Energy Ltd. News #55
Copyright 2012 Craig Carmichael - September 3rd, 2012
= www.ElectricWeel.com = www.MushroomOutboard.com
Hilight: Magnetic Refrigeration & Home Heating: vast energy saving potential?
Month In Brief (Summaries)
- 1912 Detroit Electric Car - 12V Fridge - Magnetic Refrigeration
car tests - 3D printer
In Passing (Miscellaneous topics
and editorial comments)
- Wind versus wave power on the sea: immense work to obtain the lesser
Our dysfunctional, failing society:
- Einstein Refrigerator 1930. Even
Albert Einstein fell into the "technology death by patent" trap - Self
centered versus love centered training? - Democraship, the kleptocracy,
and state terrorism - an effective cancer cure and why it'll go nowhere
- Widespread crop
failures magnify financial crisis problems - Victims of Bank Fraud.
Electric Hubcap System
* Motor Assembly
* Brake Vacuum Assist - pump & reservoir
* Safety: "Stealth" electric cars should make warning noise when in
proximity to people?
* Sprint car: motor, controller and battery installed. It moved a
little: the motor controller seems to be the problem rather than the
torque converter... makes the motor "wimpy".
* The Planetary Gear Torque Converter Looks Like a Keeper!
* A new type of control modulation: combined CRM + PWM.
Planetary Gear Torque Converter
Project: Converter seems to work acceptably in car!
* 3D Printed plastic planetary gear/torque converter parts?
Solar Electricity Project
* Electricity is already 40¢/KWH in Hawaii, making solar a 3-5
* Super battery stick proves less
than super - also beware voltage drops to batteries!
Thermoelectric Fridge &
* Single Peltier cooling works best after all. (Also can work well
without cold-side fan.)
* Better fridge shapes can provide more internal volume with less
exterior wall surface to keep cool - octagon, Low walls and terraced
* Dividing fridge with a wall(s) allows sufficient cooling of a smaller
space in summer heat, ice compartment.
* Magnetic Refrigeration & Home Heating - world's best
research into it: Victoria BC - a design implementation -
rare earths ordered - potential for thermomagnetic motion machines
* Regular Chest Freezer used as low energy Refrigerator
Turquoise Battery Project
* Project devolves to assembling 3D printer to make electrode pockets
* Battery Making Video planned.
No Project Reports on: Magnetic Motion Devices, Weel
solar cells, LED Lighting, Pulsejet steel
Construction Manuals and information:
Electric Hubcap Motor - Turquoise Motor
Controller - 36 Volt Electric
- Nanocrystalline glass to enhance Solar
Cell performance - Ersatz 'powder coating' home process for
- Electric Hubcap Motor Kit
- Sodium Sulfate - Lead-Acid battery longevity/renewal
- NiMH Handy Battery Sticks, Dry Cells
- LED Light Fixtures
...all at: http://www.TurquoiseEnergy.com/
(orders: e-mail email@example.com)
August in Brief
was asked to run the August first meeting of
the VEVA - Islands electric car club, as Cam Rawlinson was on the
mainland picking up the 1912 Detroit Electric car for a Victoria tour
to celebrate it's 100th year, making it now a genuine antique. This was
appropriate as the car was bought by a Victorian and it spent most of
life here. Running the meeting consisted of telling people that the
Detroit had missed the 5PM ferry but would arrive about 8:15. To my
surprise the midsummer meeting was the best attended so far - owing to
Its drive arrangement with the motor underneath connected
to the rear differential gears, is essentially similar to today's iMiEV
(available off the lot at Mitsubishi Motors on Cloverdale St). Of
course, performance and style are almost incomparable. The NiFe
batteries regrettably at long last wore out, and have been replaced by
short lived but cheap PbPb's, under the hood and in the trunk.
This Detroit Electric is in admirable condition, and it
won the Oak Bay Mayor's pick at the Classic Car Show on Oak Bay Avenue.
On top of several
more sundry things on my list for the month, I really wanted to get the
motor together and into the Sprint car, get the Peltier cooler fridge
finished, make the magnet machine that's been in
my head for 2 or 3 months now,
and put the 3D printer kit together.
Later on I started thinking
about grinding up some of the
borosilicate glaze I made a while back and melting it into 'pebble
some solar collector cover glass (using the broken solar collector),
about making a video about making batteries, which someone requested I
do a month or two ago, and about ...yet another... new project:
magnetic heat pumping.
Things seemed to
move at a crawl in early August, with one interruption to the work
after another. I didn't get much
done around the house either, and
the days were into double digits before I got going on much of anything
after the July TE News. I guess
that's summer! Even ordering some electronic parts - mostly for LED
lighting - was put off until it didn't get done at all. Then there were
exciting things on the 30th and 31st that kept me from getting this
newsletter out quickly as September started.
On the 9th someone phoned from Pennsylvania asking about
thermoelectric fridges. Since all I had was the July newsletter
writeup, this seemed to
indicate a hungry market for them, or perhaps for a ready-made
system for doing them DIY.
cooling system went back to a single Peltier element, which works
after all than double. (Oops!) A foam divider wall inside the fridge
allowed it to provide
temperatures in a smaller space in the heat of summer. With no fan
inside, the inner heatsink would frost up and the temperatures dropped
almost 20º to 4-6... at least near the bottom of the 1/2 fridge.
But I didn't
finish up the fridge itself (yet). Instead, an exciting new type of
refrigeration crossed my radar screen and occupied my time in study:
magnetic heat pumping.
Following TE News #54,
someone sent me a Wikipedia link to magnetic
refrigeration. This has been experimental for quite some time, but
seemed to me the
way it was being done, by pumping a fluid or a gas to move heat around,
was (as is so often the case) needlessly complicating the entire idea.
Later I found out that the fluid friction also severely limits the
theoretically very high coefficient of performance.
I soon thought up what I believe should be a better "solid
state" design. As I see it, scientists have created the "vacuum tubes"
of magnetic heat pumping and I have (in concept) the "transistors".
(That analogy is appropriate, but probably overstating the case
As conceived by month's end, the first main component is
supermagnets aimed down on a horizontal rotor turned by a motor, to
provide the on-off
magnetic field that heats and then cools gadolinium as it enters and
then leaves a magnetic field, by the magneto-caloric
effect (MCE). The other component is stacks of thin, flat solid
plates: stationary copper heat absorber and releaser sheets in a
plastic frame, alternating with loose gadolinium "wafers". Each wafer
may get warmer and cooler by 1/2, 1 or more degrees, and the number of
stacks and wafers required for the desired temperature spread will
depend on the actual
values attained. The magnetic force as a rotor magnet passes by, pulls
each gadolinium wafer up into contact with its upper copper heatsink
within the stack, and also heats it, thus heating the heatsink. When
the magnet has passed, the wafers fall down and contact the lower
copper heatsink, and they cool. It seems to me this should be very
effective at pumping
heat from the bottom of the stack to the top, and [almost] everything
looks simple and straightforward to make.
On the 21st I ordered some
erbium and some
gadolinium - the magic element with a large MCE at 293ºK (room
magnetic Curie temperature. An alloy
of Gd and Erbium (85:15%) has a Curie temperature of 278ºK -
temperature. [Gadolinium is
"malleable and ductile", so I hoped making ingots into flat wafers
be too hard. Erbium is "malleable".]
