Turquoise Energy News #133
covering June 2019 (Posted July 3rd 2019)
Lawnhill BC Canada
by Craig Carmichael

www.TurquoiseEnergy.com = www.ElectricCaik.com = www.ElectricHubcap.com

Feature: "No Dendrites!" Coated Zinc Electrodes
for Long Life Nickel-Zinc or Manganese-Zinc Batteries
 (See Month in Brief, Electricity Storage)

Month In "Brief" (Project Summaries etc.)
 - "New chemistry" battery Breakthough and Progress - Beaten to it!: an "Ultra Efficient" Electric Car! - the Lightyear - Workable Unipolar BLDC Motor Concepts (at long last)! - Better Variable Transmission Concept

In Passing (Miscellaneous topics, editorial comments & opinionated rants)
 - World Unity - Witch Hazel & Red Spots - "Rivers from the Sky"? - ESD... egg whites for burns?

- Detailed Project Reports -
Electric Transport - Electric Hubcap Motor Systems
* Ground Effect Vehicle ("GEV") (scant progress)
* More Unipolar BLDC Motor Ideas
* Jim Harrington's Latest Electric Outboard Motor

Other "Green" Electric Equipment Projects
* 36 Volt DC Wiring & Infrastructure
* Laser Engraver

Electricity Generation
* Solar Car Charging Trailer(!)
* My Solar Power System: Monthly Solar Production log et cetera

Electricity Storage - Turquoise Battery Project (Mn-Zn, Ni-Zn or Pb-Zn in Methyl Hydroxide electrolyte)
* Electrode "Pocket" with Nafion Ion-Selective Membrane Face - What Next? - Making Barium Hydroxide? - Nafion Delamination - Film-On-Electrode? - Lead doesn't work in alkali! - Nickel-Zinc in KOH - Some Conclusions... and some Dissolved Oxygen? - Nafion Versus Osmium Doped Film - The Filmed Zinc Electrode - Next: jelling the electrode... egg albumen?

June in Brief

   Where do the days go? Increasingly it just seemed impossible to get anything done on projects. I don't have the drive any more for 3 work sessions a day, and the one or two that I do are somehow consumed with other things. Then (having finally planted it all) there's watering the garden and greenhouse. And just driving around. Where does the rest go? For a few days I started writing it down ("Diary" below). It seems most days I actually wasn't idle after all. Then I seemed to start finding times here and there.

   June was literally a very dull month. By the 20th there had only been a total of maybe an hour or two of bright sunlight, on 3 or 4 days when the overcast had a few gaps in it that happened to cross the sun. And a couple of clear nights. There was some rain, but not a lot. Mostly just clouds and overcast. It didn't do the solar power any good as the daily readings show. (see Electricity Generation) And, how are crops supposed to grow without light? I was almost ready to start up the "indoor LED garden" - in the summer. Finally from the 20th on we got some sun - even 3 or 4 whole sunny days.

   But soon June did become especially interesting because I did manage a few battery experiments, one of which looks like a breakthrough - a real winner. Whatever possessed me to keep trying new things after 11-1/2 years of not coming up with a practical cell, I'm not sure. But this time there was a good result, and the focus will shift to how to hopefully improve, optimize and maybe even produce the fabulous new cell. It's not that there aren't probably other good combos of things that might work, but having got an excellent one to actually work well, it'll do just fine!

A Breakthrough?

   My mother thought I should patent it. I explained that the chief use of patents is to scam them off the inventor by business ruses or if necessary to buy them up, to kill new technologies so they won't compete with existing products that the invention makes obsolete. I don't want that no matter what I might be paid. Or else, a big company will see the patent, use the invention, and not tell the inventor. After you find it's in the big box stores, no one will help you try to collect anything from the big corporation that has more money than they'll ever need to stall and fight you forever in court - partly from selling the product you invented.
   Then she thought I shouldn't put it on line; that I shouldn't put it in the newsletter. As with the majority of inventors, my number one priority in inventing a better product is that it be used, that it will in some way contribute to a better future. What's the point of spending years on a project, succeeding, and then hiding it under a pillow? Those who think most inventors' primary motivation is financial are mistaken. Getting a good sum of money is always in the mind because everyone needs money to live, and they've sacrificed time and energy to get to where they are and new inventions don't happen every day. Obviously any sense of fair play says a successful inventor should be fairly and generously compensated for risking years of his time - that may yield nothing valuable - to make the world a better place. Unfortunately there is no mechanism for that in today's society and over 99% of inventors never do well off their inventions. Even Nicola Tesla who (among several other things of note) invented the electrical machinery to create the whole ubiquitous power grid we all enjoy, and the principles and techniques of radio that his one-time apprentice(?) associate(?) Marconi used to send the first wireless signals across the Atlantic, died, and mostly lived, in poverty.

   Then she thought I should start making and selling them. That I hope to do. A fellow who moved up here about the same time I did once ran the first recycling business in Calgary where he created various production jigs and tools. He is very interested in solar energy and electric transport, and in helping and seeing these batteries being produced. The biggest challenge is to automate each process so the batteries don't each take hours of labor to make.

   My other thought is that should some person of means ever take an interest in my work or in any one of my creations, it will be because my works get known and come to his/her attention. I promote my work - whether badly or well - by publishing everything I do in this newsletter, "open source" as it were. When I come up with something the entire process of getting there is well documented. It seems to me such a person is most likely to come to the one who's been doing it all along as being the person most likely and most motivated to be able to use the funds most effectively to take the same products from design or prototype stage to production.
   While I have uncovered a key, I also have in mind further important improvements before producing batteries. Would a potential investor be wise to invest instead in someone else, or in some company, that hasn't been thinking through the various aspects for years? Well, maybe if they're in China... Either way, let's have better batteries become available. EVERYBODY has wanted really good, lower cost batteries, for SO LONG!

Very Long Life Nickel-Zinc Batteries!

   I wanted to try out the nafion "ion selective membrane". I tried it with the methyl hydroxide electrolyte, with a lead-zinc cell. It didn't seem to work. And the nafion delaminated into 3 thinner sheets. Apparently it didn't like alkaline! When I did some re-reading I realized nafion allows "+" "cations" to pass but not "-" "anions". Wasn't that really the opposite to the desired effect in alkaline solution, and to block zinc "Zn++" ions from migrating?

   I tried it without the nafion, as a straight lead-zinc cell. It still didn't work. I thought the electrolyte must be at fault, so I tried potassium hydroxide (KOH). Contrary to all expectations that didn't work either! I discovered it was actually lead dioxide (or lead tetrahydroxide) that doesn't work in an alkaline environment. It dissolves or sheds off as particles into the electrolyte. When it touches the zinc, it turns to metallic lead and plates itself coarsely on. How blind I've been! I had seen crud on the zinc in previous experiments but I didn't catch on to what it was or what was happening.
   Since I was using KOH anyway I broke open a dead NiMH "D" dry cell and stuck in a convenient piece of the nickel hydroxide electrode for a positive. (And I almost took the trash with dozens of them in it out to the road for pickup, just days previously!) It worked but the (also new) zinc electrode collected more lead crud. And it had my usual much-too-high self discharge. At least it worked!

   If the nafion was an ion selective membrane intended for an acid environment, what I needed was an ion selective membrane that would work in an alkaline environment: The osmium doped film. But what to paint it onto/into? Cellophane maybe. But again unless whatever it was was very well sealed around the edges, the zinc ions would get past it and build dendrites.
   Maybe wrap the cellophane all around the zinc electrode? Then from there I got the idea to coat the zinc electrode plate itself with the osmium doped coating. Would that work? It's implicit in battery design that the electrolyte has to be in contact with the electrode for it to work. ...But if it worked right the film should pass "OH-" ions directly to and from the zinc. And if the electrode isn't in contact with the electrolyte, it can't dissolve into it. I've said it before: a zinc electrode that doesn't degrade is the "holy grail" of battery making. It would change the whole ballgame. Here might be a wholly new means of accomplishing that. It was worth a try. (and I still don't understand what if anything was wrong with the previous "moderately alkaline" pH technique!)

   I made a new small cell (in an old ABS plastic case with a new, better fitting lid) with fresh (KOH) electrolyte. Again I used [some other] chunks of the "D" cell for the nickel oxides side, with a piece of cupro-nickel sheet as a current collector plate. I painted the osmium doped film on the zinc with a small brush.
   The cell worked! And it had very good current for the size of the electrodes. So either the coating worked great, or it had dissolved in the caustic alkali. Furthermore the self discharge was quite low - probably the good fitting lid was keeping fresh oxygen out.
   Next the question was what would the cycle life be? Would the zinc grow dendrites and short the cell as usual, or did the film keep it out of the electrolyte entirely?
   Seven charge-discharge cycles worked without dendrites shorting the cell. That was more than I ever got before. After the fifth and eighth charge I opened the cell and inspected the separator grill and the zinc electrode under a microscope. No dendrites! And the cell continued to work.

   It looked like a winner! And after all these years and all this research, the only thing different from so many other short-lived or relatively short lived nickel-zinc cells was the coating on the zinc - which I had formulated a decade ago but not put to its best use! I may still experiment occasionally with milder electrolyte and different metals. There may be a number of things that will work well. But even if nothing comes of anything else I try, this seemed to be an excellent cell!
   But the zinc, while not forming dendrites, still degraded. Eventually the first one disintegrated at the water line. Next will be figuring out what to use to jell the zinc electrode to prevent that, to make it "permanent".

   I took many pictures of the several surfaces under the microscope during making, before use, and after being used, to the point that when I came to use them, I had lost track of what many of them were.

A "fluffy" electroplating of porous zinc on the zinc electrode to give it
more substance and surface area came out much thicker near the edges
than in the middle, so improvements are needed to the plating process.

   Once the design - every aspect of each component of the cell - is optimized, fast and automated manufacturing techniques and production will become the main focus.

   But in spite of the progress, somewhere heading toward the end of the month I must have "burned out" on chemistry & materials sciences. I did some studying, but didn't get any more actual work done.


   Of course one continually learns more about topics previously mentioned. I wondered in a recent issue why it seemed to take more energy to recharge the car than the indicator said it had used. One reason: it was mentioned in a video that recharging a car with a lithium battery is only 90 even down to 80 % efficient. Lower charge efficiency would account for a good share of the discrepancy.


   There was something odd with the solar panel system I had put in for the off-grid lady across the road, which was supposed to have been finished in May. When I first tried it, the 1500 watt inverter had run her 1000 watt 'shop vac' vacuum cleaner. Then it wouldn't. I attributed it to the initial overcharging of the 'frame 27' lead acid battery I had given her (owing to the charge controller unexpectedly being "positive ground"). I finally brought over a four 100 amp-hour lithium cells battery, a spare from the old Suzuki Swift EV. It still wouldn't run! It was all the same: turn it on and after about 2 seconds the low voltage alarm would go off on the inverter. In a couple more seconds the 100 amp breaker I had put in for safety would blow. I realized the #10 wires I had used, tho quite short (2 feet), were pretty light for such a heavy load and replaced them with #6. No change! I finally thought: it worked the first day.... what?... until I had installed the safety breaker. I took it out and taped the wires together. Sure enough, it worked fine! The 100 amp circuit breaker itself was high resistance. I found the in-line battery fuse from the old Mazda RX7-EV and installed it where the breaker had been, and sold her the spare lithium battery, which would doubtless run the vacuum cleaner longer in a session, and last longer in years, than the old "27" lead-acid. (notwithstanding that I had put sodium sulfate in the "27" in 2016 to make it last longer. I got it as a scrap battery about 4 years ago and it's still fine.)


   Toward the end of the month, on the 24th, I returned to milling my spruce after neglecting it all winter and spring when it was cold and the days were short. I had purchased an Alaska mill off Amazon for my big chainsaw. (one might say belatedly - previously I had borrowed one.) After much assembly, setup and chainsaw issues (putting on the very dull chain that had hit a screw... on backward at first, a terminally dull chain sharpening file, mixing gasoline...) I sliced into one of the smaller logs now lying there almost two years and made a couple of slabs. It was hot, loud, stinky and exhausting. I had to repeatedly stop to step back and breathe some clean air. Oh well, just a few chainsaw cuts to make it into big cants, and then I could use the fine electric handheld bandmill I had invented last year [TE News issues of 2018] to make lumber. It made eleven 12 foot 2"x6"s and one 2-1/2" thick slab.
   As I worked, woodbugs and termites and millipedes were trying to find new places to get back under the bark. The wood was spalted. And it had cracks (but they were likely formed when it was felled). Yes, it was high time to get the rest of the spruce milled up. Another winter and it might not be worth milling.

Wiring the HAT35V-50A Ceramic socket            
   I could see the ground effect vehicle, sitting there waiting all month, wasn't going to get its wing in June! And I fired the ceramic HAT36V-50A socket in the kiln and wired it, but I hadn't even made the HAT36V-50A plug to plug the kitchen water tank into it.

