Project: Self-build alternator for a Lister 6/1

[AdeV]

Calling this thread a "work in progress" at this stage would be a bit like calling a government department "efficient" - i.e. it's completely wrong. There's no work yet, nor progress... Hence, "Project".

The purpose of this thread will be to document the design & building of a simple homebrew alternator that can be made by any fool. I propose to be that fool, to make the mistakes and frankly, to have the most fun along the way. Bob's excellent white papers, most of which flew right over my head, seem to suggest a claw pole alternator (whatever one of those is). But there was a great picture of an Oerlikon unit, and I quite fancy something that looks a bit like that. Only smaller, obviously. Another thing which was mentioned was the possibility of using a claw pole alternator as a starter, given the right electronics. This also appeals to me, because if (if I'm honest) I'm a lazy bugger & don't fancy hand-cranking for the rest of my life.

Anyway... on to the real meat & potatoes:


At this stage in the game, I need to get a handle on a basic design. The spec is as follows:

- I'd quite like the whirly bit in the middle (rotor) to have permanant magnets on it, so I don't have to mess around with brushes & commutators (remember: this is an alternator which should be constructable by fools).
- I'd like to be able to attach the alternator directly to the crankshaft (or via a simple 2- or 3-cog "gearbox") of the Lister, maybe via a Lovejoy (homemade version) or some kind of sliding universal joint.
- It may need to be clutched (solenoid operated, eventually, a lever will do for now) if it cogs too hard to allow the engine to be spun over by hand when starting
- or some other means of avoiding cogging may need to be found.
- It would ideally be amenable to serving as a starter motor

Now.... the (potentially) impossible:

- It would be really nice if it output 415volts 3-phase at 50hz
- But if not, single phase 240v at 50hz will do
- It should be configurable for 110v @ 60Hz - either at build time, or run time
- If none of the above are sensible, feasible or possible, it should be arranged to produce a sensible DC voltage (24, 48 or 96), 48v is probably the best.
- Fully loaded, it should pull around the full 6hp of the engine. If I can drag 4KW out of it, I'll be very happy.

A direct drive system will rotate at around 650rpm. Obviously, a simple gearbox would allow any speed to be chosen, at the expense of mechanical simplicity and reliability, not to mention cost. Whilst I should be able to make my own gears, many who might attempt a simple alternator won't have the necessary mill & dividing head, so readily available and cheap gears would make sense. Maybe a couple of bicycle gears & a chain would do as an alternative?

The alternator needs to be as efficient as is reasonably possible. I'm not shooting for 95% or 98% here; but 80%+ should be possible?

The ultimate goal will be:
1) A useful alternator to replace my broken start-o-matic unit
2) A white paper with simple step-by-step construction plans to enable anyone with the will & a few tools to construct a similar machine

So........ help! 

[mobile_bob]

personally i would not be so quick to discount the use of a wound field rotor in favor the use of permanent magnets,
at least myself.

there is the arguement that field current takes away from efficiency, which it might? but a well designed unit should only
require about 1-2% of the generated power for excitation anyway and the ability to regulated the field and by extention
the alternator output in my opinion swings the balance toward the wound field.

yes you will need two slip rings and a brush holder, but that is really not such a problem as one could borrow both from
a suitable automotive donor, some of which are easily removed from the rotor shaft for refit to your project.

another member and i have discussed the possibility of building a large scale version of this design, but the stumbling block
is a suitable stator, we need a rather large donor for that. it could be a stator core from a larger motor for that matter.

the rotor could be made fairly easily to be mounted directly to the engine shaft via a taperlock,

the rotor could be made up of steel plate either laser, water jet or plasma cut, or for that matter torch cut
chucked and trued in a lathe of sufficient size. a jib built to position the clawpoles (also precut) on each plate
with a slight tip outward (perhaps 5 degree's) so that when the rotor is reassembled and remounted in the lathe
the claws can be leveled and have the necessary taper in the process, and the diameter reduced to provide the
designed airgap (rotor to stator clearance)

the number of poles (claw pole pieces) will have to be determined by the slot count of the donor stator.

