Turquoise Energy News #194 - July 2024 Report
Turquoise Energy News Report #194
Covering Research & Development Activities of July 2024
(Posted August 9th 2024)
Lawnhill BC Canada - by Craig Carmichael

[Subscribe: email to  CraigXC at Post dot com ; request subscription]
Website: TurquoiseEnergy.com

Feature: Some Thoughts on DC Houses & DC Appliances (Under Other Green & Electric Projects -- also more on the subject under Cabin Construction & 36V DC Wiring immediately following that, and in August in Brief)

Month In "Brief" (Project Summaries etc.)
* New Chemie Batteries - DC Wiring Demo Panel - Cabin DC Wiring & Construction

In Passing (Miscellaneous topics, editorial comments & opinionated rants)
*  Scattered Thots: What Makes an Inventor? - AC Body Voltage Levels - ESD

- Detailed Project Reports -

Electric Transport - Electric Hubcap Motor Systems (no reports)

Other "Green" & Electric Equipment Projects
* Some Thoughts on DC Building Wiring & DC Appliances
* Cabin Construction & 36V DC Wiring
* Gardening

Electricity Storage: Batteries
* Various-Zincate cells - gold plating - Electrode Powder Compaction

Electricity Generation
* New Grid Tie(s) - My Solar Power System: - The Usual Latest Daily/Monthly Solar Production log et cetera - Monthly/Annual Summaries, Estimates, Notes


July in Brief

New Chemie Batteries

   One keeps seeing new battery developments in the news and wonders whether a new "everlasting zinc" chemistry is still worth it after all the years I've spent on it and all the developments in other directions. Then one sees articles about burning cars and e-bike batteries burning down homes or even blocks of connected homes. I suppose such accidents are rare, but they can be devastating. In addition to being intrinsicly cheap, a zinc based aqueous battery is much safer. And a zinc-air battery would have the energy density per weight to power quite long range electric aircraft.

   Now that the zinc side works, I'm still having a crazy hard time with a positive electrode - which is known chemistry and manufacture. It started to seem like I could only make a reasonably good electrode from the nickel substance taken from commercial Ni-MH dry cells, not from my own mixes. But why should that be? I've become convinced that at least any of my nickel mixes or the nickel-manganese oxide probably work fine chemicly. It must be all in the compaction. I figured that with insufficient compaction one would simply get lower current capacities. This seems to be a wrong assumption. Instead it seems that at a certain minimum point, current capacity becomes good, but there is still very low utilization of the electrode substance. To get it all connected together and contributing to the amp-hours takes much more pressure. At least, that is now my present supposition, since only the zinc side seems to have much capacity regardless of there being plenty of positive substance theoreticly to run a light load for hours instead of the minutes I've been getting. (The zinc-zincate electrode needs no pressure.)
   Early on battery developers noted that for both copper and nickel oxides, tremendous pressure is needed to force the particles together tightly enough for good connections - and those are the substances I've been using. An Ovonics patent notes pressing their nickel substance together to achieve a density of 3 grams per cc instead of 2 and a bit. That must take tons per square centimeter of pressure. The commercial dry cell nickel was doubtless pressed with that sort of pressure, and the pieces retain some of that density in my electrodes, so it works.
   I've been pressing with much less force, and I discover that with more force my supposedly compacted powder can be made much denser. But I've always had a hard time trying to get the high pressures I've wanted into an electrode. Flat electrodes bulged too easily for any sort of home prototyping. 3D printed tube electrodes are also too flimsy. The PVC pipes with drilled holes are probably strong enough to hold material adequately - the electrode of commercial dry cell nickel hydroxide continues more or less to work - but they certainly won't take the pressure of sufficient initial compaction. And if material is squeezed down in the tube (however inadequately), the separator paper gets munched down with it.

   One can make a compacted blob fairly easily, but to make it with a hole for a current collector rod, and to slide it into into an electrode tube without destroying the separator paper, are whole additional challenges that I haven't mastered. Not only the 3D printed tubes but the PVC ones with holes will burst at a much lower pressure than seems to be required. I tried to make a set of telescoping stainless steel pipes to press tall nickel oxide "donuts" to 5 tons of force, but at about 3 tons the pipes bent and the ends mushroomed a bit. And the piece was stuck inside the smooth tube. It just wouldn't press out.

   But a couple of days later I fixed up the 'tap & die' and tried again, and got better results.

"Telecoping" compactor to make "tall donuts" of electrode material.
The inside of the large tube is the outer diameter of the donut.
The bolt fits in the bottom to form the hole for the carbon rod.
The smaller tube fits inside the larger and over the bolt,
and is pressed in to compact the material.
The hole in the plate is just slightly larger than the inner pipe.
The piece is pressed out of the pipe through the hole.
Ha ha.


At 3 tons the pressure stopped going up. The inner pipe was bending over.
And the pipe ends mushroomed a bit.
Apparently something heavier than "light" stainless steel pipes is required?


Later with small improvements & just 2 tons force, I successfully
made a small "donut" of compacted nickel-manganese-oxide mix
that fit smoothly on a 5/16" carbon rod. (Another on the tapered
3/8" rod behind started breaking because the inner hole was 5/16".)
The solid cylinder at the bottom, with a bolt turned on the lathe
for a plunger, was successful but it would break up if drilling a
hole in it was attempted. (Rats)


   2 tons seemed to be exactly right - at least for my equipment and the size of the donuts. Any less and the donut crumbles (I'm sure there's a pun there somewhere). Any more and the stainless steel pipes start to bend. With the small surface area of the donut hopefully it's about right for the electrode. (Calculated density of the compacted nickel-manganese oxide based on the measured dimensions and weights is about 1.9 g/cc, which is probably a bit under ideal but in the right ballpark according to literature.) I discovered that luckily the small donuts slide easily off the inner bolt and onto a carbon rod. Yay! (I hope they make a good connection to the rod!)
   Now to make the next electrode I need to make new separator papers and a new tube, and it's high time to end this 'July' report (August 8th). I think I'm on the right track.


DC Wiring Demo Panel

   In addition to continuing to wire the cabin, I decided to make DC demo panel for the Swilawiid green energy conference in September. It progressed in a couple of work sessions. As I made it it occurred to me that it was just about the same as actually wirng the cabin except the wires were shorter.

Sample house circuits on "wall" at right; "Wiring closet" wall mounting of equipment at left (gyproc for
fire safety, over plywood to hold screws securely) starting with a couple of surface mount circuit breakers and a ground bus bar.


Added charge controller and power monitor, "main" battery circuit breaker,
 a 12V lamp plugged in with DC to DC down converter.


Cabin Construction & DC Wiring

   This is mostly covered in the feature article as well as the detailed report in "Other Green Projects". Here is the brief view.

   I continued adding to the surface mounted wiring board. I installed the ground bus made in June. I 3D printed a four-outlet wall plate to plug several ceiling lights into rather than hard-wiring them. I suppose one could fit 5 or 6 T-sockets on one wall plate. No "power bars" to get enough outlets!
   This box/plate is on the charge controller's "output" circuit, which is set to shut off if the battery gets the least bit low - which would probably be from leaving too many lights on in the winter when solar power is scarce.




36 V DC Wall Receptacle

   On the 10th I wired up a wall receptacle in the upstairs room. I took off a piece of gyproc from under the window and screwed in an electrical box. Underneath, the walls are still open and stringing the wire through the 2 by 4s by drilling holes was easy.


   I used #14-3 cable, which is actually four wires counting the #14 bare "ground" wire. I tied red and white together ("+"), and the black and the bare after insulating it (two blacks, "-"), and tied them to the pigtails soldered onto the T-Plug (36V DC) wall plate. Double #14 AWG is the same as #11 - good for about 25 amps. #12-2 (rated 20 amps) would be fine for 36 volt - 20 amp outlets, too. #14-2 is a bit thin.

   Enclosing wire joins in metal electrical boxes is still a good idea. 36V may not electrocute, but arcs and sparks or heat from poor connections, can still start fires if next to something flammable. (But no need to ground the box.)


   In early August I used the new outlet and a "double T-Plug" I made a while back to power a 2500 W PSW inverter (36 VDC to 120 VAC) and vacuumed the room on solar/battery power. The double plug wanted to pull out at the slightest stress on the heavy wires. (The "giant T-Plug" for high currents is the real answer... but they need to be manufactured.)

   Most vacuum cleaners are somewhere near 1500 watts. Theoreticly that should have blown the 20 amp (36V) breaker with around 40 amps. In a brief run it didn't. I used a vac that had 'up' & 'down' buttons to increase or decrease power & suction. At full power the inverter (2500W?!?) balked after about half a minute, quitting and saying "Fault", but if I turned the vac to setting 3 of 5, it ran fine and I did the job.





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

Scattered Thots

What Makes an Inventor?

* What is it that makes an inventor? "IQ" and "Cognitive ability" are certainly measures, but I think it's "associative ability" that sets inventors apart. That's the mental quality of seeing "obscure" connections between things that are apparently unrelated.
   Just for example, in open loop air heat pumping, how are indoor-outdoor heat exchange ventilators related to compressing air (or refrigerant) to pump heat? These have always been separate, unrelated concepts and systems. But it's combining the two that is the key. The compressor heats already passively warmed air coming from outdoors, radiates heat to the room, then sends it (still slightly warmer than the room, still compressed) into the indoor-outdoor exchanger to cool by warming the incoming air, then it decompresses (through a turbine to extract even more energy) and finally it exhausts outdoors colder than the outdoor air. The combined system should give a coefficient of performance that heat pumping "experts" would deem "theoreticly impossible", perhaps 10 to 1 or better, eg, 5000 watts of heat from 500 watts of electricity, even at freezing outdoor temperatures. (I must get back to that project!)
    I was once profiled as being in the top 1% of the population in "associative ability". Others doubtless have more balanced or other leaning mental abilities, but that one is surely what marks inventors.
   Then of course, to invent one must begin to REALize the concepts. It is an iterative process where starting by DOing one thing, even without a complete or concrete idea of what the result will be. This causes more dots to connect and the creation to improve. Back to the example, the first idea was that if an air compressor compressed room air, it wouldn't have to heat it very much to have it run a heat radiator, so the COP could be high. But how could room air heat room air? The cooled air from the radiator, having actually lost heat to the room, would be much cooler once it was decompressed and would have to be vented outside. So cold air would have to come from outside to replace it. At that point in the idea, it sounded like it might be a "zero sum game". There some might think "Yeah, that figures!" and go on to something else.
   But while still compressed, that air still has compression energy. So came the idea of having it somehow help to spin the compressor - "ROVACS" and other compressor-decompressor ideas, "ending" (as far as the concept has gone so far) with the idea of having it turn a turbine that mechanicly helps turns the compressor. (probably by being on the same shaft)
   Also this compressed air from the radiator unit is still a little warmer than the room air - more "waste" energy! The idea of the passive heat exchanger came to me along with the other while putting together the basic "compressor plus radiator" unit. Most of this evolving concept can be seen in TE News issues from about January to April 2020, with conceptual improvements to the rotary compressor, and then the decompressor turbine idea, in more recent issues.


AC Body Voltage Levels

* It seems to me - speaking only for myself - that the ringing in my ears is only just starting to fade perceptibly after 8 hours sleeping in my "Faraday cage" cabin (without AC wiring and where I generally measure AC body voltages under around 15mV). And "perceptibly" is only because I'm listening for it. But it gets louder again in under half an hour in the house. (...where fields are mostly well over a volt and again because I'm listening for it.) I might almost say - for me - "days to fade, minutes to resume". If others are experiencing anything like that, it's no wonder it's little understood where tinnitus comes from since nearly all of us are so rarely out of electric AC voltage gradient fields for any length of time.

   The conductive fabric tuque/beanie actually helps more than I realized, but it's not the same as just being away from electric fields. I didn't realize how much it was helping until the day I washed it and set it out to dry. The ringing was much worse after a day without it -- back to how it used to be. Even overnight in the cabin only brought it down somewhat. The tinfoil over the car's sun visor also seems to reduce the ringing I notice after highway drives.