Twice the Gd and no Er arrived on the 31st. Oh well, good
enough to start with, and I got the better bargain in price, Gd for the
price of cheaper Er. It proved magnetic enough to be picked up (about
equally at room or fridge temperatures), but not clamping strongly to
the supermagnet like iron or steel. Perfect! "Rolling resistance" of
the magnet rotor should be low, so a low power motor should work fine
(hmm... where's some old audio cassette deck?)... Just how much cooling
can be effected, from how little power?
Temperature of a chunk went up or down about .3 degrees
entering or leaving a supermagnet's field. Obviously I'll be wanting to
get the strongest possible field and I might review prior designs again
and take note of the ways others got the most flux.
Gadolinium also seemed harder than mild steel when I went
cut a piece. Since it also oxidizes readily and has a high melting
point, it may be tough to form it into the required metallic 'wafer'
It seems to me magnetic heat pumping has far
more potential than just refrigerators. A 1500 watt electric
baseboard heater might be
replaced, not with a 1000-1200 watt peltier elements heat pump, but
instead with a 300 to 600 watt magnetic one. The
global potential for winter energy saving from this new technology
staggering! It probably has excellent potential for hot weather air
This was the only 'magnetic machine' that got looked at in
Arriving with the Gd metal was Sm2O3 powder, samarium
oxide, which I think should be about the best substance for raising
oxygen overvoltage in the batteries. It was convenient to get some
while I was placing a rare earths order anyway.
Aside from there being all the other things I was doing, I
acquired almost some mental block against working on motors and cars. I
tackled each small step on the motor without enthusiasm and then a day
or two would
slip by before I did one more. Perhaps it comes from 3-1/2 years of not
even getting a car to move. In spite of
ongoing improvements to the motors and the controllers over that time,
finally having what ought to be a working torque converter, and a
lightweight car with a 4 to 1 final reduction gear, I had little faith
that the car would actually move. What would go wrong this time - some
new variety of catastrophic failure, or simple refusal to budge when I
should be up to maybe 200 foot-pounds of available torque at the wheels?
Notwithstanding, the motor was finally done and
the Sprint car on the 27th. It looked good and seemed to fit well,
inspiring, so then I put in and wired the controller (29th), followed
by the 36V, 20 AH
NiMH battery box I made a while back (30th), and tried it out.
The car moved, but not much. The motor controller seemed
to be the main culprit rather than the torque converter. My present
analysis is that the torque converter seems to work, but it takes more
power going to it from the motor than I expected to budge the car.
At one point it did roll forward a ways, but on a slight
downhill slope. It only needed 20-30
foot-pounds at the wheels to do that. The rope tended to jam and not
let the pulley slip, even when it was loose, especially in one
direction. The driver controls in the car worked fine. The motor ran
fine, but in reverse the helical planetary gears caused the
planets assembly with the pulley to push over sideways and rub on the
housing. With the test controls by the side of the hood, I disconnected
the clutch rope and pulled it by hand. At one point, it looked like it
was about to move backwards, which needed 50-60 foot-pounds, but the
motor stalled. Finally, trying to repeat this, I put on near maximum
the motor and pulled hard on the rope. The motor slowed to a crawl and
then the controller made a bang and smelled like blown
transistors. A 'minor' design problem was fixed and the controller
repaired by Sept. 2nd.
On reviewing the videos of the tests on September 2nd, I noticed that
the motor seemed wimpy at times, and that the cause would be the CRM
'direct torque control' modulation. Each time the car started moving or
I put more tension on the slip rope, the motor started stalling, and
one had to be really fast on the 'throttle' to give it more torque. I
conceived of a new combined CRM + PWM control modulation that would
deliver the best of both types. I hope to make an external PWM
circuit and try this out very soon with further car tests - while the
Video of tests uploaded to youtube (click image for video; 'Back' to
Here the car started moving as I applied tension to the rope,
but at the same time, the motor started slowing and stalling.
The 3D printer assembly absorbed quite a few hours I would
otherwise have put into the car conversion and other things. It was
well advanced but not complete when I stopped working on it to get
the car installation ready to test.
Incidental news, editorials
Wind Versus Wave Power
I watched an episode of "Mega Builders" on Netflix where a
vast field of windplants was being installed on the North Sea off of
Holland. The plants were colossal stuctures, sticking up out of the
ocean on immense pillars. With such immense pieces, the project was
very large in scale. One piller held an electrical substation with a
very high voltage cable to the shore. Much was said about the roughness
of the sea and the hazards and delays it caused in construction.
I think it's amazing how so much effort and expense was
expended to capture wind power, on the sea, when the power of the sea
beneath was doubtless much the greater, and probably more easily
extracted. And if one is going to erect pillars sticking up from the
sea bottom anyway, they'd make better, simpler platforms for wave power
floats, mechanics and generators than floating units as proposed by me
Perhaps I wasn't thinking big enough in designing floating
wave power structures that could simply be towed into place, the anchor
dropped, and the cable plugged in on shore.
Invention death by Patent: even Einstein fell for it!
Here is just one more
instance of the most common use of patents: to kill valuable
new technologies. (Wikipedia article Einstein.) What a
crooked system the patent system is! If even
Einstein fell into the trap, whoever has a hope of evading it?
In 1926, Einstein and his former student Leó Szilárd
co-invented (and in
1930, patented) the Einstein refrigerator. This absorption refrigerator
revolutionary for having no moving parts and using only heat as an
On 11 November 1930, U.S. Patent 1,781,541
was awarded to Albert Einstein and Leó Szilárd for the
Their invention was not immediately put into commercial production, as
the most promising of their patents were quickly bought up by the
Swedish company Electrolux to protect its refrigeration
technology from competition.
The patents would have expired about 1948, and I suppose my dad was
lucky to find a propane fridge for our cabin in about 1960. By that
time, electricity was available most places and the value of the
technology to the market was very limited. I've only seen 3 or 4 of
these fridges in my life, and two of them were the ones my dad bought.
(The second one was bigger and easier to light.)
Teaching youth: focus on self centered, versus social centered
I remember the class being told in grade 1 (or was it 3 or
a couple of times, that we needed to get an education, or we'd "end up
digging ditches". We needed an education so we could get a better job
and be more affluent. There was an implicit, and common, hidden message
that everybody was in life for themself and each should look out for
"number one". (At the same time, we daily recited the Lord's prayer
asking for "Thy will be done on Earth" - but that was just a jumble of
words that didn't even make any grammatical sense, and which were never
explained to us. "Thy what will be done on Earth"? "Our Father
that art in heaven":... did he spend his time painting
pictures? Why wasn't the prayer made of proper sentences?)
Material man must "serve two masters", looking after
material needs while trying to attain meaningful intellectual and
spiritual goals beyond. But "the self" is an insidious focus to instill
in young people. "How can I make more money?" - materialism,
consumerism - is a hollow, unsatisfying philosophy to live by. Often
the best answers found to this shallow question have little to do with
serving one's fellows or improving society, and contribute little to
growth of a dynamic character. Instead the focus may become how to get
the most the easiest way, which may devolve essentially to a mentality
of robbery. At worst people may
progress to the greed and avarice
that are presently running rampant and destroying civilization.