(Shoulduna used that camera for a close-up!)

   The idea of putting up more solar panels remained just an idea even tho I had the panels. I became aware of some obscure but important safety considerations about HE ray energy and set that aside while I consider whether or how to continue. Still, successful new chemistry battery results atone for much lack of progress in other areas.

   But in a late night work session on the 27th I at least cut foam ribs for the 'GEV's wing. (At the back they end where the steerable elevator flap begins.) And the powerful 90mm ducted fans ("electric jets") had arrived - but not the "ESC" motor controllers for them.

Beaten to an "Ultra Efficient" Electric Car! - the Lightyear One

   On youtube and in news articles on the web I found the "Lightyear One", a Dutch prototype for an ultra-efficient (their term as well as mine) production vehicle complete with built-in solar panels, by the same team that had won the cross-Australia solar powered car race a few years ago against some big-name competition ("Honda" was mentioned). In the video it went for its very first low speed drive. I saw them pushing it backward - maybe like my Sprint in its current configuration it doesn't back up yet!
   It is expected to use 2/3 of the energy of a Tesla (or Leaf, etc.) to go the same distance - a similar figure to what I was trying to achieve for in-town driving. A big difference was that theirs was designed and made with the efficiency goal in mind from the ground up, whereas I simply had a car conversion in mind. They built it light, incorporated the best solar cells right into the roof, hood and hatchback, and made it extremely low wind drag. (As good as the GM EV-1?) So its highway efficiency would no doubt be highest, and doubtless better than a conversion could achieve. (But the Sprint was pretty much the lightest mass-produced car, and not so much wind drag either. I wonder if the original Suzuki equipment for manufacturing the Swifts/Sprints/Fireflys is all rotting in some junk yard(s) somewhere? But I suppose resurrecting that assembly line is a ridiculous idea!)

   Perhaps I should just admit someone has (at long last) beat me to it and give up. OTOH it's a prototype and they may not get it into production. If they do it will be a very costly car. I couldn't find any info on the four in-wheel motors, but somehow it got me thinking about ultra-efficient motors, controllers and transmissions again for a few days. On July 1st I came up with inspirations for better solutions for both unipolar BLDC motors and variable transmissions. They must have been long festering in my subconscious mind to both pop up like that on the same day!

Bonnet, lid and liftback are all solar panels, all cells
independently contributing even if others are in shade.
(1.5 KW total solar? IIRC)
Solar on a car using many thousands of watts may sound almost trivial, but it's always
charging and the makers estimate as high as 10,000 to 20,000 kilometers free travel per year
 owing to the solar recharging. (the higher figure is if you live in a cloudless desert at lower latitude)

I suspect money helps get the layout and cosmetic details looked after.
(What must all these custom components have cost?)

Unipolar BLDC Motor Concept

   Any lingering thoughts of maybe getting the newsletter out on Canada Day (July 1st) were dashed when that morning I got to the "Unipolar BLDC Motors" idea section (Electric Transport, below), and started realizing exciting new possibilities and exploring them "on paper" as it were.
   I hadn't been able to work out a unipolar magnet rotor and stator design that would work properly before, but now I discovered that if one went from "typical" 3-phase motor design to more phases and activating coils individually instead of in pairs on opposite sides of the stator, certain configurations appeared to work. If one could be made to work well, it should be worth doing. Of course a custom motor controller would be required too.

   Even if I still came up with nothing I could manage to build for a variable transmission, with two such motors of maybe 8, 10 or 12 coils, each driving one front wheel, together they would probably have enough torque for fixed 3 to 1 ratio chains, belts or gears, allowing both sufficient startup torque and highway speeds at safe motor RPM.s.
   Perhaps that would be at least as worthy of pursuit as the reluctance motor, given the "ultra-efficiency" of BLDC motors and their high mechanical tolerance for small dimensional variations. And I've already made good BLDC motors and controllers that at least work.

Variable Transmission Concept

   Then toward evening (still July 1st) inspiration struck for a better variable transmission, too. This torque converter would combine two previously used components: the planetary gear and the centrifugal clutch.
   What the planetary gear - or any variable ratio coupling component - lacked was a means to control the third "control" element to variably change the ratio between the other two elements coupling the motor (ring gear) to the wheels (planets assembly). (That was where I had tried using a rope around a large pulley. That wasted energy. But it worked with a flywheel to transfer revved up motor energy into car motion and get the car moving.)

   A centrifugal clutch unit such as I had already made could provide the means to control the control element: the faster the motor (and centrifugal shoes) turned, the more stongly the centrifugal drum would want to turn with the motor rather than just turn opposite to it. By having the control element be the one that spun fastest with the least torque (sun gear), the forces to slow the drum would be maximized, which would provide highest torque to turn the output element with the least torque on the drum.
   Free spinning with the car stopped, the centrifugal outer drum on the sun gear would turn around twice as fast as the motor turned the ring gear with the centrifugal shoes on it. And it would take 1/2 the torque to slow or stop it. (Ratio depending on the actual planetary gear chosen.) So if the motor with the centrifugal shoes turned 1500-2500 RPM, the drum if it was free-spinning would turn 3000-5000 RPM in the opposite direction, and the output gear wouldn't move. But hopefully somewhere in that range we're looking at speeds that will sufficiently engage the centrifugal coupling. As the shoes engage the outer rim, the set would have a very large tendency to want to slow the drum, which will turn the planets assembly and drive the car ahead, slowly at first with highest torque and then graudally increased speed and reduced torque.
   While the ratio of the input (ring) to output (planets) with the sun gear (drum) held stopped might be 1.5 to 1, the sun gear goes from freely spinning before the centrifuge engages, to being loacked and turning with the shoes and so all three gears turn in unison, 1 to 1.

On the other hand (July 3rd - I'm trying to finish up the newsletter but I can't stop thinking about it!)... a better way to do it would be to have the motor drive the planet gears assembly, with the output on the ring gear instead of the other way around. The centrifugal clutch drum is still on the sun gear. That way, with the car (ring gear) stopped, the drum wants to speed up to 3 times the motor speed instead of 2 times, with only 1/3 as much torque required to bring it down to motor speed. More importantly the drum also spins in the same direction as the motor. That means that as the centrifugal shoes contact the drum to slow it down, the rotating force tries to make the motor speed up instead of slow down. The inertia of the spinning drum works for us instead of against us. This is something I have been trying to figure out how to achieve since 2012 too.
   In theory if the drum were stopped the output/ring gear/car would go backward at 1.5 times the speed of the motor. In fact, the drum will never go slower than the motor, when it is fully locked to it, and all three gears will turn in 1 to 1 unison as before.

   Another possibility is to use, say, an idler wheel and a belt between the ring gear (motor) and sun gear (control element). Pulling the idler wheel tighter couples the two more strongly, so the driver would have a stick clutch or something to do that. (That's close to something I tried before, but applied a little differently.)

   So there it is! The two components together, the planetary gear to allow variable ratios and the centrifugal (or other) clutch as the ratio control element, should make a fine variable torque converter. It seems it has taken me a decade to figure this one out, too. I started on a variable torque converter in spring 2009. This one should be more satisfactory than any previous idea.)

   The variable reduction output would feed a further fixed reduction to the differential, eg, 3 to 1, to drive the wheels.


   I am now impinging on July, so it's time to wrap it up! I'll call it June 31st, 32nd and 33rd. (Hah! - that means I got more vegetables planted by the end of June after all.) (January to June only has 181 days anyway, while July to December has 184. Like so many things on this planet, the calendar isn't the first or last that 'just grew', piecemeal, instead of being logically designed.)

Early June Diary: where does my time go?

1st to 3rd - Add photos/edit/finish up of TE News #132, and wrapping up the solar installation across the road.
4th ??? Some other writing...???
5th - Another disappointing new chemie battery experiment.
6th - Fixed lawn tractor. It broke down at the far end of the field and wouldn't move. I had to try and see what to do underneath in 15" grass, then took a trip to town for the part. (Gasp! They had it in stock!)
7th - Writing up an article for Haida Gwaii Trader magazine for July-August issue. (I said I'd do one every 2 months. It didn't seem onerous when I volunteered. -- The last one was an update on the completed Handheld Bandsaw Mill. Two others have been on aspects of Planetary Management/Social sustainability.)
8th - Picked up the new crate of 305 watt solar panels in two trips to Masset - 4 hours just driving. Car and trailer were heavily loaded to bring 16 each time.
9th - Someone phoned me before I was out of bed and yakked for almost 2 hours. I planted all the asparagus I wanted, then phoned someone to see if they wanted more of the plants. They invited me to see their shop-building project and then dinner. Well it was Sunday anyway!
10th - Finally I just HAD to put an ad in the Trader for solar stuff. Why else had I brought solar panels et al up to this island and then designed the HAT36V wiring and other infrastructure stuff? Well, I had too much to fit in a classified ad so I had to update the price list and make it HTML and put it on my web site. The link TurquoiseEnergy.com/SolarStuffPriceList.html only took me to the site, not to the page. So I had to make a link from the main page menu. The main page was way out of date so I spent some time editing that. Wrote up two years "null" tax returns for Turquoise Energy Ltd. I haven't got any R & D money out of them since 2015 anyway. Nap time. But in the evening I finally fixed my mini kiln. (For once, a job that proved to be easy and went smoothly!) Then I fired the ceramic HAT36V-50A socket.
11th - Made copper spring pins and wired the socket with short #8 wires. (But I didn't get to doing a plug.)

[12th - enough of this!]

In Passing
(Miscellaneous topics, editorial comments & opinionated rants)

World Unity

   In times past, transportation and communication were such that mainly it was close neighbors that interacted, in trade and war. Even in 1700 it took over a week to get from London to Scotland by horse carriage. Egyptian, Greek and Roman civilizations centered around the Mediterranean Sea because boat travel was better than land travel. These peoples conquered those around them to exploit the resources and make slaves of the conquered lands. There was no thought of the conquered people having a culture and society that might have value which was being destroyed, or that the people there were spiritually equal with those who managed to conquer them.
   Later mostly Europeans set sail in ships and conquered around the world to exploit and enslave, to make an empire. The British empire sought also to educate and left their forms of law and government behind, but the primary motivation was still domination and economic advantage. If there was to be unity, it was by the conquered being forcibly united with the conquerors.

   In the last 500 years the world has been entirely occupied and settled. The attempts of one culture to dominate other more or less equal cultures was rife throughout Europe and later North America over the last millennium or so. Mostly little consideration was given to the conquered. Then ending only a decade before I was born, the whole world passed through the tremendous convulsion of world war two, a clash of wills between several ideologies embodying mostly domination and enslavement of peoples - starting with one's own country's peoples.

   When I was young, countries were still much more separate and culturally isolated from one another. The so-called "United Nations" had been formed, but "the international community" wasn't a part of speech and no one spoke much of "international law". Besides Europe, other continents were mystical far-off lands that were (surely) of little concern to us in North America.

"We all drink the same water. We all breathe the same air." - John F. Kennedy

   Trade and commerce started to change the world. Suddenly everyone everywhere wanted to learn English as a means to be able to communicate worldwide, to do business and open up their markets. I'm not extolling English for its own sake, but it was the first time everyone had, implicitly, agreed on one particular language as the means to enable international communication. Strange people speaking strange tongues were suddenly no longer so strange or mysterious when they spoke in your language, or at least one you too had learned.

   We are politically separate, but through extensive trade the world has largely become one economy. While we will retain our cultural and racial identities for ages, the world will continue to become more and more equal, more and more "all one people" as international and intercontinental relationships are viewed. Wars, refugee migrations and the fierceness of competition will be in the past when the population has dropped (and is thereafter managed to maintain stability) and there is material prosperity for everyone.

Witch Hazel & Red Spots

   I had been been getting red spots on my upper legs for some time, which were proving intractable/chronic. They seem to be some kind of mite - perhaps house dust mites? probably not even a millimeter long. I have stabbed some with a needle and the spot goes away, but more often I stab myself. I don't recommend it. I have recognized for some time that they occur when something is pressed hard against my leg, as in carrying in a pail of firewood or crossing my legs and one is pressed against the underside of the desk. The area pressed is exactly where they occur. Apparently they can't penetrate the skin unless helped. But then they seem to spread.

   Antibiotic cream sort of helped. Daily baths sort of helped. The witch hazel, applied lightly (hard rub bad), got rid of them. And if a new spot should appear, a drop of it gets rid of it. And Wikipedia says there's no proven medical use for witch hazel. Hah!

"Rivers from the Sky"?

   Here are a couple of good photos I ran across in a video, showing what is said to be the "rivers from the sky" rainfall that has never been seen until these last few years, which are now causing such devastating flash flooding around the world.