* I was guessing that the AC induced body voltage may tend to cause tinnitus if it is over 20-50mV. One night in the cabin I felt by 3 AM the ringing didn't seem to be fading at all. It felt like my ears were still "under attack". Why should that be? I had been working on the siding and had plugged the angle grinder into the cord that ran around outside the building. But I had brought it in in case it rained, still plugged in. There was about 10 feet of energized cord on the ground about nine feet directly under my bed. A meter said it was just around 40mV of body voltage lying on my bed. Apparently that was enough to be a problem! I unplugged it and it was still over 20mV coming from the few, short cables with the plug-in grid ties in the corner over 20 feet from the bed. If 40mV was noticeably too much, even 20 probably was having a mild but continuous deleterious effect. I unplugged the building - the one plug in the far corner - and tossed the extension cord outside. It went down to around 5mV.
   Of course that's just me. I'm not sure where it might start for others and for those most affected. Some are doubtless unaffected for years at much higher levels, but it's starting to look to me like many or most older people have ringing ears. It's just not talked about unless I bring up the subject. But I think the Greenhome Institute is right: "no exposure" is under 10 mV, not 20 or 50 mV.

Greenhome Institute gives the following induced AC body voltages for AC electric field exposure:

0-10mV - no exposure
10-100mV - mild exposure
.1-1V - strong exposure
over 1V - extreme exposure

   Induced AC Body Voltage is easily measured by connecting one probe of an ordinary AC voltmeter (precision preferably to .001 V AC) to ground, wetting one's thumb and finger, and gripping the other probe, and not moving during the reading. (Don't touch the ground connection.) Ground may be actual damp ground or the grounding pin of a power socket (can be an extension cord ground pin). Position is important in stronger fields. Outdoors at least, measured voltage is much higher when standing up than crouched or sitting.



ESD
(Eccentric Silliness Department)

* In their protracted march of capturing or liberating (depending on your point of view) Donbass towns and villages one by one (...more like "liquidating" them where the fighting is hard), the Russians have achieved pobieda (victory) in Pobieda, have progressed through Progress, and now are in the middle of New York. (Wait, what?!?)



   "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 just thought of and not tried... 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, consistency, completeness and elimination of duplications before publication. I hope they may 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

No Reports




Other "Green" & Electric Equipment Projects

Some Thoughts on DC Houses & DC Appliances


36 V DC is the optimum voltage for most in-home/in-building wiring. IMHO. Since the electrification of everywhere a century ago we have become so used to wiring with 120 VAC everywhere (230V in much of the world) that we never think of doing it any other way. The idea it's less than ideal and the whole thing ought to be re-thought will probably come as a shock to most people. The advantages of 36 volts DC haven't been thought out by most of us, but working with solar, and then discovering health reasons, I was gradually led to this logical conclusion and am wiring my cabin with it instead of 120 V AC.

What's good about it?:

1. DC doesn't induce oscillating electric fields into the body, which may cause tinnitus, ongoing vague feelings of stress and unease, and other health effects whose cause is usually unrecognized. (See GreenHomeInstitute.org for more on this topic.) So AC wiring is to be avoided where practical, including AC generated from a DC system with an inverter. Runs of AC wiring should be kept very short and few, or made with shielded cables and cords, or run inside (grounded) metallic conduit pipe. In theory, it's easy to avoid the need for AC except for higher powered equipment. Lighting in particular is now very low power with LED's, which emitters are already DC, and it is almost absurd to run 120 or 240 V AC wiring to then be reduced to low voltage DC at the fixture or "bulb".

2. 36V DC nominal is the highest electricly "safe" voltage. Electrocutions from 36V and below are not unheard of, but they are very rare. 48V DC or higher building wiring would come with some electrocution death statistics, as do the present 120V and 240V AC. (Think that if three 12V batteries are under charge at 15V, nominal 36V would hit 45 actual volts, and nominal 48V with four can reach 60 V. High DC voltages can be more dangerous than AC.)

3. Line losses at 36V are much lower than with 12V, and substantially lower than 24V. Wire gauges for higher power appliances at 12V are huge. Consider running an electric heater/kettle/toaster/burner of 1000 watts. This requires a circuit of: 12v = 83 amps = #4 AWG wire = very heavy and unwieldy; 24v = 42 A = #8 wire (better, #6); 36V = 28 amps = #10 AWG wire = much more managable. (For off grid situations, one probably would only use such high power for short intervals, EG boiling water for coffee or tea, heating something in a microwave or on a stove, vacuuming...)

4. From 36V it's easy to get any lower voltage, such as 12 or 24V for 12 or 24V DC appliances. There are low-cost (5$?) 20 amp screwdriver adjustable DC to DC converters to go from any voltage up to 40, to any lower voltage. These - just one device - can replace most any and all "wall wart" 120V AC to DC power converters to power a multitude of electronic devices. And there are even cheaper ones (2$?) with lower current ratings. So there would seem to be little reason to go to lower voltage wiring across a home or building. (Prices are for the electronics board only - as currently manufactured, 36V plug & 12V socket, case (3D printed?) need to be added.

5. Ease of Mind. No worry about getting electrocuted by touching a damp power cord in the rain, laundry or bathroom, or by flipping a light switch with wet hands. Kids aren't going to shock themselves by touching the wrong thing.

6. Today there are confusing choices for DC power of 12, 24, 36, 48 volts and beyond. It is important to establish a "standard" wiring voltage for DC in order to induce manufacturers to mass produce DC appliances. 36V, while presently somewhat overlooked, is the best voltage to do that for the reasons above. It should be encouraged, promoted and adopted in new power & building projects. It's only an uphill struggle until there's enough momentum to reinforce the logic of it, then everyone will jump aboard.

   I myself have "adopted" T-Plugs/T-Sockets (AKA Deans Plugs) as the standard for 36V house wiring. I have 3D printed triple and quadruple wall receptacle plates, plug and socket shells and more to facilitate adoption of 36 V. They should be good for up to 20 amps. ("Plus" is the "cross" blade on the right in the image.)
(I do recommend an improved version T-Plug be manufactured with 10mm long blades instead of 8mm, because they are too easily pulled out of the socket. Also other choices of plug styles are needed besides with "solder pads" to attach wires to the plugs.)


   And I created a "giant T-Plug" & Socket for currents to 50 amps. These are harder because everything is made from scratch - they need to be manufactured. But manufacturing keeps getting easier as new techniques and tools are created. (...and I must upload all the 3D print designs to thingiverse.com for all to use!)


   Various power supplies can be had to convert 120V AC from the power company to DC. Of course, DC is an even better choice for "off grid" wiring compared to 120V, which requires an inverter that itself uses extra power. Lithium-iron phosphate batteries (LiFePO3 - 12 cells) with a "balance charger" attached presently seem to be the best and safest choice, now available with ratings in hundreds of amp-hours. "Golf cart batteries" or other Pb-Pb cell costs are rising and they have much shorter life spans, while the price of LiFePO3 has come down substantially and is still dropping. (With those, "36V" is generally close to 40 volts if they are charged, and I'm tempted to call LiFePO3 systems "40V" instead of "36V". I think it sounds better! They can be charged to 42+ volts at room temperature, but little energy is gained, they drop to ~40V at low temperatures anyway and DC to DC converters are mostly rated only up to 40.0 volts.)

   Having created wall outlets for 12 VDC ("micro T-Plugs" & sockets) and 36 VDC ("T-Plugs" & sockets), and with wiring the cabin for 36, I thought I'd look for DC appliances to match. Being that there is no "consensus" DC wiring voltage, the confusion of choices from 12 to 48 volts and beyond has made manufacturers hesitant to make DC appliances.

Note for 24V DC: With no "standard" wall receptacles, sockets or plugs being available, the 36V T-Plug outlets, cords and plugs could well be used for 24 V as well. A number of useful circuits such as LED "light bulbs" and DC to DC down converters to make 12V can be run from either voltage or in between. It does however make a "user beware" situation since one could easily plug in a 24V appliance into 36V or vice versa, possibly destructively.

Lighting
This lamp works the same, looks the same as any other but
has a DC LED bulb, "T-plug" plug, and runs off 36 V DC  
   Now that LED's have replaced most other lighting technologies, DC lights are simple. LED's are inherently low voltage DC components (generally 2.9V for lighting LED emitters. A fast, simple way to get nice DC lamps is to take a nice AC lamp, cut the plug off, and put on a 12V or 36V plug. Then one buys a DC LED "light bulb" that screws into a regular ("E27") lamp socket. There are "12V" bulbs, and there are "12 to 72V" bulbs, which are ideal at 36V. (Note: These last are dimmer - lower power than labeled - when run at 12V.)


   There are various 12V LED lights (for automotive, marine, off-grid...) which can be run off 36V using an adjustable DC to DC converter set to 12-14V output. (A 20 amp adjustable DC to DC down converter running off 36 or 24 volts can replace any "AC Power Adapter".)
   I bought some beautiful, bright, 30x60cm (1x2 feet) flat LED light panels that work admirably at 36 volts, with just a 5Ω resistor to limit the current and set them near or under their 24 watt rating.

   Blue emitters for LED lighting were invented in the 1990s by a Japanese man (I forget his name) who pursued it for decades even after everyone else had given up. Without him we would still be using incandescent and fluorescent lighting. His Nobel prize was well earned. They make lighting colors the same way fluorescent tubes do, by shining the single frequency LED emitter on a phosphor that glows in a broad spectrum centered at the desired color temperature. But unlike fluorescent lights, they don't have a strong spike of retina-degrading intensity at the mercury vapor arc wavelength that comes right through the phosphor, and so they are far more pleasant and beautiful. And incredibly efficient, requiring perhaps 10-15% as much power as an incandescent light bulb of equivalent brightness. Typical "color temperatures" (in degrees Kelvin) available for LED lighting of any style include:

2700°K - orangeish/warm
3000-3300°K - yellowish/"warm white"
~4000°K - "neutral white"/mid-spectrum (sometimes seem faintly greenish IMHO when seen next to yellowish & blueish. My favorite for most lighting)
5000+°K - blueish/"daylight" (plant lights; some like even 6000K for work lights)

   Wiring specificly for DC LED lighting (ie, wires not going to outlets or heavier loads) need not be heavy. I'm using #18 AWG speaker wire. It looks patheticly thin, but #18 is rated for at least 5 amps, and even a bright 36 watt LED light draws only 1 amp. It's easy to run and cheaper than heavy house wire. (I do wish I had a cable stapler for it - even simpler.) For 36 volts, insulation need not be heavy as it is with #18 "lamp cord" wire.

Fridges, Freezers, Washers, Dryers, et Cetera

   For such actual high powered appliances, the pickings seem quite slim. The are 12 & 24 volt fridges, freezers, washers and dryers, but the usual price tag of several hundred dollars has a 1000 or 2000 $ amount added on the front. There will surely come a day when 36 volt units with highly efficient BLDC motors will come on the market for similar prices as 120/240 volt AC units, but it's not yet. I did however buy a 24 volt BLDC deep well pump for just 300+$ - a bargain for any deep well pump - so there is progress. DC power tools aren't available except those that have their own battery - which is more convenient than a cord anyway when it works for your work.
   Accepting a DC to AC inverter with its small but extra continuous power draw and what may someday be called "legacy" AC appliances seems to be the only practical route today. One can of course turn off the inverter except when an "on demand" use appliance is to be used (ie, vacuum cleaner, not fridge or freezer), and if the cords are kept short, they will generate little EMF into the space except near and between the inverter and the appliance(s).

   Of course, having 36 V wall outlets doesn't preclude running 24 or 12 volt appliances such as fridges and freezers by using a DC to DC down converter "power adapter". But IMHO appliances would be better made with 12 & 36 volt options rather than 12 & 24. Or be made more flexible to handle 24 to 36 volts supply.

Electric Heaters

   Electric heaters and hotplates having nothing but a resistance heating element and mechanical switching controls can be run at their rated voltage or any lower one. At 38 volts, each unit will run at 1/10th of its 120 volt power rating. So a 1500 W heater will run at 36V as a roughly 150 W heater.
   I made a little "cheater cord" (from an only extension cord) with a 36 V T-Plug on the end, to plug such appliances into 36 V withut chopping of their regular plugs. This past spring I was heating my bedroom at night with two or three such heaters: 150, 80 & 50 watts - just about 200, 250 or 300 W on the consumption meter. The 290 AH, 36V battery is recharged daily (usually by around noon) by five solar panels. This certainly isn't enough heat for winter, and the solar panels probably wouldn't charge it back up daily in winter with this large all-night drain, but it's good in the "shoulder seasons" when one only needs a little heat.
   Okay... On AliExpress I've found low cost 12/24 V electric heaters. Some are portable car heaters/window defoggers. There might be some suitable; I'd have to check further. A 24V heater run at 36V would be 2.25x the rated power and might burn out. A 48V heater run at 36V would be 56% of the stated power. They seem to offer similar lower power ranges to those I've been using.