In corporate terms it's called "externalizing": We're here
to make money. The problems our activities inflict on the society we
all live in aren't our problems; somebody else can deal with them -
we'll pay a tax or a fine to allow someone else to deal with them. (but
if you charge us
much, we'll move offshore.) A corrupt culture of scarcity, greed,
abrogation of responsibility and
gangster elitism has arisen.
The attitude that advances true civilization and the
human brotherhood is of course "How can I best serve my fellow man?";
"How can I
do the most good?". Since one can't contribute while trying to find
one's next meal, making a decent living, hopefully doing things one
enjoys doing, is implicit. So is safeguarding this co-operative culture
by eliminating opportunities for greed and corruption as well as prompt
prosecution and demoting of anyone engaged in it, _especially_ by those
who have been granted positions of power and influence over others. And
is contributing or at least prevented from working mischief, everyone
is safe from violence and want, and lives better.
Our present civilization makes no organized effort
to visualize how we want to live as the future unfolds. There are no
"Departments of Progress" committed to and administering working
towards worthy goals. Instead,
people and "for profit companies" are in it 'for themselves' and no one
is left with any power or resources to solve the 'externalized'
problems, or to contribute to the greater good. The "status
quo" powers simply try to put out fires and maintain themselves,
without even trying to remove the
kindling or prevent the fires from starting, without an overriding
plan. Those who conceive plans are locked out of the societal
structures of political and economic control.
Someone likened what's been happening for many decades,
even a century, to a frog in a pan. If he's tossed into a pan of hot
water, he jumps out. If he's tossed into cold water which is gradually
heated, there is no point where the frog realizes the danger and jumps
out, and he dies. Twentieth century civilization, tolerating every
fresh outrage without taking meaningful action, has reached the boiling
point of unbridled corruption and is unraveling before our eyes. In
the last analysis, its unworkable and unsustainable societal
configurations aren't worth saving.
Moving to a brighter future requires a new all-inclusive
morality and higher loyalties, real religious living where the
fatherhood of God and the brotherhood of Man are implicit in one's
daily living, and real courage to face and manage the serious and
perplexing problems just beginning to engulf our planet -- the sort of
courage, perhaps, to risk if necessary becoming eligible for the Julian
Award for News Reporting, or the Bradley Manning Award for Loyalty to a
Higher Cause. Every person who
stands up and embodies a righteous cause counts for more than he knows
the angels await such decisions so they can pitch in and help
co-ordinate people and circumstances in unseen "synchronistic" ways.
BTW I never saw a "ditchdigger". Even in 1960 the few
ditches in town were being dug with machinery.
Democraship, Kleptocracy: the internal enemy
In this century, we have rapidly transitioned from having
marginally useful civil governments to living in "democraships" where
have a vote (so far), but no more control or influence over our
governments than in a dictatorship. The USA's "kleptocracy" is leading
in this. Any
dissent or criticism of the US government is now branded "terrorism" by
agencies of that government, while that government itself is committing
far more acts of terrorism than any other organization in the world.
But the USA isn't alone. Other "western" nations are following right
An Israeli posted a Facebook message "We Love You Iran!"
which was lavishly responded in kind by many more people in both
nations. It touched me. It showed where the sentiments of the majority
really lie. But foreign governments are making economic war on
Iran, starving the people and driving it to desperation. Israel and the
USA especially want to attack it and pillage it, and Iranian society is
being held back.
The US "government" is now also talking about invading
Africa. The pretext is to take out some inconsequential villain,
"Kona"(?), who, according to a Ugandan-American girl in high school,
her relatives in Uganda, was most probably killed in a US bombing raid
on his headquarters 6 years ago. He hasn't been seen or heard of since.
In the interview, when the girl mentioned that oil has been found in
Uganda along the border with Congo, the real motive became clear. And
it seems that although Ugandan oil and Iranian oil aren't really needed
with other existing sources filling the demand, and with Canada selling
the USA oil from the Alberta tar sands cheaper than any third world
country sells its oil, the USA in its paranoia doesn't want China
getting oil, so they want to control it all.
US military personnel are also victims: as they start to
understand why they're really there in foreign lands, killing on a
genocidal scale, pillaging and despoiling, they're committing suicide
in record numbers. There are more suicides than battle casualties. The
hierarchy is calling it "stress", but "feeling guilt", not wishing to
participate in these crimes against humanity, might be a more common
I expect we are today probably looking at the last US
I've noted the laws are now set up for a dictatorship. The "Republican"
faction "primaries" have shown how little Romney and his gang care for
the democratic process, even making it illegal for Paul supporters to
cast their votes. It appears he'll win the election, and if he does, I
think he'll "delay" - call off - the next one on some "emergency"
It's becoming increasingly true, and increasingly
apparent, that it isn't foreign countries but our own
economic-industrial institutions and
governments that have become our worst enemies, the enemies of
peace, prosperity and liberty. They've lost most any merit they once
had, and soon they'll be falling apart along with the global financial
New, effective cancer cure
Dichloroacetic acid (DCA) has evidently been used for
decades to treat a rare disease or two. Last year a doctor in Edmonton
found a previously unnoticed effect and tested it in mice: it's
amazingly effective against cancerous tumors, with few side effects.
Acetic acid is of course better known as vinegar. (Perhaps
instructions for how to chlorinate it can be found on the web?) DCA is
cheap and has been around too long to patent. In our corrupt society,
it'll go nowhere. It might even be outlawed. It's a diabolical process
that eliminates medically effective and cost effective solutions:
* You can't sell something and claim it's good for any medical purpose
unless you can clinically prove it. Even if you sell common things
known to reduce cancer like almonds, strawberries and vitamin D, if you
claim they're effective against cancer without clinical tests, you'll
be shut down.
* The clinical tests cost cost tens to hundreds of millions of dollars.
* If you somehow went to all that trouble and expense for an
unpatentable substance, it gives you no benefit over others selling the
same substance. They can sell it much cheaper because they didn't have
to finance your testing, and would also be able to say "clinically
* Furthermore, if even a patented substance proves too effective, the
patient will be quickly cured and won't be buying it for long enough to
help pay back your costs.
* The best business is thus obtained by marginally effective patented
drugs that keep the patient alive but without affording a real cure.
* In addition, it's in "big pharma's" extreme interest to ban, or at
least stamp out information about, things that will actually prevent or
cure disease better than their offerings. Thus we find many insidious
- The US banning of Laetrille (SP?), derived from almonds in
1969. It's now
available only in Mexico, where Americans in the know go to get it for
- The effective banning of DMSO, a 'home remedy' that was
starting to become popular, used effectively by many for various
conditions. (It's not even a drug, it's just a solvent. If paint
thinner cured medical conditions, it'd disappear too.)
- Disinformation campaigns: For example, in a short PNW summer
where hot sunny days can often be counted on the fingers and the
population is chronically seriously deficient in vitamin D, making
cancer virtually epidemic here, on that first rare hot, sunny day that
the public heads for the beaches, "public service" ads appear on TV
reminding everyone how bad sunburn is and advising we all slather on
sunscreen (to avoid getting any of those healthy, vitamin D forming
rays on our skin). Few die of skin cancers - they're mostly easily
treated. Meanwhile, millions die of other cancers because they're too
vitamin D deficient to resist tumor growth.
