(pdf) high output/efficiency alternator white paper

[mobile_bob]

This is a paper i have been thinking of putting together and publishing for several months now
and up to this point i had no avenue to use to publish it that i felt would be around for any length of time.

This attached paper is not for everyone, many folks will have little or no interest, however
those that are looking toward the use of batteries and inverters, or building an inverter/generator should
find this paper useful.

I welcome constructive comment and of course questions, and would be willing to argue points of contention
as well should anyone feel the need to do so.

Not knowing how to copyright this work, i would only ask fellow forum members to keep and use for your own benefit
and not copy/paste this work to other forums, i have no problems if anyone wants to link back to here from another forum
i just don't want this taken elsewhere.  

This work comprises many hundreds of hours of research and development over the course of several years, every effort
has been made to design a suitable alternative using off the shelf parts to accomplish the goals set out for controlled battery
charging higher voltage battery systems efficiently. I have a significant investment in test equipment, many different controllers
by various manufactures and stacks of alternator cores that i have worked with to arrive at as simple a solution as is possible
and still get a quality result.

nuff said

bob g

excerpt from the paper

ABSTRACT

   The typical offgrid home will often be powered by various alternative means, with solar, wind,
sometimes hydro being the predominate prime sources of electrical power. In order to use these alternate source of power over the full range of a day batteries are the most common method of storage and buffering the power generated to provide for the home loads over the hours when either the wind is not blowing or the sun is not shining.

   In order to assure proper charging for longer periods of time that there is no power generation available,  the most common approach is that of a standby generator. Standby generators can either provide for the loads or provide for battery charging or as is most common a combination of both.

   Generator systems that provide for both loads and battery charging most often use either the built in charger of the inverter system or a standalone battery charger that is simply plugged into the generator. In fewer cases the system will have a generator to provide for the loads and also an automotive alternator to do the battery charging. This paper will explore the use of both methods and illustrate the overall efficiency compared to a new approach of the  "reapplication" of an automotive
alternator that will be shown to operate at higher power density and with dramatically higher efficiency than either of the prior methods.

   This paper will investigate if the typical automotive alternator herein referred to as the clawpole
alternator can be operated at a much higher output power level and at dramatically higher efficiency.
A brief history of the machine will be presented, its theory of operation, what is known about the machine when used as designed, and what can be accomplished by changing the machines operating parameters.

   In order for the reader to fully appreciate what is presented in this paper he will need to effectively forget everything he thought he knew about the automotive alternators theory of operation, and he will need to also put out of mind any assumptions based on prior experience or what has been published by the original manufactures of the machine.

   While the machine was built by and for the automotive industry our intended use is not concerned with the constraints place on it by that use. For our intended purpose of charging a battery bank at a fixed speed the machine should simply be viewed as any alternator might be, the only similarity is visual  to its automotive counterpart.

for complete text download attachment (available only to forum members)

[Henry W]

Hello Bob,

I liked what I read. Reading it on paper gave me a better understanding on how higher efficiency can be reached. as you know I have been trying to decide what my next project would be.

It looks like this it the perfect project for the small Cat engine.

Henry

[mobile_bob]

Thanks Henry:

one hopes when he gets heavily into research/developing and then testing something
that he thinks might be useful, will indeed be useful to others?  who knows?

with ongoing discussions relating to inverter/gensets and offgrid use of battery inverter systems
it just seems very reasonable to me one would really want to look into how to do it as efficiently as
possible.

the difference is efficiency is not a small number that would make a significant upfront expense hard
to justify, the difference equates to in some cases enough fuel savings in a year or two to pay for the
total generation system, engine and alternators

there are a hand full of folks that have been following the testing progress, and i just figured that a paper
needed to be written on the subject. there is a lot of misleading info out there, and to my knowlege no one
that has explored this method.

hopefully we pickup a few more members because of the paper?

that would be a good trade in my opinion

bob g

[SteveU.]

Hi BobG.
I read your paper too.
I liked your keep it simple concept. You were right; I did have to set aside my automotive preconceptions to appreciate this.
Thanks for the early claw pole history. I do recall seeing late 50's, early 60's big heavy claw pole Leece-Neville emergency service vehicle 3 phase claw pole alternators that were pre-solid state diode using a remotely mounted "fish plate" style battery charger rectifier plates. I cannot remember the proper rectifier active material name. They did have to use a generator type cut in and out relay do to the rectifier plate reverse current power bleed.

To achieve 42+ volts output do you think the rotor coil winding will need resizing? Did you intend to up the buss rotor supply voltage to 24 volts to get these higher output voltages?

I have been told the reason for the 42 volt standard was to keep the charging voltage below 50 volts to keep Union shop equipment users in many areas from running afoul with  Unions rules requiring over 50 volt systems to have only Certified Electrician service. Urban myth, you think?

Regards
SteveU.

