Turquoise Energy Ltd. News #46
Victoria BC
Copyright 2011 Craig Carmichael - December 1st 2011

http://www.TurquoiseEnergy.com = http://www.ElectricHubcap.com = http://www.ElectricWeel.com

Month In Brief (Summaries)
* I'm to speak at Discover Tectoria conference Dec 8th, about Electric Hubcap motor system
* In passing: Nuclear Industry Melting Down; Buying Police Loyalty; Yet Another GM Dirty Trick

Electric Hubcap System
* Motor repairs (working again) & improvements to rotor compartment
* Polyurethane spray paint "crust" to keep ilmenite coil coatings from flaking off.
* Motor controller: V2 boards, CRM = direct torque control, great current control during regen... but how??
* Cutting shaft keyway slots with a lathe - lathe as milling machine?

Electric Weel Motor - no report (I did coat the coils with ilmenite. Very flaky, unfortunately.)

Mechanical Torque Converter Project
* New design: Magnetic Impulse torque converter
   - simple, robust, no wear, no internal moving parts!
* Combo magnetic/mechanical design offers higher torque

Sprint Car Conversion Project - no report
* Working on components - see Electric Hubcap System, torque converter and NiMH battery projects

New Electric Hubcap Outboard Project?
* An outboard from scratch might be just as easy as a conversion - and better?
* Production kit version?

LED Lighting Project
* Bright, cold white emitter fixtures: 1100 lumen/15W
* Electric bills continue lower than expected into winter (yay!)
* Energy Star partnership agreement form file format problems

NiMH Battery Project
* Ultimate Battery Stick? - 4" PVC pipe holds 7 sets of D cells... a 4" x 25", 12V, 70 amp-hour battery
* Car(s): Musings about using combos of NiMH and lead-acid

Turquoise Battery Project
* Zinc powder?
* Zinc powder from zinc oxide
* Best quality zinc electrodes!

Newsletters Index/Highlights: http://www.TurquoiseEnergy.com/news/index.html

Construction Manuals and information:
Electric Hubcap Motor - Turquoise Motor Controller
- 36 Volt Electric Fan-Heater
- Nanocrystalline glass to enhance Solar Cell performance - Ersatz 'powder coating' home process for protecting/painting metal

Products Catalog:
 - Electric Hubcap Motor Kit
 - Sodium Sulfate battery longevity/renewal
 - NiMH Handy Battery Sticks
, Dry Cells
 - LED Lighting Products
Motor Building Workshops

...all at:  http://www.TurquoiseEnergy.com/
(orders: e-mail craig@saers.com)

November in Brief

I'm to Speak at Discover Tectoria, December 8th

   Late Breaking News for local Victoria BC people: On December 1st I was invited by phone to speak next Thursday (Dec. 8th) at the all-day Discover Tectoria conference and trade show [http://discovertectoria.com/], known in previous years as Island Tech, on the Electric Hubcap motor system. I'm among several speakers expected to speak for about 10 minutes each sometime between 2 PM and 3 PM. I don't have all the details yet, but I'd like to bring in a motor, controller, a couple of battery sticks and C-clamps, and run it on a table as a demo and point out features.

Discover Tectoria will take place:

Thursday, December 8th at:
Crystal Garden
713 Douglas Street.
Victoria, B.C.

   The entry fee to Discover Tectoria is $20, but there are supposed to be some free entry coupons in the newspaper this week.

The Month

   In some ways, the first half of November felt like one of those dreams where you seem to be trapped in slow motion and can't get anywhere. There were days when virtually nothing seemed to move forward. On the other hand, I spent a lot of time on the internet doing various promotions. Then from the 16th or 17th to about the 20th good things started coming together bang bang bang. Then it went back into slow motion.

   Everything in its place... I decided that on the Sprint conversion that the next thing to do was to fix the motor. In the process I'd also further the motor development: I'd make a mold to improve the rotor compartment rim. It's safer: not only much thicker and tougher, but the attachment bolts won't be exposed inside, to be hammered, and bent and ripped out by any magnet(s) that might potentially come loose. The last arrangement was too easily damaged with potential for inflicting serious injury.
   By then maybe the IR2133 motor controller V2 boards would arrive and I could make the motor controller. With a working - improved - motor system again, it would be time to tackle the torque converter.

   Considering installing the Sprint batteries, I came up with the idea of Super Battery Sticks: a single 4" PVC drain pipe, with 1-3 struts glued in for battery alignment, will hold 7 parallel strings of D cells, eg, 840 WH at 12 volts. And they'll only weigh about 30 pounds, mostly the 70 D cells themselves. Three of these would fit where the car's radiator was for 36 volt, 2520 watt-hours and plenty of amps, for about 1400$ at the current cost of the cells.
   And I'd see if I could convert three of the ten 12.0 volt, 5 amp power adapters I got (at XS Cargo for 3.99$ each) to 13.8 volts, to constant-voltage charge them.
   It would be even cooler to put three 65 watt solar collectors on the roof, regulate them to 13.8 volts, and not even have to plug the car in when I'm not doing a lot of driving -- but that'll have to await some time when I have money.

   For a couple of days I considered that maybe I should do some contract work for someone to earn some money. I was thinking something short term, part-time, but I knew it was more likely either to come to naught or grow into a lengthy commitment that would pretty much displace the inventing. Personally I consider the work I'm doing to be more valuable to all than anything I could do for some one concern, but it would be nice not to have to wonder how far into debt I'll be this year before I get the annual pittance next spring or summer that society will actually give inventors to work on new things, or wonder about how to do increasingly urgent repairs on a low budget, such as going up on my high, 45º roof myself to replace "30 year" shingles that are falling apart and blowing off after only 18.

   On the 17th, the day I (finally) finished the rotor compartment rim mold, I thought up a Magnetic Impulse Torque Converter. This is in principle a magnetic version of what I was planning to make mechanically, with supermagnets tugging at aluminum 'spokes' or 'arms' for only a short part of each rotation. It's an exciting design: no moving parts, nothing touches anything else to wear out - it's simple and looks easy to make! An experiment indicated the magnetic forces strength should be in the right ballpark -- unless things saturated electromagnetically somewhere below car moving force. I decided to make one for the Sprint instead of finishing the mechanical one. But the torque increase wouldn't be all that high, and by the end of the month, still before I was ready to start working on it, I decided on several changes, and then to make it a combo magnetic-mechanical one that would have more kick to it. I'm keeping the pieces for the original mechanical converter design just in case.