It's small wonder then that so many are chronically ill
and medical expenses go up and up and up. Health and nutrition are
areas where the individual is well advised to do enough of their own
study to learn some fundamentals. Health care with doctors, healers,
drugs and nutrition have a vital role in society, but ignorant reliance
on an overblown health care system after a careless, unhealthy
lifestyle empowers the greedy to do the things they do to extract
millions from everyone.
Global Bank Fraud, coming food shortages
A UK financial analyst estimated that a hundred million
been driven into poverty since 2007 as a result of bank fraud. Since
poverty aren't living well and often die, he suspected that more people
have been killed by bank fraud than by all the wars since world war
two - perhaps seven million.
(My own very meager RRSP investment was worth 33% less now
than in 1998, 15% less than at the start of 2011, and 6% less than this
January. And that doesn't take inflation into account. It didn't
average 1% a year over 17 years. If I had bought a chunk of gold, it
would now be worth 4x what I put in; silver - 6x. Both have been rising
rapidly in this century and won't suddenly drop like stocks. At 57, my
plan was now 'unlocked' and tho "sell
low" is bad advice in normal times, the writing was on the wall and I
cashed it out. In hindsight it would have been better to have done this
in 2010 or 2011 -- but better now than before it
gets worse or even disappears as many pension funds and futures
It looks like worse is to come. Corruption
and greed will magnify the problems of poor or failed 2012 year crops
all over the world:
severe droughts across the USA, in Gujarat India, the Sahel and in
typhoons in China, Taiwan and the Philippines, wildfires in Russia, and
an infestation of crop eating mice in Germany (50% losses) are ones
I've heard of.
Seems to me that at one time the USA kept food reserves in
case of bad crop
seasons. There have been other bad years, but today the larder is bare.
Today there are almost no family
farms to "go back to", and any failure in the food supply chain, or
petroleum to transport food (thanks to corruption preventing and
eliminating electrification of the transportation network), can quickly
result in shortages, hunger, and perhaps famine and epidemics. With the
default (and indeed evident disdain) of governments, responsibility for
stocking enough food to ride out times of shortage devolves on the
man caused as well as
natural disasters are obviously in store, and combined with financial
collapse, whole nations, large or
small, could potentially cease to exist as organized entities overnight.
But it's all part of a cosmic plan for progress - a
necessary evil. This world must move forward from its present
evolutionary dead end. The pillars of the corruption are being cast
and in a disorganized time, unhindered by powerful reprobates, the more
enlightened and increasingly spiritual people will learn of and
implement sustainable societal
structures that can adapt and grow. (I expect that my writings, Fundamental
Government and the Department of
Progress idea will be among the historic documents and proposals
considered in the processes.)
Hubcap Motor System
By the 13th I had finally cut the keyslots in the shaft
for the sun gear and for the SD bushing to hold the magnet rotor on,
turned a large center hole in a bearing mount plate for the larger
bearing on the
fat part of the shaft, and put together a magnet sensor PCB.
The new, thicker rotor outside body had increased the
diameter to 11.6". With 10" rotors and bolt holes through the walls
instead of within the rotor compartment, those bolt holes had to be
moved out about .1", so a new hole drilling template was needed. On the
15th I made the CNC g-code using a spreadsheet, drilled the template,
and drilled the motor.
I still needed a way to lock the sun gear in place, to
keep it from sliding off the end of the shaft. I didn't want to put a
set screw through it because the only place it could go was through the
gear teeth. Instead I needed to put some fitting on the end of the
shaft. I ended up with this cludjed arrangement:
Motor shaft with SD coupling to hold magnet rotor (hiding stator end of
and sun gear held in place by a shaft key, a sleeve on the shaft, and a
flat-head bolt and bent washer on the outside.
(It may be a saving grace
that the outer washer is only stressed when the car is in reverse. How
can I reinforce it?)
As I was busy with other things, it was the 26th before I
put everything together. Since the rotor holes had been moved outwards,
the holes in the Sprint car mounting plate had to be likewise reamed
out to match,
and I made good use of the mini milling machine for this (still with
cranks, not CNC control - luckily).
I thought of what to do to cover the coils: wrap a narrow
strip of furnace type air filter around the motor and hold it in place
with some small wires. However, I didn't get around to getting the
I thought of testing the motor, and I thought of checking out its
back EMF voltage levels as this still hasn't been done on the modern
versions. Instead I decided to go for the gusto and mount
it in the Sprint car and try it in situ. Hopefully I might
still get the Sprint running before month's end, if all went well.
After some further fiddling and reaming of mounting holes, it seemed to
fit well, the sun
gear on the shaft centering on the torque converter planet gears.
This is what replaces the car's engine, transmission et al!
Brake Vacuum Assist
One can only get so concerned with "minor details" before
the thing is running, but hey... they're easier to work on than some
aspects of the original R & Development work, and they do need to
be done before hitting the streets. (And the brakes, preferably before
hitting anything anywhere. Although, the Sprint is so small and light
I don't see why it should need a vacuum assist. But it's made with it -
Someone gave me a link to cheap Chinese 12 volt vacuum
pumps, so I bought one. It doesn't have the speed of suction to do car
vacuum assist directly, so the thing to do is to hook it to a vacuum
storage (anti-storage?) tank. What would hold a vacuum without
imploding? It suddenly occurred to me that an empty small propane tank
could work, the kind used for propane torches and campstoves. A great
feature is the price.
The guy who gave me the link thought that would be
overkill on the pressure capacity, and also that it needed to be about
2 liters. Hmm, that would mean two tanks. Well, maybe a 3" ABS plumbing
drain pipe a foot long or so, or a longer 2" one? Or is one liter is
big enough for the Sprint?
I didn't get any of it hooked up, except to mount the pump
under the hood. The mass of wiring that went to the engine will
need sorting out. Perhaps the "fuel pump" wires can be "vacuum pump"
Electric cars are beautifully quiet, however there has
been some concern about their "stealth". A friend recently said he
stepped onto the street and was almost run over by one. He hadn't
bothered to look since he didn't hear any traffic. Another friend
almost hit one on his bicycle. And there's very few around so far.
(Reminds me of Ministry of Transport stations: whenever they added a
second vehicle on a small island, there was always a collision.)
I don't want to make noise unnecessarily, but it occurs to
me that some sort of sensor might be used to start some humming or
buzzing noise if an anything is too close to where the car is headed.
(Something sounding like a hive of angry hornets should grab anyone's
It could get louder as the danger from the proximity increases. This
would doubtless require computer control to calculate where the danger
zone is and how far it was penetrated based on vehicle speed, power,
Motor Controller: Operator controls wiring
The motor & controller under the hood. There was lots of room, but
many obstructions and no flat surfaces to mount anything on.
I bolted an aluminum bar where there were two threaded holes together
on the fender and bolted the motor controller to the bar.
I had decided
on a single 10 pin "operator
header connector in the controller. But what would be on the other end?
The Tercel and two arrangements in the Sprint were a bit of a
hodgepodge of wires, even with just the Fwd-Off-Rev switch and the
'electron pedal' wired.
I got an idea to screw on a 10 position screw-down
terminal block under the dash in the car, and run a single 10 pin cable
to the connector in the motor controller. That way, a mechanically
solid connection point in the cab was an exact extension of what was in
the controller. The individual 'device' wires could go straight from
the device to the terminal block, ending with crimped (& soldered)
fork lugs to be screwed to the block. In the Sprint I left the rubber
Pico trailer plugs I'd already put in, but they should be a bit
superfluous in future installations.
