Human Powered Recumbent Generator ...
"Bent Genny" project...
I
think it may have been about 10 months ago that I shelved the HD-Generator
project in disgust (Feb/Mar 2006)...
The PM
Windmill is still flying after a year, so I thought that I ought to at least
dig a shallow trench and run a length of AWG #8 the 40 odd feet to the Foundry.
Given that the PM
Genny typically supplies only 10 to 20 watts peak I can't be bothered to carry a
battery to the tower as I may be consuming more energy moving the battery then
what the generator can provide.
The trench isn't
really for the existing mill, it was more so to provide a measure of inspiration
to revisit the idea of building a usable Wind Mill from scratch.
To the left is the
aftermath of 125 Hard Discs that were husked for the Magnets, as well as the
aluminum for the foundry.
This is an ongoing
chore that is best done over a few days as a dis-assembly line process.
By allowing a
sufficient passage of time and reviewing my last attempt against a number of
proven designs and "build-diaries" on-line @ Otherpower.com, I had a
clear idea of where I'd gone astray...
The tipping point
to rekindling my interest was the log of Mick S's HD generator that was able to
produce 500 Watts. (to locate this search the OtherPower.com discussion brd with
"MickS hd gen" and it should be at the top of the list).
Bare Magnets separated
from mounting plates.
One of the primary
mistakes I made previously was to disregard almost everybody's urging cut the
magnets in half.
The magnets are
"Face Polarized" so this creates a single N on one flat with a
corresponding S pole on the opposite side.
I fashioned this
simple but sturdy chop-saw of sorts for cutting the mags with a thin Walter
Zip-Cut.
The key is to cut
the magnets as fast as possible to keep the heat build-up to a minimum as
excessive heat (>200C) lowers the strength of the magnet markedly.
This is one boring
fucking job, and unfortunately its not one that you can do watching TV due to
the noise and attention that it demands.
Regardless I was
able to fill a couple of plates as shown to the right with single face poled
magnets.
Along with moving
toward a "Split HD Mag" design I finally buckled down over the summer
and came to grips with the realities of 3-Phase generator principles.
The main source
for understanding the basics came from windstuffnow.com (Browse around for an
excellent 3-Phase tutorial page), as well as numerous perspectives from
Otherpower.com and Hugh Piggott's scoraigwind.com site. As an aside Hugh is the
"grand-daddy" of on-line wind power info with OtherPower a close
second for years on the web (for DIY'ers anyway).
The above being
said, I've read so much info over the last couple of years I could never do
justice to it and have mentioned them for your reference, should you become
inspired before the shit really hits the fan.
With a reasonable
qty of magnets now on hand, I scanned a single faced pole (into Corel Draw),
duplicated it and flipped it such that they formed a block rather than an arc.
The scanned magnet is used as a reference to determine how many poles can
be squeezed into a given sized rotor.
After a bit of
trial and error I settled on 16 poles arranged on a 10" diameter disk. By
using Corel Draw I was able rotate the mag placement at 22.5 degrees (360
degrees divided by 16 poles).
The coil count for
3-phase ends up being 12 coils spaced at 30 degree intervals as shown below
(each phase is designated by colour).
With a much
clearer idea of where this project is headed than the last attempt, I moved from
the theoretical onto the practical.
Another issue that
I had concerns about from the previous project was the thin rotor backing, as it
was simply a 7 1/4" circular saw blade, approx 1/8th" thick.
To the left is my
first attempt at hacking out a reasonable disc of 1/4" plate.
I drilled a
3/16th" hole in a piece of band steel for a pivot, and drilled another hole
for the plasma nozzle to sit into at approx 5 1/4".
This guide yields
a very rough 10 1/2" disc.
Given the course
nature of the disc as shown above, I made a more accurate jig for cutting or
truing discs out of some scrap pine and a bicycle hub.
The unit also has
an accompanying nozzle holder that is firmly clamped into place to create an
infinite number of disc sizes up to about 12" Max.
This is far
superior to the previous cutting jig as the nozzle can be positioned more
accurately for far cleaner cuts. The disc is fed by hand (with heavy gloves on).
With a 1/4"
disc now in hand, the Corel Draw template is stripped down to create a toolpath
for the mill to machine out a Mag-Spacer for the rotor.
The CNC Mill is
running an aggressive 0.075" up-cut router bit that was sent to
theworkshop.ca by Bernard B. from Ireland (to whom I am most grateful).
These bits seem to
be able to run at a 40 IPM feedrate through a 50 to 60 thou cut with a spindle
speed of 20,000 RPM.
The finish is just
astounding with no loading up of the cutting flutes even down to 1/2" depth
of cut. The Dremel bits that I had been using up until this point would have
lost their edge and/or broken given the parameters as listed above.
The cut magnets
are loaded onto the rotor plate such that the poles are alternating between N
and S orientation.
(the images are
marginally out of sequence as this disc has yet to be trued on the newer cutting
jig shown above).
The magnets are
stacked between 5 to 6 thick to achieve a total of 1/2" height.
I say
"between 5 to 6" as they are not uniform in thickness.
Unfortunately it
took 4 attempts at shuffling the mags about before they were consistent in
height.
The rotor is
finally trued, and assembled with magnets ready for being cast in resin.
The mold is made
from a section of 3/4" MDF and cut out on a bandsaw.
This is secured to
a 2" thick MDF backing (obscured by the newspaper).
All surfaces of
the mold are waxed and painted with Vaseline as a releasing agent. The newspaper
is inserted between the mold and backing to aid in separation.