Hotplates/Stove Burners

   I put a cup of water in a pot, put it on an "1150 W, 120 V" simple burner hot plate, and plugged it into 36 V with a "cheater cord", making it only ~115 watts. It took the better part of an hour before the water was simmering softly. It's just not enough power - it takes so long that much of the heat is radiated to the room while the water is heating.
   But I do see 12 & 24 volt hotplates ("DC Stoves") of several hundred watts on AliExpress. There was a single burner induction cooker "12/24/48 V". There was just one model, so it would almost doubtless work also at 36V too. It looked promising as induction transfers the heat really well to the pan (which has to be magnetic metal), but it was over 200$C.

Hot Water Heaters

   A 3000 W, 230 V water heater plugged into 36 volts would heat unbearably slowly, at under 3% of its usual power. My own house hot water tank is powered by 120 V instead of 240 and is hence 750 watts instead of 3000. When I turned it off and went away for a week, it took 8 hours to be hot enough for a shower. Except for that, with one or two people I have never been short of hot water. By using a lower power, if I use hot water during the day, there's often enough power (ie, 750 W) coming from the solar panels through the grid ties that it's being reheated from them instead of much of it coming from the power company. Water tank elements for various low voltages are available too (eg, "Dernord" elements). I have a 15 liter "under sink" tank presently with its original 1200 W, 120 V heating element. At 38 V that would be 120 watts. That will eventually heat the tank, so it works well provided you don't need to use the hot water often, or use more water than is held in the tank. (I un-installed it from my kitchen owing to a well water chemical issue - the hot water reeked.) At one time I had changed the element for a 36 volt one (it had 3 x 400 W elements), but found 1200 or 400 watts was a bit much for my solar system with small batteries at the time, so I ended up at around 150 W anyway by putting the three elements in series instead of parallel, or 200 W with two in series. One should be able to come up with almost any desired wattage range.



Cabin Construction & DC Wiring

Main Wiring Board/Closet

   I continued adding to the surface mounted wiring board. I installed the ground bus made in June. I 3D printed a four-outlet wall plate to plug several ceiling lights into rather than hard-wiring them. I suppose one could fit 5 or 6 T-sockets on one wall plate. No "power bars" to get enough outlets! This box is on the charge controller's "output" circuit, which is set to shut off if the battery gets the least bit low - which would probably be from leaving too many lights on in the winter.
   A problem I've run into is that I always want to have a negative ground, but in most charge controllers it's the negative side that switches, both for the solar panels and the load output. They are "positive ground". The reason for this is that n-channel MOSFETs are better for switching power than p-channel, so it's easier or cheaper to manufacture. That doesn't mean it doesn't create headaches. Does one ground the solar panels, the batteries, or the main ground bus? What about the lights on the low-power switch-off circuit? They can't be grounded. The 36 volts "+" stays live when it's off, and the "-" becomes 36 volts as well. If it was higher voltage it could be dangerous.


36 V DC Wall Receptacle

   On the 10th I wired up a wall receptacle in the upstairs room. I took off a piece of gyproc from under the window and screwed in an electrical box.



DC Wall Outlets/Wiring: Wire Colors and Connections

   Since AC wiring uses different colors than DC wiring and is the most commonly available wire, there is often no red wire. (What bozo decided black should be the "hot" wire in AC wiring, anyway? It's the color of no energy, of earth, absorbing all light -- and somebody made it "hot"?!?)
   For DC wiring we'll stick to black, or bare or green per AC standard, for negative or ground. If using a bare wire, be sure it can't short to anything.

   The other normally available color is white, so that will have to substitute for red in many situations. If deviating from this, wires should be wrapped with colored tape to make their usage clear.

Black, bare, green: minus
Red, white: plus

   What is the purpose of grounding the metal box with AC wiring? My understanding is that it is so that if a "hot" wire touches the box, it will blow the circuit breaker to avoid a shock hazard if the box or a screw is touched. At 36 volts, there is no significant shock hazard, so it's better to leave the box unconnected. If a +36V connection (and no other) touches the box, no harm is done and the breaker isn't needlessly blown.

   As I see it, there are two distinct types of circuits for DC wiring: "LED lighting only" circuits and "other appliance" circuits. "Other appliances" using considerable power may need heavier wires than for 120/240 V AC. (And allowing thinner wires is the main reason such high voltages are commonly used.)

   A single LED light of 24 watts is a pretty bright light, yet the current to drive it at 36 volts is just 2/3 of an amp. Even light #18 AWG "lamp cord" will handle 5 amps - enough for any typical lighting. Plus, with no need for higher voltage insulation, even #18 "speaker wire" is fine. (Note: Watch out for cheap non-copper speaker wire. It is coated to look like copper wire, but the resistance is much higher. Silvery color on cut ends betrays this "fake" wire. Even that is probably sufficient for a lower power LED light or two.)
   Wires for outlets or other applications that will or might have higher power appliances connected need much heavier wires. (But not so heavy as for 24 or 12 volts!)

   We don't really know what appliances are or will become available for 36 V DC power supplies. (We do know that 36 V DC can be converted to any lower voltage for lower powered circuits with a low cost power adapter, and that there are some 12 to 24 volt appliances presently available. And or course 120 V AC appliances can be run with a DC to AC inverter. 36 V to 120 V inverters are available.)

   We typicly run circuits at up to 85% of the circuit breaker current. #14 AWG wire running 12.5 amps at 120 volts is 1500 watts - the heaviest appliances commonly plugged into a 120 V AC wall outlet. That current at 36 volts is just 450 watts. That's just enough for a toaster, kettle, hot plate, microwave oven or coffee maker, or a vacuum cleaner, but it will be a slow one. #12 wire (w. 20 amp breaker) or heavier is preferable for wall outlets. Running 17 amps at 36 volts would give appliances of just over 600 watts.
   #10 wire (rated 30 amps, 30 amp breaker) allows 25.5 amps, allowing just over 900 watt appliances.

   Appliances over 1000 watts will need #8 AWG wire and heavier plugs, like the "high current T-Plug" & socket I made just samples of by hand some years back. ( TE News #161 ) To my chagrin the size specs are (presumably) in some earlier issue I didn't readily locate. They are: blade protrusion, 2mm x 8 mm x 15 mm; spacing between blade centers 12mm, making a 7mm gap.

   3D printed nylon socket bodies should be acceptable. As I didn't have nylon print filament when I was doing them, I made them of porcelain. (On the plug the contacts are outside. On the socket they are inside and heat from a poor connection could melt many plastics.)

(I really must upload all my designs to thingiverse.com for anyone to use!)


Most off-grid installations will have trouble supplying such higher-powered devices for extended periods. (Of course "on grid" buildings could use the AC to drive a heavy 36 V DC power supply and run all the DC things they wish at low voltage DC.) It is especially desirable to keep AC wiring away from bedrooms (along with WiFi and cell phones) for health reasons. Many people don't give their bodies many breaks from such mild but eventually deleterious emissions.



   I also made a "double T-Plug" idea for heavier circuits, maybe 15 or 20 to about 30 amps. I thought it would be good because it uses the same T-Plugs & Sockets. Frankly I don't think I recommend this, but it consists of two T-Plugs/T-Sockets beside each other exactly 12mm apart. I spaced some wall outlets that way and made shells to hold the T-Plugs at that spacing.
   The "giant T-Plug" is a better idea.


   In early August I used the new one upstairs in the cabin to power a 2500 W PSW inverter (36 VDC to 120 VAC) and vacuumed the room. The double plug wanted to pull out at the slightest stress on the heavy wires. (This is why I wish the T-Plugs had 10mm blades instead of 8mm, but it would have only been somewhat better. The "giant T-Plug" for high currents is the real answer... but they need to be manufactured.)

   Most vacuum cleaners are somewhere near 1500 watts. Theoreticly that should have blown the 20 amp (36V) breaker. I had a lower powered shop vac (850 W), but also a vac that had buttons to increase or decrease power. At full power the inverter (2500W?!?) balked after about half a minute, quitting and saying "Fault", but if I turned the vac to setting 3 of 5, it ran fine and I did the job.


   For said wall outlet I ran a #14 AWG - 3 wire cable from the DC "wiring closet" under the stairs straight to the outlet box. Since the bare "ground" wire is also #14, that makes four wires. The way "AWG" works, the diameter of a wire is half of one three gauges down. This means that two #14 AWG wires together have the same cross section as a single #11. Best breaker rating for #11 would be 25 amps, allowing up to 900 watt appliances to be plugged in.
   I may want to run a low wattage electric heater at night or potentially an inverter to run an AC vacuum cleaner. (I have a lower powered 850 W shop vac. Many vacs need a full 1500 watts.)
   I insulated the bare wire by stripping pieces of black insulation from a scrap of #12 AWG wire (seen in front), which are just a little fatter than #14 and slip on easily.


   I tied red and white together, and the two blacks, and tied them to the pigtails of the T-Plug (36V DC) wall plate.


   I used a moderately deep box for the wires plus the sockets to fit into.


   When I put the wall plate on I discovered the box stuck way out of the wall. Oops!


   It worked but after a couple of weeks I took the crumbling piece of gyproc off and moved the box back. Still not quite flush. But much better.  (How can it be crooked when it's screwed to the wall stud, top and bottom?)


Cabin Construction

   The end of the second set of ceiling sheets didn't quite reach the last rafter, with this one central rafter gap being more than 24 inches wide. What to do? I finally opted for screwing in a 2 by 2 onto the side of the rafter an inch above the bottom, then screwing a 1 by 8 to that, horizontal, so it was flush with the bottom of the rafter. That gave 10 inches width to screw ceiling panels to.



Horrid Fiberglass

[14th] I finally got the nerve and put on crappy old clothes to continue insulating the upstairs room ceiling. It was even worse than I expected. I finally used the last loose pieces of R28 insulation I got used, but the "new" bag was just as bad. It was hard to tell where one so-called "batt" ended and the next began - they all just stuck together almost as if one big fat piece. Whenever it was touched, clouds of very fine dust came off it, and bigger bits continually fell off like snow. Not only would it not stay put in the ceiling spaces between the rafters, but it would hang from a string like a limp towel. In the lower parts I could put in strips of plastic and then slide the "batts" in, but I had to string the top part. I used some old twine and roofing nails to hold it once in place.


   This particular insulation reminded me of some from 1976-77 when manufacturers had decided to used "recycled glass" to make fiberglass. It turned out that many fibers were the same shape and size as asbestos, hence were carcinogenic. I had used that horrid stuff under my house in 1977-78, and this "new" stuff didn't seem any better. What luck! Hmm... had it lain around all these years since that period? I finally thought to look for a URL on the packages. If there wasn't one, it could potentially be that old. There wasn't, but there was a date: Dec. 2001. Maybe it had degraded over a couple of decades of imperfect storage?



I had to vacuum up the "snow" that fell in simply cutting
one batt into four short pieces with a clean, sharp knife.

   I tried my very best to get almost any nicer insulation than fiberglass. I finally decided dried lawn grass (~R3 per inch - investigated in TE News #170 & previous, two years ago) just wasn't good enough for the walls, no one in northern BC would sell me common cellulose fiber loose fill (R4 per inch), and nice looking organic fiber batts (cellulose, wool...) that I saw on line in Europe likewise aren't to be had around here. The only other choice is mineral wool batts, which cost substantially more instead of less. So I ended up with fiberglass, bad enough any time, and someone who was moving sold me his old R28 batts for the ceiling, cheap. (It wouldn't have been so bad under the floor, or at least in the walls! I'll have to seal the ceiling really well.) I opened a "new" bag for two more batts and found big rat holes in them. Yetch!

   Having managed to put in thirteen fiberglass batts and three sheets of coroplast ceiling, my endurance completely ran out and I had to leave the last three batts and last sheet for the next morning.

[15th] I was very glad to finish the job and clean up and vacuum the room. I went for a walk then started checking out the window frame.