   The Turquoise battery project seemed to devolve down to looking for zinc powder. It's there on e-bay and other places, but American sources don't want to ship it outside USA, and Canadian sources don't return e-mails. I ended up grinding a little myself, which was tedious, and then got onto the other things without making an electrode.

   I tried another type of LED emitter, making an 18 watt, 1100 lumen fixture. These ones are 'cold white', which looks rather bluish. They'll be about the cheapest per lumen price-wise. They're also 12 volts, so no matter how many emitters I use in a fixture, it will be that voltage with no need for 6V, 9V or other power adapters. I may do these, and nothing but four-emitter Cree fixtures (3V each), and use 12 volts exclusively. Only on the last day of the month, after a couple of weeks of hassles with some new PDF file format that my older computer and software wouldn't handle, did I get the Energy Star Partnership application form off in the mail to begin the process of obtaining Energy Star ratings and BC Hydro rebates for potential LED light fixture customers.

In passing: Nuclear Industry Melting Down; Buying Police Loyalty; Yet Another GM Dirty Trick

   Switzerland is next to eliminate their potentially devastating ticking time bombs - they've legislated shutdowns from 2019 to 2034 to bring nuclear power generation to an end in that country. (Now there's a country that could well string windplants like cable cars between mountains!)

   Michael Moore spoke on RT Times - I watched it on youtube. He said he was setting up cameras and (IIRC) crime scene tape in front of the bank (a week before "Occupy Wall Street") to do a piece on it, when a police officer came up. Michael tried to tell him they were just shooting for a sci-fi, but the officer wasn't fooled. He said, "Go right ahead Michael, they've taken our pension fund too."
   Evidently recognizing they'd committed one crime too many, "the 1%" made a big "gift" to the NYC police dept. When an outraged public is howling at your door - or occupying Wall street - it's best not to have the police on the wrong side.

   On the 25th I met a retired mechanic who had worked at GM car dealerships up island in the 1960s-1970s. He told me that one day a customer brought a brand new car back in, saying the gas gauge wasn't working, because it still said "full" after driving 250 miles. He looked under the hood, and where the carburetor usually was was a big carburetor or something (he motioned over a foot or more square or in diameter) with all kinds of hoses and connections. They phoned the factory to find out more about it. Someone flew out from the factory all the way to Vancouver Island that same day, removed the unusual carburetor and took it away after installing the regular one.
    There was nothing wrong with the gas gauge - the car had just used far less fuel than usual. It was one of many demonstrations that answers are all around us, but are being witheld. Needless to say, the new, super economical carburetor never made its debut in production cars. This disproves GM's story that it wants complex cars that need lots of servicing because that's its "business model". This carburetor was more complex than the regular one, not less, so according to their own humbug excuse, it should have been favored. In reality of course the same gangster families control GM and the oil companies, and maximizing oil consumption was and still is their goal - contrary to all interests but their own lust to wrest the money from other peoples' pockets.
   The active undermining and sabotage of every transportation energy economy and alternative has held up vital technology for a century now and is one of the major forces gradually ripping civilization apart. It seems obvious to me that natural resources should be owned by the public and administered on its behalf by the government or by a public trust, not owned by private interests for monopolistic profit at the public's expense. That step and philosophy would get things under public control, and I think we'd find that would get us switching off oil - in years, not decades.
   I think as the population levels off and competition for land, housing and goods diminishes, as supplies become sustainably adequate to the demand, we have what we need to all be well-to-do with about 4 or 5 hours toil a day. Presently we can't achieve it because of the grossly inequitable and rapidly worsening distribution of the wealth. That too should be greatly improved by public ownership of natural resources.

Electric Hubcap Motor System

Motor Renewal and Improvement

   On the 6th I repaired the magnet rotor. I made sure there was plenty of epoxy and that the polypropylene strapping was well stuck down, including brushing more epoxy onto the remaining seven magnets and the material around them along with the five that came off like bullets.
   It will likely withstand the RPMs that 42 volts can give it now, but I think I'll stick to 36 volts unless I find that has insufficient power in an actual use. Besides, it's easier charging three identical 12 volt sections without an extra 6 volt section thrown in.

   I was surprised in October's motor failure that the outside rim of the magnet rotor had ripped open. In the second week of the month, I started putting together a mold to make molded rims for the rotor compartments. The molded rims are sturdier because of thick, solidly molded walls. They're also safer because the wall encloses the bolts that hold the rotor compartment to the stator, so even in the event of a rotor failure, loose magnets won't hammer the bolts to bend them and pry them loose from the stator plate. As one more bonus, the compartment will be 1/4" thicker, 1.75" inside, and things will fit more easily. I couldn't readily do that using the PP strapping because it only came 1.5" or 2" wide.
   The mold has several pieces. Some of them had to be machined, and what with forgetfully trying to shop for polyethylene on Rememberance Day and other things (house and porch repairs), it wasn't ready until the 17th.

   The next day I 'retrofitted' an existing rotor cover with a new rim in the mold. It was tedious, but it was virtually a snap-on fit onto a stator center ring. Great! It couldn't shift sideways a bit and line the shaft up slightly crooked like before. It's now pretty much "well done" to sell in kits, where it had been a bit "cludjy" before. That didn't mean it didn't need a bit of touch-up as there were spots of dry polypropylene here and there. That should happen less for a new rim than with stuffing cloth in thin spaces beside an existing thinner one - and the work will go faster. The outside wall couldn't be filled much and mostly looked the same as before.

   I have the idea it'll take about 300 grams of epoxy resin and 75 of PP cloth shreds to do a rim properly from scratch.

   A minor remaining concern is the ragged top of the lip that fits over the stator center plate. I sanded it down but there are gaps. For the next one think I'll try getting a 1/8" thick piece of polyethylene to stuff into the top of the mold around the edge, and fill the mold well... in other words, mainly just improve my technique and see how that works.

The thickened rotor compartment rim (before drilling bolt holes through wall),
and the magnet rotor side of the center plate.
(Previous image shows stator side, with buttons to hold toroid core coils in exact position.)

   With various little things like cutting a keyway slot in the shaft, needing different length bolts, etc, it wasn't until the 29th I finally reassembled the motor, and it ran well the next morning (after finding the coils were rotated one position from the sensors and swapping phase wires).

   I was surprised to find that it hit 1100 RPM with only a 12 volt supply. At 42 volts, that would be headed well up towards 4000 RPM instead of 2000. Yow! I'm starting to see why magnets might rip off the rotor. I think I'll take it apart and reduce the flux gap.

