project X alternator progress report

[mobile_bob]

it looks like they end up being about 6X the cost of a good lead acid battery, and about 3X the cost of a good agm

narrow that down to maybe 1.5X the cost of a good agm, and have a dealer network that i can walk into for warrantee
issues, and then maybe i would get interested.

ok, actually i would be very interested at that price point if there was a local reputable distributor.

i suppose if that every happens i will have to do a bunch more studying on the technology.

bob g

[BruceM]

Bob, since you're rolling your own charging system, you'd just want to plan on how you could accomodate these LiFePo4 batteries.  The scheme is very similar to the AGM/EV approach of individual battery shunt regulators, which can also be used on wet lead batteries.  A design that can accommodate a serial string shunt regulator system would work for all three options. The only real difference is that no temperature compensation is needed and that there are a whole lot more regulators.

For a minimalist solution you need need at least 2 digital inputs to control the charging process.  One input is the OR of all 16 battery shunt regulator "on" switches, the second input is the AND of all 16.  When the OR is high, then charge rate must be reduced to the low rate ( 1-3 amps  depending on the battery shunt regulator current capability).  When the AND is high, all batteries are full, and charge is complete.

Since you have our own custom controller, with 16 (48V) inputs a processor could track the variation in time between first battery full and the subsequent batteries, so that a bad battery could be identified early.  I think that would be a nice feature, though just watching the shunt regulator LED's and a stopwatch during a charge completion occasionally would work.

For my setup I am now writing the software to transmit battery bank info to a remote serial LCD display (4x20 $26 Digikey) in my home office. It will display the last "full charge cycle" time variance for the slowest three batteries, along with SOC, status, charge and discharge amps.  I had planned on a more elaborate data logging scheme but have decided to defer that software effort. (KISS)

[mobile_bob]

didn't get to the testing phase of the water cooled stator today, but at least i got a stator built with the water jacket
and it is ready to be assembled into the project x alternator, never enough time it seems.

i also found what i think will turn out to be the perfect water circulating pump for the stator, it is a dc motor driven pump
that is made to work at 6-12volt dc and move 4-7liters per minute, it draws 2-4amps

i plan on driving the pump paralleled to the field so that the regulator can control both field current and also pump capacity.
the thinking is as the field current goes up, so does the stator heat, and the pump will move more water. when the charging
rate reduces, the heat level goes down in the stator and the pump will move less water.

my regulator is sufficient to handle both the field current and the peak current of the pump motor so i think this will work out
quite nicely.

hopefully the pump will arrive later this week, so i can get the thing setup for testing by next weekend.

after doing a bit more research and number crunching, the overall efficiency of the alternator should be over 90% (electric + heat)
and my earlier calcs of 5 gallons of hot water per hour looks to be a conservative estimate. depending on loading/battery bank size
and the need for high rates of charge, the unit might be capable of producing approx 8-10 gallons per hour of 140 degree water
from a 60 degree startup source.

i have some really nice pharma cold shipping units, that look to hold about 2gallons or so. i plan on using a plastic bag liner to hold a measured
amount of water and the pump, then i can stick a thermometer through the lid and monitor heat gain vs loading/time and get some accurate
numbers of the heat harvested from the  stator.

after doing more research last night, it would appear that there is a very significant amount of heat given off an alternator, in one example
the unit with a water cooled stator (150amp 12volt) produced enough hot water to heat the cabin space of an automobile that was powered by
a diesel engine where the engine at idle could not produce sufficient heat to cover the heating load of the car.

the source did not break down the btu's recovered, but the emphasis was focused on the diesel being too cold in cold climates at idle to
produce enough heat to even defrost the windows, so
i am thinking that there may well be 2500watts of heat given of the project x alternator at 7.5kwatt electrical output
2.5kwatts x 3415 = 8537btu's  if i can get half of that recovered that is still over 4200btu's