   I stepped into the house just in time to hear the last ring before the answering machine came on. It was McKenzie furniture with the "Ashley Pillowtop 2.0" mattress, and they were just down the road! I opened the gate and walked down to the highway to see the truck just coming around the bend! The big mattress came in a small, very heavy box. (14" x 14" x queen mattress width) Was that really it? All of it? It was all vacuum packed and rolled up to an amazingly small size. When I got it out of the box, unwrapped the wrap and cut into the plastic bag, air hissed in and it started expanding. Then it started to look like a real, thick, plush mattress. "Allow 24-48 hours for full recovery." Amazing!
   Owing to it being an ucky construction site I left it in the plastic. But I brought down some old bedding and had a nap there, and then started "camping" in it at night. I put a plastic tarp over the whole bed during the days.

   It could almost be my imagination that my tinnitus is quieter by morning, but it's definitely not imagination that it gets louder fairly rapidly when I get back into "normal" electric fields. The conductive fabric tuque/beanie actually helps a lot, but it's not the same as just being away from electric fields. (I find this especially on the day I've washed it and set it out to dry. The ringing is much worse after a day without it.) Lots of older people say they have ringing in their ears when I bring up the subject. "Oh, ya, my ears are whistling away as we speak." The more I know and experience, the more I think 120/240 VAC wiring should be eliminated throughout most of most dwellings.


[17th] I had a double pane window which had been two parts, one sliding, one fixed. I had split the double pane and got it out of the frame successfully (to clean the inside -- in spite of the sticky black tape/goo). Now I cut the frame down to use just the fixed part and spent the day roughing it into place for the bedroom window. I had to cut a small notch in the rafter and trim the ceiling a bit, so it was certainly the biggest window that could fit. I put the frame in backwards so that I could put in or remove (to clean) the window glass from inside instead of from the outside on a tall ladder.
   Now the midges ("no-see-ums") would have a much harder time freely zipping in and biting me, which makes me horribly itchy.
   Only a little triangle space was left above the window to place a vent that would open. I guess it will have to do!



[19th] I continued sleeping in the cabin. This gave me great incentive to finish the last wall. I made a bizarre mistake on the framing under the center beam: the space was 49 inches tall. Four inches had to be subtracted from the studs for the 2 by 4's, and somewhere overnight my mind had imprinted "39 inches". I never questioned or verified it so of course it came out 6 inches too short. I put it up anyway. The top part with the angles and every post a different length didn't go perfectly either, but it ended up right.


   I ran the gyproc verticly, floor to ceiling, which essentially nullified the framing mistake. I also found I could remove the heavy beam joiner on the bedroom side, leaving just the one on the other side and here just bolt ends to be plastered over.

(Only one sheet of badly stored gyproc that got wet left to use up. Yay!)




Gardening

   The raspberries are producing far more than I wanted to devote the time to pick. Still, I've filled some margarine tubs and put them in the freezer in addition to frequently having a good helping of them with cereal. Friends who dropped by picked a bread bag full one day. Judging by the canes growing for next year, I may have to get a ladder to pick them!


Corn

   By mid July the corn was starting to grow ears! To my surprise, the ones nearer the back were doing much better than the ones at the front, some of which had actually died in the spring. Was not enough light coming in near the front... or did it get colder around the edges in the spring? Hmm... There were pretty big cracks around the edges. I thought they'd be fine, but the stalks in the middle and toward the back probably stayed warmer. Next year I'll reuse the same 'greenhouse box' [TE News #191 - scroll to bottom of "April in Brief"] but close around the edges better. (And maybe find something that lets more light through for the sides?)
   On August 7th I picked and cooked my first delicious cob of corn.


Sweet Potatos

   My friend Kamil said he had some sweet potatos from the store that were starting to sprout. I've never seen them sprouting before, so that was interesting. My brother says he grows them in his greenhouse in Toronto. But I was sure it was too cool to grow them here. He reminded me that I had some all-styrofoam cooler box cubes. If I covered them with a sheet of glass and put dirt in them, the sun would heat up the dirt and they would stay warm inside.
   That sounded like a plan. So I cut a hole in the bottoms of two and lined them with clear LDPE, and put in a few inches of old compost, and half a tuber with some good eyes in each. I put the boxes in the greenhouse, watered them and put a couple of broken glass or acrylic plastic over them.


   In a week they were both sprouted! The dirt is pretty shallow but can be deepened as they grow, and it will hopefully fill with tubers. (Now, what about yams? I think they need it even warmer, but might it be worth a try? I guess I should see how these do first.)

  In a couple of weeks the leaves were pressing on the glass. Someone suggested just using soft plastic so they could lift it and keep growing. That seemed to work well, and then I added some dirt to each box.



   The coffee bushes flowered and started growing new beans. I saw some nice flower petals but when I finally went to get a picture they were gone.
   I've given them bone meal, lime, nitrogen and so on, but they could really use bigger pots. Then what when they get too tall for inside under lights in winter?





Electricity Storage

The
Copper Oxyhydroxide
Nickel-Manganates
Nickel Hydroxide
Nickel Oxide - Zinc Zincate Cells

Zincate

   I have already made a hugely important advance in battery technology by creating a zinc electrode that will last 'forever'. (How could it possibly have taken 16 years?) I think I'll call it the Zincate electrode to distinguish it from all the "failed" (single use or short cycle life) Zinc electrodes out there. After all, the key is that when discharged the zinc substance remains as a "supersaturated" zincate solution that never turns into zinc oxide.


Nickel Manganese Oxides

[12th] After months and months with little progress I decided I was fed up with trying to get copper hydroxides to work - for the moment - and to try a tube electrode with nickel-manganates again. (eg, NiMn2O4... oxidation state varies with state of charge.) I had always tried charging these cells up to nickel oxyhydroxide voltage (w. zinc, almost 2.0V) instead of to manganese dioxide voltage (1.4-1.5V), and now I'm pretty sure that was wrong. I may have been charging the manganese into soluble forms, MnO4-, MnO4--, which would gradually migrate to the minus electrode, reduce, and coat it with Mn(OH)2 to block the electrolyte.

   In the cupboard I found a jar with a good quantity (187g) of mix I had made in December 2022 (TE News #175), which apparently was made with Mn substance from old dry cells using the following percentages:

Ni(OH)2 - 29 g
Dry cell Powder ("MnO2" + CCB) - 63g
Sm2O3 - 3 g
Additional Graphite or CCB power - 5 g

(proportions of Ni to Mn are approximate)

   There was no mention of acetone in #175, so I must have discovered acetone after that. Powders simply mixed together aren't the same as when chemicly combined. I have the impression that's another reason the cells degraded with cycling. Acetone dissolves the oxide powders and as it evaporates they combine them into epitaxial co-precipitated crystalline forms. At least, this is what I think happens - the supposition under which I'm doing it.
   So that was the first step. It's amazing how crunchy powders dissolve into a smooth paste in acetone. But it would need a couple of days to evaporate, so I wasn't going to slap together a NiMn2O4 electrode in an hour in the morning as I had unthinkingly thought.

   Being solid substances with no dissolving ions, probably this electrode would work fine without osmium and SDBS. OTOH, if I used SDBS in the separator, hopefully anything that did dissolve would stay within the electrode anyway, and the 'forever' life of a gelled electrode might apply.

[14th] I keep stirring it. Getting to be more of a loose powder again. Still a faint smell of acetone... I spread some over the bottom of a shallow container. By night the smell there was little.

[15th] The tube with the carbon rod weighed 30.90 grams. I added NiMn2O4(?) powder mix and it was 42.75, so 11.85 grams of mix. Plunging in the tapered rod and then the tapered carbon rod... and scraping some out with a screwdriver because the rod wouldn't go in far enough... I think I got it pretty compacted, and about 10mm below the top. I topped it up with heat glue.
   I put it in the cell and after bubbling a while it read about 1.28V. That seemed promising. I briefly tried a couple of loads: with 50 Ω it dropped by about 100mV and with 10 Ω, 350mV. Most of the drop was the NiMn2O4 side. The two old zinc tubes were pretty good. I set the power supply to 1.6V to charge it knowing that Mn-Zn is about 1.5V almost regardless of pH, and that I am expecting to get a rechargeable version of MnO2 with the mixed NiMn2O4 oxide. It started charging at ~50mA [9:35 PM PST], limited by crappy alligator clip connections and other resistances, since the voltage only read 1.55V instead of 1.6V. This rose toward 1.6 as the charge current dropped. (If I start soldering connections to the cell, it gets hard to replace electrodes.)


   (Note: from here, ' = minutes, " = seconds...) In 15' of charging (which dropped to 20mA) I tried a 20 Ω load test. It started out at 1.45V O/C and dropped to 1.1V in 1'. After 4' it was down to .977V and I stopped. This seemed about right since it was discharging at over 50mA after such a short charge. It recovered to 1.206V in one minute and started recharging at 60mA. Would it continue performing similarly but run longer with more charge, or disappoint me like my copper "+" trodes? After another 30' of charging (down to 15mA) I tried again and it ran 5', staying over 1.1V for 2'. Recovery 1.184V. Recharge started at 82mA (momentary max.). Another test... only slightly better. ...Mañana is good enough for me!

[16th] A test early in the mañana yielded only slightly better results than the ones charged just half an hour. Huh? I did another run with 50 Ω instead of 20, and it ran down to 1.000V in 20.0'. As anticipated, it doesn't seem to be as good as the copper when that is working well. [Later: But it also just didn't seem much good. It was disheartening.]

   How many square centimeters of interface area was there, anyway? The last CNC cut tube had 19 rows of 18 holes, 342 holes. I measure the drill bit as about 2.75mm (.275cm).

.275 ^sq. * π/4 = .0594 sq.cm. - per hole
.0594 * 342 = 20.3sq.cm. ... The tube was just about the same as the 50x50mm square electrodes. That is assuming of course that the holes in the "+" tube are the greatest bottleneck, the point of least area. It's harder to calculate for the 3D printed "minus" 'trodes but there are two zinc "minuses" in the cell.

   But this (earliest) tube had many smaller holes. Maybe 15 sq.cm? Then, it put out 300mA into ".1Ω", which is just 20 mA/sq.cm. Of course, I probably should try it without soduim sulfate in the electrolyte. I had added that since it seemed to do a good job of changing copper into CuOH. ...or does that make any significant difference?

   Surely I'm stumbling around in the dark. None of the positive electrodes of whatever chemistry I try seem to have even a fraction of the amp-hours that I think they ought to, deteriorating performance, and (except for using cupro-nickel sheet metal), very low current drive as well. The commercial dry cell nickel oxyhydroxide worked better than my homemade "+" 'trodes, but not nearly as well as it did in the original Ni-MH dry cell. What am I missing? Could it simply be that I couldn't seem to seal the lids to keep air out? Could a bit of air getting into the electrolyte really matter? After all, most of the early researchers had open top beakers and jars. But what if it affects the charging, hour after hour? There was the next thing to try!
   This jar had a plastic lid. I drilled appropriate hole in it, filled the jar almost to the brim, poked the terminals through the holes and screwed the lid on. Then set it on to charge again.

   That done, it was much weaker. It seemed the carbon rod just wasn't making as good a connection inside the electrode. When I pushed on it, performance improved again. When I was done, it was about what it was before. The idea behind tapered rods was that when pushed in, they would compact against the electrode substance and connect better. The flip side is if it comes out a bit, it loses connection all around. Is there some better way to do it? (How about carbon rods with screw threads? Well, they'd probably break putting them in.) But there seems to be a limit beyond which pushing the rod in farther or harder is no help. I guess at that point it's already as connected as it gets.

[17th] As I rather expected since I couldn't think of any chemical problem except very, very gradual CO2 contamination of the KOH, the lid on the jar seems to do nothing but keep the electrolyte from evaporating... which anyway is a blessing. After taking out the "+" trode and pushing the rod in pretty hard, there was no improvement over the previous day. Once again I'm baffled as to why it doesn't perform far better than it does.


   I had the CNC machine drill out another electrode tube. I adjusted a setting and got 18 holes and 26 rows, 468 holes times .0594 sq.cm per hole - about 28 sq.cm. Except maybe the drill didn't go quite deep enough and left many 'bottoms' attached. They only went back flat into the holes when rubbed. I poked ?150? "hanging chads" with a paper clip to get them standing up inside again and then scraped them off with a curved file. I put in a separator paper with all including osmium.
   When the tube was ready I put in 100% the old monel mix from long ago - the baked/fried mix. It didn't work any better.