Cutting shaft keyway slots with a lathe - Lathe as milling machine

   Uncertain about the long-term ability of the SDS taper-lock shaft bushings not to move under conditions of high torques and loads with vehicle vibration, and especially when the torque comes in bursts, I decided it would be wise to install shaft keys. The SDS bushings and the sprocket gear have keyway slots, but of course the 1" round shaft, bought and then cut to length, is plain round.
   In October(?) I ground a slot into a shaft for the torque converter output and sprocket gear, and gouged it to final size with a 1/4" square "lathe tool", which itself became the final 1/4" square key. But I wanted some way to do it properly, more easily, and repeatably. I thought that properly, this needed to be done on a milling machine with an end mill bit.

   I hadn't even found a 1/4" end mill on my previous shopping trip looking for one. But in a "miscellaneous" can on my own shelves I found a genuine milling bit, that had a 1/8" section and then a 1/4" section, bought cheap a few years ago at a pawn shop to use just as a two-step "countersink" drill bit in wood. I cut off the 1/8" section with the angle grinder to leave a 1/4" end mill. The end was a bit rough, but the sides would mill.

   I phoned my friend who I'd bought the CNC machine from - never having done any milling, I wasn't sure of myself. The speed of the router was way too fast, but I wondered about the drill, if the drill chuck would hold it square properly. Instead he suggested gouging out the slot with the lathe using the tailstock and a 1/4" square lathe tool. That sounded hard to set up and tedious, but I walked into the shop as we were talking. Looking at it, I suddenly realized how it might be done "properly" on the lathe.

   I put the end mill bit into the main 3-jaw chuck on the lathe spindle. This would hold it securely and turn it at a (selectable) lower speed appropriate for milling. Then instead of a tool, I put the shaft itself in the tool holder on the turret, square, lengthwise front to back. I put a 1/8" flat piece of steel under it to shim it up to line the center of the shaft up with the bit. Left-right adjustment would set the depth of cut, and screwing the turret backward-forward would mill along the shaft.
   After trying a few things, I found it seemed best to start at the end of the shaft and mill out the whole 1/8" depth at once. An end mill that would plunge (drill) properly might use a different technique, and cut a blind slot that a key can't slide out of. I had to hammer a crimp into the slot to ensure that couldn't happen. Later I managed to sharpen the ends. Tho the four flute ends weren't even, this would allow the mill to drill in for blind slots.

   So there's one more motor making thing I don't need to buy more equipment to produce!

   It occurs to me that one might contrive to affix a small drill press vise that adjusts X, Y position with two screws, sideways on the lathe carriage facing the chuck. Parts could then be fed up, down, forward and back. With the turret to move left and right, one would essentially have a small, sideways oriented milling machine. I must remember that in case I have a need for one... maybe to make an injection mold for making battery cases or LED light fixture bases?

Motor Controller

   The V2 circuit boards from iteadstudio.com didn't arrive in November, unlike the fast turnaround from AP Circuits. The low price makes it worth a considerable wait: assuming they come. After a month I'm getting nervous.
   On a discussion list, it appeared other motor controllers take their time going from forward to reverse, and some were talking about them failing during braking - so much the better for mine! Someone wanted one with fast directional response for a computer controlled unicycle, so mine was of interest.
   It occurs to me that the "Current Ramp Modulation" (CRM) control is also "direct torque control", since torque is directly proportional to current.
   On the instant reversal, I've been idly puzzling in my head why the control potentiometer works properly when the motor is braking. CRM works by sensing when the current has risen to the set level. But when it's braking, the current is flowing backwards, into the batteries. It never occurred to me it might not work when I was testing it, but the sense voltage should be reversed, negative... shouldn't it? Yet the control works reliably and smoothly, seemingly about the same as going forward.

Mechanical Torque Converter Project

New design: Magnetic Impulse torque converter!

   On the 17th (just after I finished the Hubcap motor's rotor compartment rim mold) a friend came by and we talked about some things... including the resistance of aluminum to magnetic fields and the potential uses of that. As he drew a thick piece of aluminum across the magnets on a rotor, feeling the resistance to the motion: Eureka! I had the flash of an idea that's surely been on the fringes of my consciousness for ages.
   As I considered, experimented and calculated over the next day and some, it became evident that it should indeed be practical -- it appears to be the ultimate torque converter concept that I've been groping around looking for for the last 2-1/2 years.

   For a mechanical torque converter, I've been trying to have something impart 'hits' of torque to an output rotor. (I had good hopes for the latest design.) For a magnetic one, I've been trying to have magnets interact with aluminum rotors, other magnets and pivoting magnets. I found various problems with magnetic "hits" with magnets on magnets, and magnets on an aluminum rotor causes a steady drag that never gets up to vehicle moving torque.
    I've thought at times of various 'out there' ideas like incomplete-circle gears, so that the motor can pick up momentum in the gaps, then impart that momentum to the output in the areas where the teeth mesh. But if the mechanism stops where the teeth engage, the motor won't be able to start turning. That leads to adding a centrifugal clutch. It all promised bizarre, complex mechanical units that would (at best) wear out quickly. (It's how so much of our technology is cobbled together...)

   What finally hit me was the idea of having rotating aluminum blocks rather than a complete round rotor. The rest of the rotation, the motor would spin freely and pick up speed. Since aluminum isn't attracted to magnets, the arms would only be pulled forward, just as the magnets passed by them. This gives the "torque hits" of the mechanical designs, in a smooth magnetic form, with no moving parts except the rotors.
    It can't be 100% efficient because the magnets do slip past the arms, generating some heat. The higher the torque needed, the more the heat. But I expect it would be very good overall. (It would be really hard to make something half as lossy as a fluid torque converter.)

   The strength of the interaction would vary directly with the relative speed between the motor and the output rotor. Since it drops to nothing at zero speed, there's no force to prevent the motor from starting to turn. (Along with making the converter work, this goes a long way to ensure against motor and controller burnouts.)

   If (for example) the output had two opposite arms ending in big blocks of aluminum, and there were magnets at opposite sides of the motor's rotor, the motor could freely pick up speed and momentum for almost 180º of rotation. If it tugged on the arms for about 2 inches of rim distance as the magnets passed the aluminum block, and then had 12 inches of free travel, it might potentially have about 6 to 1 torque increase.

