Turquoise Energy Ltd. News #65
Victoria BC
by Craig Carmichael - July 5th, 2013


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

Month In Brief (Project Summaries)
- Next Projects possibilities, Building the Green Economy event, Electric Mazda, Solar 'DC grid tie', 3D printer repair (sigh!), 'Ultra-efficient Transmission' work, new chemie batteries - MnMn still looks best

In Passing (Miscellaneous topics, editorial comments & opinionated rants)
- extended supply chain breakdown

Electric Transport - Electric Hubcap Motor Systems
* PGTC Transmission: Clutch mechanism figured out and built. Now motor controller has quit - sigh!
* Idea: A simple 2-speed transmission with one clutch pedal shifting? - pedal down is low; pedal up is high; pedal centered is neutral.
* Electric Weel motor design change: 24 coils as 2 sets of 12, 2 motor controllers @ 48V, 150+amps.
* Homemade "Segways"! from Germany - with schematics, source code, ...
* Mazda RX7 stuff - maximum range 13Km, individual battery float charging system, NiMH batteries are charging more slowly than lead-acids - and not to very full

Other "Green" Electric Equipment Projects
* Solar PV main distribution panel things & lead-acid batteries

Electricity Generating (no reports)

Electricity Storage - Turquoise Battery Project etc.
* Production Prototype Cell #4 (PP#4): MnNi - a nickel negode to try to isolate MnMn electrode/cell problems
* Worked crappy - rebuilt.
* Low pH (6) causes high self discharge! (Well, duh!)
* PP#5: Ni-Ni. Even with "traditional" Ni(OH)2 posode and pH (near?) 14, same self-discharge problem and even lower conductivity.
* Carbon fiber cloth separator sheets to improve electrode conductivity
* Mn-Mn is better. I'm returning to it - with carbon fiber cloth, paper and nylon separator sheets, and any other new tricks I can think of. With a 5-fold improvement (over PP#3) it would be a practical battery, whereas the NiNi would need a 20-fold improvement over PP#5 - and even then it would be just 1/2 the voltage.
* Perforating metal for 'porous' electrode conductors: a silkscreen with tiny holes and acid? ...and faster PCB etching?

No Project Reports on: DSSC solar cells, LED Lighting, Pulsejet steel plate cutter, Magnetic Motion Machine, CNC Gardening/Farming Machine (sigh, maybe summer 2014?), Woodstove/Thermal Electricity Generator, Peltier & vacuum pipe heat pumping.



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

Construction Manuals and information:
Electric Hubcap Motor - Turquoise Motor Controller
- 36 Volt Electric Fan-Heater (say, this heater is now obsolete! Use Peltier module heat pump!)
- Nanocrystalline glaze to enhance Solar Cell performance - Ersatz 'powder coating' home process for protecting/painting metal

Products Catalog:
 - Electric Hubcap Motor Kit - also please inquire about Electric Caik 3KW Motor Kit
 - Sodium Sulfate - Lead-Acid battery longevity/renewal
 - NiMH Handy Battery Sticks, 12v battery trays
& Dry Cells (cheapest NiMH prices in Victoria BC)
 - LED Light Fixtures

Will accept BITCOIN digital currency

...all at:  http://www.TurquoiseEnergy.com/
(orders: e-mail [email protected])



June in Brief


Mazda RX7 parked at Westburne Electric, early June
They had another one of those fancy EV charging stations that I can't use
Tom did a video here of me driving it: www.youtube.com/v/vSX3JQWIKng

   The electric Mazda RX7 continued to occupy time into mid month. Now I'm driving it and with most of its components now functional, where should I next turn my attention? The possibilities were and are endless:

* Fix the 3D Printer (sigh!)
* the Referendum Web Site to be developed
* some exciting new ideas as well as unexplored old ones for new chemistry battery making
* the Chev Sprint ultra efficient electric transmission/planetary gear torque converter... that sat idle all winter after exciting progress in September
* then, a variant of that transmission to finally make the Electric Hubcap car hybridizing system
* motors and motor controllers to be made including:
   - the 13.5KW high torque Electric Weel motor that must be about 3 years old and hasn't been finished
   - the wheel rim/arc motor for bikes
   - more Electric Hubcap and Electric Caik motors, to have kits on hand to sell or to use
   - more BLDC motor controllers. It also makes me nervous that only the V1 controller and not the V2 is running reliably.
   - Optimize component values to make the V2 more reliable and robust.
   - other motor controllers such as VFD for AC motors, and controllers for DC motors with brushes. These can be variants of the existing controllers, but some high power ones, like 300 or 400 amps at 36 or 48 volts, would be good to have for cars, even with very high efficiency.
* The "Outboard from Scratch" - and casting the more efficient-looking prop (plastic prop from a scuba unit).
* lots more 12 volt DC house wiring, LED globe lighting and other solar equipment projects that could be done. I should make a production version of the circuit board for the LED globe lights: with CAT plug pins - the little round power adapter plugs are frustratingly unreliable, especially if you have to wire one yourself, and of course you can freely plug in the wrong voltage or polarity of adapter.
* the peltier/thermoelectric heat pumping and thermoelectric electricity generator projects (All these projects use almost exactly the same components.)
* now having reliable means to do simple circuit boards and now also microcontroller programming, there's various microcontroller-electronic projects:
   - the fridge control
   - peltier heat pump control
   - solar PV system 'dump load' and other ancilliary equipment control
   - motor controller 'overcontrollers' with RPM limiting
   - the individual battery monitor for EVs of all types

...and in case those weren't enough possibilities, then moving just a little farther afield I could try getting back to other projects I haven't had time for in quite some time or haven't got very much started on:

* high refraction, low reflection 'pebble glaze' (nanocrystalline titanium dioxide borosilicate glaze) solar panel cover glass and DSSC solar panels
* pulsejet steel plate cutter
* magnetic motion machine
* the ocean wave powered electrical generator of the recently described type
* CNC gardening/farming machine

   On the 4th I decided that with the onset of good summer weather, I should 'make hay while the sun shines' and do the things with outdoor activity. That means physical EV things, and especially the PGTC transmission/Sprint. That car has now sat over two winters in one place, and I fear things may start seizing up like on the Mazda.
   Also, I decided to work on the long 'overdue' Electric Weel, which a couple of people are interested in, and try and make molds to cast the stator body parts. Of course, people have expressed interest in my other motors too and yet no one has bought any kit yet. I got as far as a 'life size' drawing of an octagonal section.
   And I also decided to try a few battery chemistry things, because surely it's pretty close, and the short range Mazda shows the glaring need for better yet economical batteries. Other things could be worked on on occasional evenings or put off until inclement weather. Even if one achieved only the 50 watt-hours per kilogram of nickel-iron flooded cells - with of course high current capacity and long cycle life - they'd be better than lead acid. With modern techniques and materials one could certainly make much more economical and better performing NiFe batteries than Edison did 100 years ago. But MnMn seems to be the superior choice.

Electric Mazda

   The first thing I actually did, on the 4th, was to put in a plywood and carpet cover over the battery well in the rear of the Madza RX7, with a few more bits of matching carpet surround. It looks nice, provides a cargo space, and muffles the noise coming through the battery well from the rear axle and wheels.
   The next day was the VEVA electric car club, and I finally got to show off something besides semi-functional battery prototypes, and once the electric outboard motor, even if the car wasn't using motors and components I'd made myself. But there was a free, outdoor Buffy St. Marie concert that evening. Everyone rushed off to that and paid little attention to the car, except I gave someone a ride.
   After it was over we went to his place and he showed me his 'new' Morris Minor, a 1960's British car produced with a 948cc engine, that he wanted to make electric. (See Morris Minor)

   On the 14th, I swapped out the weakest lead-acid battery in the RX7 for a new one Tom Sawyer loaned me. With that and upping the NiMH charger voltages a bit, the car's range went up to about 8 miles (13Km) or better, with the somewhat undercharged NiMHs becoming the limiting factor. It's a long way from the 70Km or so the original owner got with 18 new 8-volt high current golf cart batteries.
   Later Tom came over and insisted on helping me wash the RX7 (which he doubtless realized I would never have gotten around to myself) in preparation. The car works better through my efforts; that it looks rather good must largely be credited to Tom's enthusiastic 'detailing' efforts.