One improvement would be to have 2 or 3 extra positions on
the block to clamp more ground wires, since most everything has a
Sprint Car Tests
On the 30th I installed a 36 volt battery in the Sprint,
having put in the motor and controller in the preceding days, and tried
it out. On the lawn where and as it was sitting (slight slope), a
wrench on the front wheel nuts showed it took around 20 foot-pounds to
move it forward, and 60 to move it backwards.
The motor ran fine. But there was notable hesitation and
little torque starting from a stop - a persistent controller problem.
The slip clutch rope, tho loose, also tended to jam and not let the
motor start turning. Evidently it needs a bit less rope angle wrapped
around the pulley... or maybe the motor just needs its starting torque
improved. In the reverse
direction, the helical gears pushed the planets assembly sideways
(about 3/16") where the attached slip clutch pulley rubbed on the
housing. Simple to fix. The driver controls in the car
worked fine. The motor temperature sensor, properly hooked up for the
first time ever, gave no reading.
I had to disconnect the clutch rope and pull it by hand to
avoid the jamming, and use the test controls at the hood. The car
didn't seem to move as easily as was hoped. It consented at one
point to roll forward, with much force required. It looked a couple of
times as though it was was about to start rolling backward, but it
didn't. I wasn't very impressed by the torque converter at the time,
but on reviewing the videos I took, I think it works well enough, and
that the motor controller is the real problem.
Theoretically the car should have
rolled even with the jammed slip clutch. The motor theoretically has
around 10 foot pounds torque with 100+ amps of current, and with 1.8x
speed reduction in the planetary gear and 4x in the chain drive, that's
72 foot-pounds to the wheels to overcome 20 to roll forward. But since
the motor is held stopped with no clutch, this would
presume good motor torque at startup instead of a hesitancy to start
spinning even with no load.
I put together a short video typical of the tests from
clips I took, and
uploaded it to Youtube.
Here car is rolling forward as I pull on the slip clutch rope to apply
Motor Controller Improvements (sigh!)
Then in one test, giving the motor high "throttle" and
pulling hard on the rope, the heavily loaded motor (probably drawing
over 100 amps) slowed to a crawl and there was a BANG! from the
controller and the smell of toasted transistors. (luckily just 2, phase
B low.) On inspection, the phase A power wires ran right alongside
the phase B high gates control wire for about 4cm, an ideal layout for
spurious phase B high-side MOSFET turnons and shoot through currents in
phase B -
but only at very high current loads owing to most of the gate run
now being a twisted pair, a 'balanced line'. I repaired it and moved
the gate wires farther away.
Additionally, the hesitancy of the motor to start is
obviously a separate controller problem that needs addressing. I've run
into it before. It's caused by the
capacitors of the floating high side gate drivers discharging in less
than a second in the phase being driven high if the motor doesn't
quickly get turning. Once it's going they get recharged as the motor
turns. I switched from .5uF to 2uF ceramic capacitors. It
doesn't seem to have done the trick. And the higher the power, the
faster they discharge, so 'giving it the boot' doesn't seem to help
New type of motor modulation control
I started to
think of going back
to something I did with the motor controller that moved the Tercel with
direct drive in 2008: pulsing the controller on and off at a few hertz
until it reaches some set RPM, then allow normal control. The charge
pumps will get reset in the
'off' periods, and the highest currents at very low RPMs won't be
Reviewing the videos, I noticed also that the motor seemed
wimpy under load. Then I divined the cause.
"CRM" modulation - "direct torque control" - is much more prone
than "PWM" to letting the motor stall under load, since the torque
doesn't automatically increase as the speed drops. Once the
motor starts slowing down, it keeps doing so and rapidly comes to a
stop unless you're very fast to "hit the gas". This in fact seemed to
be the main reason the results weren't better in trying to move the car.
This suggested making a combined control: Use the CRM,
perhaps always at "full throttle", and modulate that with low frequency
PWM. The CRM will come back on after the short fixed off period, so the
PWM can be as slow as desired, even 20, 10 or 5 Hz, without loss of
power. Switching transients and losses (= controller heat) are still
minimized with the low PWM frequency. Hopefully the PWM "off" period
will allow the charge pumps to recharge and eliminate the low starting
torque problem. I suspect the combined control will probably make the
best system ever. 4-1/2 years later, I'm still finding better ways to
do things that most makers think are already optimized, 'standard
practice'... and then having to redesign my circuit boards yet again!
But I can try out the system with an external single chip
PWM generator tied into the 'electron pedal' in the car, sending a
pulsed 'max. throttle' setting to the CRM controller.
If everything works well in the controller, motor, and
converter, it seems
to me the 4 to 1 chain reduction could be reduced, even to 1 to 1. That
would allow the Sprint to do 100+ Km/Hr without over-revving the motor,
can get going that fast.
Planetary Gear Torque
The first converter tests
are covered above as a component of the total system in the Sprint car.
3D-Printed Plastic Planetary Gears?
At the end of July (28th) I went to the Victoria Maker
Fair. I was mostly interested in the 3D printing, and I got the
names of some software and a person or two doing home 3D printing in
the displays was a planetary gear with straight teeth, about 2" wide
and 5" in diameter. The parts seemed huge compared to their metal
counterparts. I was about what I had envisioned to replace metal gears
with, to spread the loads over large areas and to avoid needing an oil
drip. Would such a construction stand up to automotive stresses?
(If 2" wasn't enough, maybe 3", or 4"?) I was told its ratio was about
5 to 1. I should have got the name of the person who made it (who
wasn't the person at the table). Maybe the design is on line somewhere.
I think my ideal would actually be aluminum sun and ring
gears and a steel planet gears assembly with plastic planet gears -
probably of the higher temperature nylon-like plastic whose name
escapes me at the moment. With the slight porosity of printed plastic
parts, the plastic gears could hold lubrication and some grease should
do for lubrication. I doubt I'd find 2 or 3" wide, large diameter
aluminum planetary gears, and that's too wide for abrasive waterjet to
cut, so it would probably devolve down to a nasty machining job needing
fairly special CNC & cutting tools. If printed plastic for all the
gears works, it'll be the simple choice.
Furthermore, for under hood conversions, a plastic part
could easily be printed as a ring gear on the inside and toothed belt
pulley on the outside, to drive a similar toothed pulley
at the differential. This would make the torque converter/transmission
unit very compact.
Solar Electricity Project
Price is already there some places...
A couple I know bought a half finished cabin in Hawaii
(Oahu?). Evidently the price of electricity there is already
40¢/KWH, and it gets a lot of sun year round. Payback on a solar
system might be 3-5 years. This would already be a good place to market
solar system, LED lighting and refrigeration equipment... for which I
but not as yet products.
Super battery Stick not so Super
I hooked my 70 cell "Super Battery Stick" in the 4" PCV
pipe, with 7
rows of D cells in parallel, into the solar collector system. I thought
it should run the 4 amp fridge all night. Instead, it was crapping out
after 4 hours. I shook it and the voltage went up. Of those 7 rows, how
many were actually making reliable contact? And there was no way to
test individual rows.