   The idea is amenable to a number of different configurations. First, it's reversible. Whether it's the arms or the magnets that are driven, the coupling effect is the same. The driving side needs some 'flywheel' inertia. On the motor side, this must be added. Going the other way, the inertia is provided by the moving car for regenerative braking.
   Second, it can be done radially or axially. Flat magnet surfaces naturally suggest axial configuration, but with a radial layout it might be easier to get a very small flux gap without danger of the rotors hitting each other. And magnets on the inside of a rim aren't likely to fly off it.
   Then of course the number of arms, the weights, the unit diameter, the width, the magnets and the interaction materials, etcetera can all be adjusted.

   All the parameters of operation can be adjusted in various ways. If 6 to 1 isn't enough torque increase, one could increase the diameter. That would provide more free turning inches before the arm and magnet met again, or in other words reduce the angle where interaction occurred. It would also increase the torque by putting the force farther from the axle. This ups the torque required of the motor as well. Another way would be to have one arm and one magnet instead of two, so that the motor had a full rotation to speed up instead of half. Still another way would be to use narrower magnets and aluminum pieces that would effectively pass by and interact in one inch instead of two.
   Then, the heavier the motor rotors, the less the motor would speed up and slow down in each rotation. At the extreme of a weightless motor shaft, each interaction would just bring the motor down to low speed without imparting momentum to the output. Then it would speed right back up again. At the other extreme, a heavy flywheel would impart energy to the output while hardly slowing at all. It would of course be equally difficult to regain the bit of lost speed. This is closer to what is wanted, but how much weight is needed and desirable? The motor already has a fairly heavy magnet rotor, but at high power this can be accelerated and decelerated rather quickly. A heavy steel magnet drum as the torque converter input would about double it - I'm going to try that as likely to be a good motor flywheel weight.
   The amount of interaction between the arms and the magnets is important. If it's too strong compared to the motor torque, the motor will again be dragged back down to low speed every half rotation, and it won't be able to pick up good speed and store up sufficient flywheel energy between hits. If it's too weak, there might not be enough interaction to start the vehicle moving even with the motor running full speed. In that case, the motor wouldn't be working hard; it might spin rather freely. There has to be a happy medium somewhere. I think I want to see the vehicle start moving at around 500-700 motor RPM on level ground and maybe 1200-1400 up a fair hill.
  I'm not sure how much coupling is useful or desirable. Reducing the interaction between arms and magnets is simple: just increase the flux gap. Or use a smaller aluminum block on the end of the arm. This is what's needed if the motor just can't get up much RPM even at a high 'throttle'.
   But it seems more likely that even with a small flux gap, there just won't be enough interaction at reasonable motor RPM. The output rotor doesn't get vehicle moving torque imparted to it. In that case, other parameters need to be varied.
   Easiest, one can put more magnets at each position for more magnetic poles, with bigger aluminum blocks. If the motor is still over-revving without making enough output force, more arms and magnet positions can be tried: four of each instead of two will double the interaction strength, but will also only allow the motor 1/4 turn to accelerate instead of 1/2 turn. All these steps reduce the maximum torque multiplication.

   Another way is to increase the diameter - if the diameter is doubled, so is the linear speed at the rim at the same RPM, doubling the force. Of course, a new diameter means all new parts.

   Finally there's putting multiple elements at each arm. This would best be done with an axial configuration - interleaved 'fingers' of motor magnets and output rotor aluminum passing each other. Two pieces of aluminum, one on each side of the magnet, would cause twice the interaction. If even that fails, a second magnet and a third aluminum block will further increase it. Any number of magnet and aluminum 'fingers' can be added until either it works as desired or the whole unit becomes too wide to fit where it's wanted. I don't expect it would require such extreme proportions as to be impractical, but possibly it might need more space than I've allowed for in the Sprint mechanism.

   I did a drawing (below) using a radial flux design with a brake drum rotor bought for a previous torque converter design that will fit on an SDS bushing on the motor shaft, and a piece of 5/16" x 5" aluminum bar for the output.

   I wanted to do experiments, and I was thinking of making jigs for them, but I came up with a simple way to test the interactions. I set up the motor's magnet rotor (the motor still being disassembled) in its casing so it could spin on one bearing with the magnet side facing up. I put a piece of white tape on one of the 12 magnets so I could see when it had done one revolution. With C-clamps and a piece of wood, I set up a fat bar of aluminum (.75" x [1" to 2.5" tapered] x 10") to sit right over the magnets on the rotor. I used the fish scale to measure the pull on the bar, and the noisy wall clock, ticking off seconds, to estimate 60 RPM - one rotation per tick. I spun the rotor by hand at roughly that speed and measured the forces.

   I found that for good effect, the aluminum bar had to span about two magnets width (~2.5"), presumably to span the north-south transitions for rapid magnetic field changes as it passed over. The thin end of the somewhat wedge shaped bar, still an inch wide, caused surprisingly little drag. This suggests that two magnets should be placed in each position in the drum rather than one, and the aluminum blocks need to extend a little farther around the arc than I'd expected. And the smaller the flux gap, the more the drag - it needs to be pretty small, eg, preferably .05" rather than .15".

   With the thick (2.5") end of the bar and a smallish gap, the pull was about 2 pounds. It didn't seem like much. But that would be 20 pounds at 600 RPM, and if there are two sets pulling at opposite sides of the rotors, 40 pounds - we'll just round that off to 39. The intended radius being 4", 1/3 of a foot, that would be 13 foot-pounds. 13 foot-pounds times four with the chain reduction is 52 foot-pounds torque to the wheels.
   That's enough to move the Sprint on level ground -- the torque wrench test indicated 30-40 would do it. At 1800 RPM - if the motor will deliver that speed under the load - the calculated 156 foot pounds should overcome most any normal obstacle and get the car moving.

   The conclusion to the experiment is that the forces of interaction appear to be in the right ballpark with the 10" steel rotor on the motor if paired magnets, a somewhat larger block of aluminum than expected, and the smallest flux gap that doesn't cause collisions, are employed. Because copper is more conductive than aluminum (1.72*10-8 ohm-meters vs 2.82*10-8 Ω-m), switching to annealed copper blocks should provide 282/172 = 1.64 times the forces.

   Two conceivable problems could be (a) if there's some electromagnetic saturation level that the interactions hit, preventing the development of force beyond a certain point that's less than what's needed with the intended components, or (b), if the flywheel effect is insufficient and the motor gets slowed down considerably in the interaction zone, limiting forces by reducing the effective speed in the zone. However, I'm optimistic that neither of these conditions will occur with the selected components up to the limits of the required torque. Even if they do, it only means up-sizing the converter and its active elements, and again I'm confident that it wouldn't have to grow beyond a size that makes a practical converter -- certainly smaller and lighter than a vehicle transmission even if it won't fit into the allocated space in the Sprint.
   A third possible problem (c) could be that the torque multiplication isn't high enough for the torque of the motor. It's not overly high unless the interaction zone can be made thinner than I expect. It seems workable but it's not overkill by any means.