Finally, after some great black 'rubber' texture paint (bumper, rusty battery well steel under),
we had seen "Electric" nameplates on a converted car at the April VEVA meeting.
Tom ordered a couple from Canadian Electric Vehicles and we put one on the back.
Not only many conversions but the manufactured electric cars seem to want to hide the fact that they're electric, giving the general
public the impression as they travel that there are essentially no electric cars on the road even when there's one here, one there...
I think I want to paint the black parts between letters white, because it doesn't show up very well.

   On the 26th I decided I needed the 6-tray NiMH battery from the Mazda for the Chevy Sprint experiments. I decided to first see how much "ultimate" range the Mazda had with the 11 batteries. I drove it 13Km. The CycleAnalyst showed 17.5 amp-hours - 2310 watt-hours. 17.5 amp-hours doesn't seem like a lot of juice to get from batteries rated from 60 to 95 amp-hours. Part of the problem is doubtless that continuing currents of 50 to 150 or 200 amps are very high for regular marine-RV type batteries. The reason for using higher current golf cart batteries becomes more apparent.

   Two of the NiMH batteries were below 12 volts, eg 11.8. They surely had little life left. The other two were just over 12v, eg 12.10. The lead-acids were all above 12.2 volts. If I beefed up the charging of the NiMHs to get them up to 14.0 volts it should surely be good for 10 miles/16Km. Then again running lead-acids down too far with no sodium sulfate in them is worse than running down NiMHs.
   The best battery appeared to be the "reconditioned" one I had purchased at Battery Doctor. It was still at 12.6 volts and so apparently could well have driven at least as far again, total 16 miles/26Km or more.

   Of course, while the ever decreasing voltages can be seen by glancing at the CycleAnalyst while driving, there's no indication of which batteries are dropping by how much when you hit the pedal. Is it mostly the NiMHs are at the end of their charge and going way down?, or are they doing okay while the weaker lead-acids are going down to 8 or 9 volts?, or are they all dropping more or less equally? These are the sort of questions my individual battery monitor is intended to answer.

Solar House Equipment

   Also that day I repaired a 20 amp, 13.8 volt power supply I'd been given some months ago. (blown transistor from foolishly placed, clamped, output wires that melted and shorted together.) It's exactly what I've been wanting for the solar PV system as the backup "DC grid tie" power supply. I fit it in, replacing a 12 amp battery charger I'd been using. It gets pretty hot doing 100-200 watts if the car is plugged in and it's cloudy. It has its own 20 amp breaker in the main panel.

Building the Green Economy Event

Showing EV mixed battery system to Murray Rankin, MP.
   On Saturday the 15th, I had a display table at Building the Green Economy, an event put on by Murray Rankin, member of parliament for Victoria BC. Before the event started, I presented him with a printed copy of my booklet Fundamental Principles of Democratic Government - Towards Utopian Systems of Governance, which he seemed interested in. He said he would read it, and asked me to e-mail him to discuss it later. There was an hour long panel talk with short talks by several people. Guy Dauncey, who started the BC Sustainable Energy Association, spoke last and gave a surprising, brilliant talk mainly about about one of the broadest issues that nobody talks about except on youtube: "locally has to be in a global context", the impoverishment of people and nations by the bloated financial sector's intrigue and fraud, the need for more local, accountable public financial institutions and for local or regional currencies, and the need for everyone to learn more about how finances work - and how they ought to work. I just hope many of the audience of perhaps 130 understood the background of systemic, unsustainable global financial problems which he was addressing. Sustainability has to begin with sustainable societal institutions, and we ain't got 'em.
   I put the electric Mazda RX7 on display with the hood up out in the parking lot, with a sign on my table saying it was there. I showed it and explained what I was doing with mixing non-identical batteries and separate charging to a number of people. Jim Harrington brought some economical LED house light bulbs to add to my small 12V collection, but as usual no one bought any, even once the amazing electrical savings were explained. Perhaps everyone's been made skeptical by the claims of compact fluorescent bulbs, which were heralded as being better than they proved to be in most peoples' experience.