   To see each week or two's growing crescendo of calamities from all around the world, go to "World of Signs" and "Nared King".

Calgary, Alberta


(Eccentric Silliness Department)

* I searched on eggwhite as something to put on burns, having heard of it and noticed that it wasn't mentioned in the Wikipedia article on albumen. There were some hits where it was obvious that it wasn't wholly in favor in the medical community. A lady on youtube said, with reasonable arguments, that it was an urban myth and that eggwhite was no good for burns - and it might cause a salmonella infection.
   But there were exactly 21 comments under the video, 17 from people who had used it on various burns mild to severe including sunburn that said she didn't know what she was talking about, that there was nothing better for burns than eggwhite. Not one of the commenters supported her. One person said eggwhite has the amino acids glycine and proline and that these give the body the building blocks to build collagen when eaten, and that it seemed to work well applied on burns too.

* The issue reminds me of "planting by the phases of the moon." This doesn't fit with scientists' narrow logic, so they are quick to call it "superstition". But apparently any farmers' almanac discloses that there is statistically more rainfall at certain phases of the moon than between them. (IIRC it was more around full moon and new moon.) The moon makes atmospheric tides as well as oceanic. Depending on climate, if you don't have running water for the crop or "anytime" irrigation it could be important or at least helpful to plant according to the phases of the moon.

* They say there's such a thing as "50-50" chance. When I open my microwave oven door, the coffee cup handle is almost invariably rotated around the back somewhere. It's been that way at least for a year or two. But the last couple of weeks sometimes it's been at the front or at the side. That's really spooky to find the coffee cup handle at the front when I open the door! (Is the rotator platter in my microwave wearing out? or am I just selecting the right amounts of time sometimes now, instead of invariably the wrong lengths?)

* A store owner tries to guess what to order that he hopes his customers will want to buy. With some items it's like throwing a dice - people may buy lots, or none at all. That's why it's called "merchantdice".

* People have said that the last living relative of the dinosaurs is the tuatara of New Zealand. But they forgot the Thesaurus. Oh, wait... that's probably extinct now, too.

* Something with four square corners is a rectangle. If one side gets shoved too hard the corners buckle and it becomes four wrecked angles.

* Crazy English spelling conventions yield a seafood favorite: everyone loves ghoti! ("gh" as in "enough", "o" as in "women", "ti" as in "vacation")

* Nilliamps: current flows too small to measure.

   "in depth reports" for each project are below. I hope they may be useful to anyone who wants to get into a similar project, to glean ideas for how something might be done, as well as things that might have been tried or thought of... and even of how not to do something - why it didn't work or proved impractical. Sometimes they set out inventive thoughts almost as they occur - and are the actual organization and elaboration in writing of those thoughts. They are thus partly a diary and are not extensively proof-read for literary perfection and consistency before publication. I hope they add to the body of wisdom for other researchers and developers to help them find more productive paths and avoid potential pitfalls and dead ends.

Electric Transport

(Note: Don't miss the Solar Car Charging Trailer under Electricity Generation!)

Ground Effect Vehicle (first the R/C Model)

   I didn't get much done on it, but the 90mm ducted fans arrived early in the month, and I did cut some styrene foam ribs for the wing late at night on the 27th.
   A control problem occurs to me: the "ESC" motor controller I have presently seems to have no control to put it into reverse and I expect I'll be very lucky if the new ones (where are they?) have reverse either. And the radio control also has no control for reversing, much less for reversing two motors separately. By using the "elevator" control it could thrust two motors separately. Then how do I control the elevator? With the "ailerons" control?
   It will be very tricky docking if one can't reverse the thrusts. I'll have to either make some elaborate on-board electronics or accept that the RC model will have some aggravating limitations - all in the control system.

The ribs for the wing

The left-and-right ducted fan "electric jet" motors will mount just above,
and may shrink (or replace?) the "dorsal fin" rudder idea.
The elevator goes behind the wing to almost the back of the hulls.
One might say the wing profile will be somewhat akin to a plane with its "flaps down".

More Unipolar BLDC Motor Ideas

   I was closing windows on the computer screen when I ran across TurquoiseEnergy.com from editing it last month. The Electric Hubcap motors were of course there, with the word "unipolar". I see a definite advantage of unipolar as being reliability: there is no path from B+ to B- except through a coil. Spurious turn-on of the wrong transistor can't create a short circuit and blow up the controller. If the permanent magnet assist idea works, reduces energy consumption, then that will also be a major advantage. So far I have been unable to test out that feature in a motor.

   Unipolar coils can work easily for a reluctance motor because magnetic polarity doesn't matter - either the rotor metal is attracted or it isn't.
   But I thought about how I had been unable to figure out a BLDC rotor magnet & coil configuration that would actually work. There had to be north and south poles on the rotor and in the stator for good magnetic circuits, or the thrust was much reduced. Then, any arrangement seemed to reverse the drive half way around the rotation. One would have to reverse the polarity of the drive to finish the circle. Then it wouldn't be "unipolar".
   But that was with turning on coils in pairs, one north and one south, across the diameter of the motor, and using typical 3 phase power.

   Out of the blue I thought I had an inspiration. By the time I had drawn it all out and figured out how it would work, it required a different and more complex operation of the coils, and very specific numbers of coils and magnets, not the usual 3-to-2 of 3-phase motors. But it could use the regular Electric Caik motor. That would mean I wouldn't have to make a new motor to try it out - just change the wiring in the Electric Caik and connect the unipolar motor controller.
Instead of activating pairs of coils across the motor from each other, each of the 6 coils would be separately actuated. Half the motor would still have reverse thrust at any given rotor rotation, but that half wouldn't be activated.

   They wouldn't be actuated in fixed "pairs", but in rotating pairs. Only two coils would be on at a time. Thus to rotate clockwise, the coil pair activation sequence would be: B, E, A, D, C, F - counter-rotating compared to the rotor.

"Electric Caik" version

   But an 8 coil, 6 magnet "Electric Hubcap" size motor would probably be a more valuable unit. It would have more power and torque, and instead of only two coils on at a time, it could have three or four - up to half of them - to further increase the torque.

8 coil, 6 magnet "Electric Hubcap" size version.
To rotate, individual coils are activated [always in pairs for magnetic equality?]
Start: 6 & 7 are activated. If more torque is needed, 5 & 8 are activated.
When magnet F reaches coil 8: coils 5 & 6; add 4 & 7 for more torque.
Then 4 & 5, then 3 & 4, then 2 & 3...

    Then I checked out what would happen with 10 coils and a 6 magnet rotor. It seemed much better. It looked like one could after all have symmetry around the motor. But after some considerable figuring in my head, I realized I had once again traced out the path for only 1/2 of a magnetic cycle, and it didn't work equally for the other half, when the north and south magnets were in opposite positions.

   But the 6 coil, 4 magnet, version looks like it should work for a full rotation. That gives me hope that there must be other workable combinations too. Obviously there must be an even number of magnets on the rotor and an even number of coils on the stator to avoid two in a row of like polarity. It looks like if the magnets are N-S on opposite sides of the rotor, the coils opposite must be N-N, or vice versa. I need to think further on that! Apparently it's trickier than it looks. I think I need to cut out a paper rotor and actually turn it in a paper stator for each set of coils and magnets, to properly plot out coil activation sequences.

   With unipolar coils, one can try out the promising "permanent magnetic assist" idea (in some 2016(?) issues of TE News). There just might be some 'free energy' to be had in that!

Jim Harrington's Latest Electric Outboard Motor

   Jim sold his previous electric outboard and has made/converted a new one. (3 HP IIRC) It is again a 3-phase induction motor with variable frequency drive, running off 36 volts of batteries. He says he ordered the components including the motor from China and they cost far less. He isn't making innovative motors, controllers and new chemistry batteries, but unlike mine his projects get done and work in good time!

36 VDC to 240 VAC Inverter, and controls

Other "Green" Electric Equipment Projects

36 V DC "Off Grid" Infrastructure

HAT36V-50A Ceramic Sockets

   On the evening of the 10th I repaired my mini kiln and then fired the two clay
socket halves to 'cone 05', which my scribbled piece of paper said was the temperature the kiln reaches after 75 minutes, after which time I unplugged it. The red clay looked the same after firing as before, but it had that ceramic "clink" to it when the pieces touched together. Then I went hunting all over for the plug, on a long black cable. How do I keep misplacing things? In this case the cable had kept me from putting it in the "HAT36V" drawer and I found it draped over a chair in the workshop. Somewhat to my amazement, it had shrunk to virtually the ideal size in drying and then firing. (I had scaled the mold something like 18% oversize - apparently a good guess... for this clay, at this stiffness, at this firing temperature. I hope it's duplicatable.)
   The next morning I made the "hairpins" and attached two short "pigtail" wires. The plug seemed to fit pretty snugly.

How to hold the two 'hairpins' in line while soldering the wire on?

Since it worked, I molded two more. I may
do a couple more yet before I fire them.
(My only #4-40 bolts are a bit too short, 1/2"
so I made the square indents.)

Another small 36 V solar water heater

   An off-grid friend had bought some of my 305 watt solar panels. I saw he had connected one straight to a ~15 liter hot water tank that had come out of a travel trailer, now lying loose. I knew from my own experiment that he wouldn't heat water very fast with a 120 volt heating element with a panel under 40 volts. My next two Dernord 36 volt water heater elements arrived on May 31st and I took them to his place on June 2nd. It turned out the 1 inch thread model fit his tank. With considerable difficulty we (he) got the rustic old element out and we put the new one in. (rustic = rust; ick!) I should have checked the resistance/power of the old element. It didn't occur to me. It probably wasn't even 1000 watts.
   He filled the tank from a hose and we hooked it up. There was just one problem: it was raining and there was virtually no solar power to be had. When we've had some sunny weather and I get a chance to go up there again, I'll see how that's faring. (As of the 14th, still not one nice sunny day.)
   Near the end of the month I saw him and asked. He said it gets "pretty warm". I guess for "really hot", especially on cloudy days, one would want two solar panels instead of just the one he has connected (which isn't well aimed, either - almost vertical against a westerly wall, held in place with a rope).

Rong Inverter? ...Hmm, No!

   I had got a 2500 watt, 36 volt to 230 volt inverter to run my well pump if the power was off. I decided to order a similar 36 to 120 volt one to have one for the fridge and another (I already have) for the freezer. It arrived on the 10th but when I took it out of the box, it looked like another 36 volt to 230 volt one! What to do... send it back? keep it to sell? and then, order another one?

   The next day I looked at the label on the bottom to see if anything said the input was 36 volts. The tag had everything with boxes to tick off. The numbers ticked off were 36 VDC, 120 VAC, 60 Hz, 2500 watts. Huh? 120 volts with a 230 volt socket? Did I need to make an adapter?
   I looked at the rather odd socket again and discovered it was cleverly made with extra slots so you could plug either a 120 or a 230 volt plug into it! It had pins for both! Wow! That defeats the whole purpose of having different plugs for different voltages! Perhaps only the Chinese would dare do that! (I've seen another such plug since, also on a unit Made in China.) I could plug a 120 volt appliance into the 230 volt inverter or vice versa - especially as they were identical on the outside except for the tick marks on the tiny label on the bottom. The top said simply "Power Inverter / 2500 Watts / Pure Sine Wave". That applied to every model. (I wondered if the components inside were actually different or if one merely changed jumper wires - or even just programming - for the different specs. But for that much power one suspects they'd have different spec power transistors, capacitors, and a couple of other components for the different input and output voltages. The rest might well be the same.)

   I took a felt pen and printed "120 V" on the one and "230 V" on the other on top near the socket end. And "36 V" near the input end of both for good measure. Now you can only get it rong if you space out. The trick will be to mount the 230 V one near the well pump plug. (Yet to be installed.) Then the position will indicate what it's for. It seems to me the manufacturer could fix the socket pretty easily by inserting one of two plastic plugs, flat on the outside, that would plug in and block the wrong holes so that only the right plug would fit. That could be easily defeated by removing it, but why would one do so?

Laser Engraver

   I was given a very small laser engraver. (NEJE model DK-8-KZ) The owner couldn't get it to work for him. It specifically prints black and white images (no grays) of size 520x520 pixels, at 350 pixels per inch. Each pixel is either burned or not burned. Burn time for each pixel can be set from 0 to 230 - I think that's milliseconds.
   I understand the size is because it's made with obsolete hard drive stepper motor systems. Clever to find such a cool new use for them instead of the garbage can!

   I spent much of the 20th and got a computer set up to run it. (Why are these things never simple?) I recall that once upon a time I got laser diodes and was going to set up my CNC machine to burn holes to perforate plastic for plastic pocket battery electrodes. I never got it set up. This might do that job if I ever want it done.