then the stator will have to be rewound, after first winding in a test coil to determine the operating characterisitics
of the new machine, (output voltage at your designed rpm at your designed field current), then you will have the turn
count after which you rewind for 3 phase and your off to the races.

i would expect the rotor diameter would need to be somewhere in the 22" range to be effective at 650rpm, however
that is a guess because i haven't run the math... but i think i am close

finding that large of a stator might be a problem, unless you can find a old motor rewind shop with a scrap pile of large frame
stuff lieing about?  a suitable stator ought to provide the iron for 3 -4 of the machines you are proposing, so there might be
others that would go in with you on the purchase of a surplus stator?

the next major issue will be mounting that stator rigidly so that the rotor can run within it without contact of course, that might
be quite a job in itself.

Then there is the Torogen design, i will look and see if i can find my notes and drawings on that one
it might be easier than what you have in mind and it uses a permanent magnet rotor.

in any event this project would be more of a labor of love in my opinion, as there are several alternatives that might
be much easier to work with?  but nothing would be cooler in my opinion that an alternator of your own manufacture.

bob g

[Jedon]

Jens, sounds like you are talking about an Axial Flux alternator?
http://www.otherpower.com/listeraxialflux.shtml

[AdeV]

Hi Jens,

An axial flux generator was, initially, what I was thinking of. It's still definitely on the cards as an idea, but I wasn't sure if it could be turned into a motor with any sort of power, especially using neo magnets. Plus I have to say I think the Oerlikon-style alternator just looks so cool... but I suppose I could always hide the axial flux machine in a similar shroud, if I wanted to  I guess the only other possible fly in the axial ointment is efficiency; the otherpower unit which is frequently linked to failed due to overheating (due to its own inefficiency); I'd like to overcome that if I can.

Bob - I will have to read your post a few more times before I reply to anything in it.... your grasp of things alternatorish is clearly far above my own...

[mobile_bob]

i need to get my scanner working, maybe tomorrow while i am home i can accomplish task
the old book that i got the oerlikon alternator from has several plates in the back illustrating all
sorts of really awesome looking alternators of that time frame.

they are downright beautiful and look like an art form at least to me, and should you want to build
an alternator from scratch for your own use, you could do worse by not using one of these designs.

i am thinking a lister coupled directly to a reproduction of one of those alternators would be about the
coolest thing on the planet, i would gladly part with some efficiency for a faithful reproduction of one of
those alternator,, and Brown's Oerlikon alternator ranks right up there to me.

As much as i like the looks of the Oerlikon alternator, i have to admit there are others that are as interesting
and as artistic or even moreso.

if i live long enough, i would love to reproduce scale models of those alternators, along with scale steam engines
or diesel engine's that would be fully functional. can you imagine a tiny steam powered oerlikon alternator powering
lights around a model train set?  that would be awesome to look at in my opinion.

the plates are all bi and tri fold in the back of the book and most are in terrible condition having been folded incorrectly
many times over the last 100 plus years, so taking them out for scanning i can justify as a conservation measure.

maybe i will take them to kinko's and get some of my favorites blown up to poster size, and make them available to
anyone that wants one for their engine room.