Hmm... Where'd the image rez go? You were supposed to be able to read this!
Outside, the 10" O.D. drum rotor rim with magnets, on shaft "A"
Inside, the aluminum 'rotor' and pieces, with a small flux gap to the magnets, on shaft "B"

   About the 22rd I decided axial flux would be simpler. The problem with radial is the the two shafts would have to be pretty much perfectly concentric. I'd have to add a common bearing or something to ensure that. With axial, approximate shaft alignment is good enough as long as the plates are parallel. I could use any one of the rotors I'd had made for motor magnet rotors.
   I cleaned and torched one, sprayed it with zinc, heated it to 225ºc to sinter the zinc (zinc melts at only 420ºc), and then sprayed on a coat of polyurethane for a finish.

   On the 23rd and 24th I started thinking about the limited torque increase the converter would make. If I put 4 magnets on the rotor in the usual places for a motor rotor, that would be 8 magnets not installed, and thus, in simple terms, the motor would be free to accelerate for 2/3 of its travel, for a 3 to 1 torque increase. It would probably be better than that since the maximum force transfer is only achieved where the magnets and arms are aligned, but I'd prefer to see larger figures that would ensure the car can start moving even if it has to climb out of a pothole or over a board or a rock.

   I then started thinking of making the arms on the output have a limited pivot angle, with springs, instead of being fixed. They would pick up a speed and travel (with some slip) along with the magnets on the input rotor, freely at first and then against increasing spring pressure, which would be pushing the output rotor body and the car. If the torque increase required was higher than the maximum spring pressure, they'd reach the end of their travel and strike an end stop - perhaps a rubber pad. Then either they would be brought to an abrupt halt, or the rotor would turn from the sudden force of impact (moving the car), or of course, part of each. After hitting and after the magnets had passed, the arms would spring back to center, ready for the next pass.
   Then I considered whether or not, despite unbalanced forces, to have a single set of elements instead of two - just one position per rotation where force is developed. This gives the motor almost a whole turn to regain its speed after imparting some torque to the output.
   After all the failures I've had, for the first tries I'd rather have excess torque coming too seldom to give good street speed than insufficient torque to move the car.

   On the 28th I did another experiment with some supermagnets and a block of 3/4" thick aluminum. It seems maximum force is generated with opposite magnets either touching each other at the edges or with a small gap. That is to say, one magnet has its north face to the aluminum and the adjacent one its south face. A 3/4" gap between the two magnets considerably reduced the force. For 3/4" thick aluminum, the 1" x .5" x 2" seemed to be a good size - Four .5" x .5" x 2" would probably have had a shallower field depth and less force. But three 1" x .5" x 2" (eg N-S-N) would be almost double two owing to the two field reversals.
   3" long magnets, or two 2" long touching ends for 4", might help a bit, but they and their forces would be getting rather close to the center of the rotor.

   Magnets clamped together on their faces with the edges facing the aluminum delivered very weak force. Presumably, most of the magnetic field lines went straight between the two magnets and had no depth into the aluminum.
   Four 1" x 1" x .5 magnets, tho cut from 1x2 magnets by angle grinder and considerably weakened thereby, had a very good effect with N-S-N-S faces - three field reversals. However, that made the length four inches across, which would only allow a small (2 or 2-1/2 to 1?) torque increase if they covered 4" at opposite ends of the rotor, as so much of the passage would be against drag with little free-spinning angle for the motor.
   A single strong 2" x 2" x .5" magnet, having no field reversals, exerted surprisingly light force on the aluminum.
   A bunch of small magnets, in several different arrangements, also created very light forces. They would have had less depth of field.

   Then I decided that I should probably use copper instead of aluminum for the output side to maximize the force from whatever magnets were used. On December first, finding no large blocks of scrap copper, I bought 4 pounds of bare wire from Ellice Recycling for 11$, thinking of melting it into blocks in my mini-kiln. I was a bit worried about certain aspects of the idea, from getting big air bubbles inside the blocks, to the container cracking and having liquid copper pouring out, to potentially ending up with just copper oxide. Then I decided to check at Smith Bros Foundry. A 1/2" x 3" x 6" block of copper (3 pounds) was "only" 50$, so I bought it.
   To my surprise, the effect wasn't 1.64 times the force of aluminum: it was more like five times. Wow! If you drop the magnets flat onto the copper from an inch or two above, the effect is so great that they slow as if on a cushion of air, and do a soft silent landing. The aluminum then seemed like a cheap, third rate imitation. Now I'm certain the coupling between the rotors will be adequate, and indeed will probably have to be reduced, with a larger flux gap or by cutting the copper pieces smaller than 3" x 3".

   On November 30th I started getting the parts together. The plan is:

* axial flux with two pairs of magnets on opposite ends of the 10" diameter input rotor disk,
* acting on two copper blocks on opposite ends of a single 3" x 10" steel output "arm" (acting as two opposite arms in unison), which will pivot on a center bearing on the shaft.
* The arm will be free to pivot about 45º forward or backward with weaker springs returning it to center
* The ends of the arm will strike two heavy springs at the ends of their travel, compressing them and bouncing back.
* The heavy springs are heavy enough that compressing them provides sufficient torque to turn the part of the output rotor connected to the shaft, starting the car wheels rolling.

(Details "subject to change without notice" as they say. ...The copper is so much more interactive with the magnets than the aluminum.)

   Thus, the magnetic interaction will pull on the pivoting arms and fling them forward. Their weight, hitting the heavy springs, will knock the output rotor ahead, moving the car. Here is a very large torque multiplication. The rotating magnets will continue past the arms as they stop, and the arms will return to center. This repeats each 1/2 revolution of the motor relative to the output. As the output rotor (and the car) speeds up, the hits will become fewer and less forceful as less torque is needed. (Phew!)

A New Electric Hubcap Outboard Project?

   The Honda outboard motor that I converted to electric had a built-in two-to-one gear reduction down at the propeller shaft in the foot. That's fine for a displacement hull boat and 5 or 6 knots, but it's a frustration for a powerboat, because at half the motor RPM, the propeller RPM is too low to get the boat moving very fast. I was also quite surprised by the amount of noise the bottom gear made: the electric outboard was certainly quieter than gas, but it still had that characteristic outboard whine.
   The RPM would be about right at one to one, but an axial flux pancake motor is exactly the wrong shape to try to mount in a sealed pod under the water to direct drive the propeller. For an outboard, it has to be above the water at the top. (An inboard Electric Hubcap motor to a fixed prop shaft should be easy.)