   The only present use I can see for this machine is to engrave labels onto things like my HAT series plugs and sockets. (and it doesn't seem to work on white ABS.) Perhaps something more in keeping with its suitabilities will come along.
   Someone on youtube said he tried painting a printed circuit board with (?)black nail polish(?) and burning it off the parts of the board that were to be etched. It wasn't successful, but he hadn't yet tried a longer burn time or going over it twice or more, which (looking at his board) just might make the difference. (I wonder what other coating might work. Black spray paint?) It might be a cool technique, but the board size would be quite limited. 520 pixels at 350 pixels per inch is only about 1.5 inches square.

   I view this software installation as practice for getting the ANYCUBIC I3 MEGA 3D printer software working as well as the GECKODRIVE CNC Table Stepper Motor Driver. So many things to do!

Electricity Generation

Solar Car Charging Trailer

   Jehu Garcia is a solar and EV buff who reports everything he sees and does to Youtube. He has a LOT of videos. He went to the 2019 EVWest EV show and found a trailer with fifteen big 72-cell solar panels, batteries, charging system electronics, a wind sensor and a sun tracker.
The panels are in 3 rows. The middle row of 5 panels is the roof of the trailer, fixed in position. The sides of the trailer hinge up from the top to make the outer two rows. These track the sun to unfold to the best angles. Sometime during the show it was charging a car at 4.2 KW. It seems like a great system!

   It's a pity commercial EVs have no facility to allow charging while driving. Even the 5 roof panels might let a car drive for free in city traffic.

   From it I have the idea (a) to do something similar with my utility trailer, or (b) to put three panels on top of my Chevy Sprint instead of one, and have two of them unfold to the sides. The downside would be that they probably couldn't charge the car while driving, which I would arrange for with a single panel. Furthermore there are lots of places one would be unable to unfold the panels. Then again, many arrangements are possible both physically and electrically. Perhaps it could charge with one panel while driving or parked without room to deploy, and with all three when parked and unfolded. (I was just going to have them facing up and not try for the solar tracking either automatic or manual.)

   Am I headed into "impractical" territory here?

My Solar Power System

Observations of System Operation

On the 4th the rain stopped and the sun came out for a few minutes. Of course the solar panels were still cold. I happened
into the garage and saw that the house panels alone (6 plus the 7th one on the lawn) were making 1550 watts! Gotta love those cool panels. By the time I got to the trailer to see what new record it might hold, the clouds had returned. And that was it for sun for the day.
   But the next morning it was sunny and well before noon the trailer was putting out upwards of 940 watts. That's slightly above the 75% "realism factor" for 1220 rated watts of panels, and more than ever before. The maximum with one grid tie microinverter seemed to be 860-870, so having two with two panels each was helping more than last month's few readings would have indicated. It was at least 70 watts or 7.5% extra and a 5 year payback on the second microinverter. Later it was down to 915 - still more than before.
   Another consideration besides simple electrical payback math is that if the microinverters are more lightly loaded the thermal stress will be less, so the cooling fans will run less often and they may last longer. And if one does fail, the system is still putting out lots of power. I don't think I'd put more than ~900 watts of solar panels on one of these 1000 watt rated microinverters again. (I suppose I should install the second one properly in the trailer. The freezer can't be used with the open door and wires coming out of it!)

   OTOH, if the majority of days are cloudy anyway as they were in June, one microinverter or charge controller could run more panels and have higher output without being pushed to its limits on all those cloudy days.

Month of June Log of Solar Power Generated [and grid power consumed]

(All times are in PST: clock 48 minutes ahead of sun.)

   As I had been logging them, the power consuming activities noted didn't always give an indication of whether they occurred before or after a power meter reading. On the 20th I got the idea to put them in the square brackets with the power readings, before or after them. I backdated this to the 15th but didn't attempt to reorder the whole month.

Note: On the 7th I thought to start including the DC power as well as that to the grid. DC power measured is that consumed - usually under one kilowatt hour per day - mostly by the 36 volt kitchen water heater with a bit by LED lights. The DC "power produced" reading disappears when the sun goes down and restarts the next day. I virtually never look at it at right before dark, so I can't use it. I reset the "consumed" meter each evening after reading it.

Note FWIW: On July 2nd I turned the battery charge controller voltage down from 40.35 volts to 39.6 volts. That's 13.2 volts for each 12 volt section or 1.32 volts per cell. That's probably only about 80-85% charge, but it should be easy on the NiMH "D" dry cell batteries. One can float charge them as high as 42.0/14.0/1.40 volts, but experience with them in the Mazda RX7-EV says not to push them so hard. It's probably a reason some had dried out. (With 5000 watt-hours of refurbished battery storage, under 1000 watt-hours from DC loads doesn't draw them down too far. Long winter nights may need conservation measures like turning the water heater off.)

Date  House solar KWH(Grid+DC), Trailer Roof solar KWH - day total KWH made [power co. meter readings] weather, usage...

May 31st 53.10, 393.27 - 5.26 KWH [66469 KWH @9:00, [email protected]:30] Lt.Rain. (no BR heat) 85 Km drv. fast charged.

June 1st 56.94, 397.18 - 7.75 [[email protected]:30] cloudy, occasional sunny spots in PM. 60 Km drv.Fast Chj.
2nd 58.56, 399.19 -    3.63 [[email protected]:30] RAIN, more RAIN! (We REALLY needed rain!) 49 Km Drv.- chjd.after meter reading. The 4 older poly panels went to the DC system only, to keep kitchen hot water hot, so real total might be ~4.4 KWH (with a little power going to waste from the 4 panels).
3rd 61.14, 401.68 -   5.07 [[email protected]:30] Rain AM, cloudy PM. 55 Km Drv, fast Chj.
4th 64.08, 404.41 -   5.67 [[email protected]:00] Rain, clouds.
5th 70.02, 409.40 - 10.93 [[email protected]:00] Sun & clouds.
6th 75.29, 413.34 -   9.21 [[email protected]:00, [email protected]:00] Clouds, Lt.clouds. 55Km drv.& fast chj.
7th 79.08+.60, 415.90 - 6.93 [[email protected]:30, [email protected]:30] Clouds. 55Km slow chj.
8th 83.00+.65, 418.93 - 7.60 [[email protected]:00] Cloudz.
9th 86.20+.42, 421.88 - 6.57 [[email protected]:30 & 21:00] Cloudy, some rain.
10th 88.10+.91, 424.04 - 4.97 [[email protected]:00] Clouds, rain. Car part [email protected] after 55Km. (Left BR heat on & door open, so it was heating house - yowr!)
11th 92.27+.70, 427.57 - 9.40 [[email protected]:30] Light Clouds & chemtrails. Finished chj.car. Didn't turn off water heater for night.
12th 96.38+.50, 430.88 - 7.92 [[email protected]:00; [email protected]:30] Lt.cloud AM, Cloud rain PM. Bath AM; car slow chj. from 15:30-20:30 (How did the water heater use LESS power by being left turned on all night? Ah... the water isn't as hot. The top fell off the thermostat yesterday and it must have turned the dial. I'm only turning it part way back up because the water was really too hot - scalding.)
13th 98.86+.74, 433.28 -    5.62 [[email protected]:00] Cloudy. Finished slow charging car during day.
14th 103.88+.58 , 437.16 - 9.48 [[email protected]:30, [email protected]:00] Clouds 85Km drv. slow chj. 2-6 PM. HWH was off for night.
15th 106.65+.71, 439.21 -  5.53 [[email protected]:30] Lt.Rain. Car not full, but 55Km drv. 4 KW car charge.
16th 108.80+.66, 441.42 -  5.02 [ Laundry; [email protected]:30] Cloudy all day -- again.
17th 113.57+.69, 444.77 -  8.81 [[email protected]:00] Cloudy with a short sunny break.
18th 116.16+1.09,446.95-  5.86 [Laundry, Part car chj.slow after 55Km; [email protected]:30(2:30)] Cloudy. (did extra dishes)
19th 121.68+.53, 451.11- 10.21 [[email protected]:30, Finish car chj; [email protected]:30] Cloudy, some rain, couple short sunny breaks.
20th 128.61+.75, 456.18 - 12.75 [a little BR heat, bath; [email protected]:30] Sunny AM (Yay!), cloudy PM
21st 138.88+.52, 463.74 - 18.27 [[email protected]:30; 85 Km drv. & fast chj; [email protected]:00] Sunny; jets making cloud trails all day from one end of the sky to the other.
22nd 142.65+.61, 466.66 -  7.30 [55 Km-Chj car slow 56%=>80%-not much solar; [email protected]:30] Cloudy
23rd 147.77+.73, 470.35 -  9.54 [bath, br heat;[email protected]:00;finish chj.car;[email protected]:30] Cloudy, some rain later
24th 154.21+.57, 475.26 - 11.92 [BR heat; [email protected]:00; slow chj car; [email protected]:00; bath] Cloudy AM; Sunny PM (not even any jet trails).
25th 164.52+.57, 482.60 - 18.22 [NO heat,finished car chj;[email protected]:00] Sun. No jet trails. But warmer than 21st - solar panels don't like heat. And, the 21st was the solstice, the longest day of the year.
26th 174.57+.55, 489.83 - 17.83 [NO heat,laundry,bath; [email protected]:30] Sun, warm, some jet trails.
27th 184.55+.60, 497.11 - 17.86 [chj.car slow in sun; [email protected]:00] Sunny, warm again! Slow car chj: I plugged it into the house solar outputs outlet for a while. (less than an hour?) Then I walked by and noticed that the meter read ~500 W instead of 1100-1200...then I realized that was extra going IN to charge the car (1500 W), and that the solar output was going straight to the car without being recorded. I plugged the car in elsewhere. So actual production was surely well over 18 KWH - maybe a record? Only 5 KWH from the power grid for the whole day [surely at night], when charging the car alone would have used almost 10.
28th 187.91+.50,499.60- 6.35 [[email protected]:00; then charge car @3.8 KW after 85 Km] Cloudy
29th 3.71+.69, 502.54 -   7.34 [bath; 55Km part [email protected]; [email protected]:30] Cloudy.House meter reset in 9-10AM power failure
30th 11.82+.68,508.53 - 14.78 [laundry; 45Km, finish chj;  [email protected]:00] Cloudy AM, Sunny PM

July1st 15.90+.63,512.76- 8.94 [[email protected]:30] Clouds, dim sun all day... or are those heavy chemtrails? Oops, grid tie on 1000 W panels was turned off until afternoon. (shoodabin over 10 KWH!)
2nd 22.47+.78, 417.66 - 13.25 [[email protected]:30] cloudy, dim sun. again. (not chemtrails, I think?)
3rd 25.74+.66, 420.24  -  6.61 [65 Km, chj.car 3.8KW; [email protected]:00] cloudy

KWH-  # of Days (in June)
3.xx  - 1 (Now that's HEAVY overcast!)
4.xx  - 1
5.xx  - 6 days
6.xx  - 3
7.xx  - 5
8.xx  - 1
9.xx  - 4
10.xx- 2
11.xx- 1
12.xx- 1
13.xx- 0
14.xx- 1
15,16- 0
17.xx- 2
18.xx- 2 (just four sunny-all-day days in June!)

   June was certainly a good month to see how solar power does in clouds and rain. Production was under 8 KWH on 16 of the 30 days, and under 10 KWH on 21 of 30. One might say half the month was at 1/3 production and only 9 days were over 1/2 production. Until the 21st I really didn't even know how much the system should be giving in full sunlight because there was hardly a day with much sun, let alone a full sunny day. In May it was over 16 KWH on rare sunny days. All the clouds and cool certainly didn't help for growing vegetables either. Oh well, there are plenty of worse places to be living these days! There's been no flooding, snow, giant hail, tornados or hurricanes.
   The morning of  the 20th, and the whole day of the summer solstice (June 21st) were sunny. But even then, jets filled the sky with a thin haze of trails all day. The 11 panels (with the one still propped up on the lawn) made a record 18.27 KWH of energy that day - the day with the most daylight of the entire year. (What would it have been without the jets? 20 KWH? How much climate chaos does there have to be before this madness ends? I'm sure that much of the cataclysmic weather and the horrendous destruction of the whole 2019 crop season globally is owed to it. Well, not to mention... 7.5 billion people, burning fossil fuels, surely has a lot to do with it, too.) On the 22nd it went back to being cloudy and sprinkling rain, with some sun on remaining days as noted in the log.

Low light Performance

I tried again to measure lower-light performance of the old polycrystalline panels compared to the new and supposedly better-in-low-light monocrystalline ones. This time there were still the 4 polys (~1000W), and now there was the 3rd mono sitting on the lawn - at a steeper angle than the 2 on the roof (total 915 nameplate watts). And I added the watts going to the DC system from the polys to the total watts. The lesser reading is just the 3 mono panels, which go only to the grid tie, after turning off the polys' grid tie. All the readings jump around by 10-20 watts, so they are an average as I estimated it at the time.