As for the otherpower alternator, it is an aircore that also dates back surprizingly to about 1888 or so, i will be sure and
scan it as well. the original had a wound rotor instead of neo magnets, but clearly it is the grandpa of the otherpower aircore
alternators.

there are at least 2 of the otherpower alternators running direct drive on listeroids, one of which is doing very well as it is feeding
an mx60 controller in buck converter mode to allow the alternator to run up to fairly high voltage and then rectify, buck convert
and feed a 48 volt bank,, his efficiency ought to be quite good, but
we have not seen good testing done on it yet, at least i have not

he reports something ~4kwatts output from a 6/1 which if true (and i have no real reason to doubt him, but can't quantify at this time)
makes the alternator either quite efficient or he has one of the strongest 6/1 engine's i have heard of.

my hope was he would tell us more about his machine and agree to some standard testing wherein we could accurately measure his
output (amps x volts) and not take what the mx60 says as gospel and also weigh the fuel consumed with a gram scale so that
we could get an apples to apples comparison. It takes a bit of work to do good testing, something most folks don't care to do, most
especially if they are met with ridicule from the start ( i have seen that happen on one of the other aircore units). hard to get any info
if we shoot the messenger before we try to get more data out of him.

so at this point the jury is out on the air core alternator, i don't know how good they are as implimented by the guy with the mx60 attached.  they may be excellent? i just don't know and probably won't build one to find out. Neo magnets are expensive and it takes
a fair bit of labor to build any alternator, so i would have to believe the unit to be near 85% or better for me to want to jump onboard
and put out the effort,,, unless...

it was a faithful copy of the original design, then maybe??? 

of course the original design did not use neo magnets

bob g

[SHIPCHIEF]

I like the idea of a direct drive alternator for my Listeroid 25/2. Those axial flux equipt 6/1's look cool, but they don't put out 60 Hz and 120/240 1 Phase  or the 120/208 3 Phase that we all know and love..
900 RPM is right near the sweet spot for my engine (as tested).
So that would be an 8 pole machine. Not knowing much about Neo magnets etc. and having much experience with regulated alternators, I think I would go regulated.
My ST-12 seems to work OK, but the oscilloscope does not lie. I would like a nice clean sine-wave.

[mobile_bob]

before a made a move away from the st12 i would have to ask
1. is it harmonic winding excited? or
2. does it have other excitation, and
3. are you using an AVR?

setting out to design and build an alternator is a serious project in my opinion,
having said that i do think a guy can design one or adapt an existing design and
get a good result.

be forewarned though, there will be several folks of the electrical engineering persuation that would
differ with my position on that most vehemently.

but thats ok! after reading about all the alternator designs from the early days, and finding most were
in excess of 90% efficiency, i figure why not give it a try? 
even edison with his incredibly poorly designed early DC unit was a bit better than 80% and it used solid
iron in places his contemporaries were using laminates.

building a functional direct drive alternator for a lister or similar engine that could beat 80% efficiency would
place its designer/builder on a very high pedestal in my eyes,, most especially if it turned out to be an effort
by one or more of our members.

and if it could be made to produce 60hz sinewave with a reasonable (better than a standard ST head), well
now that would surely make the one(s) responsible and this forum a very popular place.

lofty goal? perhaps? 

never going to know until somebody tries it i guess

bob g

[AdeV]

Well, I'm tapping my feet waiting for my lathe to arrive (not till next week now  ), so I can cut round stuff... The dividing head in the mill just ain't no substitute.

Hokay, I suppose it would be an idea to do some maths...

If I'm shooting for 3-phase at 50hz, from 650rpm, how many coils do I need & how big will the stator be to accomodate them? To answer this, I need to know how each coil generates. It's been a few years since I did Physics, but I think that as a magnetic north passes the centre of the coil, the sine wave peaks at the top; and the south pole peaks at the bottom. Therefore, one coil generates one complete cycle in two magnet passes (1 north, 1 south). This would be true if the magnet irrespective of whether the magnet is a neo PM, or an electromagnet.

Presumably, where you have multiple coils, the same pattern is repeated every 4th coil (every coil being joined to the 4th one around from itself), building current; and the number of turns in each coil determines the voltage. Please, someone stop me if I'm wrong here. Also, stop me if the basic physics is 8-year-old schoolchild stuff which I really ought to go read up in private...

Anyway, assuming I'm on track here...