The rotor and
center mold are held in place by a 3/8th" carriage bolt that is counter
sunk into the backing brd.
The mold assembly
is tested for level on both the X & Y plane to ensure that it fills flat.
The volume of the
mold is calculated in total, subtract the volume of the rotor disk, magnets and
the center mold to determine how much resin to mix.
The blue is
acrylic enamel paint added to the polyester resin and thoroughly mixed prior to
adding the catalyst. The paint does inhibit the curing process (overnight) but
the end result is worth the wait.
The mold released
like a charm the next day yielding a very slick 16 pole rotor.
I knew after I'd
rough-cut the disc initially that there would be some head scratching once the
rotor was completed.
The issue was how
to spin the disc for coil testing and ultimately stator testing.
Previously I'd
used the drill press but that was with a much smaller and lighter rotor.
This unit will fit
within the throw of the drill press (between the chuck and support post @ the
back), but there is no way that a completed stator and support jig for setting
the air-gap would.
This dilemma wore
on me for a couple of days while I considered the options available to me. Along
with the objective of building a generator from scrap Hard Disks, I wanted to
incorporate parts from scraped bicycles as well.
The bicycle is a
marvel of high-quality mechanical parts that is essentially free to any one
willing to check the land-fill or steal one. Recently I've come to an
arrangement with a local scrap yard to separate any bikes from the recycling
stream for me to pick-up for a flat rate price per bike regardless of shape or
style. Hopefully this should yield at least 100 bikes over the next year.
(As an aside to my
local associates that frequent theworkshop.ca, I'm most keen for full suspension
type bikes, frame size is irrelevant as they will be chopped into parts and
re-purposed... As always I'm most appreciative of the junk of others)
Human
Powered Recumbent Generator - "Bent Genny"
The previous
HD-Gen if successful was to be spun on a cast alloy frame that had a bicycle
wheel bearing assembly integrated into the casting. Seeing the weight of the new
rotor, it is obvious that it would be too weak.
The next beefier
bearing is the crank assembly, and that was where I wanted to start.
Above a crank is unceremoniously
liberated from the oppressive frame that has been keeping it under it's heel for
too long. The fuckin' hob-nail boot of the man!!!
The recent success
of the LEV-1 project over the past summer, and the
easing of the Highway Traffic Act within the Province of Ontario to finally
recognize Electric Bicycles on our roadways, has had me doing some back-ground
work on Recumbent trikes for a future project.
In the course of
that line of thought it occurred to me that I could gain some basic experience
in mechanical fabrication by making my "Test Jig" for the wind
generator into a Human Powered Recumbent Generator.
I know, I thought
it was a stupid idea at first too... but once I started to look at the pros
& cons, it didn't seem like such a bad idea. Like most folks sucking at the teat
of the OECD nations I could stand a good work-out on a regular basis through the
winter months as I don't cycle in the cold.
To the left is a
second crank, similarly freed of it's pedestrian yoke.
The bearings on
both were disassembled, cleaned and re-packed with fresh grease.
The two cranks
were tacked together to ensure that the booms would be aligned as they got cut
and fitted with mounting plates.
The crank
assemblies are later cut apart for greater flexibility of chain alignment and
tensioning.
So far so good...
As this is in a
very preliminary stage of evolution, I opted to go with just a minimalist wood
frame, for fear that I lose a couple days cutting and welding a metal one that
may have to be scraped a few times before I'd worked everything out.
I scavenged a
couple of the seat cushions from the boat and screwed together the arrangement
pictured to the right.
This had to be
done prior to securing the cranks to the base so that the spacing would accommodate
an ergonomic fit for pedaling.
This was the easy
part...
Aligning the gears
and chain proved to be more challenging.
Ultimately it took
over a week of farting around with parts from 4 (four) different bikes
before I arrived at the configuration shown below.
Typical pedal RPMs
are in the range of 60 to 80 with some short bursts to 100, this is considerably
lower than what I'd wanted to aim for as a test jig for the wind generator. The
cut-in speed for the wind mill would have to be closer to 120 to 200 RPM.
The
"Drive" gear is from a mid to late 80's 10 speed (Not 18 or 21 speed,
Ten - count'em Speed) because it was the beefiest metal of the bikes I currently
have on hand. I'd bent both of the original drive gears which were inferior in a
order of magnitude that smacks of "Made in China".
The
"Driven" gear also under went a series of revisions that included
welding a 7-gear cassette directly to a shaved and shaped crank stub. To the
right is a far more elegant solution.
A small disc of
3/16th" plate is cut to size, and a single-speed coaster gear is carefully
welded such that normal forward pedaling drives the rotor, and it can free-wheel
when the pedals are stopped.
The rotor is
fitted and torqued down in place.
The tape is a
marker to aid in determining the drive to driven ratio
1 : 2.66, for
future calculations I think that I'll use 2.5 to stay on the conservative side.
The RPM of the
driven shaft was measured a number of times between myself and one of the twins,
although Brittany clocked in closer to 300 RPM, I aimed for a more sustainable
220 to 230 RPM.
As these are
"No-Load" values, and the numbers work-backward to a cadence of 88 to
90 on the drive side, the next hurdles will be to determine whether this will be
12, 24 or even a 36V unit, cut-in RPM, and total power.
Given the
limitations of this as a "Human Powered" generator, or at least
test-jig, I think 100 to 150 Watts would be a target to aim for given the low
RPM situation.
The next update of
this project will hopefully be posted early in the new year...