   All else being equal, 915/1915=46.7%, so theoretically 46.7% of the power should come from the monocrystalline panels. If the percentage is higher in low light compared to in full sunlight, then it would seem they perform better in low light. The amount of load on the grid tie inverters affects their efficiency, and it would be difficult to compensate for that. However, if the DC load (via the charge controller) is 10% of the grid load, which is not unusual, then the mono and poly inverters should be doing rather similar power output. Then again, as little as a couple of percent to the DC isn't unusual either.

6th cloudy morning 10AM PST: 460W, 212W, 212/460=46%
7th more cloudy 10:20: 340 157, 46%

24th 15:50: SUN! 1175W, 594W 51%
25th 9AM Sun, 55°(?) angled sunlight: 865, 412 - 47.6%
11AM still sunny 1137, 575 - 50.5%

   If anything, the monocrystalline panels seemed to give more in higher light. We saw last month that a light load made a significant improvement to the efficiency of the grid tie microinverter. Perhaps another possible conclusion is that in lower light, the less loaded inverters performed more equally. and more efficiently. This would contribute to a somewhat higher output in cloudy skies per the actual percentage of light coming through. Or more accurately, to a lower output proportional to light on bright days. Then again, the 305 watt panels are only a few months old, whereas the ~250 watt ones are from 2012 and may have slightly reduced output at maximum light levels(?) Heating of the panels also reduces their output on warm sunny days. All the panels are thus in fact operating at higher efficiency in lower light levels, and higher in cold winter than in warmer summer.
   The conclusion: There are too many variables that can't be controlled to assess the actual panels.

More Panels? - Thoughts

   What good is it to get 17-18 KWH on four sunny days in June if most days only give 5 to 9 KWH? With all the cloudy days providing under half the capacity of a sunny day, and knowing how much less daylight there is in winter regardless, probably the best way to improve things would be to have more panels. Perhaps even double - 20 instead of 10. Even just another 4 or 5 would help. If the grid was down, the house would fare better. And one needs a full 1500 watts to charge the Nissan Leaf off solar only (ie, using the 2500 W inverter) - anything less and the charger will choke. That requires more than 10 panels except on a good sunny day. One doesn't necessarily have to buy anything but more panels - if the panels were potentially putting out 5000 watts but the equipment can only handle 2500, well, 2500 is enough. On the cloudy days and winter days, the output would still be under 2500 anyway, but proportionately higher. OTOH if one was properly connected to the power grid and selling power to the utility, one would spring for adequate equipment to take advantage of whatever sunny summer days came along.
   From videos I've seen on youtube, the Y-Solar 1000 watt grid tie "micro"inverter must be the world's most popular. At least with DIYers. I wonder: even if the equipment and installation doesn't meet standards for "approval" by the utility, what if all the microinverters tied into one approved switching circuit with an approved shutoff before entering the grid system?

   Something I would consider with more panels is that the new ones should be at a steeper angle to capture more than 50% of what little winter sun there is. That's when the least power is available and the most is wanted. Obviously sun tracking, AM to PM left to right, and Summer solstice to Winter solstice vertically is optimum, but the mountings have to be very strong since the panels make a big sail in high winds. A few extra panels is probably cheaper than an adequately strong tracking system with sensors and motors. (I guarantee one would very quickly get tired of moving panels around by hand for tracking AM to PM manually. Changing winter to summer tilt angle by hand is probably practical.)

   For a fixed mounting, the angle of latitude (here 53.4°) is the best for all year, and that's so steep here there's probably little point in standing them even more upright. (For winter solstice 76° would be optimum, but 53° still gives 92% as much output - and works better most of the year. The 15° slope of the house and trailer roofs is much less than ideal for any time of year. A vertical south wall is almost ideal in December!)

   So I was thinking to make frames to put up on the roof(s) to mount some new panels at latitude angle. Then I was thinking of orienting panels on the west end of the roof to face partly eastward and take advantage of the morning sun with the least early morning shadows, and vice versa at the east end of the roof, where the afternoon shadows hit last.
   Then my ever scheming mind though of putting two panels, one on each side of a single "upright" pole (at 53°) with two other poles behind forming a tripod. This should be stronger than a single pole with both/all panels attached to the top by an angled mounting. The bottoms of the panels would be up off the roof so they could turn side to side on this angle. Note that rotating systems are for panels sufficiently separated from others, since any such panels next to each other would cast shadows on each other at different times of day. I have sufficient roof space not to "stuff" the whole south slope with panels.

   Entering July I noticed I had a stiff metal post with a stout mounting at the 53.4° angle sticking out of the ground. It was a mounting for an old 1980s era satellite TV dish. All the bolts and joints were rusted solid, but they were already set as required. There was just one catch: it was under the edge of the tree line west of the house. I wondered if I could dig it up and move it to a better location... somewhere.

   These are just thoughts. Not today! Well, if I do do more panels, it's easiest to just bolt them onto the roof, poor angle or not.

Electricity Storage (Batteries)

Electrode "Pocket" with Nafion Ion-Selective Membrane Face

   The zinc electrode still looked pretty smooth and shiny, so I etched it in ferric chloride and then cleaned it with varsol, then dried it off with tissue. I put it in the nafion pocket. I looked at the manganese oxide electrode and didn't think much of it. It's probably pretty heavily contaminated with zinc oxide by now. So I took a lead positive electrode from the commercial lead-acid battery and made it lead-zinc. I used the mixture of methyl hydroxide and potassium hydroxide previously mixed for electrolyte.
   IIRC this should have yielded a theoretical 1.68 volt cell. Both electrodes were supposedly already in their charged form. But the cell started out just over 1.4 volts and started dropping. It current capacity was just milliamps. With a 1000 ohm load, it held just over 1.1 volts. And it hardly came back with the load off - just to 1.2 volts.
   The cell should have worked great (perhaps with limited cycle life) without the nafion membrane. But I took it out and it didn't work any better. The current capacity, which one would expect to be quite high, didn't even go up. As I tried things the voltage got lower, but it wouldn't hold any further charge. A few milliamps would go in, but when the charge was removed, the voltage would go back to where it was before and continue dropping. In the morning it was sitting at .70 volts. But the zinc sheet didn't look oxidized - hardly changed since it was put in.

   The barium carbonate "glue" was still on the plastic and the nafion sheet stuck to it, but it hadn't set in any way and the sheet could be slid off. If one put thin pieces of plastic over the nafion and screwed them to the frame I could see that it might still form a seal and block ions from getting around the edges, but it wasn't my idea of glue.
   All I could think was that perhaps the lead oxide electrode was contaminated from some previous try, or that the barium had added something bad to the electrolyte.

   Nothing about this experiment seemed to have worked, not even as a straight, ordinary lead-zinc cell.

What Next?

   The next thing was the whole day of the 6th fixing the lawn tractor. A bearing seized, an idler pulley froze, and the main drive belt melted through the plastic pulley and fell off. (Grr!) At least the store in town actually had in stock that pulley with the bearing for my model tractor.

   The first thing I could try was changing the electrolyte, putting in a new lead dioxide electrode, cleaning off the zinc again, and putting it together as a simple lead-zinc cell - no nafion or barium carbonate 'glue'. If it still wouldn't perform, it surely must be the electrolyte? I could then try it with just potassium hydroxide electrolyte, much as I dislike the idea. Theoretically if one has a truly effective ion blocker to keep ions in their own electrode space, one can allow electrode reactions that make ions. This would enable a bunch of new battery chemistries that have very limited cycle life without it. One enough ions of a particular metal and composition are in solution on their own side of the separator sheet and not crossing it, the solution would be saturated and no more would dissolve.
   Thus the zinc would work in the negative side, and a metal like copper could work for the positive current collector because once enough had dissolved to saturate the solution on its own side, it would stop dissolving.
Obtaining Lead positive plates from 'motorcycle' battery

   On the 7th I chiseled a new PbO2 electrode free from the "motorcycle" battery. As I started to insert it into the electrolyte in the cell I heard a slight fizzing sound. Then I remembered hearing that the previous time, too, when I had filled the cell with electrolyte. I had hoped it didn't mean much of anything, but given the results, it probably meant that this electrolyte was spontaneously discharging the PbO2 to PbO, which wouldn't further react with the zinc. Without going any farther I pulled it out. The electrolyte was probably the problem. Rats! There went yet another of my seemingly "ideal" electrolyte ideas.

   So I dumped out the cell and went and got a jar of potassium hydroxide, and mixed up a batch of this dangerous stuff. If it had to be, it had to be. If it worked, and then if an ion exchange membrane worked, then one might even go for something that would really dissolve the zinc like potassium chloride. First to make sure the KOH worked. It's what everyone else uses. How could it not?

Making Barium Hydroxide?

   On another note, if the barium carbonate won't react to 'set' as glue, perhaps converting it to a more reactive substance would work. The temperature to "calcine" barium carbonate to barium oxide is apparently 1237°C (2259°F). This would be a tough job for the mini kiln (which as I recall takes about 4 hours to hit 'cone 6', 2300°F and it doesn't get any hotter), but theoretically it should just work. Success would be determined by weight loss of the material:

BaCO3 + heat => BaO + CO2

From the atomic weights of these elements:

137 + 12 + 3*16  =>  137 + 16 [ + 12 +2*16 -- gas]

197 => 153 [ + 44 that floats away]

I might just try half that much (3.01*10^23 molecules according to Avogadro, and probably about half an avocado too): 98.5 grams of carbonate should make 76.5 grams of oxide. The oxide will convert to hydroxide in water. Whether that is something useful to work with I don't know. Perhaps to make barium metasilicate it could undergo an exchange reaction with some other silicate... like soluble sodium silicate, perhaps, to make sodium hydroxide and barium silcate?
(...Another reason to fix the mini kiln.)

   When I fixed the kiln and went to fire the first HAT36V-50A socket, I looked at the temperatures table. Cone 6 is only 1235°C. So it'll be dicey whether or not I can make the hydroxide. It could be like trying to boil water at 98°. Or thaw ice at -1. But soon a new idea occurred to me and the question was: did I need it anyway? It didn't look like it.

Nafion Delamination

   When I looked at the cell pocket in the sink I discovered that the fabled nafion membrane had delaminated into 3 thinner sheets. Egads! Perhaps the methyl electrolyte was responsible for that, too? Additionally, the barium 'glue' was still soft and the inner sheet readily peeled off it. It was only "sort of" glued on.
   I had heard nafion had surface and interior "layers", but I thought they were a surface treatment within the single sheet. Apparently it's 3 entire separate sheets. How do you glue them back together? Will they work that way, unglued, or is there some magic trick occurs at the boundary between two layers? ...Or are the outer layers just some permeable protection for the inner layer?

   I read up on it again and found things I'd glossed over previously. Nafion doesn't just let tiny ions like H+ (protons) pass through as I had previously understood. Really it lets any positively charged ions through, but not negative ions. It was therefore useless for any alkaline purpose. It would lets the zinc ions pass, but not the hydroxide ions that make alkaline reactions work.

   What I needed was just the opposite: a membrane that let negatively charged ions through but not positive. That would stop the zinc ions from migrating, but allow the hydroxide ions to pass. Then I thought: the osmium doped acetal ester film!
   Then, how to employ it? I could paint it on a sheet of flimsy cellophane, but like the nafion, the slightest leak around an edge somewhere would let the zinc ions through and nullify the effect.

   What about using a nickel positive electrode and painting the film directly on that? The zinc ions could enter the solution until it was saturated, but they couldn't penetrate the nickel electrode. Since the nickel oxyhydroxide to nickel hydroxide reaction is NiOOH ==> NiOHOH, seemingly only a hydrogen ion need move through the membrane. That sounded like nafion again. Or did the water have to get to the nickel for it to grab off the H+ ion and ozidize, releasing an OH- ion? It seemed like a choice unlikely to give good results.


   Then by the 11th I started to wonder, what would happen if I painted a film of my osmium doped acetaldehyde directly on the zinc electrode? Then it would be an ion selective film that would keep the zinc out of the water entirely. Could it let through OH- ions (atomic weight 17) to form Zn(OH)2, while preventing Zn++ ions from even contacting, much less entering, the solution?
   That too seemed a little dubious. It's simply understood that an electrode surface has to be in contact with the electrolyte to work. Still... the film should let the reactive ions through wherever it was placed, right? If it did work, it might just solve all the problems with zinc. It seemed worth trying.

Lead doesn't work!