   Having done a chain drive for the Sprint car conversion, it occured to me that an outboard might have a chain drive with small sprocket gears, avoiding the use of meshing tooth gears entirely. This would allow adjusting the gear ratio, and bring construction of the entire drivetrain into the realm of DIY potential. Perhaps the chain drive would have no more losses than the usual 90º gears down at the propeller? (It could hardly make as much noise!)
   This would mean making the whole outboard from scratch. How hard would that really be? - an outboard isn't that complicated, especially an electric one where forward, neutral and reverse are just a switch. Main drive components would be:

1. motor - it would be mounted vertically with the sprocket gear at the rear.
2. cover for motor (Aha! 30cm PVC culvert pipe should just fit over the motor!)
3. pipe for leg. If the sprocket gears were 2" diameter, the pipe might need to be 3".
4. cowling to streamline the pipe/leg
5. cavitation plate
6. bottom housing - 3" pipe "T" fitting, holding: bearings, bottom sprocket gear, prop shaft, some oil, rear seal, and the propeller

Mounting components would include:

7. transom clamp
8. pivoting leg holder with prop insert to hold leg up
9. leg swivel
10. steering handle

   The first thoughts of aluminum or stainless steel pipe bring visions of welding, cost, lots of difficult and time consuming work, and problems with seals. But what about plastic pipe, perhaps with a flat plastic strut front to back, inserted for stiffness? Or as a wide fin glued to the back on the outside? The propeller foot could be attached with - or could simply be - a 90º "T" joint piece, with a dome endcap at the front and reinforcing 'fairings' to the leg. Plastic is easy to cut, the pieces glue together, and sealing... well, it's made to seal. Only where the propeller shaft comes through would the seal be any sort of issue - all the rest below the motor would be glued and sealed.
   A short section (7"?) of turquoise 30cm PVC culvert pipe should just fit over the motor, and it should glue to the other pipes, tho the join would need reinforcement. A plain piece of flat ABS or acrylic plastic could cover the back side - tho it might need to be bolted instead of glued to allow access to the top sprocket and motor shaft. Somewhere above the waterline the pipes should slide to afford chain tension adjustment.

   An inch or two of oil in the bottom would keep the bearings and chain well lubricated - no different from most outboards, and better lubrication than most chain drives get. There would have to be some sort of cover at the top end to keep it from spraying around at the motor end and being used up.

   This should be much lighter than an equivalent gas outboard. At the top, it wouldn't stick out backwards beyond the leg - mainly just to the sides except for the steering handle. The motor would have to be above the clamp and transom in order to have room to turn. (unless it was mounted ~7" behind the transom to miss it when pivoting, in which case it could be placed lower.) It would have a small "nuisance footprint" compared to a gas outboard.

   With high energy density batteries, such a motor could be used on quite a small boat without making it too stern-heavy, and give it lots of oomf and range. Or it could be an extended running unit for a larger displacement boat, even with heavy batteries.

   Another thought, thinking the plastic might not be strong enough, would be to add a shell to the outboard body of tough polypropylene-epoxy, painted over the plastic. Or (or 'and'), an aluminum pipe might fit just inside the plastic one for strength, and extend into the foot to hold the front and rear bearings.
   More thoughts: use 4" ABS for the leg, soften it to "limp" in the oven at 300ºf, and form it into the streamlined teardrop shape. That would at least be stronger than round 3". If that still wasn't strong enough: use 4" aluminum pipe and bash it into the teardrop shape with a maul on an anvil (or a big rock). Of course, this would make aluminum welding of an aluminum foot necessary, and that's out of my DIY department.

   A steel pipe of the right diameter inside would prevent forming the pipe (ABS or aluminum) too small to fit the chain. The top of the pipe above the waterline might be left round as the steering pivot.

   Hmm... that's a far cry from a conversion and having to work around an existing outboard with inappropriate shapes and mechanisms.

   I think I like this! Where is my team of production engineers to work out the details?

LED Lighting Project

Two LED lights hung up in the kitchen: on the right is the original. The strobe effect so common with fluorescents, here interacting with the camera to make dark stripes, is what I've since avoided by adding a filter capacitor or using a 'modern' regulated power adapter. (I do plan to add a filter capacitor to this one - running water in the sink below looks weird.)

The left hand light is the new 1100 lumen "cold white" one. Tho I reduced the exposure, the bluish color only shows at the top where it's not quite so bright, and a bit in the dark strobe areas on the other light.


   I got my big order from Deal Extreme and on the 16th(?) put together a two emitter light with the "20 watt, 1500 lumen, 15000K cold white" emitters. These cost about 40% more than the Cree XM-LAWTs but give 50% more light, using somewhat higher energy. They have an added attraction of being 12 volts, so fixtures with any number of emitters (in parallel) can be run from a 12 volt system. With the usual deratings from the maximums to run cool and get or exceed "energy star" spec performance, two emitters at 7 watts each should be giving about 1100 lumens.
   The 'cold white' light seemed bright and rather bluish, but it's funny how the eyes adapt. After a while of that, all my 'cool white' lights, that I think of as being relatively pure white, seemed yellowish.

Still, 3 and 4 emitter fixtures of this type (~1700 and 2200 lumens, 24 and 32 watts) will need 3 amp power adapters instead of about 1.5 amp, so it doesn't let me order only one type of power adapter.
   Of the several types I've tried so far, these and the Cree "X-Lamps" seem the most suitable. These appear to be made the same as the 32 volt one I bought and used for the shop light, but smaller.
   I found a spool of #26 Ni-Chrome resistance wire at Queale electronics to use instead of buying 5 watt resistors in packs of two. I thought I could cut the pieces a bit long, test them, and add solder along the wire to effectively shorten them to suit. A second objective of using resistance wire is to try and get a positive temperature coefficient of resistance to counteract the strong negative change of forward junction drop of the LEDs with temperature. I find I'm having to put in considerably larger resistances than work well at room temperature, and the lights get brighter and draw more current as they warm up. If the resistances increased substantially with temperature, more of a steady state brightness might be achieved.
   To my chagrin however, I discovered that nichrome wire can't be soldered. I guess the solder won't stick to the chrome. This makes connecting the strong but fine wire difficult. I might just go back to resistors.