4200btu / (140-60) = 52.5lbs of water, or
the amount of heat to raise 52.5lbs of water from 60F to 140F = 4200btu's

a gallon of water weight about 8.35lbs so
52.5 / 8.35 = 6.28 gallons

thats not a lot of hot water, but it is enough to take a quick shower, or wash some dishes, or make a big pot of coffee 

i am really looking forward to getting some testing done, and have some numbers to crunch to see how close my forecast calculations
are.

bob g

[mobile_bob]

update on the water cooled stator

i got it reassembled into another core alternator, so that i could keep the unit that was used to do testing on the long distance
transmission and transformer rectifier intact in case  there is a need to do more testing.

got it assembled, connected the stator in wye configuration, setup a 3 gallon plastic bucket with about a gallon of water in
it, and mounted the little pump to draw out of the bucket, feed the stator cooler and return to the bucket.

this unit is setup for 48volt nominal, and for now i have it controlled with a field reostat (bad idea by the way) and connected
the pump in parallel with the field so that the pump flow varies with field current and load, more load = more pump flow.

i ran it for about an hour at 60amp load 57.6 volts, which is a bit less than half capacity as a trial run, and was very impressed
with the cooling capacity of the cooler, the alternator remained cool to the touch so apparently significant amounts of heat are
drawn off very easily with the system.  that is about  3400-3500 watts output on the alternator.

at shutdown the field and pump are de-energized and it is amazing how quickly the stator will heat soak the cooler and you can
feel the heat rise quite rapidly, turn the pump back on and it is gone "pronto"

the next step is to remove the unit, and install a bridge rectifier set so that i can run more tests at full output

still shooting for ~135-150amps, not sure if i can get there or not, but i am sure the stator will remain cool at whatever the upper
power limit turns out to be.

my trigenerator is beginning to look like franken-gen

all i need to do is add some red food coloring to the water and it would look like blood coursing through the 1/4 inch tubing going
to and from the stator.



more to come

bob g

[mike90045]

I think the TurboEncabulator with a Reciprocating Dingle Arm would be a good work-around
see http://home.comcast.net/~steveham21/turbo.mpg for the explanation.

[mobile_bob]

ok, that is just not funny!



its hysterical!

actually here is what i have done, and a bit of history for those new to the table

i have been working on an alternator conversion to produce charging current for a 48volt nominal battery
this requires 57.6 volts dc for an agm or flooded lead acid battery at 77 degrees F

i am using a leece neville 175amp j180 mount alternator because i have access to several cores and the parts
interchangability is huge with this alternator.

in earlier testing i attained approx 4.8kwatts from the machine, and i knew i could get more with the right stator
and more testing, my goal is 130-150amps at 57.6 volts which is well in excess of what my engine can deliver.

as can be imagined the alternator runs quite hot, and it occurred to me that i might be able to water cool the stator
and thought i would give it a whirl.  the stator OD is exposed and is 2" thick, so i did the following

i wrapped 8 turns of .25" copper tubing around the stator OD tightly, fluxed and soldered the turns together
in the process of doing so the solder gets between and fills the gaps under the coils to the stator OD, and as the cooler
coils cool they draw down very tightly to the stator core.

it was then a simple matter of finding a suitable pump, and i chose one that delivers between 4-7 liters per minute and runs
between 6 and 12 volts, i power it in parallel with the field current of the alternator so that as the output of the alternator goes
up, so does its heat and also the flowrate of the pump.

the results are nothing less than amazing, where formerly the stator would attain temperatures in excess of 240 degree F, it
and nothing else on the alternator will go over 75-80 degree's, unless you unplug the pump at which time the heat rises quite quickly.

i haven't done anymore testing yet, to see if there is a change in efficiency or if there is an increase in output capability, but
it stands to reason the alternator should at least last longer running cooler.

bob g

[mobile_bob]

ok, the following is of no scientific or anything useful, but i did it anyway

i added a bit of red food cooling to the coolant water, lmao!!

too cool seeing blood course through my alternators veins!

kind of satisfies that kid in all of us, or at least me anyway.

"its ALIVE"!!!

(now all i need is a pwm drive for the pump and an oscope to track the pulse of the pump,, might look like a heartbeat if
i set it up right)

bob g

lmao, still having fun here!