[BruceM]

Wonderful paper, Bob G!  I had no idea that the claw pole design had seen use in serious power production, and that the efficiency could be so high.  It makes rotor cost low, and also is a design which is well suited to high rotor rpm.

When I eventually get around to wind power,  I think I'll take a close look that the "Mallard" made by a local area craftsman. http://mikeswindmillshop.com  They are using an truck alternator frame, rewound and rebuilt.  I friend has several of these, in service for the last 6 years. He's very impressed with the performance and reliability.  Perhaps with custom windings, I could get some output suitable for my 120V battery bank.

[mobile_bob]

Thanks Bruce and Steve:

Steve:

the excitation voltage never goes over ~10volt dc at 3.5amps at 100amps output and 28.8volts
there is no need to overdrive the field which would only serve to build heat and reduce overall efficiency.

because of the isolated and seperate excitation system used we maintain the 12volt rotor as is without modification
this was a huge benefit to the overall concept.

i have spent well  over 25 years rewinding stators, overdriving rotors, swapping parts and everything you can imagine
trying to find the holy grail with mixed results, usually i could get the voltage, but always ended up making more heat
and took a serious hit in both output amps and overall efficiency.

it wasn't until i started to really dig into all the theory and math that is used to design these machines that it occurred to
me to take a look at the automotive alternator (in this case the 110-555jho) as simply an alternator without concerns about
it being useful in an automotive application that things started to take on a different light.

when you seperate the excitation from the machine, and allow the unit to operate freely at your target voltage you end up
with a unit that will do what you need it to do very well, but
the end result after all is said and done is the machine is now worthless for automotive use, such a machine would not even begin
to charge until you got up to mid throttle and above, making it not at all useful for automotive use.

i figured that we need an alternator to make 28.8volts and do it well, with common parts at a specific speed, because our stationary
engines operate at a fixed speed anyway.

btw, i think the rectifier you are referring to from the 50's were selenium rectifiers they looked like a stack of metal plates
with a center bolts running down the center sort of things
and stunk like hell if you short or overheated them,, i find out now 40 years later the smoke off of a selenium rectier is some deadly
stuff,, who knew?

lastly because of the isolated excitation the project x alternator will make ~130amps at 57.6volts (48volt charging) and
still remain under 12volts for the field.

that is probably the biggest thing to get past because typical automotive alterantors rotor coils are matched to the output voltage
ie 12 volt coil for a 12 volt system, 24 for a 24 and 32 for a 32,, and who knows what they plan for the 42volt systems?

we can do anything from a 12volt excitation within reason, certainly 12, 24, 48, 60 and up to 480v is possible
much like the big boys that use 240vdc excitation on multikvolt output generators. 

i like 12 volts for excitation because we have a lot of rotors to choose from, large numbers of regulators to work with,
and often times a 12volt starting system in place on the machine to take our excitation buss power from.

bob g

[mobile_bob]

something came to me after reading some text on alternator testing, increasing alternator efficiency,
and other concerns relating to high output alternators being driven together from a common prime mover.

when i am testing the st head for BSFC (gr/kw/hr) i have the excitation current shut off on the
24volt alternator, so there are no core losses from the alternator being attributed over onto the ST head

however, i forgot to disconnect the excitation to the ST head while doing the same BSFC on the 24volt
alternator so the core losses of the ST are being attributed to the 24v alternator.

this is flawed and skews the numbers in favor of the ST head, by perhaps as much as a 2 percentage point
spread.

if this is the case, which i suspect it is, then the reapplication of this alternator for 24 volt nominal operation is
more efficient than the ST head by 1 to as much a 2%

more testing will confirm this

what is interesting to me is this
if the numbers support the 2% skewing is the case, then the 24volt alternator is more efficient than the st head
by 2 percent, and
the 48 volt "project x" alternator should be between 2-3% better than the 24volt version
for a total of as much as 5% more efficient than the ST head.

this gets me very close to parity for produceing power that could be fed into an inverter and have very close to
the same cost in BSFC in gr/kw/hr as what is produced by the st head

whats important here is the resulting power would be dead on 60hz, no flicker with superior voltage regulation, very
low harmonic distortion , basically very clean power.

and that is the holy grail as far as i am concerned

it would appear that i am very close to parity, close enough to push onward to try and attain this goal.
which btw going into this i would have never expected to get within 10 percentage points of the ST head.

we are inside the moat on this one and storming the gate.

bob g

[Jedon]

Very very interesting! Thanks for all your work and the write up Bob.

[NoSpark]

Bob, this is mostly over my head, but why couldn't the exciter alternator be of the permanent magnet type?