The Path of Lowest Resistance: Tried and True Nickel Oxyhydroxide

[18th] OKAY... The zincate was the vital thing. I've been unable to improve on the other side of the cell, the positive electrode, with a new and better chemistry.

   The nickel manganates doesn't seems to work well - at least, not if only charged to MnO2 voltage. In previous experiments at nickel oxides voltages it had seemed to deteriorate. The copper hydroxides/copper ions electrode seems to deteriorate and I don't see why and haven't found a solution.

   There is almost surely some way or multiple ways to make either and both of these things work, but for myself, I think it's time to give up on them. I've simply had enough of failures.


At this point, I think there are two choices:

1. Delve into the uncharted realms of zinc-air cells. This probably has even more hurdles and unknowables than what I've already been doing. I think I'll pass. Hopefully the U of Waterloo is still carrying on their zinc-air research and I've written to them with my good 'everlasting zinc' suggestions.

2. Take the path of least resistance on the positive electrode... Tried and Tested, well known and long used Nickel Oxyhydroxide, NiOOH. (AKA nickel hydroxide. Oxyhydroxide is the charged form.) A nickel-zincate cell that lasts "forever", or at least for many thousands of cycles, would be a fabulous battery. It would be notably lighter, cheaper, higher voltage and higher energy than nickel-metal hydride, and surely less costly and safer than lithium types. The PVC plastic pipes with holes should be able to sufficiently compact the electrode substance (NiOOH needs to be held under quite strong pressure for the substances to conduct well). Now... can I make ones that work? I've done reasonably well with nickel electrode substance from dry cells... and I actually still have quite a few Ni-MH "D" cells to recycle. Plus I have plenty of nickel hydroxide and other chemicals used for that chemistry.

   Going from valence 2 [nickel hydroxide] to 3 [nickel oxyhydroxide] moves one electron per nickel atom for a theoretical figure of 289 mAH/g. The oldest nickel oxyhydroxide mixes as used by Edison in the early 1900s probably yielded an actual, practical figure of around 90. It was hard to attain a total nickel valence above (?)2.8, and if the valence falls below about 2.2, the conductivity becomes very poor, so the actual available valence change between charged and discharged was (theoreticly) 60% of the theoretical. In the Ovonics patent with various trace additives they claimed to be attaining a mixed valence above 3, with some Ni centers being valence 4, thus going from valence 2.2 [fully discharged] to about 3.2 [fully charged] and a figure of around 200 mAH/g. There's a lot more to the chemistry and crystalogy(?) in all this, but the result is that 'modern' nickel electrodes are now twice as good as they used to be.
   It occurs to me that Chinese manufacturers probably wouldn't do much research and would simply copy whatever existing chemical mix was best - probably the Ovonics mix. To make dry cells they would probably have just shrugged about the patents that prevented anyone in the USA and Europe from making more Ni-MH EV batteries (which of course has led to the worldwide adoption of lithium battery types instead). And anyway Ovshinsky's key patents expired in 2014.

   I added some 'ovonics' mix into the center of the dry cell NiOOH tube (from last month) and stuck a new rod into it, and glued the top. It didn't want to charge at all fast. But even without much charging, on discharge it would - momentarily - put out 150mA at 1.6V into a 10 ohm load. (Then charge pretty fast to make up what it had just lost, then go back to a few mA. All presumably because the zinc side was already charged and had no OH-'es to supply to the nickel.) Hopefully discharge would stay up there with the fabled 100 or 1000 hours of charging.

   Then let's see if I can actually make one 'from scratch'? I had a little of my 'ovonics'+graphite+Sm2O3 mix (acetoned) left; maybe not enough to fill a tube. I emptied the short tube of copper substance and refilled it with the 'ovonics'. Pipe 14.40g +carbon rod 9.30g = 23.70g. Filled it was 32.10g, 8.4g of NiOOH mix. [Wait, I actually used NiO, not Ni(OH)2! - more on that further down.] Less than an amp-hour I suppose? Well, I can only try out one at a time, so this one will have to wait.

[19th] Wait... the short one fits into the square cell with the lid on, which has a square zinc 'trode in it. Anyway, the large tube with the "D" cell stuff in it plus a bit of 'ovonics' mix around the carbon rod put out good current but it didn't run long. [It got better]
   I started testing by running it down to 1.400 volts with a 50 ohm load. This occurred in ~4'30" a couple of times, then with a few more hours of charging, in the afternoon it went for 6'. It seems to be improving instead of getting weaker like so many (all?) of my other "+" trodes. By another 94 hours (or is it 994 hours?) it should acquire a good charge. If only it would charge faster! But it charged quickly enough after driving a load, then went back to vey low currents, doubtless as the zinc side became fully charged again. A couple of hours later: 7'30".


   I remember that once I tried "D" cell nickel substance and got good results with 20% KOH and 20% KCl electrolyte. Now I'm using 5% KOH and 10% KCl. Might it be worth adding a few percent more nasty KOH? The cell is around 400cc so 8 grams should add 2%, making it 7%. I did that. Short circuit had been around 400mA. Now I got figures of 330mA-440mA that seemed to depend mainly on the relative positions of the electrode tubes in the jar (and maybe the alligator clip connections) rather than the concentration of KOH. (Checking with an ampmeter straight across the terminals instead of through all the alligator clips, switch, wiring & .1Ω resistor it read over 650mA. Good enough - it's a real battery!)


Best NiOOH Mix?

   I re-read some of the Ovonics European patent for nickel hydroxide electrode. (EP 1 672 724 A2 .  If anyone thinks some of my chemie stuff is involved and technical, this  document seems well written with good explanations, but it's a really heavy read. And in some places even they say "We think it works like this...") There was something I had missed before: the particles of nickel [oxy]hydroxide were coated with cobalt oxyhydroxide. This made for good electron conduction between particles without other additives such as graphite - and somehow still allowed proton transfer to the electrolyte. Also, they stressed small quantities of 'dopant' metallic atoms made for improved proton conductivity within crystallite particles - from the interior to the edges where the electrolyte is. (You don't get NiOHOH <=> NiOOH without that H+, and its associated e-, moving around somewhere!)
   These features, and perhaps even more the technology employed to achieve them, is very impressive. However I'm going to venture a guess that my technique of simply dissolving the final compounds (not precursor compounds) in acetone and letting them form into epitaxial crystals that contain all of the very same substances Ovonics ends up with, will work just as well. Most of the 'dopants' will be the same. The cobalt will go in as a 'dopant' instead of being a skin around the particle. That should improve electron and proton conductivity through each particle. It might still want a bit of graphite to improve electron conductivity between particles, but with the conductive cobalt in the particle edge here and there, it should be quite a small percentage of lightweight graphite.
   Anyway, that's my theory. Making working positive electrodes hasn't been my strong suit so far.


Nickel Oxide?

   Another thought is that nickel oxide (NiO) has been identified (as seen briefly in a web search) as a good potential battery substance - for lithium cells. Might it work better than hydroxide (Ni(OH)2) in a moderately alkaline nickel-zinc cell? Would that reaction go to oxyhydroxide on charge and then back to oxide on discharge?:

NiO + OH- <=> NiOOH + e- ? or would it simply convert to hydroxide on discharge?:
NiOOH + H2O + e- => NiOHOH + OH-
   or would it not convert at all?:
2 NiO + 2 OH- <=> Ni2O3 + H2O + 2 e- ?

   I had thought it would just convert to hydroxide, but now as far as I can find, copper, the next element over, doesn't ever seem to convert from oxide to hydroxide. Why have I never heard of nickel oxide as an alkaline battery substance? I would think this would be worth trying. Perhaps it's because there is no convenient process for converting the usual nickel sulfate precursor into nickel oxide instead of into the hydroxide?

But the oxide is denser than the hydroxide also with higher percentage of nickel per gram:

NiO :      58.7+16=74.7
Ni(OH)2: 58.7+34=92.7

289 mAH/g * 92.7/74.7 = 359 mAH/g. (per single valence change) If it works per the last reaction (NiO <=> Ni2O3) it would pack a lot more punch than the hydroxide. And it might sidestep the whole question of beta versus alpha and gamma and swelling/shrinking of the different forms.

   I suspect it's all a matter of crystallography how it actually works. If it starts as a nickel oxide crystalline substance, perhaps it charges to a different crystalline form of oxyhydroxide than if it starts as hydroxide, or perhaps to a form of nickel "peroxide" instead.

   Definitely worth a try - a substantial improvement if it works!

[20th] Another night of charging. Charging current is up to 11mA instead of unit mA. (Improving - or leakage between trodes?) 50 Ω test down to 1.4V ran for 9'30". That's still improving.
   Info says NiO is green. When I made it by roasting Ni(OH)2, it was black. Maybe I made Ni2O3, or a nonstoichiometric mix of NiO:Ni2O3? This time I decided to buy some. Best choice seemed to be an AliExpress store, ~130$C for a kilogram of "high purity nickel oxide powder". The one review said it was very fine - 5 stars. (Sigma Alderich wanted 900$C for 500g. Pottery supplies saying "nickel oxide" actually had "mixed" nickel compounds for glazes, oxide/hydroxide plus carbonate - but that they "will turn into NiO when fired". That sounded the same as making my own but starting with a supply of dubious purity.)
   A 50 Ω load test ran for 10 minutes. With 20 Ω it only ran 3' down to 1.4V. OTOH it was putting out 70mA instead of 30.

   I can't find anything on line about using NiO for alkaline cells. Oh no, am I breaking into new, untried territory again?!?

This nickel pourbaix diagram suggests that Ni(OH)3- is formed at pH 13-14. This isn't mentioned in battery literature. One expects it would saturate at very, very low concentration. Even so, it could limit the 'forever' life of the cell if not contained within the electrode. So even with nickel, my positives are going to get the full ion blocking treatments of SDBS in the separator and osmium doping.

    Nor is NiO nor Ni2O3 mentioned anywhere, and I havent found anything about convertibility between nickel oxides and hydroxides.


[21st] Charge current is over 20mA. When turned off, the cell drops quickly to 1.90V instead of drifting slowly down through 1.95V. (Is there a low resistance path somewhere? But I can move the individual electrodes around, so they can't stay shorted and if there is, it's in the electrolyte itself. Hmm, it keeps dropping if left idle, below 1.7V.) With a 50 Ω load it drops immediately below 1.8V spending no time in the 1.8xx region (not that it stayed long there before). But it ran 11' total down to 1.4V. (Originally I thought I should be running it down to 1.3V. 1.2V nominal cells are usually tested down to .9V, so a 1.6V should go at least down to 1.3V? But in fact the voltage starts dropping really fast below about 1.46V.) Recharge current after each test has been steadily rising with initial peak currents starting around 60-70mA. It's now over 120mA.
   Replacing the electrolyte had no effect. Instead, what seemed to be happening was that the cupro-nickel clamp around the carbon rod was resting on the jar cap, and electrolyte had seeped up and made contact. This had the effect of trying to oxidize the clamp, at a lower voltage than nickel and so discharging the nickel electrode. Yikes! I took it apart and reset the clamp, then put a little plastic disk under the clamp so it was above the jar lid.

[22nd] Nope, that wasn't it either! A load test didn't run long. I fiddled with the connections. One of the zinc trodes was disconnected and the voltage went up. Was one of them somehow bad? Then I noticed a bit of something - powdery, and no doubt damp - on top of the lid, which seemed to run between one of the zinc "-" terminals and the carbon "+" rod. I wiped it off. That seemed to solve the problem - for  few moments! Then it occurred to me that I had filled the jar, and when the electrodes were put in it rose right to the brim. Another turn of the lid, if there had been one, would probably have brought it squishing out the edges. That meant it was above the top of the electrodes. If the glue there wasn't perfect or maybe if the separator papers didn't come to the top, that made a path for whatever (ions?) to come out of the electrode. I hadn't intended them to be totally immersed. I poured out about 30cc so I could see the level line inside.
   With just one zinc connected, after another hour of recharging a 50 Ω load test ran all but 15'. Still not long. Later I came back, thought it wasn't quite right so I moved the electrodes around a bit, and this time it took nearly 7-8 minutes to drop under 1.7V. At 20' it was still 1.622V, putting out 30mA. This was seriously cutting into my nap time! What did I finally do right on something I was getting pretty disappointed with?