The engine develops 6hp @ 650rpm, or 10.833 revs per second.
50Hz generates a complete waveform 50 times per second, or 4.1654 times per revolution.

So that ain't going to work... which means we need a transmission....

Gears or belts, what does the panel think? Or have I ballsed up a calculation somewhere?

[mobile_bob]

Ade:

i think a 10 pole machine would get you down to 600rpm, but there is not a combination that will
provide 50hz at 650rpm alternator speed.

now if you go belt/gear/chain drive of course you can do about anything you want.

building a 50hz 3 phase machine might be an ambitious endevour, i don't know.

now if you can find a stator from a large 3phase 50hz machine, the project goes from
very difficult to something that is doable in my opinion

what are the chances of finding an large stator over on that side of the pond?
you might check any old surplus or salvage/junk yards, or maybe some industrial motor
rewind shops, or similar,

maybe a suitable old stator could be found that would then be the keystone of the project.

there ought to be someone that has a large old motor kicking around that is just too old
and of some oddball voltage or otherwise unsuitable for modern installation, as long as it
is a 3phase stator, it would be good enough to work with in my opinion,, and we don't care
about the stator windings,
although a good stator with windings would provide test bed to see what you can get out of it.

it would however be much easier to make this thing into an alternator to rectify and charge
a battery bank with, i think.

i like the idea of a direct drive alternator, one that mounts directly onto the lister crankshaft
such a machine would then be very easy to determine the efficiency of, because you could simply
measure the torque delivered to the stator case and the rpm and calculate the exact power input
and factor it against the power output...

that would certainly stop all arguement about efficiency of the alternator.

i would like to see you do this project, and write up a white paper on it.

bob g

[mobile_bob]

dammit Jens

quit peeking over my shoulder!!!

eyes on your own paper Jens!



bob g

ps.  i am waiting for your white paper on the "jens stirling cogen"

[mobile_bob]

i think i just completed the 32nd edition of that one!

lmao

bob g

[AdeV]

I'd like 3-phase power so I can run my mill & lathe - and any other 3-phase stuff I end up with - without modification. However, I think this may be a different paper, some other time...

So, I'll shoot for 48v DC I think. I can still wire this as a 3-phase, to make the rectification smoother without giant capacitors.

I have a pair of old brake drums off a Capri. I think these might make good claw pole rotor components



I'm assuming they're cast iron. A magnet is attracted to them. If I turned the rim off them, then milled out the slots, would they do the job?

Hmm, maybe I should do a white paper on the appropriate colour of paper for a project 

[rcavictim]

I have a pair of old brake drums off a Capri. I think these might make good claw pole rotor components



I'm assuming they're cast iron. A magnet is attracted to them. If I turned the rim off them, then milled out the slots, would they do the job?

Hmm, maybe I should do a white paper on the appropriate colour of paper for a project 

I suspect there is way insufficient 'meat' in the flat end caps on those automotive brake rotors to carry the desired magnetic field lines out from the shaft wound solenoid to the claw poles.

If you look at the huge heavy rims from commercial transport tractor rear wheels you will find something with possibly enough meat, although the central solenoid shaft will have to be huge, and bolted to the holes in the rims.  I have three of these out in my work yard (salvaged from the dump) providing ballast for a 30 foot lifting gantry to keep the wind from toppling it.  It is pretty light made from TV tower. The original design with concrete blocks in the corners almost flipped over once but got hung up and this saved from destruction by getting tangled in the elevated end of my buried in concrete, elevated railway track. That's wandering off-topic but the point here is that they are really heavy and contain a thick cross section everywhere. I think this would be doable and one would have a rotor suitable for a large ~22" diameter stator containing many poles and thus slow RPM.  Cut the claw poles in a mill as you suggested.  Hmmmm..........  Too bad my lathe only swings 16", and I don't have a mill.