Electric bills continue low

   According to my electric bill, from mid October to mid November I used 32 KWH of electricity per day, including a certain amount of electric heat. Last year it was 49 KWH/day. It was 20-30$ less than I anticipated. Taking a median figure of 25$ and figuring the darkest winter months are yet to come, and that there are no more bulbs to replace, that's probably around 300$/year being saved by having mainly LED lights instead of some incandescent (2/3 of them?) and some compact fluorescent (which I found dim and didn't like the light of).
   My investment in various parts and LED emitters for doing the house lights myself (estimating the various lights without actually adding up all the bills) was around 400$.

   I don't leave every light on, but I do have a large house and I don't flip them on and off as I go from room to room, either. The kitchen lights (now around 25W) are often on, and either the machine shop, the livingroom or my office lights, or maybe two of them, will be on depending what I've been doing. The lights in the dark hallways (now 8W and 5W) are usually on.
   Replacing at least lights that are turned on a lot with LEDs is a good investment!

Energy Star

   After a couple of weeks I finally looked into getting Energy Star approval for the LED light fixtures. Then their PDF application form on the web wouldn't open properly. I e-mailed and eventually got sent another copy of the same unusable form.
   Several days went by between each response from Energy Star. Selling LED light fixtures had promise, but the cost seems too high without the Energy Star approval and consequent BC Hydro rebate to purchasers.
   I finally asked for a copy on paper in the mail. While waiting for a reply to that request, I thought to send a friend the file and have them try. It worked fine for them, and they sent me back a copy re-saved using an older PDF program. This worked fine. By then it was the end of the month.
   I should have been onto the application earlier, but I wanted to have at least "pre production" fixtures made and be sure what I was doing first. I didn't expect it to take weeks just to get the forms.

Nickel - Metal hydride Battery Project

Ultimate Battery Stick?

   As I considered mounting the batteries in the Sprint, I thought it would be good to enclose them in one big housing so there weren't a bunch of little pipes with live ends sticking out. But that would be putting batteries in cases into a case.
   Then I thought of making making long thin square ABS plastic boxes to enclose 4 or 8 strings of ten cells, with the contacts on the inside except for two terminal posts.
   Then I thought of using a bigger pipe. A quick mental calculation suggested seven strings might fit nicely into a 4" I.D. pipe. I tried it out, and this proved true.
   At 6.50$ per cell, the cost of each pipe works out to about 475$ (my cost). I preferred last July's sale price of 5$ - they'd be under 400$. (Perhaps I could make and sell the pipes as cases, and let other people buy their own D cells to fill them.)

   With a strut or two to prevent the strings from twisting, one end with a flat metal plate, and one end with seven bolts that could be adjusted for contacts, a 70 amp-hour, 6 or 12 volt battery stick should be easily made. The 12 volt one would weigh about 30 pounds. Three 12 volters should put out plenty of current to run an Electric Hubcap motor, and would fit nicely in the Sprint radiator area.

4" diameter pipe holds the batteries of 7 small pipes:
70 D cells for 70 amp-hours at 12 volts.

   The 2.52 KWH should be good enough for testing and short trips - plenty of amps. But it won't give a car (even a Sprint) a lot of range, so I'll doubtless want to expand on that once it's on the road. Doubling it would be great, but owing to the cost, I might just add a single string or two at a time. Or maybe there's a good pipe size for 70 x AA cell (17.5 AH/210 WH) pipes?

   Come to think of it, the offer to borrow 3 KWH @ 36 volts of lithium ion batteries is doubtless still good. They might fit in under the hood somewhere, with a battery switch to select between them and the NiMHs.

Lead-acid EV battery options

   Other than that, I might put in some lead-acids, with a battery switch to select NiMH, or PbPb if the NiMHs run out of juice. Lead-acids with sodium sulfate should last quite a while if they're only occasionally used, essentially as backup batteries. It looks like there'll still be room for 3 under the hood. Of course, even three 100 amp-hour units would be 135-150 pounds - plus 3 more (PbPb) chargers. That's like lugging around an extra passenger, but valuable when needed. And having to switch gives warning that you're hitting the range limit and need to recharge soon.
   Of course, you won't actually get anything like 100 amp-hours out of them. Derating for PbPb, and at the higher discharge rates of electric driving, suggests about 25 to 30. (How "cheap" is lead-acid when everything's taken into account?)

   Arbitrarily continuing this somewhat off-topic train of thought, if one were to use lead-acid more, or entirely, two banks in parallel would drop the currents in half and each bank would be more productive, perhaps 30 to 35 amp-hours for a total of 60 to 70, thus 270 to 300 pounds yielding similarly to the 70 AH (probably should derate the NiMHs to 60 for high currents as well?) from the 90 pounds of NiMH dry cells.

   Six 6V golf cart batteries should do somewhat better than the six 100AH/12V cells, maybe 80 to 90 AH... but they weigh and cost proportionately more, 370 pounds and 960$. Three parallel banks of 100AH/12V (405-450 pounds) should yield still better results (with the currents being divided by three), 35 to 40 AH each, 105 to 120 AH total... and also make the Sprint heavier than it was with the gas engine system in it.

   Two 70 AH banks of NiMH D cell pipes would be at least as much available energy as that, much longer lasting, and just 180 pounds - reminding us why lead-acid isn't a very popular EV choice. They would also cost 3000$ instead of 1000$ - reminding us why the owners of big oil should be relieved of their strangleholds over better economical battery types.

   Yet another - and simpler - way to use lead-acids to get sufficient batteries with less expense would be to use them in series with NiMHs. Charging each 12 volt section independently allows this. One would want the lead-acids to have at least as many effective amp-hours as the NiMHs if they are to last at all. One could put in (eg) two NiMH "Super Battery Sticks" of 70 amp-hours for 24 volts, then finish with two 100 amp-hour lead-acids in parallel for the third 12 volts, making 36V, 70 AH and 150-160 pounds. Or if three NiMH supersticks were used initially as I'm planning, one could later add one more (paralleling two pairs for 24 volts) and use four lead-acids in the last position to get double the above energy and almost double the range.

   Still better of course would be to get my own economical chemistry batteries working to the point where I could use them in the car.

Turquoise Battery Project

Making Zinc Powder

   I was sure using some zinc metal powder rather than just starting with yellowish zinc oxide would make a better electrode. I can't seem to find it in town. I can't think who might have it. I tried ordering some on the 6th, but the e-bay seller wasn't sure he could ship it outside the USA. I e-mailed two potential Canadian sources but got no replies. Sigh!
   I tried grinding some from a zinc rod I had. It was very slow going. At one point (foto) I made a block to hold the zinc rod, but held like that it quickly got too hot. By the 10th (after several 10-15 minute grinding sessions) I eventually decided I had enough powder/flakes for an electrode, but it was more flakes than powder - the powder would build up on the grinding wheel and then flake off. Doubtless the particles comprising the flakes were all fused - melted - together. I was dubious about using it.