   A 50 Ω load test later ran about 40'. Into the evening, there were definitely bad connections which had doubtless been messing with the results all along. I took the lid off the jar, and I sanded and filed the clamp connecting the carbon rod externally. After considerable charging (with higher currents!) it looked stronger than ever and like a 50 Ω test would run much longer than I wanted to sit around for. A 20 Ω test with (both zincs connected) ran 6'30" to 1.4V. Momentary short circuit current was as high as 800+mA.

[23rd] My chief problems seemed to boil down to two: poor connections and one of the zinc tubes was poor. The zinc tube (if not both of them) was probably made back last fall before I had everything completely worked out. It worked better with that one removed. There was virtually no extra voltage drop under load, saying that it's limited almost entirely by the nickel side. (Theory said it would take several nickel tubes to balance one zinc tube anyway.)
   The worst of the recurring poor connections was that between the carbon rod and its clamp. It doesn't seem one can use anything but carbon/graphite as a current collector inside positive electrodes unless they are full caustic pH 14, in which case nickel forms a thin, solid skin of NiOOH that protects the underlying metal. But even that doesn't seem to last 'forever' - maybe some years, a couple of decades or even a few. Carbon seems to be the best (if not the only) choice. The cupro-nickel or monel I was trying at lower pH'es doesn't seem to work as well as I had thought for a while. It is however much better than copper or nickel alone. Perhaps someone will figure out an alloy that works 'perfectly' some day.


Gold for Connections?

   Or!... come to think of it... gold! The alkaline reaction voltage of nickel to oxyhydroxide is around .5V, gold to gold hydroxide is about .6V. It doesn't seem like a very big margin of safety, but if it works and doesn't degrade it would certainly provide high current capacity. Or even if it oxidizes, if an impenetrable Au(OH)3 layer will form on the surface at pH 12-13, that would protect it while charging. When charging stopped that surface would discharge back to Au metal - at least, as soon as there was a load. (Wait, where was Jungner when he was testing all the metals and found that "only nickel" didn't oxidize away? Did he try gold? But he was doing pH 14, and there can still be a difference between pH12-13 and pH 14.)
   Of course if I end up picking a lower voltage positive electrode (NiMn2O4 [same as MnO2] or CuO), gold would safely remain in solid metallic form. (Platinum and all[?] other metals would oxidize.)

   Gold is a bit pricey, but gold plating might work. (It hasn't worked for me so far, but then I wasn't very happy with the results I got from the gold plating solution I bought the one time I tried it.) For the clamp connection to the rod, gold is the one metal that springs to mind. Gold plating. Surely I can at least do a tiny area of good plating for a clamp.
   And while I'm at it, maybe I could gold plate a few of those ever-troublesome alligator clip test leeds. (All non-soldered, non-crimped or not-screwed-down-solidly connections seem to be troublesome. Even the contacts inside the switch, and it's the second switch I've used.)

[25th] To continue this theme, my chief concern with electroplating is that if the metal underneath is susceptible, anywhere there's the slightest gap in the plating - whether made during the plating process or from stresses afterward - the metal will be eaten away and eventually cause  failure. From all I've heard, eventual corrosion appears to have been the fate of most long life nickel-iron flooded cells.
   So I wondered if gold could be electroplated over graphite/carbon. Then it could contribute its conductivity and if there were any gaps, nothing would happen to the graphite. The answer is yes, and where I read it mentioned carbon fiber. Somehow I had forgotten about carbon fiber. It would be a good alternative to a carbon rod, or an electrode addition, allowing connections to be dispersed through the electrode substance.  BTW Technicly "carbon" of course is the element and it may have forms such as graphite, CCB, diamond, amorphous carbon and charcoal. But "carbon" seems to be used almost interchangeably with "graphite" in solid "graphitic" forms. I'm unclear about the difference between "carbon rod" or "carbon fiber" and "graphite rod" or "graphite fiber". It would seem that graphite is pure carbon where "carbon" might have epoxy in it. or not. (But how could "carbon fiber" be anything but graphite?) Graphite fiber would surely be an excellent current collector - or conductivity additive (something I've done before). And still better if gold plated.

   I found my gold plating solution (from 2019) and tried plating a small piece of graphite foil. It worked.


I rubbed one face to shine it up - and most of the gold rubbed off! The yellow color was pretty much gone and it only remained in the recessed spaces. I guess being able to electroplate it doesn't stop graphite from being slippery and greasy! There's probably enough bits left to improve conductivity, but it's disappointing that it comes off so easily. I can't imagine there being much left if a graphite rod is jammed into an electrode and twisted to compact the powders... and to get good connection.
   But for electrode production, I can imagine gold plating carbon fibers and having something like an "iris" tube where the nickel oxide mix and the fibers are laid in with the fibers oriented top to bottom sticking out the top, and then it's squeezed down to the correct diameter. A closed-end gold plated metal tube (like a very small bullet casing?) presses the fibers together at the top to form a terminal. Hmm... or they are glued together into a terminal. A prepared separator paper is wrapped around, and then it is inserted into a holey tube and a is top pressed on.




Left: gold plated versus unplated
Right: gold plating wiped (much less left) & unplated

Next: Improving Performance

   At this point I have a working nickel-zincate cell. At long last I'm through finding two basic battery chemies (a plus and a minus). But the "plus" electrodes, even the nickel, surely aren't performing even colse to capacity so next I'll be into performance improvements. Now I can see there's lots of further experimentation and development work to be done on that - ug! But the resulting batteries are bound to be fantastic!


Continuing Tests

[Back to 23rd] When connections &c are good, charging day after day seems to be improving it. With a 50 Ω test in the afternoon, in the 1.75X range, the drop rate was down to about 8 mV/minute, way under half what it had been a couple of days previous. Then it slowed down. It finally hit 1.700V after 22'. 40' - 1.661V. I noticed current was a bit low for the voltage - not by much, maybe 6-10%. I fiddled with an alligator clip and flipped the switch a couple of times. Yup, that was it. 27mA went to 29. 1.6V, 50Ω = you'd expect 32mA, but it's never quite as high as expected. 50' - 1.642V. So it's only dropping about 2mV/minute. Great! This is cutting into my nap time again!
   It's good that it's running more the sort of lengths of time one might expect. Nickel hydroxide is certainly a breath of fresh air after all the frustrations with promising but not quite delivering copper hydroxide! 60' - 1.632V: now just 1mV/minute drop rate. If this rate doesn't change it should run for 232 more minutes before hitting 1.400V. (No doubt it will change, but there a ballpark duration - adding in the first hour, that's 5 hours. Much longer than any previous test, probably owing to removal of the 'bad' zinc tube as well as - for the moment - pretty good connections.) 90' - 1.581V [50mV in 30' = 1.67 mV/'] 130' - 1.255V oops. (I thought I could go away half an hour! Obviously must have started dropping off quickly at some point.)
   In the evening I ran a 20 Ω load test, which started in the 1.7XX V range and was soon down into 1.6XX, starting around 71mA. It stayed above 1.6V for about 10'. 15': 1.58+V, 64mA; 20': 1.558V, 64mA; 30': 1.510V, 61mA. It ran for 39', rapidly dropping through 1.4XX V, especially below 1.45X V. It seemed like very good performance, especially compared to the previous 6'30".


   A footnote is that after changing the electrolyte, and with trodes that are no longer brand new, there's still black stuff accumulating on the bottom of the jar. Either a tube is leaking or substance is dissolving out and then solidifying and precipitating out. I see greenish (& black) nickel oxide or hydroxide in the holes in the pipe wall. Probably it's outside the separator sheet. Likely Ni(OH)3- ions, spontaneously converting to NiOOH(black)+H2O? Hopefully if I put osmium dopant on the inside surface of the sheet it'll prevent the penetration and the conversion. Hopefully. (The zinc also has a bit of black around the holes, but I don't think it's emitting anything. If it is, well, it's not the latest design.)


[25th] Another 20 Ω load test, this time running 12' (even) to 1.600V instead of 10'. It ran throughout with slightly higher voltages (& currents) but still pooped out at about the 40' mark.

[26th] I left it off charge overnight. It fell from ~1.9V to 1.78V. Then I ran a load test. It dropped through 1.600V in about 3' instead of 10+. It ran for 35' instead of 40, so it had lost around 3/8 of its charge in around 9 hours. That certainly needs improvement but it's better than a lot of my cells that lost most or all their charge overnight. The "bad" zinc tube had caused very high self discharge. Perhaps my newest ones are okay?
   Since the top was off the jar I stuck in a reference electrode and verified during the test that it was the nickel side losing voltage while the zinc stayed pretty constant and obviously would have run for hours. (It would seem a graphite reference electrode is about zero volts, since the measurements of zinc as -1.32 V and nickel oxy. as +.52 (charged over theoretical), when the theoretical values are -1.28 and +.48. The one with gold plating seemed to be ~~80mV more positive, giving a larger value for the zinc and smaller for the nickel (eg, -1.33, +.43.)


The Nickel Oxide Trode

   I decided to try the other one, with my 'Ovonics' completely 'homemade' nickel oxide mix (TE News #186 - or as it was there calculated, next...) instead of nickel hydroxide coming from dry cells.

NiO ----- 74.7 * .95 = 70.95 g
Co2O3 - 82.9 * .03 = 2.50 g
CaO ---- 56.1 * .01 = .55 g
MgO ---- 40.3 * .005 = .20 g
ZnO ---- 81.4 * .005 = .40 g
                               --------
                          100g, 100%

   I thought when I made it that the nickel content would just turn into hydroxide anyway (NiO => Ni(OH)2, in a more compacted way), so the idea that it may be better will get a bit of a test. (Too bad the rod won't come out easily. I could try gold plating it!)
   Current started out at 0 mA and I thought some connection was wrong. Then it rose to 1. then 2, then 3. There it stayed, probably again because the zinc was already charged and had no OH-'es to contribute? But it was charging, however gradually. Once again, it looks like days, not hours, will tell.
   In the evening I got tired of waiting. I took the electrode from the cell and put it in a small bottle with a teaspoon of bleach to help oxidize the nickel from NiO oxide to (best guess) Ni2O3, or to NiOOH - the charged form. I let it sit for 5' then filled the jar with water and rinsed it for 5' (hmm, maybe 10'+).
   I put it back in the cell. It seemed to have no charge remaining, but the initial charging current seemed higher. First run times were [even] shorter. I got the impression that the bleaching had accomplished nothing. Momentary max short circuit current was still ~~250mA.

[27th] First 50 Ω load test of the morning ran for 50" (seconds!) instead of 30. It's an improvement, but it doesn't look like it's ever going to perform well. I'll chalk it up to my homemade nickel oxide being black instead of green and wait for the real thing, now on order. Actually, maybe I'll try opening the tube, see if I can squeeze in any more powder to make it more compacted, gold plate the terminal rod, and try it again.

[28th] I did that and it ran for 1'15" instead of just a minute. It would seem I'm on the right track, but that it needs to be compacted to another whole level. That would mean opening the top of the tube and pressing the powder in with a press. (Then drilling the center out so I can put a rod back in.) How much more can it be packed down from its already "pretty packed" height?
   I pulled out the rod and filled the hole with loose powder. I took the tube out to the hydraulic press and using a 1/2" shaft as a plunger, compacted it quite a bit further. After some fiddling around and drilling, I pushed the tapered rod back in. But it didn't go in so far. I used the press to push it in notably further (maybe to the bottom of the tube). I didn't use a lot of force for fear of cracking the rod, but it was notably harder than I could push it in by hand.
   Back in the cell, the charge started around 80mA instead of around 40. So far so good! When charged, it drifted down from 1.999V through 1.95V more slowly, where before it would soon be 1.90. With the 50 Ω load it went through the 1.7XX V range (rapidly) instead of dropping right to 1.6XX. There the good news ended because it ran down to 1.400 V just as fast and dropped down under 5 mA charge current just as fast.
   ???