   I put one lead dioxide electrode and the same strip of zinc in distilled water to rinse off anything that might be there. I heard a fizzing sound again - apparently just the porous lead electrode letting out its air? Then I put them into a cell space in the commercial lead-acid battery. I filled it with 40cc of 20 wt% KOH solution. The voltage read just .683V. What? A charged zinc electrode (-1.24V in KOH) and a charged lead dioxide electrode (+.48V in KOH) should be 1.72 volts. At least when shorted they put out almost 2 amps, but where was the voltage? I've seen no indication anywhere that lead dioxide wouldn't work in alkaline, KOH. Zinc is well known to do so.

   I pulled out the zinc plate. It seemed to be coated with something, especially on the side facing the lead... it looked like maybe fine particles of lead or lead oxide. No lead should be coming off the lead oxide electrode?!? Apparently lead dioxide [more likely lead tetra-hydroxide in alkali] actually doesn't work in KOH electrolyte - one of few things I've heard of that doesn't. That's probably why it didn't work in the pH 13 mix of KOH and CH3OH, too.

   Looking at the chart, if the PbO2 even at just .47V spontaneously discharged itself to PbO then we'd have: -.54V - -1.24V = .7 volts. Just about the voltage it actually had! The fizzing sound then probably really was the second oxygen bubbling off -- even in distilled water? Maybe there really was no mistake in the diagram that I "corrected"?: +2.47V would definitely explain it fizzing off almost instantly, as it did.

OTOH, this pourbaix diagram indicates the +.47 at pH 14 is surely right:

   I thought a .47 volt lead dioxide/tetra-hydroxide electrode seemed like a no-brainer. The diagram doesn't say Pb(OH)4 is dissolved. OTOH... it doesn't say it's solid, either. (Then again, it claims it dissolves in acid (Pb++), which we know it doesn't in sufuric aicd.) Maybe it just disintegrates and turns to powder? Lead as an electrode substance seems to just leave me more confused all the time.

Ni-MH "D" cell take-apart

   I decided I should go to nickel for a plus side. Nickel-zinc in alkali is well known. My technique would be simpler than making a nickel electrode since in KOH it could be a typical commercial one: Take apart one of my dead NiMH "D" cells and unroll the nickel oxyhydroxide electrode from it. (If that one didn't work, try a different one having a different voltage reading. Some of the dead cells say .4 volts, some are .7, and only a couple are 0. Some of those electrodes if not all of them have to be good!)
   The one I picked said .22V. The metal can was pretty tough but I peeled it open eventually. When I had the wrapping off and unrolled the electrodes, they were dry, but as I unrolled further, the middle area was quite moist. I've long suspected that's why they go bad: they dry out inside. Nothing else wrong! Maybe I should try throwing "dead" ones in a tub of water after all!

I peeled off the label then got a small pair of diagonal cutters and started peeling off the outer case. It was tough going, but I got it down eventually to where the innards fell out. Even a simple little dry cell had a lot of intricate pieces. The can was 24 grams and all the case parts together were 35. (A whole cell IIRC is 169g, so that's only 20% non-contributing.)

   At the top of the rolled-up cell, a fine nickel mesh extended a little above the oxide powder impregnated section. This contacted (by pressure only) a round piece of nickel mesh, which contacted an elaborate nickel plated top cover. Thin plastic insulating rings kept the cover (positive terminal) from touching the outer can (negative terminal).
   At the bottom the rolled-up metal hydride electrode protruded from the separator, akin to the nickel at the top, and touched the bottom of the can to make the connection.

   The separator seemed to be some sort of very thin fabric - it ripped more like woven cloth than paper, along a straight line. But under the microscope it looked non-woven.

Measuring the separator, it was
50cm long (x2 = 100cm since it has to be double). The electrodes were about 4.8cm tall, so the interface surface area was 4.8 x 50 x 2 or 480 sq.cm. At 50 mA per sq.cm that would be 24 amps. The manufacturer claimed... I think it was... 30 amps and 50 momentary, but of course they wouldn't run long with such currents. 30 amps would be 62.5mA/sq.cm. That is high current.


The edge of the nickel hydroxide electrode strip, showing the fine nickel mesh.
Where the electrode powder is compacted onto the mesh, little silvery flecks
of mesh can be seen at the surface. (40x)

   The electrodes both broke into short sections as they were unrolled. The metal hydride one was hardly more than flakes on the separator sheet.
   In the .7mm thick nickel hydroxide one the porous oxide powder was compacted onto the flimsiest imaginable nickel mesh, which broke apart without any fight except where it stuck out from the powder section a bit at the top and bottom. A thin strip of nickel crimped along the top edge of the mesh touched against a thin round sheet of nickel that itself pressed against the top button.  Under a microscope bits of the mesh were visible at the surface here and there through the powder.
   One suspects the powder would be a similar nickel hydroxide/manganese oxide mixture to the ones I've formulated, as described in much earlier issues of TE News.

   I used one section of it and attached an aligator clip leed with a very light wire to the delicate strip. Then I put in the sheet of zinc. The voltage measured 1.2V but there was only a few milliamps of current drive.
   I didn't much like charging a charged zinc electrode - as I've seen, it forms brittle zinc hydride. But obviously the nickel was discharged to hydroxide and wouldn't do much until charged. (That's probably why it didn't catch fire as I was opening the cell!) I started charging it at 1.9 volts (at about ~15mA through a milliamp meter that had some resistance). It didn't seem like much. But after 15 minutes it seemed to be over 1.7 volts and would put out over 150mA. (Later even 1/2 an amp momentarily.)
   While I was charging it I remembered that a nickel hydroxide electrode can be charged by itself by putting it in bleach. Well, never mind!
   NiZn @pH 14: +.49V - -1.24V = 1.73V. And it seemed to like to stay above that voltage for an hour or more. Self discharge seemed lower than many of my cells, but it was still there. In a couple of hours it was down to 1.68 volts. Overnight it was down a little more and only put out 60 mA if shorted. Since this is a chemistry known to work fine and to hold a charge, I'll go with my earlier surmise and conclude it's because of air getting into the cell (which it certainly is) and discharging the negative. Evidently lead-acid is about the only rechargeable type that doesn't have to be rather well sealed. (And that would be owing to the very low reaction voltage of metallic lead in acid.)

   12th: Actually, I could test which electrode it is by putting in a new zinc electrode. If it's only the negative that's discharging, the voltage and current will come up again. The old zinc plate was covered in crud again,
this time more evenly on both sides of the plate. It would seem lead hydroxide or lead tetrahydroxide is probably soluble in alkali. So as dissolved lead (or possibly as a powder in suspension?) enters the electrolyte, when it touches the zinc (its reaction voltage being lower than that of zinc) it is discharged to metallic lead as a rough, coarse plating on the zinc. And there was enough left in the electrolyte to coat the zinc plate all over again.
   With the new electrode, voltage rose to 1.664. Current drive stayed at about 60 mA. Was that because it was a smaller piece of zinc, or because current was being limited by the other side? It looked more like it was the nickel side that was losing voltage and current drive. I don't remember anything about that side being affected by air.
   I put it on charge and the current was much higher, about 45 mA instead of 15. Perhaps that was because the new electrode didn't get coated with lead particles? - it seemed to stay smooth. And after 1/2 hour of charging, discharge current went up to 1/2 an amp, with a momentary reading of over an amp. And afterward the voltage recovered much faster. At 11:00 PST it was at 1.755v; at 12:00 1.748v; 1.730v at 15:30; 1.722v at 18:15. A brief short circuit test then showed it still started at 1/3 of an amp and settled at 150 mA after a few seconds. It recovered to 1.717v. ... [email protected]:30. But a small voltage drop equates to a large loss of stored energy.

   My little chunk of the nickel oxide electrode was about 4.5cm * 4.8cm = 21.6 sq.cm. At 62.5 mA/sq.cm we would have 1.08 amps. That's ignoring that the zinc strip was only 3.5cm wide, the tops of the electrodes were out of the liquid, and there was considerable separation between the electrodes. (The separator sheet in the dry cell was very thin.) Only for the briefest of moments when the cell was well charged were the short circuit currents in excess of this figure. The 150-500 mA it put out seems reasonable.

Some Conclusions... and some Dissolved Oxygen?

   So... I've shown that Pb-Zn in alkali, while it sounds good, actually doesn't work. The methyl hydroxide electrolyte has been neither proven or disproven, since I only tried it with lead oxide, which doesn't work. And I've proven that Ni-Zn in KOH works fine. As if we didn't already know that. But it has that notoriously short cycle life. And since it has the similar unacceptable high self-discharge as all my other cells when Ni-Zn is not noted for unacceptably high self discharge, the problem must really be because oxygen continuously absorbs into the water and causes discharge.

   Hmm... Flash! Not to mention, that there is probably an atmospherically balanced amount of oxygen already dissolved in the water when I mix the electrolyte. I had never considered that there might be oxygen already dissolved in the water and no way to get it out of the cell! This topic has never been mentioned in any of the battery literature I've read. Hmm, is there a way to purge the dissolved oxygen from water/electrolyte? Then I thought, that flooded alkaline cells always have air inside the cell, and the buyer fills them with regular water. Maybe the amount isn't enough to worry about unless fresh air keeps entering? I decided to go with that idea.

Nafion Versus Osmium Doped Film

   When I was young I would study every specification for an integrated circuit in great detail. If this was connected to that, and that to the other, would all the circuits together meet the needed rise and delay times at every temperature and in every situation? Today I'm a great one for glossing over vital points and making unwarranted assumptions. And "the devil is in the details", as they say.

   After the nafion delaminated I re-read about it on Wikipedia. Somehow I had got the idea that the sulfonic acid parts passed only small ions like H+, and therefore Zn++ was too large to go through it. On re-reading I discover that it passes any "+" cat-ion, but won't pass "-" an-ions. Therefore the Zn++ ions would go right through it. It might block the OH- ions, but it's probably not as useful as it would seem.
   Also it is apparently made for use in an acidic environment rather than alkaline - as implied by its intended purpose of passing 'H+' ions and the fact that it itself is acidic. And by its use of some 'glue' that falls apart in even moderately alkaline electrolyte (the CH3OH [+ trace KOH yielding pH 13]). (I suspect the flimsy center sheet is the actual nafion and that the outer pieces are just there to protect it and give it strength.)

   On the contrary, what I hope is that the osmium doped film will do is allow anions to pass and block cations, in an alkaline environment. And that by painting it directly on the zinc sheet electrode, I won't have to find some way to glue a sheet to completely block every edge.
   There is some hopeful if not wishful thinking in all this, but if it works, it will be the zinc electrode protector that makes zinc electrodes last and hence changes the whole battery ballgame.

The Filmed Zinc Electrode

   Meanwhile I had decided that as the zinc was still picking up bits of lead in the lead-acid cell chamber I should use another piece, fresh electrolyte and some other container. The big cell seemed too big for just using bits of the NiMH'es nickel electrode. I had a couple of small ones I had made long ago (2012) and I used one of them. I made a new lid, trying to eliminate air gaps as much as possible. I tapered the edges so it would fit like a cork in a bottle. But I had to file in slots for the electrode terminals. I did my best to keep gaps to a minimum. It wouldn't be perfect unless I glued it. ...But would the air at the top of the cell, and the electrolyte with (no doubt some) dissolved oxygen in it, cause self discharge anyway even if it was perfectly sealed? Hopefully that wouldn't be significant.

   I decided to use the same plastic separator grill material I had used earlier, on which the deposition of dendrites was clearly visible, as shown in TE News #128 If there weren't any, it would be in stark contrast to the previous test.

A dendrite growing across the white separator grill in previous experiments.
(TE News #128. This one is apparently mostly copper - some silvery zinc)

   And I used several more chunks of the "positrode" from the NiMH cell. I took a piece of cupro-nickel sheet to make a current collector, which they would just press against. Their nickel edges should make good contact.

Finally there was too much space, so I stuck in a chunk of PE foam plastic so the electrodes would lightly press together - especially so the plastic grill was in contact with the zinc 'trode and the nickel oxide bits were in contact with the cupro-nickel sheet.

I filled the cell with 50cc(?) of 20% KOH. The voltage read 1.422 steady. I put 1.9v charge on it and it popped up over 1.8v drawing ~60mA and slowly started rising. Current dropped to ~20mA. After just a few minutes I stopped the charge and shorted the terminals. After momentarily giving 3/4, 1/2 and 1/3 of an amp, it put out .20 amps steady for several seconds until I stopped. It recovered to 1.689v then started dropping. I put it back on charge at 2.0v (60mA, dropping to 40) and the cell voltage passed 1.9 and climbed slowly. With another 15 minutes of charging it was .25 amps steady.

   Now... how long do I want to charge the already charged zinc electrode for to charge the discharged nickel side to oxyhydroxide? And what is the relative storage of each side? Probably way more nickel than zinc, but I didn't even weigh either side to know how much there was.
   OTOH, the first important thing of note was that the zinc electrode, although painted with the osmium doped film, was WORKING! That was one of the questions in the process. Either that or the film had dissolved in the caustic alkali. If it had done that, I could still try the methyl hydroxide electrolyte. But for the moment I wasn't about to open it and look. (Could I even tell if it was there or not?)