[mobile_bob]

the exciter can be any dc power source that is a stable 12volts dc nominal, beit battery that is charged by
a wallwart charger, a pm alternator that has some form of controller to stabilize the output voltage to no more
than ~15 volts, or another standard wound field alternator such as even a very small lawnmower alternator such
as one that is part of the engine charging system.

the only reason i used the second 110-555jho was because it was already mounted and will be used for heavy charging
a 12volt battery bank that other things will be powered from that need 12vdc power.

bob g

[Geno]

Hi Bob, I've been following your experiments closely and very much look forward to some info on the 48v modification. I now have a 110-555jho, a 48v battery bank and a 48v, 5.3kva UPS to plug them into. I'm hoping to be able to use a remote rectifier as the batteries will be 75' away. I also want to use an Outback Flexmax 60 because of its versatility.

My second application of the alternator/controller will be to charge batteries at camp. Probably no more than 4 t-105's at a time @ 12v. For this I have a 4hp air cooled Yanmar clone and an old generator frame to mount it in.

Another thought I had (your idea perhaps??) was to rectify my current AC and feed the Outback with that. I believe that will give me 170VDC (120x1.41) The only problem I see is that the DC voltage will be higher than the Outback will accept. Is there an easy, efficient way to reduce the DC to 140v or less?

I very much appreciate the research you've done and your willingness to share it.

Thanks, Geno

[mobile_bob]

Geno:

let me clear the air a bit,
the idea of using the outback mx60 was not mine, the first talk of using such a scheme was first postulated
by Bill Rogers a few years ago, and the first real application of its use with an engine driven generator was the aircore
alternator that is mentioned in another thread here on the forum.

as for reducing voltage to the point that the mx60 can handle it, perhaps a special built regulator could be made
to control the ST voltage output to something on the order of 90volts or so, allowing a bit of overshoot and still stay
under the 140volt dc limit of the mx60?

what i don't know is how stable the st head is at lower voltage, but with a good regulator that should not be a problem.
however i think i might suggest some sort of voltage clamp to catch any load dump situation that might cause the voltage
to overshoot or spike if a load was dropped quickly, so maybe some sort of MOV?  there are electronic guru's like BruceM that
are better at addressing that concern. Actually i think he is working on a 120vdc system, so he may well have a very good
regulator design already that can be used to do what you have in mind.

another option is to design and build a buck converter that can handle much higher voltages, perhaps as high as 240 x 1.41
or 338 or lets call it 350volts or more. then you can use the st head at its full capacity? just a thought.

as for a remote rectifier you must use some fast recovery high power rectifiers, such as the press mount automotive types
the 600 plus hz coming off the alternator will lock up a standard set of rectifiers into full time conduction, making a very effective
electric brake for your motor. probably hard on batteries too, and likely really hard on the mx60

fun stuff huh?

bob g

[BruceM]

Standard recovery power diodes should be fine at 600Hz, I think. Fast recovery is more of an issue at 30KHz and above (switching power supplies).

For my 120VDC battery bank I'm using 240VAC (regulated by my AVR) as input to two toroidal transformers (surplus, otherwise one to my specs would suffice), with their output going to some zero cross switched capacitors , then a 1100 watt toroid core used as the primary filter choke.  The caps switched in before the filter choke allow the battery bank charge controller to adjust charge current.  I can also switch windings on the transformer for an even wider range of applied voltage and current.  I've attached a schematic for the capacitor switcher.  This is a novel and highly efficient linear charger, with digitally controlled output regulation.  It avoids the huge EMI emissions of a switching power supply, which is why it was important for me. It is my original design, and I've never seen such a thing before, but in electronics there's not much that really new.


It would be feasible to regulate the ST output to match my battery bank, but since I'd have to use a transformer anyway I'd rather not.  I can then use the excess AC power capacity while battery charging.

It would not be hard to modify either of my AVR designs to provide regulation at some other voltage, or a pair of voltages.

Bruce M

[mobile_bob]

well from experience a common 3 phase bridge recifier is made to operate at 50-60 hz, many start to go into
avalanche at 100 hz, and most will certainly do so at 600hz

from the manufactures engineering staff i spoke with, an i can't remember which one just now, they told
me most common high power bridge rectifiers are made to operate at 60hz in this part of the world and will
not work at 600hz, however they did refer me to a manufacture that does make fast recovery rectifier bridges
that can tolerate 600 hz at 150amps

perhaps it is the combination of high current and 600hz, that causes a problem,, i don't know.

but they certainly make  for converting your alternator into a very effective electric brake!

the black smoke sure rolls off the changfa when the rectifier faults and goes into avalanche
interestingly the bridges i have don't seem to be harmed by it, but they don't work for what i need either.



as for regulating an st head down to 48, 24 or whatever, you are still limited to the stator winding current capacity
unless you regroup the stator pole windings and parallel them then you can get a pretty decent output for a 48volt system
anyway.  otherwise ~40amps at 48 volts is ok, but not very much for a 24 volt system from an st5 head, it would seem
to me anyway.

bob g

[BruceM]

Sorry Bob,
I didn't realize you were talking 150 amps.  I know nothing about diodes at that much current, except to avoid it.