[mobile_bob]

the plate will need to be at least 1/4" thick and better to be 3/8 or even 1/2"

if you need a 24" finished rotor, you turn the disks to 23.5"
the lay them down onto a jig, where precut claws of .5" thick are to be
attached by welding them with a mig.

the jig would have two functions, one is the proper spacing placement
the other would be tilting them out at approx 5 degrees. they would be placed
to where the claw is set with approx a 1/4" overhang from the edge of the disk
this would provide for a root for the outer weld and give you the OD dimension of
approx 24" after the ourside is turned down or ground to size.

after the OD is turned or ground the claw profile will then be thickest at the root
and thinner at the tips as it should be.

the would then be tac welded in place, after which the claws would be fully welded.
the weld is important not just for the obvious concern of strength, but the flux has
to make the transition from the disk through the weld into the claw poles.

going with a DC machine allows many more poles, which makes "making" power much easier.

i would not waste my time using car or truck wheels, they are unsuitable for this project in my
opinion, better to give the project every advantage from the start, compromise is something you
do only when you must. my opinion and your mileage may vary.

the Torogen (patent unpending)

the torogen is a single disk rotor pm machine
where in you make up a disk with as many hole around the periphery as you want poles, this single disk is
attached to the crankshaft

you now have a spinning flat rotor with  many poles, if you used a 1/4 inch disk and inserted into it 3/8 or 1/2"
round neo magnets they will be mounted a bit proud of the disk, with a 1/4 plate i would recommend a 3/8" magnet
so that they protrude approx 1/16" from the disk face. each side of the disk is then covered in resin and fibeglass cloth
to encapsulate the magnets.

the rotor runs inside of specially made toroids, the manufacture of toroid cores can cut and polish them, so given that
they can also cut a slot within the toroid. we simply have a 3/8" slot cut (plus your airgap dimension) and we wind each
toroid with sufficient turns as needed based on a test coil.

these toroids are suspended around the periphery so that the disk runs through the slots of the toroids

the advantages of such a machine

1. the toroid provides the perfect magnetic path

2. the stator is now modular, and each toroid can be replaced easily should one coil burn out at some point

3. interconnections for various voltages could be obtained

to further improve the design

mount the rotor disk to a shaft collar, and turn the outer portions to fit the id of a pair of ball brgs
the endplates are bored to receive the ball brgs, so that the total alternator is mounted on the crankshaft
this make mounting much easier, and requires only one torque arm to hold the stator stationary
(the stator of course is the end plates and the toroids mounted between them)

if this torque arm were calibrated for length from the center of the crankshaft, and a spring scale were attached
with the torque arm comins out parallel with the floor, one could then measure the torque delivered to the alternator
and factored against rpm of the engine we could determine the exact amount of power delivered to the alternator

with that and the measured rectified DC output (volts x amps) we could determine very quickly the efficiency of the
alternator.

so there is the Torogen in rough outline, of course there are finer details as one could imagine, things like welded in
webs to stiffen the rotor, leaving cutouts in the end plates so that the webbed rotor could pull cooling air flow in and push
it out past the toroids. mounting of the toroids themselves, winding of the toroids by hand (made much easier with the slot)
encapsulation of the toroids and their windings, etc etc.

also the type of tape wound toroid is important, the orientation of the laminates is important, the laminates need to be
more like a stack of washers than simply a wound ribbon, (although a wound ribbon would work with higher eddy current
losses)

when i originally designed the torogen, my first thought of prototype toroids was to use wound steel strapping tape that had
been fully annealed, after talking to a toroid core manufacture i found that the toroids of the size i wanted were about 7 bucks
each so i figured it might be best just to have them produce them, cut the slots and grind/polish the slots as needed.

bob g

[mobile_bob]

btw, if you are interested in a variable induction system for the torogen let me know
it would allow for zero cogging for ease of startup, and an infinite control of the amount
of flux in the toroid airgap, from zero to maximum flux. the method is much more complex
from a mechaical standpoint but certainly doable in my opinion.

bob g