Cycle Test #1
   The next part of the experiment would be to do discharges and recharges and see it the cell shorted out after one or several cycles. Then it would be time to look. First it would have to charge at least for a while. I left it on for about 3 hours. It got up to 1.87v and it was drawing .02A. When I disconnected it,usually the voltage does a gradual fall toward some target. This time it dove down to 1.781v and I feared I had turned the zinc into hydride and wrecked it. But then it stopped and just sat there. Instead of a "\_" down, it was like an "L".
   It continued to drop, but at the snail's pace of just half a millivolt per minute. That's as low as self discharge ever seems to get in my cells - and never so soon after taking one off charge. When it was at 1.770 I shorted it. It started at over an amp, hovered at 3/4 amp for 3 or 4 seconds, then went down to 1/2 an amp.
   That seemed like very good current from two small electrodes with about 3 or 4 mm separation. The zinc electrode was about 3.3cm wide, and the liquid came up to maybe 5.5cm in the cell.
1.1A/18.15 sq.cm = .055 A/sq.cm
.75A/18.15 sq.cm = .041 A/sq.cm
.5 A/18.15 sq.cm = . .028 A/sq.cm

   Those were notably better figures than I usually get. It recovered to 1.773 V -- slightly higher than it started out from! ...and then resumed its leisurely dropping. I decided it was time for a load test. I used a 120 ohm resistor. That was probably needlessly light, because the voltage only dropped 32mV in the first 5 minutes, and then went down at the leisurely pace of about 1mV per minute. (That's a slower drop than my most usual self-discharge figures!) Current would have been: 1.7v/120 ohms = 14.2mA. hmm... charged it at 30mA for 3 or 4 hours? If the cell was working as well as it seemed, this could be a very long test.

Start at 19:15:00:
15 1.763 V (just prior to connecting load)
16 1.739 V (after 1 minute)
20 1.731 (after 5 minutes)
25 1.726
30 1.721
35 1.717
40 1.710 (Dropping faster? - or was it just because I took off and then reconnected an alligator clip?)
45 1.705 V (1/2 hour mark.)
50 1.699
55 1.694
00 1.689 V (20:00:00 - 3/4 of an hour)
15  (1 hour)

Very close to the 2 hour mark the voltage started dropping rapidly. That was good, saying that it was behaving just like a battery run out of juice. OTOH it also said it had only 28 mA-hours of juice. I'm going to guess that's all the small, simple zinc sheet had in it, as I used maybe a couple of amp-hours worth of NiOOH 'trode pieces from the "D" cell and it had charged substantially more amp-hours than discharged.
   Next and rather vital question: would it grow dendrites and short out the cell on recharge?

Cycle Test #2
   I recharged it for almost 2 hours with no evident problems (yay!), and then did another load test, this time with 60 ohms. It ran for an hour and 15 minutes at mostly a little higher voltages and delivering twice the current for a total of 35 mA-H. Would the amp-hours improve more? Would dendrites short the cell? The way to find out was more testing. It was well after midnight.
(14th) With the briefest of charges, the cell held 1.73+ volts overnight. (I guess the top fit pretty well!) The charging currents seemed to be gradually staying higher longer. But the low self discharge said nothing was shorting or forming a low resistance between electrodes. That would seem to be in keeping with finding it had increasing amp-hours.

Cycle Test #3
Still no internal shorts after the third charge. A quick external short gave 1.8 amps - 100 mA/sq.cm. That was good current. The voltage drops during discharge weren't high so I decided to try a 40 ohm load this time to shorten the test some more. Current would be 1.7v/40 ohms = 42.5mA. The voltages started dropping much faster than with the 60 ohm test - or even both tests previous. I wonder if there was a bad alligator clip connection in that test - a bad one may have made it more than 60 ohms. I'm always struggling trying to get good connections from alligator clip leeds, and this time I had been having trouble with one during charging - maybe it was already high resistance. I took it out and shortened the connection for the discharge. Say it was 30 ohms in that clip leed: it would have made it 150 ohms and 90 ohms in the first two tests. That would explain too the "increased performance" in the second test since the ratio of the two wasn't 2 to 1 but 15 to 9.
   40 ohms was definitely too much load for it and the discharge only lasted 12 minutes. That made the milliamp-hours look even worse: 8.6.

Cycle Test #4

Here I unraveled a mystery. I used 120 ohms again. It seemed very hard to get good contact with the zinc terminal tab. (And it was actually the cupro-nickel tab I was fiddling with.) I switched to yet another "poor" aligator clip test leed. And it seems to get worse as the discharge proceeds. By rubbing the meter probe within the aligator clip directly on the terminal, the voltage would jump around considerably. After over an hour, of seemingly falling voltages, it could occasionally be brought back up to the same voltage it had just one minute into the discharge, 1.733 volts! Suspicious, I checked right on the resistor leads themselves with another meter. But the full voltage was indeed across the resistor.
   The falling voltages in the discharges weren't from the battery getting lower at all. They were, apparently, from gradually increasing resistance at the cupro-nickel terminal tab as the test proceeded. Then after a lot of frustration and inconsistent results I finally discovered that it wasn't the terminal at all. As I wiggled the clips I was slightly moving the whole current collector plate. The bad connection was inside the cell, between the plate and the chunks of nickel oxyhydroxide electrode. Internal reactions doubtless explained why they got worse as discharge proceeded. A better arrangement would be required for the next cell.

   The battery had more capacity than I had thought it had. The small amp-hour values in the previous tests were wrong. For a while it looked like it could drive the 120 ohm resistor all day, at least! Or at this point, maybe it would be all night. There was certainly little point in checking it every 5 minutes as I had been doing! But after 3 hours the highest voltage I could coax out of it was 1.700, and gradually dropping. Was the zinc plate getting discharged, or was it that I hadn't charged it long enough and the nickel was spent? After all, it was discharging now longer than I had charged it for, albeit at maybe at 1/3 the current.
   Obviously it could run longer but it was midnight. I disconnected the load resistor.
(15th) The cell had 1.744 volts in the morning, so it had lost little to nothing overnight. YAY! I put it back on charge at 9:30 but after an hour I opened the cell and inspected the separator grill. There were some little dark specks on it but no dendrites. The specks do doubt fell off the crumbly nickel side.
   When I put it back together I stuck a piece of plastic behind the foam piece to press the electrode chunks more firmly against the collector plate. It started charging at 110 mA instead of 90 and when shorted a bit later it put out 2.47 A for a moment and took a bit longer working its way downward. 2.47 A / 18.12 sq.cm = 136 mA/sq.cm. That was starting to look quite impressive.

   (Let's see... my "full size" electrodes are to be about 6.5cm * 12.5 cm; 81.25 sq.cm. At .1 A/sq.cm  that would be about 8 amps. for a moment. with a short circuit. of course. That's per electrode pair. If there were 3 nickel electrodes and 4 zincs, that would be 6 such surface areas for ~50 amps. At a more reasonable 50 mA/sq.cm that would be 25 amps. For 300 amps for a light car like my Sprint you'd want at least 12 in parallel then, which would also be around 600 amp-hours. Then multiply by 24 for nominal 36 volts would be 288 batteries. 22 KW of storage. All still pretty speculative, but starting to enter some ballpark. And how much would all those weigh?...)

Cycle Test #5

   Having stuck in the extra piece of plastic, and seeing the higher currents, I made the next test a 15 ohm load. It put out about 110mA for just 30 minutes before the voltage went for a dive. 55mA-H.
   Now I'm wondering... in charging for just a couple of hours, did I actually put more into the cell than was pulled out?

Cycle test #6

   15 ohms seemed like too heavy a load, so this time I tried 27 ohms. At 1.7 volts, that's 63mA. It ran for exactly 3 hours, with the voltage dropping from just under 1.8 volts to below 1.6 volts only in the last 16 minutes and then it rapidly fell below 1.5v in the last 3 minutes (to 1.450v at the 3 hour mark). That's about 190 milliamp-hours - almost a fifth of an amp-hour from a very small zinc sheet electrode. (3 grams? - again somewhere a little under 10% utilization.)
   Well! From 1/2 hour to 3 hours by halfing the load. And from putting in a chunk of plastic to push the nickel oxide chunks better against the current collector.

   But in order for that to happen, at the two hour mark I fiddled with the positive current collector. I could wiggle it a bit from the tab, and obviously it still wasn't making a really solid connection inside. The first time I got the voltage to go up ~10 millivolts, and and not too long after that, ~7 mV. It would continue down from the higher value. I got the impression it was the nickel that was running out of charge rather than the zinc, and that it helped to move the current collector to better connect some other area. After that it stopped helping - I couldn't get any more out of it.

   The mere fact that I could do 6 charge-discharge cycles on this cell without apparent degradation (in fact with apparent improvement) doubtless makes it better than any of my previous cells.

Take-Apart & Inspection

(16th) I opened the cell to examine the components. I took pictures under a microscope (@ 40x unless one or two were @ 100x):

- The separator grill once again had some bits on it, but no apparent dendrite growth. I rinsed off the grill and then tamped a piece of the nickel electrode on it. Again it had similar looking particles on it. (It later occurred to me to put in a separator paper next time to block their movement and verify for sure that they are coming from the nickel oxide electrode side. But they may just as easily be zinc bits.)

- The cupro-nickel still just looked like a sheet of metal. Later I took pictures of a never immersed piece for comparison.

- Zinc Electrode front side after initial use

- And the back side, which hadn't changed much since being coated with the film.

- The next time the zinc electrode looked much like it had before except now the back side looked much like the front side. Probably the longer (esp.#6) discharges had activated both sides of the electrode. Nothing that looked like dendrites - YAY! Later I took pictures of a sanded and etched but uncoated sheet of zinc that had never been immersed for comparison.

   When I put it back together I put in a second separator sheet and another piece of plastic behind the nickel electrode chunks to press them together more. It didn't seem to want to put out much more than 2 amps when shorted (the extra separator sheet?), but it continued putting out around 1.5 amps for several seconds (10?) instead of dropping quickly to 1/2 an amp. With a 1 ohm load it put out around 1.2 volts which only dropped quite slowly. (The current meter said it was only putting out .9 amps. HOW much resistance in the test connections? Hmm, .3 ohms? Actually that doesn't seem either unexpected or unreasonable.)

Cycle Test #7

   Since it was handling current so well, and not wanting this next test to take another 3 hours, I gave it a 15 ohm load again for this test. The previous test with 15 ohms had ended when the voltage dove off a cliff in 1/2 an hour. This time it was still going strong at the 1/2 hour mark, putting out 1.665 volts. It ran for an hour and 25 minutes. The only things to attribute this to are (a) it wasn't fully charged last time, or more likely (b) the nickel oxide sheets being pressed together better, making better connections in the positive electrode.
   Following a few hours recharge after this, I found that it held over 2 amps short circuited until I stopped at about 10 seconds.

(17th) I had had enough of lengthy cycle tests and went on to experiment with getting more amp-hours from the zinc electrode.

More Amp-Hours from Zinc?

   ...Okay, I wrote the following before I discovered the zinc sheet had higher capacity than it showed before cycle test #4. A plain sheet of zinc is still a low surface area electrode with low utilization and this still looks like a great way to get excellent high capacity from a coated zinc electrode - the sort of specs wanted for commercial manufacturing.

   If the coated zinc was working, but had low capacity, how to get more amp-hours from it? It weighed 2 or 3 grams, a theoretical capacity of over an amp-hour or even two. But a simple piece of sheet metal, if it's not dissolving during discharge (as in a Mn-Zn cell in chloride electrolyte), doesn't have much surface area.
   The obvious thing to increase the surface area is to use zinc powder and grit of various grades. But then, how are those compatible with the osmium doped film? Instead of a thin surface film, perhaps all the interstitial spaces would have to be filled. Osmium isn't cheap. I've been trying to make cheaper batteries, not more costly ones! Perhaps one might fill all the interstitial space with electrolyte, and then (somehow) just coat the entire outer surface with the film? The zinc particles inside could perhaps dissolve, but they couldn't much go anywhere even if they did.

   An interesting thought (without doing much to solve the above problem) was that the lead particles had made a very rough plating with a lot of surface area. If a zinc sheet was plated with lead, and then placed in zincate solution and the lead plated with zinc, that would give a high surface area zinc sheet. (But then the lead would add weight without adding capacity. I guess I won't try it.) Hmm... One could instead sprinkle zinc oxide particles on a zinc sheet layed down flat and plate them on for a similar effect. Probably smoother (taking less osmium film material) but still porous. If one plated it and then coated it with the film, the particles couldn't come loose and fall off on discharge, either. That should be well worth a try!