Zinc from zinc oxide

    Then I got the idea to try and reduce some zinc oxide to zinc powder electrically. I got a round "basket" of perforated brass sheet from an older experiment.
   But wouldn't it just electroplate the brass with zinc? I mixed 60g ZnO and 25g graphite powder to keep the zinc from conglomerating into a solid piece or plating. About half of it tamped into the basket.
   Knowing salt electrolyte would corrode the basket, I used potassium hydroxide for the electrolyte, and I used a piece of graphite sheet for the positrode.
   I attached a power adapter and got 2.3 volts with a current of .27 amps. Since there was no real positive electrode to oxidize, the oxygen from the zinc oxide should bubble up from the graphite sheet. Soon it was in fact bubbling merrily with small, rapid bubbles.
   If I had used a nickel or grafpoxy plated positive electrode with nickel hydroxide, I'd have made a nickel-zinc alkaline battery. Many have tried to get good cycle life out of such a cell. I only needed it to charge once.
   Unfortunately, I left it a few minutes, and when I came back, the substance in the basket had frothed up and spilled all over. Of course! - 2.3 volts was too high and the zinc filled negatrode would be generating hydrogen. The trapped gas puffed up my nicely compacted basket of zinc/graphite. I limited the current to reduce the voltage to about 1.5 or less, but by then it was a mess in a tub of caustic hydroxide. It did however seem to be charging.

   I read over means of making zinc electrodes in Alkaline Storage Batteries by Falk and Salkind. I noted that they were usually immersed in a tank of electrolyte, left to sit for a day, and then charged against a 'dummy' positive electrode, and maybe run 3 charge-discharge cycles. Soluble impurities would leach out of them. Then they were dried. It occurred to me that even if I was making salty batteries, I could use the same procedure. The mobility of zinc ions when cycling in alkaline solution should be an advantage for a few cycles, forming conductive zinc pathways within the electrode to improve the current capacity.
   I also noted that zinc electrodes were said to be very fragile before 'forming' with initial charges and discharges - exactly as I had discovered. Another possible means of making them was by sintering zinc oxide at around 750ºc for 20-80 minutes (I presume with a collector screen enclosed). Those would be more solid, but might be less amp-hours by weight. Seemingly, it would be an easy thing to try out - simple ingredients! But for use in salty electrolyte, the collector screen needs to be grafpoxy coated. That would burn up at 750º. Not so simple after all! Zinc metal powder sinters at kitchen oven temperatures (over 200ºc), as I had found doing the 'ersatz powder coating'. That might make nice pre-charged electrodes, but it's still too hot for the grafpoxy.
   Another thing mentioned was 1-4% mercury oxide to raise the hydrogen overvoltage. No thanks. I'll stick with the environmentally friendly antimony oxide, and maybe the eggwhite, maybe a rare earth. Anyway, it should be less prone to hydrogen generation in salt solution - after all, standard dry cells stay charged for years with nothing but zinc sheet metal... unless that's actually an alloy with some small additive for overvoltage.

   On the 27th I took the remaining zinc oxide and graphite and tried again, adding a bit of antimony oxide to the mix. This time I limited the current, and the voltage started a slow rise from 1/2 a volt. On a whim I replaced the graphite "+" with a flat nickel pocket electrode from a nickel-iron pocket cell. A Ni-Zn alkaline cell should charge up to about 1.75 volts, without the oxygen bubbles unless the nickel became completely charged. In fact, after 3 hours or so the zinc was making dirty hydrogen bubbles with the cell at 1.72 volts. The zinc couldn't possibly all be charged so soon. Of course!, I had just tamped it into the basket: it wasn't properly compacted, so resistances away from the basket metal were bound to be high, allowing some zinc to charge fully without getting to the rest.
   I tried shorting the cell, twice. It put out just over an amp, which quickly dropped off. Not bad, considering the poor interface between electrodes, and all the work I go to to get an amp out of a salty cell! No wonder everyone went to alkaline cells and abandoned the the harder job of trying to make rechargeable salty cells.
   I decreased the current to eliminate the bubbling, and went over things from the battery book... oh ya, 'immersed and left to sit for a day'. I removed the charge.
   Even 2 or 3 hours later, when the voltage had dropped to 1.1 volts, it still put out over half an amp when shorted.

   Another item to consider was that afterwards I'd have to dig the mix out of the basket and properly compact it into the desired electrode form. Why wouldn't I make the desired electrode first, then 'form' and charge it in the alkaline solution, and then simply rinse it and use it in the salty battery? The zinc should migrate to form great dendrite 'tentacle' connections during forming in the alkali, and they'd become fixed in place in the salt battery with every prospect for a super long life, highly conductive electrode. I could make some sort of little basket or compartment so the electrode could be inserted and removed with little stress.

   All this seemed very good. But I was still wondering why I couldn't seem to buy something so simple as zinc powder. However, I did a bit more research, and found that something I was doing was making it into better oxide than it started as. Not what I thought, tho...

Best Zinc Electrode!

   Back to the basics... it seems that not only zinc powder, but zinc oxide powder will absorb carbon dioxide from the air and form zinc carbonate. Zinc carbonate is passive in the battery. Perhaps this is partly why results from my first zinc powder electrode didn't seem so great? One can heat zinc carbonate to produce "calcined" zinc oxide and CO2, but it seems that for batteries there's a better way.
   Simply putting zinc carbonate into alkaline solution causes it to become fine, high surface area zinc oxide. This is sometimes termed 'active' zinc oxide. 'Fluffy' high surface area chemical is ideal for batteries, yielding the closest to the theoretical amp-hours of the substance.

"Active" high surface area zinc oxide

   So all I have to do to get an ideal zinc battery electrode is to make the electrode, immerse it in potassium hydroxide solution for a while (like the alkaline battery makers), then rinse it off - probably no electrolyzing or anything. And then - presumably - no dendrites in salt solution to deteriorate the electrode and limit the battery's cycle life. It appears I've stumbled onto the best form of zinc negatrodes!

   It looks like the only thing that might be a better electrode is if the hydrogen overvoltage with manganese can be raised that extra .3 volts so that a manganese negatrode works properly. The eggwhite didn't seem to do it. I think that's worth more research, but I'll go with the zinc because it's well known to work well as is.

Victoria BC