out more current, maybe I can reduce the electrical load of the bike and try to make the best use of the available current that the dynamo will produce.  That means replacing some of the high load incandescent bulbs with some modern LED technology.  Of course 6 volt positive ground LED lamps are not commonplace, so I will have to create them.

Here's the general idea.  The trick will be to diffuse the light enough that it doesn't look like they are LEDs.  The other trick will be to separate the high and low intensity "filaments" on the tail / brake light so they don't feed back into the system.  That's easily done with some regular signal diodes, and two different resistors.  With both resistors running in parallel, I'll set their combined resistance at the maximum current of the LEDs.  With only one running, I'll see what percentage of max current still gives a decent output for the tail light "filament" circuit.

The running lamp bulb in the headlight should be even easier.  I can just use some white LEDs and pile them into the bayonet base of a dead lamp.  Diffusing this lamp will probably prove to be more difficult since there's nothing to hide the fabricated LED lamp assembly in the big headlamp housing.  Maybe a piece of plastic diffuser or reflector from something else will work.

I did sacrifice an ultra-bright LED that I had in my junk box from an old project to prove a concept.  Most LEDs have a rather narrow view angle.  I took the file to the lens of one of these LEDs and ground it flat.  Then I used some fine emory paper and polished it up.  I'll have to refine the technique, but essentially it works.  It diffuses the light into a wider area, rather than focusing it into a beam.  I did get a weird halo effect, but I think I just need to polish the edges a little more.

I don't plan to run the bike in the dark if at all possible, so If I can just reliably run the running lamp circuit without the main beam I should be fine.  The main headlamp is a HUGE 30w draw on the dynamo, and it just doesn't keep up.

---

The running light in the head lamp was easy.  6 Ultra-Bright White LED's soldered into a bayonet lamp base.  The rear tail light lamp was a little more challenging, per the above diagram.

Dang! that's BRIGHT!

Both these replacement bulbs LED assemblies draw less than half an amp and put out MORE light than the original incandescent bulbs.

S7 Dynamo MC450 (early rather than late MC45L)

Well, thanks to Steve Benson's pix of his dynamo, I took a closer look at mine.  It is likely a forest for the trees type of thing.  The dynamo fitted to my bike only has two brushes instead of four.  I had never seen a picture of one with the cover off.  And I apparently missed this diagram.  (thanks to Phil)

Lucas Model 450 fitted to KYL95.

Apparently there's supposed to be a later Lucas model MC45L.  It must be an improvement.

So the question is ... what to do about it?  Fashion brush holders for the bottom to increase output and balance the load on the armature, or try to get a different dynamo plate or different dynamo / alternator altogether?  Shame I just spent the cash on a solid state regulator, if I would have known that I'd have so much trouble with it, I'd have been more inclined to covert to 12 volts.

First step will be new brushes.  I'll bodge a set if I have to just to see I can get less voltage drop across the brushes and charge better.

Started on 3rd kick

It was glorious!!!

I simply checked the spark, squirted some extra oil on the camshaft, and put fuel in.  It started on the 3rd kick.

KYL95 is looking good here shining in the fluorescent glow of my workshop.

Now the remaining issues...  The oil light doesn't go out.  So, I checked that with a gauge... a gauge that was too small.  It produces over 10 psi and pegged the gauge, so I'm good there.  Must be the switch, although it checked good last time I checked it a year ago.  Maybe the contacts are stuck?

I'd like to put a T in the line and monitor the oil pressure with a permanent gauge.  Anyone know what kind of thread that is?  Or should I go back to Stewart Engineering and get the adapter and a modern oil switch?

Second issue.  The charge indicator stays on, which indicates more trouble.  I'll try checking the case ground and then tickling the field coil again and see if that does anything.  I recall having an earthing issue on the dynamo casing when I first started working on the bike back in October.  There is the chance that the voltage regulator isn't working right, but that will take some time to debug.

Overall I'm very happy.  It starts amazing easily.  I didn't let it warm up properly enough to turn off the strangler.  That will be the real test, because before it popped and chugged and carried on like crazy before with the old loose & leaky throttle slide.  This new throttle slide is a little on the tight side... The return spring BEARLY overcomes the friction to close the throttle.