   I took the second small ABS tall, thin cell and made electrodes to fit (cupro-nickel and zinc sheets) and a new lid for it, again making it pretty tight against air.

 On the 17th I tried plating the zinc electrode (4.20 grams) for that cell per the above. I poured some Caswell "Concentrated Zincate Solution" into a shallow plastic tray, and then I dumped in some zinc oxide powder, which proved to be lumps, and tried to mix it up. I put in a zinc positrode and the negatrode to be plated, connected them, and turned on 1.5 volts.

At first nothing seemed to be happening. But a few flakes had plated around the edges. It turned out to be a long process and patience was the key. I dumped in some more zinc oxide, and later diluted it all with some water, which seemed to help.

Top side near one edge
In addition to zincate ions, lumps of zinc
oxide resting on the plate turned to zinc and plated on.

Bottom side was less lumpy since there were no lumps of zinc oxide resting there

The middle of the faces didn't get very much coating -
mainly just some oxide lumps.

   While it was working I looked on the web. Everything came up old papers with abstracts, that various hucksters wanted to charge you quite a lot of money for to read the whole thing. There was one with a list of ways to make porous zinc electrodes.
   Then I thought to go to youtube. There I found NurdRage, who posted various chemical techniques. He had a video doing basically exactly what I was doing - zincate solution (from NaOH), zinc oxide from a pottery supply, and all. It seems the zinc oxide turns into zincate by itself to the limit of solubility. And as I surmised, so does the zinc positive electrode. So gradually any zinc oxide you throw in, and any zinc touching the positrode, will turn into zincate and plate onto the desired electrode.
   One difference was that his container was a beaker and the plate he was plating to [a copper circuit board] was almost vertical in the beaker.

   My electrode plated around the outside edges, mostly ignoring the center. I guess lying horizontal on the bottom of a tray of shallow liquid isn't ideal. His plated especially in the center, but not ignoring the edges. He left his going overnight and there were great gobs of porous zinc plating on it, even wrapping around the edges of the circuit board. (Having attained this fine result, he then scraped all that plating off and into a beaker. Horrors! I guess he wasn't making an electrode.)

   My electrode weighed 4.30 grams after plating. Theoretically I had added a whopping .1 grams, .82 amp-hours to it, but my scale only moves in .05 gram increments, which may move up or down .05 with nothing happening, so it could be more or less.

Osmium Supply

   On the 18th everything seemed rosy and settled. Everything seemed to be working great and the path ahead seemed straight. I had a suspicious thought. Production of new chemistry solar panels was sabotaged a few years ago by silicon solar panel makers, who having no use for it themselves, bought up the rarest vital ingredient: all the available gallium with options on the whole of the future supply. Having got investment and done the work and set a factory up for production, and with all the great promise for better, cheaper solar panels, the company went broke because they couldn't get gallium. This "dog in the manger" tactic is illustrative of the dirty tricks existing businesses are permitted by our society to use to prevent competition. (You've probably heard this story before in some TE News issue.) And long life nickel zinc batteries would threaten lithium battery producers with billions invested in the technology. So I thought I'd get a bit more osmium.

   The film only needs a trace of Os, but it's the second rarest element to be found in the Earth's crust and so it's very expensive. I used osmium powder before - that cost far more yet than solid pieces, so I thought I'd get a solid piece. I ordered a 20 gram ellipsoid 'droplet' for about 900$C. (So much money for less than 1cc of metal!) I trust, without looking it up for now, that there is some practical means to turn it into powder or dissolve it into the mix.
   And without trying to figure out quantities, I expect that should probably last for the foreseeable future including limited production.

Oh No - Another Problem!

   I played a bit with the cell now and then, driving a heavy load or seeing that it hadn't lost much energy overnight, and I charged it once, until the evening of the 19th. I was really pleased with the continuing good performance. Then something was wrong. No dendrites, but it turned out the sheet of zinc had fallen apart, separated into two along the waterline.
   I don't remember reading anything about anything like that in the battery literature! OTOH my previous zinc terminal tabs had become brittle zinc hydride and broken off before.

The top above water part and the lower underwater part under microscope

   I have a theory... If it was overcharged it would bubble hydrogen. The coating should prevent it from forming zinc hydride in the water (I trust), so the hydrogen would bubble up to the water line. If there was the slightest weakness in the coating there, hydrogen bubbles all around the electrode would turn the zinc to hydride and a crack would start and spread, along the waterline.

   This theory seems to fit, anyway. I figured at worst I could paint or otherwise solidly coat the zinc from below the waterline to above the point where it comes out of the case. This time at least the terminal tab seemed to be "good as new". I had wrapped it with magic transparent tape and so kept the reactions away form it. And there may be easier ways. Perhaps I'll try two coats of the ion selective film instead of one.

   The two pieces seemed considerably different from each other, and the lower one seemed thinner than originally. At first this puzzled me, but of course... the top one was in the air; the bottom one was in the liquid. Later I weighed them both together: 2.7 grams. It was very close in size to the new electrode, which weighed 4.2 grams before plating. So it had definitely lost some substance. So much for it having no degradation. If it was losing substance off the surface, where was the osmium?

   And would it stop at some point, reach an equilibrium, or would the zinc continue to get thinner until it fell apart somewhere? If the former, perhaps I just needed thicker pieces of zinc. (This would be a surprise - I has considered that most any thickness was just wasting the interior atoms. But if the surface ate in as far as half way from each side, then the zinc was being fully utilized and the "current collector backbone" portion in the middle was being totally disintegrated. It could also be  a problem as the readily available zinc sheets I had found were the "moss killer" roofing strips which were all pretty much the same. Or perhaps with a porous plating the sheets would be less affected. Perhaps I should order some thicker and specified purity sheets?)

   Was my 'breakthrough' really the final answer? A partial answer? I've had what I thought were breakthroughs before too, only to find problems or incompatibilities between different parts of the cell.
 I took the cell apart. It seemed to leak a bit anyway.

Next Cell

(20th) But with it having no apparent dendrites, it had to be an improvement over other zinc electrodes. Would the metal only be affected to a certain depth into the surface? Would the porous coated electrode fare any better because the surface was deeper? What would happen if I painted on two or three coats of the film instead of just one for better coverage? Some more experiments were in order to try to answer the questions.

   So I put together the new slightly smaller cell. First I coated the porous plated zinc electrode with three coats of the film. Perhaps it would make it better or seal the surface better or something. Of course with the plating being mostly around the edges, a lot more of the liquid penetrated the pores there than went onto the relatively smooth center area. I had to dip the brush into the liquid to get more, more often.

   I put the cell together with the nickel oxide bits from the previous cell, so the voltage started out high - all charged. It didn't seem to put out as high currents - less than 1.5 amps shorted, and higher voltage drops under load. I mostly attribute this to (I suspect) a poorer connection from the nickel to the current collector. Nonetheless it seemed to work fine. I foolishly started a discharge test with a 60 ohm load at 11 PM. It ran 3 hours, and the voltages were lower than I expected before it finally started to peter out. It turned out to be in the wiring - I hooked some skinny, lightweight clip leeds to the somewhat flimsy terminals, and the meter was on the wrong side of those. Why was the voltage getting so low, yet it wasn't rapidly dying? I finally put another meter straight on the terminals and found the battery voltage was somewhat higher than I had been reading it as.
   The cell averaged maybe 27mA, 24mA and 21mA over the 3 hours. That's 72mA-hours. (Current was separately metered, so those are the actual figures, which say the "60 ohm" load was about 65 ohms with all the crappy connections and skinny wires.) 72mA-H compares well with the first test of the first cell: 28mA-H.

   The cell seemed to have a high self discharge after the first cycle. I took it apart and rinsed some sludge out of the bottom but it didn't help. Then I changed the electrolyte... which I had poured out of the first working cell and was still using because the first cell had been working fine. To my mild surprise that solved the problem. ...For one cycle. It held charge overnight fine, but after the next discharge it didn't recover as expected and after being charged, once again it would discharge itself. I took it apart and inspected it, but nothing was crossing the separator grille. I changed the electrolyte again (I spilled it in disassembly anyway), this time without any  improvement.
   I could write this off as yet another puzzling failure of chemistry or technique and go on to try something else... except that two of the three components, the nickel oxide electrode pieces from a commercial cell and using KOH for electrolyte, were not experimental. Neither was zinc per se. Of all the many things I've tried over the years, here the coating on the zinc was the only experimental thing in the cell. And... this cell seemed to work sometimes, and the first cell with the same coating, albeit one coat on a rather plain surface instead of three coats on a more porous surface, had worked pretty well for a while and hadn't done anything like this.
   One difference to the first cell belatedly occurred to me. I had folded a piece of the separator cloth from the NiMH dry cell around the zinc electrode, to separate it better from the nickel side. (If little bits still appeared on the grille, they came from the nickel side. If not, there were probably some coming from the zinc that now couldn't get there.) When I opened it, I removed it. But it had bits of crud from the old electrodes stuck on it. And I don't know which side I had touching the zinc. Perhaps it introduced some foreign material onto the zinc surface? A new zinc electrode was doubtless called for.
   Another thing I noticed later: my first bottle of distilled water had run out. The new bottle said "ozone added". Huh? Why would one add a potent oxidizer to what was supposed to be pure water? I had rinsed the parts in it. Would it persist? Might it make some undesired reaction? I'd better go back to the first store and the other brand!

   When well charged, the second cell put out over 3 amps momentarily: 165mA/sq.cm. That would suggest that a porous electrode handles more current than a flat sheet, in case we didn't already know that.

   I decided to charge this cell, #2, overnight in spite of all the hydrogen that would bubble off the coated zinc, and see if it helped.

   On examination of the Zn electrode after use and with overcharging having bubbled hydrogen off of it, it didn't look as bad as previous zinc electrodes. The osmium/coating seems to help protect it. And the separator grille, while it had bits of crud on it, still didn't have any dendrites.
   The bits of crud were thickest near the bottom. This of course is why lead-acid cells have a space underneath the electrodes for crud to drop down into. They fail when the crud has built up to the plates, and a "better" PbPb cell is one with a deeper space underneath. The bits on the grille readily rinsed off in water.

Third Coated Electrode - Second Porous Plated One

   In the meantime I was thinking of how to set things up better to get more even platings on zinc sheets, to make a similar #3 cell (or at least a #3 electrode) with such an improved sheet, and hope that it would work without issues.
   If I made another the same size, etched it evenly, and could then plate it evenly with .5 grams of porous zinc, it should theoretically have about 1/2 an amp-hour of charge. (In this I note that in spite of the uniform black appearance when it comes out, that the zinc was much more etched around the edges than in the middle - something else to try and improve if possible and practical.)

* "Scotchbrite" does a better job of scratching up the zinc surface than sandpaper. (Neither is a substitute for etching it.)

Egg Albumen Coating?

   It seemed to me the zinc electrodes needed another coating on top of the coating to keep them from deteriorating and bits of zinc from dropping into the separator. My preferred means would be to jell them, and jelling has now been borne out by two other researchers as making electrodes "everlasting". But what to use? I started to think the Sunlight dishsoap (with the sulfonates) should be for the positive electrode, but not the negative. Then there was agar... I wasn't sure how that would stand up to hydroxide electrolyte. Okay... what about egg albumen? "Albumen paper" has been used in photography. It's "developed" or "fixed" by immersion in sodium hydroxide.
   Wikipedia says eggwhite is an alkaline solution (but probably pH8 rather than 14!), and "ovotransferren is a glycoprotien that can bind bivalent [zinc!] and trivalent metal ions in a complex."

   It seemed to be something worth trying. At best the coating will stabilize and hold everything together and prevent 'stuff' from flaking off the electrode, making it far more long lasting. The worst that can happen is it degrades (at pH 14?), is deleterious or just isn't useful and I have egg on my face. If you don't try things that might seem a little crazy, you'll never find the ones that actually work.

   A youtube video on making albumen [for foto paper, by: Will Salley, 2013] gave me more clues about albumen.

1 liter egg albumen (30 eggs - no yolks, no stringy bits, just pure liquid)
15 grams sodium citrate
15 grams NaCl
2 cc glacial acetic acid
30 cc distilled water (Surely can make from 5% vinegar and water? There was nothing special about adding each ingredient. Hmm... make that 32cc of 6.3%, to be technically correct.)

Whip with egg beater until it turns into a froth. (He blenderized for 15 minutes)
Put in fridge for 24 hours.

   Another video had a simpler process:
100 g egg white powder
700 cc water

   However, by about the 3rd week in June, after so much experimentation and trails, successful as things were becoming I had other things to do and had become a bit burned out with bettery research. The third zinc electrode with its improvements and albumen will have to wait for July. (and even then... it's summer!)

Haida Gwaii, BC Canada