I decided that the "Super Battery Stick" idea wasn't
really a very
good one. Individual pipes of 10 cells can be tested individually and
proper operation of each ascertained.
However, another factor to be considered was voltage drop.
started at 13.8 out of the DC to DC converter from the collectors, but
alligator clip wires it became 13.4 at the battery (thus it didn't stay
at full charge even during the day) and 13.1 at the plug of the fridge
unit. And, since I was measuring... the long, thin cord supplied with
the cooler lost another whole volt to 12.1 before it hit the circuit.
The moral of this is to put the batteries close to the charge
controller with short, fat wires, and to keep all wires thick enough.
Additionally, I have the idea to make compact plastic 10
volt D cell battery holders using the 3D printer. These are likely to
be the most practical general solution, at least where 26 or 14 inches
length of space isn't available for the long 12v or 6v sticks.
Superinsulated Thermoelectric Fridge & Freezer
Single Peltier Cooler Again
On looking over the datasheets, I
decided the two-stage cooler didn't really look any better than the
single stage. The heating of the warm side element by the cool side
element essentially seemed to cancel out the benefits at refrigerator
So I tried the same arrangement as had given me freezing
temperatures but with the single original peltier, and got better
results: it hit -13ºc. With the heatsink attached but no fan, it
still managed to hit freezing in a warm room. (or even -2,
depending on the orientation (& hence airflow) of the heatsink.)
Keeping the warm side a few degrees cooler would also cool the cool
side a bit, but the desired temperatures were in fact attainable by the
itself as made.
At the flip of a switch, the cooler unit could become a
warmer. That didn't seem very useful at first, but it defrosts the cold
side heatsink in minutes, before the fridge itself warms much. A
rectangular 'funnel' under the inside heatsink would carry out the
drips through a
plastic hose. I was wondering if 'defrost' should be left as a manual
operation, or if the temperature control should have it built in. Then
I put it on defrost and forgot it a couple of times, and the fridge was
warming instead of cooling. Now I'm sure it should be
funnel might well be a custom 3D printed plastic part.
It also occurred to me that the unit should shut off, or
at least the inside fan should shut off, when the lid is opened, to
avoid blowing the cool air out.
Evidently I had been wasting my time with the double
peltier. Logically then the fridge should just need more watts, eg, two
cooling units... but I still suspected a problem in the cooling unit,
perhaps in its layout, because it attained about the same temperature,
8 to 10ºc, no matter how much insulation I added to the fridge, or
even back on its original cooler. Why was it hitting a "wall" at that
temperature? A second identical unit would double the watts, but might
get it little cooler.
So I wasn't quite ready to give up. About the only thing I
could think of to try was to
separate the warm and cold sides more, to place them entirely outside
and inside the thick insulation of the fridge, by putting in a longer
block of aluminum through from the peltier to the inner heatsink. (If
I'd known I was going to do this, the fridge would have just needed a
small square hole through the insulation instead of a giant one to
mount the whole original unit as it was.) Copper transfers heat even
better, but it's both costly and heavy. If there was much temperature
rise along the length, a fatter piece of aluminum is probably a better
answer. (Seemed pretty good.)
I added a couple of chunks of aluminum and extended it
over 2", but the fridge insulation is over 3" so the inside was still
well recessed into the wall. Without a fan, the cool side heatsink went
to -6 and the fridge to about 12 (seemingly somewhat stratified -
cooler at the floor below the unit than up at the lid). With a fan, the
heatsink rose to +5 but the fridge stayed at 12, less stratified. With
the peltier cold side 10º warmer, it should have been delivering
notably more cooling. It was still hitting that "wall" where it simply
didn't want to further cool the space, for no readily apparent reason.
It would need about an added 2.5" to 3" block of ~42mm x
~44mm aluminum to have the inner heatsink sticking out in the inner
space. The next morning I was passing a recycling place. I stopped and
found a 3/4" thick slab. I cut a piece to fit. With no fan, the inside
heatsink, now exposed properly to the inside air - vertical with the
fins up-down, hovered at about zero. It seemed to reduce or eliminate
the need for an inside fan, but the fridge space didn't get any colder.
I finally concluded the obvious, that the unit simply
enough watts of cooling power to attain the desired temperature in the
space being cooled. That means the unit will inevitably run all the
time, and a temperature control is virtually superfluous. I also
finally realized that a temperature reduction on the warm side doesn't
mean an equal reduction on the cold side, just a few more
cooling watts. A bigger heatsink, stronger fan, or water cooling of the
warm side just might make 1-3º difference inside, but it would be
the 'law of diminishing returns', much effort for a small gain.
Of course, the lowest temperatures already attained, 8 to
10º, are much better for food storage than room temperatures. Some
might even say "Good enough for me." (Someone wrote that 3-5º is
optimum where I thought 2-4º... one less degree to cool to) In the
winter in my cool kitchen
it would get lower. I decided I'd consider 6-7º or under (with a
23º room temperature) to be essentially food worthy.
I decided to
replace the peltier in the cooler with the highest power 18$ peltier
element available at Digikey, tho it's only incrementally higher power
original. It might just make it 2 or 3 degrees cooler without needing
more hardware - heatsink and fan units. If that didn't hit 6º, a
bigger heatsink on the outside might just make the remaining
difference. But a new plan (below) came to light before
I got to ordering it.
Meanwhile, although it was
too late to shape the fridge
for minimum exterior surfaces per volume, eg, as an octagon shape with
shorter walls and a terraced saucer floor (so that tall objects could
fit in the middle and shorter ones around the edge), I put some wedges
of foam in the corners (making a slight octagon of the inside).
It also occurred to me I might still shorten the walls a
bit and make a domed lid, or a lid with less insulation inside in the
middle for tall objects, in lieu of the saucer floor. (The almost flat
floor is done. I could take a jigsaw to it - ugh!) Every inch
shortening of the walls is 3/4 of a square foot less surface to cool.
(Rectangular interior surface area of 18" x 36" x 12" walls is 18
I suppose the ultimate shape to minimize the exterior
surfaces of an
object with flat sides, would be a geodesic dome.
Considering it gets close to fridge temperatures, such
little things as minimizing surface per volume, sealing the lid, and
putting on the outer skin, should make notable differences to the
temperatures attained with the same tiny cooling
unit. There's another possible expedient...
Fridge Size Experiments - & without inside fan
It occurred to me on the 16th to cut and insert a piece of
2" styrofoam to divide the space. Thus I obtained a variable size
refrigerator. With a room temperature of 25-26ºc (scorching heat
for my house), the whole fridge (with the cooling unit reconfigured to
go through the 3" of insulation, but no proper duct over the cold side
fan and heatsink) was attaining 13-14º. Divided in half, the
cooled side reached 17º while the virtually uncooled side settled
to 19-20. Huh? Something was definitely rotten in the state of Denmark.