[30th] No improvement at all - still discharges in less than 90". In Inside the Nickel Metal Hydride Battery by John Kopera (2004) he described an unwanted byproduct of charging as being K(NiO2)3 [soluble], in which the valence state of the nickel somehow averages +3.67. Perhaps starting with NiO it charges to this state, which would be a soluble ion, instead of to NiOOH? IF so, perhaps we can make it a useful product instead of a problem, again with SDBS to keep within the electrode and osmium catalyst to keep it in solution. If it works, charging to nickel valence +3.67 instead of just +3 would move 5 electrons per 3 nickel atoms, instead of just 3, giving us a theoretical 289 * 5/3 = 481 mAH/g. Actually, that would be for starting with Ni(OH)2. That would have to be combined with the calculation under "Nickel Oxides?" from the 19th, above, and it should be even higher than that. Half an amp-hour per gram, combined with nickel's high "+" voltage, would be really impressive.
   And if, like the zincate ions (only opposite), dissolving bits expose more reactant as charge proceeds, more of the electrode's substance will be utilized, getting it closer to the theoretical value.

   I've done it all in this electrode except the osmium doping. There's only one way to see if that was the vital missing step. I'll take this one apart and redo it since I have no more of the "Ovonics" mix with nickel oxide instead of hydroxide.
   I put the stem of the rod in a vise and to my relief managed to get it out without snapping it. (That would have made it really hard to get the substance out.) I got the electrode substance out, but as usual with a disassembly (and especially when the powder mix was so strongly compacted into the tube), minute bits of separator paper were all through it. I pulled out many bits with tweezers, but it doesn't make the mix any more conductive! I found one last toluened, PP'ed sheet of separator paper. I cut it to fit, soaked it in SDBS for an hour, then dried it. I painted it with osmium doped acetaldehyde, but it might be a bit thin. I discovered that I was out of osmium powder. I ordered some more (in May?) But I keep getting emails saying it's being delayed, now until September. (Maybe by Christmas?) Then I put the trode together the same way it was before. It took about 13 grams of the NiO mix.

   While it was apart I reconnected the other cell. It was still holding 1.72 volts. No doubt I should have done a load test to really see how much charge it had left. But I just put it on charge. It charged for around 3 hours at over 30mA. Then I decided to do a 50 Ω load test and see how much of that was actually charge being retained in the cell. It started out around 1.8V delivering 35mA. At 100' it was still putting out 1.565V and 30mA. (By now my new NiO trode was ready. Now it would have to wait for the test to end!) After 2 hours it was at ~1.515V, 29mA. 135': 1.465V, 28mA. It dropped through 1.400 V in 153', sourcing 27mA. (Total about 90mAH. Okay, not huge.) In half an hour it only recovered to 1.55(?)V. But I wanted to try the new electrode!

   On a whim I put it into the same jar cell, simply disconnecting the Ni(OH)2 electrode and subbing the new NiO one. I left it a few minutes to soak. After the load test, the zinc couldn't be fully charged, so it couldn't be blamed if charging current was low. Sure enough, charging current soon dropped to 9mA anyway, and a quick load test was quick because it only ran for 50".
   But in spite of charging at just 5-6mA, it seemed to improve over the hours, where previously it was stuck at about 1'25". Now... 50". 1'20". and through the evening, 2'10". 2'40". 3'10". It would have to keep improving for a long time - or the rate of improvement would have to rise a lot - before it could be considered a good electrode. But it just might do it!

[31st] Charging current is up to 9mA from 5 or 6. Yet the voltage drop soon after removing charge seems to be less. Theory says that nickel valence 2 is a low conductance semiconductor whereas valence 3 is a much better conductor, so increasing charge current as it starts to charge should be the result. A 50 Ω load test in the morning ran for 7'10". Recharge started around 140mA instead of 40, 60 or 70. Short circuit current, which had been consistently 250mA before the osmiumed paper, was over 600. This shorter tube probably has half the holes at best, so that's probably the equivalent of 1.2 amps from the larger tube - great! The osmium seems to have made all the difference!
   Early PM: 8'00". Later: same. Later: same. Has it got "stuck" at 8' the way it did at under 2' before?

[August 1st] I got to it at noon, so it had about 12 hours charging. This time it ran for 9'. Is it working, or another failure? Painfully slow improvement and will it continue? Why does it seem less than 1% of the intended amp-hours (as usual) is available. I decided to try running higher currents to see if that might induce any more substance to come "on line" as it were, and I did some discharges at 10 Ω, down to 1.100 V. (Below 1.1 V the rate of voltage drop, already fast, doubled.) These ran for less than 2'30".


MORE Compaction!

[Aug 3rd] Seems to be stuck. Great current drive, not 1% of the amp-hours it ought to have. I keep running up against this wall with my positive electrodes no matter what chemistry I try. Only the ones with material coming from dry cells work - and even then only to a greater degree with nothing like the expected amp-hours. What am I missing? Maybe I'll try compacting one even harder - a couple of tons of force in a metal tube so there's hard "chunks" in the mix before I put it into a cell. That's probably the key difference with dry cell material: I'm using seemingly pretty hard chunks that I don't grind to powder.
   The more I thought about it, the more sure I was that that was the problem with all of them. I had heard copper and nickel oxides both needed tons of pressure, not just to be basicly munched together. The Ovonics patent mentioned around 3 grams per cc of their compacted hydroxide mix. I may have had half that, if, which would mean I should be able to crunch the stuff way farther down.
   How did I get here? 3 or 4 years ago I was testing with flat electrodes and pressing down with my body weight. The currents went up up to a certain point of pressure, beyond which they didn't increase. Despite having read about the high pressures required for Ni(OH)2 and CuO, I had concluded that that was sufficient pressure/compaction - and have ever since been operating on that assumption. But probably it was enough that the current became limited by the interface area and wouldn't rise further, but much or most of the substance still wasn't in good (any?) electronic contact through to the current collector, so available amp-hours was very low. At least, this fits the symptoms. I started making round "porous" plastic electrode tubes last year instead of flat, to hold some pressure so the material couldn't expand, but it couldn't be compacted enough in a plastic tube to start with.

   I made a jig with some stainless steel pipes and a bolt. I intended to press to 5 tons and get a tall "donut"/toroid. It didn't go well. Watching the gauge, after 3 tons the pressure stopped going up. Finally I looked down to see that the pipe had bent! The pipes were also mushrooming at the ends a bit. Furthermore, when everything was corrected and I tried just two tons, the "plug" stuck in the pipe. Even with the end unsupported,supposedly to simply pop out, the material just kept compacting. Ug! (Maybe try half as much material?)


   Giving up on that, I finally took the last PVC tube made with the long rod and started compacting the material in it further, inserting different diameter bolts and pressing them in - not hard - with the press. Finally I pressed the rod in harder than I could by hand.
   I thought this tube had nickel manganates. Now I see in my quick experiments I didn't take any pictures of any of three tubes and now am not sure whether it was the one of copper, the one of nickel-manganates material, or the one of the really old mix. Oh great, now I'm sure it isn't half as compacted as I wanted and I don't even know what it is! I decided to charge it at 1.5V as if it was copper. If it was the other, it wouldn't charge much.

[August 4th] The behaviour said it was copper. And it didn't work any better. I put it aside. I took the short tube with the nickel oxides out to the shop, twisted the rod out, dug out the heat glue, and started pressing the powder in better. If it didn't break the paper or the tube, or the rod, I was at least going to compact it better than it was and see if it performed, at least, better than it had. When I had finished pressing it down with various bolts of different diameters, I put the rod back in, again pressing it in with the press. The rod cracked, but it sheered verticly from the top. There was still enough to connect to. Unfortunately it didn't go in as far in the denser material - there was probably 1/2 inch at the bottom now rather distant from it. That may negate some gains. The NiO mix was down to about 1/2 the tube instead of 4/5. I didn't think the heat glue would fill that far in and, lacking more NiO mix, I stuffed in some NiMn2O4 powder - not very compacted, but just to fill the space. Then heat glue, and back into the cell. Again charge started at 0 mA. This time I was less surprised and I just waited for it to start coming up. In half an hour it was 8-9mA.

[August 5th] I decided to try pressing the graphite rod in farther by removing some material within, but I broke it trying to get it out. That was it for that trode! What next? I didn't have any more nickel oxide mix. Hopefully the new nickel oxide will arrive soon.

[August 6th] Back to the compacting jig, I turned a bolt to go into the outer tube to make a solid "plug" of electrode material. I pressed it to 2 tons. It worked great. Then I tried to make another one but drill a 1/4" hole through it before removing it from the tube. When I pushed it out of the tube it broke apart. Rats!

   I had improved the "exit hole" in the plate to hold the pipe centered properly and I went back to the previous arrangement. Again I used 2 tons, and given the small area being pressed, that's probably enough.


I extruded a nice donut and to my surprise it slid off the inner bolt easily, so it had a hole for a carbon rod. I tried again using a bigger spoonful of powder to make a longer one. This one worked but had a lot of loose powder around it. If the original scoop of powder extended past the inner bolt, that would happen. Also the tube wasn't quite small enough inside, so there was a bit of a gap around the bolt. I'd like to do something about that. Enough for today!




Successes! The solid "plug" lower left, two usable donuts on right,
crumbled tall donut mid left.

[August 8th] Production turned out to be rather exacting: less than 2 tons pressure the donut would be crumbly and fall apart. More than 2 tons the narrower stainless steel pipe would start to bend. More than about a level tiny spoonful of powder (1.5g) and the donut would be crumbly and fall apart. Also it starts getting very hard to press it out of the pipe if too much powder is put in, which again could lead to bent pipes. I had to make 10 short donuts (7-9mm tall - a couple were 10mm) to get enough for a 4 inch electrode tube, instead of 3 or 4 tall ones.

Enough short NiMn2O4 donuts for a 4 inch electrode tube, strung on on a 5/16 inch carbon rod.


   Let's see: 8mm long, 13.8mm OD, 8mm ID

8 x (13.8mm^2 * π /4) - (8mm^2 * π/4) = 794.4 cubic mm = .794 cc.

1.5g / .794cc = ~1.9 g/cc.

   If it was nickel hydroxide, that's not especially well compacted (should be somewhat over 2 g/cc). Since it's nickel-manganese oxide, we don't really know what it's "supposed" to be. But it's probably much better than I was doing before and sounds like it's in the right ballpark. The next challenges are (a) to get the string of beads into an electrode tube with paper in it without crumbling them or shredding the paper, and (b) hoping they make good contact with the 5/16" carbon rod and that they don't find room to swell up inside the tube and lose compaction and conductivity.
   I would also note that 13.8mm - 8mm = ~6mm, so 3mm thick electrodes (3mm thick donut walls) from the current collector rod to the outer separator. That's considered to be a fairly thick electrode (lower current capacities but higher amp-hours), but even 6mm for this sort (eg, in Mn-Zn "D" cells) isn't unreasonable.

   But next I need to make a new tube and some new separator papers. I expect NiMn2O4 can live without osmium doping, which I'm out of, on the separator, so I should be able to do it. But it's high time to wrap up this 'July' issue!

   It's cheering that I've been using the same 3D printed zinc tube, as well as the same square zinc trode, for months now. The "forever" zinc electrodes seem to be working!


Nickel Manganates (What, again?)

   I still figured the problem was insufficient compaction. Another chemistry variation would be fine. I took the tube of nickel-manganese oxide and got the rod out. I pressed the substance down into the bottom half (or so) of the tube. This pulled the separator paper down with it leaving the upper holes open. I decided half a tube was good enough. I drilled in a 3/8 inch hole for the rod. It didn't go in far enough, so I took it out and tapered the end. This time it pressed in most of the way and was pretty solid. Somewhere in there I burst the tube, a slit opening on one side. I glued over it with heat glue. I didn't bother trying to fill in the top of the tube. I decided to charge it to nickel voltage rather than manganese.
   I put the clamp on and put it in the cell. It started charging at 125mA. That seemed really impressive!... but it was at 2.0V charge (Ni-Zn level). Then I thought it should be at Mn-Zn level, so I dropped it to 1.6V, which cut the current to 40mA. Soon the bottom of the cell was covered with black powder. So much for leaving the top part open!
   After 2-3 hours charge short circuit current was a little under 500 mA - so a whole tube would be a bit under an amp. Good enough. I tried a 50 Ω load. It ran from 1.5V to 1.100 V, for 7-1/2'. An hour later 8'24". Would a whole tube have been 15-16'? (This tube was 20' three weeks ago.) Recharge started at about 100 mA.