I found a rusty alligator clip wasn't making connection from the
converter to the battery. The 30 AH battery, having delivered about 5
amps for nearly 5 hours, was down to 3 volts. That was serious
overdischarge, even for NiMH, and it certainly explained the poor
cooling performance. Probably on
reconnecting the low battery, I found the "25 amp" fuse from the solar
collectors to the DC to DC converter was actually 7.5 amps and was
I waited for the current
(charging the battery) to drop below about 6 amps and tried again,
making a note to replace the rusty clip that evening. This time it
reached 11, still with the uncooled side about 19. (The outer heatsink
read 40º. The inner one was 7º, but the peltier was 1 to
2º, showing poor thermal coupling from the element to its inner
heatsink.) Cooling only 1/4 of the space, the temperature hit 9 -- a
16º cooling. It wasn't a whole lot cooler than 13 considering it
was 1/4 the space, but it probably shows the physics: the higher the
temperature difference, the harder it is to attain and maintain.
Later (20th) I tried without the inside fan. Evidently
there are more air leaks than I thought or something, because it worked
better instead of worse. Cooling about 1/3 of the fridge, temperature
was 3 in a 22º room. The 19º was the largest temperature
differential attained. A still smaller space hit 2º. The inner
heatsink, at -7º, was heavily frosted up after a few hours.
either the heatsink should be enlarged or the fan is needed. I'll
probably make a small ice
compartment right around the cooling unit.
I turned the setting to "heat" and again defrosted and
cooler in 5 minutes - what a great defrost unit! The tray beneath
collected about a tablespoon of water.
A little later I
found that the temperatures were considerably higher near the top than
near the floor of the unit, the temperature stratifying. Either the
unit should be at or just below the lid instead of mid wall height, or
the inside fan is needed to circulate the air.
By the expedient of making the fridge smaller with an
inside wall, a certain amount of food can be kept cold even on warm
days, while in the winter the whole space will be available if needed.
On the other hand, around 38 KWH a month (since it runs all the time,
so far) was as much juice as a good energy star full size fridge to
cool around 1/3 to 1/6 of the space. That's about a whole solar
collector's worth and more in the summer, and unless it shut off part
time it probably couldn't be powered continuously all winter.
Well, I can't say I wasn't warned Peltier cooling wasn't
very efficient. Coefficient of performance is typically .3 to .6
(Wikipedia), or around 1/2 watt of cooling for every watt of
electricity used. A compressor based fridge, or a heat pump, might be
1 to 1.5. A home heating heat pump might be only one at sub freezing
temperatures, but the electricity used is also heat, so one becomes two
for heating but not for cooling.
But it does cool, and... "Seek and you shall find." - and
not always what you expect to find (or you'd have already found it).
Soon after the July newsletter someone sent me a Wikipedia link to
magnetic refrigeration, which I
hadn't heard of. The potential coefficient of performance is perhaps 5
higher! Seemingly in accordance with my 'thermomagnetism'
theory, some substances heat notably when exposed to a strong magnetic
field, and similarly cool when removed from the field - The
"magnetocaloric effect" (MCE). If they're
cooled back to room temperature with a heatsink while within the field,
they get notably cooler when the field is removed. The fields employed
are in the range of 1 to 7 teslas. 1 or 2 is attainable with
supermagnets. Potentially there are a number of ways the heat could be
transferred from the fridge to a heatsink via the substance being
heated and cooled.
Development is in experimental stage and it looks very
promising. The most promising research cited is in my own
town, Victoria BC. A 2009 paper said it was potentially more
efficient than compressor + gas refrigerant refrigeration: 30-60% of
theoretical carnot cycle, versus 5-10%. And it uses one of my
favorite things: magnetism! The
best substances (which change magnetic state) at about room temperature
include gadolinium (Currie point 293ºK), alone or in an alloy with
europium (I think the document is in error - it seems it's erbium, not
europium. Currie point 278ºK).
How complicated would it be
to do well? I decided to try and contact the researchers and look into
their latest system and ideas. Peltier was the only viable
refrigeration idea I initially found, but it's not very efficient per
se and I'm not wedded to it. I'll jump horses in midstream at the drop
of a hat like a scared rabbit rather than throw good money after bad if
there's some other bandwagon I can put before the horse - if it
proves reasonably simple to do. After all, much of my effort went into
building the enclosure. That can be cooled any practical way that can
be found. Perhaps adequate cooling can be accomplished with fewer watts
instead of more, by building on the UVic research?
An 'autoreply' said the chief researcher was away. He
didn't respond to my e-mail when he was supposed to have returned. Then
I noticed he had started a company looking for funding to commercialize
his complex designs. By that time, I had read a paper surveying
existing designs, worked out my own simpler idea, and ordered
If the performance is good, perhaps really efficient small
scale heat pumps and air conditioners can also be made without typical
compressors and refrigerant gasses. If a practical unit can be
produced, the potential for small size, lower energy heating and air
conditioning units is enormous. The peltier heat pump might do 1.3
times better than a regular electric heater - significant but not Earth
shattering. And it wouldn't make an effective air conditioner.
Something that could provide two to five times
the heat per watt of an electric heater, in a small plug-in size unit
(and maybe cooling too), would be a real game changer. Decentralized,
room furnace replacement units using maybe 300-700 watts instead of
would be worth putting a hole through the wall for!
But, even if it's practical and much better than peltier:
is it worth it for me to get involved in developing it? I only
started out to make a refrigerator that would work with a low voltage
DC solar power system. On the other hand, my winter electric bills are
high, and lots of people pay much more than me. To cut heating bills in
half, or by 2/3 or 4/5, is potential to drool over.
I decided to order
some Gd. I
saw it for about
1.0-1.5 $/gram in 100 gram lots on e-bay, then I phoned HEFA Rare Earth
in Richmond BC: 490$/Kg or 54¢/g for smaller amounts. Then: So an
alloy of Gd:Er
might be tailored to a specific Curie point? What about one with
a low point (erbium?) alloyed to one with a high point (iron,
nickel...?) in varying proportions? I
ordered some erbium too (35¢/g), but I'm not sure how I'll
any alloying without oxidizing any rare earths.
Gleaned from Literature
I read a 32 page paper that was a survey of all magnetic
refrigeration experiments and prototypes prior to 2010, including th
eUVic ones. (wenku.baidu.com/view/8dd4353410661ed9ad51f3f8.html###)
Evidently an alloy of Ni:Mn:Sn has a reverse
magnetocaloric effect: it cools on entering a magnetic field, and
warms again as it exits it. Some alloys (eg, Gd:Si:Ge) have a "giant
Gd - 100% - Curie = 293 K (20ºC)
Gd:Tb - 74:26 % - 285 K ??
Gd:Er - 85:15 % - 278 K (5ºC)
La:Fe:CoSiB - Bal:10.97:.78:1.05:.2 % ??
About 50 watts of cooling power is supposed to be sufficient for a
typical refrigerator. (In which case why can't I cool the
superinsulated fridge with around 20 watts?)
Typical magnetic field strengths were .9 to 2 teslas, with a few
superconducting electromagnets supplying up to 7 teslas.
Typical cooling cycles were around .2 to 4 Hz, mostly under about 1 Hz.
The gadolinium needs time to magnetize and time to transfer the heat.
(Thinner pieces at a higher frequency of operation should pump the same
amount of heat while requiring less gadolinium.)
There were heat conducting fluids employed including
helium, water, air, pH 10 'basified' water, and olive oil, but in 41
prototypes described, none used solid state heat transfer. Personally,
I suspect that the first researchers used a gas or fluid, and while
varying much else, the rest have just copied that. But solid contact
between flat copper or aluminum faces is surely the fastest way to
transfer heat, and the fact that gadolinium is attracted to the magnets
provides a very convenient means for shuffling it back and forth
between cold and hot sinks at (surely) just the right times. (If a MCE
heavy enough to make good thermal contact with the lower heatsink,
perhaps a bottom facing of copper foil could add weight.)