   I'm still convinced powerful compaction of the powder - whatever mix it is - is the piece I've been missing all this time. But it looks like achieving the goal is going to be more difficult than I expected, even in round tubes that can take a certain amount of pressure. Now I'm out of nickel oxide and out of osmium powder. Perhaps before more arrives I can figure something out to get the required pressures & compaction without breaking? I suppose the osmium isn't vital to nickel or manganese electrodes, and I can use up the NiMn2O4 mix trying to get good performance.

  



Electricity Generation

My Solar Power System


Grid Tie Exchange

   In searching for something I discovered I had an unused 1400 watt plug-in grid tie inverter on the shelf. How could I have forgotten? And I had a couple of poorly functioning ones in use.

   On August 1st I put the new one in the cabin, replacing a 1000 watt one whose fans never ran, and which consequently continually shut off on overheat. Daily energy from the cabin went up. Then I took the one without fans and put it in the carport where one of the 700 watt grid ties seemed to work rarely if at all. Daily energy from the carport went up.
   I also ordered another AC power monitor.

   The weather cooperated beautifully: August 1st to 8th were sunny and warm, even hot, with the improvements making for 25 KWH days.

   Since I'm making the 36 volt cabin pretty much "off grid" with just one wall outlet where the grid ties plug in, it would be nice to put a couple more solar panels on the roof to make it six, to work those grid ties harder and (especially) to have more charge capacity for the battery. But it's not the best place for panels as it is mostly in tree shade in the winter months. For overall collection I'd be better off to put two more on the carport instead. (I could of course do both? I do have the panels.)


The Usual Daily/Monthly/Yearly Log of Solar Power Generated [and grid power consumed]

(All times are in PST: clock 48 minutes ahead of local sun time, not PDT which is an hour and 48 minutes ahead. (DC) battery system power output readings are reset to zero daily (often just for LED lights, occasionally used with other loads: Chevy Sprint electric car, inverters in power outages or other 36V loads), while the grid tied readings are cumulative.)

Daily Figures

Notes: House Main meter (6 digits) accumulates. DC meter now accumulates until [before] it loses precision (9.999 WH => 0010 KWH), then is reset. House East and Cabin meters (4 digits) are reset to 0 when they get near 99.99 (which goes to "100.0") - owing to loss of second decimal precision.

Km = Nissan Leaf electric car drove distance, then car was charged.

New Order of Daily Solar Readings (Beginning May 2022):

Date House, House, House, Cabin => Total KWH Solar [Notable power Uses (EV); Grid power meter@time] Sky/weather
        Main      DC      East
                            (carport)

   After 5 full years (March 2019 to February 2024), I am no longer recording solar collection figures daily. (Frequently, and surely monthly.) For the chart, a figure for "n" days will usually be divided by "n" and taken evenly as "n" [more] days of whatever the result is. Oh... now with the DC system running an electric heater at night (From April 11th 2024), I'm interested to do daily again for now.

June
30th 1628.30, 6.40, 70.13, 29.06 =>   8.26 [40Km; 15776@20:30]

July
  1st 1631.76, 7.26, 72.71, 32.41 => 10.92 [45Km; 15790@21:00]
  2d  1635.72, 8.76, 76.02, 36.43 => 12.79 [15797@21:00]
  3rd 1642.16, 2.00, 80.54, 41.42 => 17.95 [15802@21:30]
  4th 1551.25, 3.94, 88.01, 47.93 => 25.03 [15808@20:39]
  5th 1560.24, 5.74, 91.55, 54.59 => 20.99 [85Km; 15824@23:00]
  6th 1669.76, 7.09, 97.11, 60.97 => 22.81 [55Km; 15836@21:00; 50Km]
  7th 1678.80, 8.93,   5.30, 66.92 => 22.13 [35Km; 15849@20:30]
  8th 1688.14, 1.90, 10.87, 73.49 => 23.38 [15854@20:30]
  9th 1697.53, 3.53, 16.42, 80.23 => 23.31 [55Km; 15866@20:30]
10th 1701.36, 5.04, 19.46, 83.96 => 12.11 [15872@21:00]
11th 1704.51, 6.64, 22.19, 87.20 => 10.72 [40Km; much laundry; 15889@20:30]'
12th 1710.42, 8.53, 26.40, 92.15 => 16.96 [15897@20:30]
13rd 1715.98, 9.78, 30.51, 96.68 => 15.45 [60Km; 15916@22:30; 55Km]
14th 1723.34, 1.71, 36.07,   6.01 => 20.64 [35Km; 15930@'24:00']
16th 1732.58, 4.10, 42.49, 13.27 => 25.31 {est as: 12.66,12.65} [55Km; 15950@21:00] 2 days
19th 1743.58, 6.20, 49.57, 20.78 => 27.69 {est: 7.60,7.60,12.49: rain, rain, cloudy} [{0,55,90Km}; 15993@22:00] 3 days
20th 1751.98, 6.34, 53.82, 25.33 => 17.34 [55Km; 16007@21:00; 50Km]
23rd 1771.87, 7.03, 64.26, 37.70 => 43.39 {14.46, 14.46, 14.47}[50Km, 55Km (21st,22nd), 3x laundry (23rd); 16045@20:30] 3 days
24th 1778.54, 7.13, 67.73, 41.93 => 10.47 [55Km; 16060@22:00]
25th 1782.60, 7.63, 69.95, 44.74 =>   9.59 [16067@20:30]
26th 1786.16, 7.76, 71.74, 47.09 =>   7.83 [90Km; 16088@20:30]
27th 1788.22, 8.22, 73.52, 49.16 =>   6.37 [55Km; 16105@20:30; 50Km]
28th 1790.37,   .60, 75.17, 51.10 =>   6.37 [35Km; 16126@20:30]
29th 1795.62,   .88, 79.02 ,54.96 => 13.18 [16136@21:00] Still not much like summer!
30th 1799.09,   .98, 81.65, 57.93 =>   9.17 [16145@22:00]
31st 1802.52, 1.24, 84.21, 61.00 =>   9.32 [55Km; 16162@23:00]

August - Wow, suddenly, sunny days!
  1st 1809.67, 1.37, 89.72, 66.98 => 18.77 [16167@20:30] Changed a couple of grid ties - carport & cabin should increase slightly.
  2d  1819.11, 1.49, 96.04, 74.75 => 23.65 [85Km; 16182@20:00] found one turned off  at house, midday!
  3d  1830.54, 1.63,   6.03, 82.68 => 25.53 [55Km; 16196@21:00; 50Km]
  4th 1841.88, 1.81, 11.80, 90.56 => 25.17 [45Km; 16208@21:30]
  5th 1852.83, 1.88, 18.25, 97.78 => 24.69 [16213@21:00]
  6th 1863.92, 1.98, 24.80,   7.39 => 25.13 [16218@20:30]
  7th 1872.92, 2.13, 30.96, 13.51 => 21.43 [55Km; 16228@21:00] Foggy AM. Ate first corn cob!
  8th 1883.93, 2.27, 36.81, 21.15 => 24.64 [16236@21:00]
  9th 1889.64, 2.40, 40.50, 24.67 => 13.05 [85Km; 16254@20:30] Not sunny. (Was great while it lasted!)


Chart of daily KWH from solar panels.   (Compare July 2024 with June 2024 & with July 2023.)
Days of
__ KWH
 July 2024
(18 Collectors)
 June 2024
(18 C's)
 July 2023
(18 C's)
0.xx


1.xx


2.xx


3.xx


4.xx


5.xx


6.xx
 2
1
7.xx
 3
1
8.xx
2
 2
9.xx
 3
 1
10.xx
 3
1
11.xx
1
 2
12.xx
 5
 3
13.xx
 1
 5
 2
14.xx
 3
 3
 2
15.xx
 1

16.xx
 1
 5
 2
17.xx
 2
 2
 1
18.xx
1
 3
19.xx
1
20.xx
 2
 1
 2
21.xx

1
22.xx
 2
 2
 1
23.xx
 2
 2
 3
24.xx

2
25.xx
 1
 1
 3
26.xx

1
27.xx

1
Total KWH
for month
 444.56
 420
 563.61
 Km Driven
on Electricity
 ~1265.8 Km
165 KWH
 ~1465 Km
190 KWH 
 1348.1 Km
(180 KWH?)


Things Noted - July 2024

* Nothing special. Weather not very nice (clouds, damp, rain) for last 1/3 of month.

Monthly Summaries: Solar Generated KWH [& Power used from grid KWH]

As these tables are getting long, I'm not repeating the log of monthly reports. The reports for the first FIVE full years (March 2019 to February 2024) may be found in TE News #189, February 2024.

2024
Jan KWH: 31.37 + 3.14 +  16.85 + 16.82 =   68.18 [grid: 909; car (very rough estimates): 160]
Feb KWH: 96.52 + 2.36 + 49.67 +  52.98 = 201.53 [grid: 791; car: 130]
FIVE full Years of solar!
Mar KWH 150.09+ 1.63 + 93.59 +  92.50 = 337.81 [grid: 717; car: 140]
Apr KWH 181.89+35.55 +123.50+142.74 = 483.68 [grid: 575; car: 140]
May KWH 129.23+67.38 +109.6  +126.32 = 432.53 [grid: 405; car: 145]
Jun KWH  152.54+51.02+118.99+141.17 = 463.72 [grid: 420; car: 190]
July KWH 174.22+30.53+111.19+128.62 = 444.56 [grid: 386; car: 165]


Annual Totals

1. March 2019-Feb. 2020: 2196.15 KWH Solar [used   7927 KWH from grid; EV use: -] 10, 11, 12 solar panels
2. March 2020-Feb. 2021: 2069.82 KWH Solar [used 11294 KWH from grid; EV use: - (More electric heat - BR, Trailer & Perry's RV)] 12 solar panels
3. March 2021-Feb. 2022: 2063.05 KWH Solar [used 10977 KWH from grid; EV use ~~1485 KWH] 12 solar panels, 14 near end of year.
4a. March 2022-August 2022: in (the best) 6 months, about 2725 KWH solar - more than in any previous entire year!
4. March2022-Feb. 2023: 3793.37 KWH Solar [used 12038 KWH from grid; EV use: ~1583 KWH] 14, 15, 18 solar panels
5. March 2023-Feb. 2024: 3891.35 KWH Solar [used 7914 KWH from power grid; EV use: ~1515 KWH] 18 solar panels

Money Saved or Earned - @ 12¢ [All BC residential elec. rate] ; @ 50¢ [2018 cost of diesel fuel to BC Hydro] ; @ 1$ per KWH [actual total cost to BC Hydro in 2022 according to an employee]:
1. 263.42$ ; 1097.58$ ; 2196.15$
2. 248.38$ ; 1034.91$ ; 2069.82$
3. 247.57$ ; 1031.53$ ; 2063.05$
4. 455.20$ ; 1896.69$ ; 3793.37$
5. 466.96$ ; 1945.68$ ; 3891.35$

   It can be seen that the benefit to the society as a whole on Haida Gwaii from solar power installations is much greater than the cost savings to the individual user of electricity, thanks to the heavy subsidization of our power owing to the BC government policy of having the same power rate across the entire province regardless of the cost of production. And it can be insurance: With some extra equipment and a battery, sufficient solar can deliver essential power in electrical outages however long. (Feb 28th 2023: And it's probably well over 1$/KWH by now the way inflation of diesel fuel and other costs is running.)
   It might also be noted that I never went into this in a big way. Instead of installing a whole palette load of 32 solar panels, I have 18, and my grid ties aren't the best, and I would be hard put to give an accurate total of my installation costs. All in all the grid tied part probably cost me (with all my own 'free' labor) around 7000$. At the actual "total savings to all" figures, they have paid for themselves twice over in five years. The 36V DC system largely cost a couple of thousand dollars for batteries. The solar panels were up. The charge controller, circuit breakers, DC combo meters [V, A, W, WH], 36V compatible LED lights and wiring cost were a few hundred dollars at most. (I did have to make my own T-Plug cables & 3D printed wall plates.) The battery cost has come down substantially in recent years and will come down a lot more if I can get cheap, "forever cycle" batteries working.



http://www.TurquoiseEnergy.com
Haida Gwaii, BC Canada