Furthermore, the report states that while the promise is
there for 20-30% efficiency improvement over compressor/gas systems,
substantially higher performance is being prevented by the pressures
needed to overcome friction and drive the fluid or gas sufficiently
through the system. Such extraneous power demands are largely
eliminated in the solid state system, so it *should* prove far more
efficient. If I get 1.5 times what compressor & gas refrigeration
gets, it'll be an advance over the fluid systems. Double would be even
Gd & Magnetic Motion Machines
I also wonder if there's some role for gadolinium in
ambient temperature powered magnet machines? Surely the heat and the
motive force are
Related to this, I found the following video, where a
paperclip hanging from a metal "string" is attracted to a magnet, which
lures it into a candle flame. The flame heats it up to the curie point
of the steel, at which point it loses its magnetism and falls back.
Soon it cools off until it's attracted to the magnet again, and the
process repeats itself repeatedly until it becomes repetitive.
The curie temperature, BTW, is also the point iron has to
be heated to in order to anneal or harden it.
"Curie Point Magnetic Heat Engine
How it works
The heat engine uses a
principle of magnetism discovered by Pierre Curie. He studied the
effects of temperature on magnetism. Ferromagnetism covers the field of
normal magnetism that people typically associate with magnets. All
normal magnets and the materials that are attracted to magnets are
ferromagnetic materials. Pierre Curie discovered that ferromagnetic
materials have a critical temperature at which the material loses its
ferromagnetic behavior. This is known as its Curie Point. As an
example, a piece of iron (Fe) at room temperature is strongly attracted
to a magnet. Heat the iron to a temperature of 770 C, which is its
Curie Point, it loses its ferromagnetism behavior and it is no longer
attracted to a magnet. If we let the iron cool, it regains its
ferromagnetic behavior and is attracted to the magnet again."
This was hardly a perpetual
motion machine. The energy of the paperclip going back and forth was
minute compared to that of the candle flame. But the curie temperature
of iron is 770ºc. Even slight heating and cooling would put
gadolinium above and below its room temperature curie point, but the
magnetism doesn't seem to change as far as I could feel.
My Plan for a Magnetic Cooling Unit
I came up with an idea for a cooling unit that uses solid parts only,
no gas or fluid:
A rotating disk of supermagnets would pull on 2 or 3 stacks of
gadolinium wafers (each faced with epoxied copper foil to protect the
reactive gadolinium from air and moisture), interleaved with aluminum
or copper heatsinks. Each stack of wafers would "shiver" up and down as
magnets passed, heating as the magnet approaches while being pulled up
to the upper, warming heatsinks, then cooling after the magnets pass
while the stack drops back to the lower heatsinks, pumping heat through
the stack from bottom to top. Each individual wafer cools by 2 or 3
In my first plan, the (epoxied, copper foil faced)
gadolinium wafer stacks would have been be held together by a framework
along two opposite edges. The interleaved copper heatsink pieces would
be framed along the other two edges, creating two interleaved stacks.
On the 24th I came up with the idea of framing the copper
heatsinks on all four sides, which would simply trap each loose
wafer in a thin box between two heatsink pieces. Nothing would hinder
the wafers from
jumping up when the magnet arrived and falling back afterwards, and
making good, square contact with both the heatsink above and the one
below. They only need to go up and down a fraction of a millimeter -
good square contact - very flat surfaces - is the key.
Naturally I thought of making the stack frames of plastic
with the 3D printer, but for a prototype that would probably just be a
luxury. They could just as easily be glued up from sheet plastic. If
each gadolinium wafer can be in a tiny sealed
compartment with minimal retained air and moisture, it shouldn't be
necessary to epoxy them and face them with copper foil, simplifying
everything, shortening the stacks, and maximizing heat transfer.
The reason for multiple stacks is to attempt to balance the magnetic
forces acting on the motor's rotation, so as one magnet is approaching
and being pulled towards a stack, another is leaving a stack and being
pulled backwards. It'll also be necessary to put stacks thermally
in series to get sufficient difference between hot and cold ends while
keeping the stacks short to keep the gadolinium wafers close enough to
* A rotating disk of supermagnets (should provide sufficient magnetism,
but stronger is probably better).
* A stack of copper plates (Cu sheet metal heatsinks) with narrow
spaces in between them, sealed around the edges to protect the Gd from
air and moisture.
* Each space is loosely filled by a wafer of gadolinium. (Even Gd dust
would work, but single flat pieces will transfer heat best.)
* There may be two or more such stacks, with copper or aluminum
connecting the cold side of one to the hot side of the other, to double
the number of cooling units in series. (Since the wafers all have to be
as close as possible to the magnets, the stacks will be at most an inch
* An outside/hot side heatsink with fins and (probably) a fan. (must be
thin at the end to allow magnet rotor to pass closely to stack - fins
aim the other way.)
* An inside/cold side heatsink (large or with fan) inside the fridge.
That seems quite simple and straightforward - in fact,
much simpler than gas refrigeration.
complication, especially for
production, is how to form the gadolinium ingots/chunks into sheets.
fudge a few some way for test prototypes. Even gadolinium powder or
should essentially work, tho not transfer heat as fast as a solid
If it needs to be heated, it would have to be done in an inert
atmosphere, and Gd's melting point, 1312ºC/2394ºF, is at the
extreme limit of my kiln temperatures. (I wonder if one can create a
sealed steel box and, say, keep argon flowing through it while heated
to high temperature? It seems welding places do have tanks of argon -
Considering the incredible potential benefits, I can
hardly see walking away from it. (Sigh!) The project is On!
Chest Freezer As Refrigerator
Someone wrote of a chest freezer being used as a
refrigerator for off-grid use. This seems like a brilliant idea! A unit
designed to cool to -10c naturally cools to fridge temperatures very
easily and stays there longer. (If it isn't super energy efficient to
start with, putting a couple or 4 inches of foam in the bottom and on
the lid as I did with my freezer should help considerably.)
Energy use is evidently minimal and the owner is quite
pleased with it. I might try it myself except instead I want to try the
magnetic cooling, which should be even better.
Turquoise Battery Project
The battery project, LED lighting project, and aspects of
other projects came down to assembling the 3D printer kit, with which
device I plan to make various battery parts, LED light bases, and PCB
artwork. I did 3 hour
sessions on several days or evenings, and assembled the frame, "Y" axis
carriage, "X" axis carriage, heated printing surface, hot plastic
extruder head, ...
About the only other things that happened were the arrival
of the osmium ordered last month, and
some samarium oxide - in an order with some rare earth metals for
magnetic refrigeration experiments. Sm2O3 should be a bit better trace
additive for raising oxygen overvoltage than Nd2O3, which I have about
a pound of, so it wasn't exactly a critical purchase. But here was the
time and place to get the best.
The osmium is indeed dense. The 10 gram blob of bluish
metal is only 1/2 a cc in size, like a fat shirt button. A kilogram
would be just 45mL. But there's no way one could afford a kilogram of
it, cool tho it would be to feel its heft.