   At the end of the event, Rankin's assistant Edward, who had invited me to the event after we chanced to meet earlier in the week, asked if he could get some pictures of me and Rankin out by my electric RX7. So when we (Rankin too!) had stacked up the chairs and tables, and I had packed up the Electric Caik Outboard and my boxes of stuff, we did the pictures and they watched me drive off burning no fossil fuel.

Magnetic Control with Light?

Video (@youtube): 405 & 980 nm photo control of diamagnetic pyrolytic graphite

   Jim Harrington and his Ecosat Satellite team have taken lasers - one violet or UV and the other IR, both in the x100's of milliwatts range - and shone them on a round disk of pyrolytic graphite floating by magnetic repulsion just above a "checkerboard" of cubic supermagnets. If the laser is shone near one edge of the graphite, it jumps in that direction, right over the magnetic boundaries between magnet cubes to the next corner boundary. Shine at the other side, it jumps back. There's recent research on the reciprocal relationship between light and magnetism, but I don't think it's ever been graphically seen and demonstrated before. Jim says it could revolutionize space flight.

   I can see a potential use for it in magnet machines to generate electricity, where the effect might help generate larger circular magnetic imbalances to get more power from a smaller magnet machine. And I'm probably not even scratching the surface of the possibilities.

3D Printer Repair

   My RepRapPro Mendel 3D printer hadn't worked since I started programming the TI MSP430 using the Launchpad board. I've had ongoing converse with their tech support. Out best guess (theirs and mine) is that somehow programming the MSP430 was also programming the AT-mega1284 chip on the printer's Melzi board, and it had wiped out the firmware and the bootloader for uploading new firmware, via the USB port. We figured I'd need a USB "AVR" "ISP" programmer to reprogram the boot loader so the Melzi's AT-Mega firmware could be reprogammed. The Arduino Uno board I got as a possible alternative to the MSP430 Launchpad appeared to have the right header pin plug on it, labeled "ICSP".
   I spent most of the 23rd on the problem. The Uno's port proved to be an input programming port rather than an output. However, only one wire was different. I cut the trace and converted it, and made a short cable to go to the Printer's identical port.
   The software was another problem. Following the instructions on the reprappro & reprap web site, I got "Burn Bootloader" to work. (Apparently.) That was a big step. However, the LED on the printer board still didn't light, and I couldn't upload the firmware. I must be very close, but something isn't quite right.

   Once it's working again, I have the 3D '.STL' design for a J1772 car charging plug. I intend to try using the 'ground' wire in the sockets as a 'neutral' to use with 120 volt charging systems - or at the very least with my charging system where I can roughly balance the load by putting 1/2 the chargers on each side of the 240V line.

Ultra Efficient Transmission/Chevy Sprint - motor controller blowout (sigh!)

   Then I went out and made the clutch linkage for the Sprint 'ultra efficient planetary gear torque converter' transmission, having done a spring loaded belt tensioner the previous day. After assembling the mechanism properly on the 25th, it seemed ready to attach the motor controller and batteries, and test.
   But now I'd used the previous battery in the Mazda and would have to find a new arrangement for that. I could use 100 more NiMH D cells, but am uncomfortably far in debt. I decided to settle for 50 D cells, and to pull the seven-3D-printed-trays battery from the Mazda, and simply have one less battery in it for now. By changing the bus bars in the tray, I made it 24 volts, 30 amp-hours. I added 3 'quintos' pipe batteries taken from the solar PV system for the third battery. That could certainly run the Sprint around the yard if it could be coaxed into moving at all.
   On the 28th I made a new wooden box to hold the batteries in the Sprint radiator area, the original one for long pipe batteries having proven a little too wide to get in and out with the motor/transmission in place, and the second one having lost its short pipes battery into the Mazda.


Under Sprint Hood
Front: 36v battery in open box,
Left: motor controller
Center: yellow motor, PGTC with slip pulley, flat belt on green pulleys,
To the right: friction cable & pulleys that tension yellow rope (from automatic gear shift lever)
Right, under: clutch linkage to slip flat belt (from clutch pedal)

   On the 30th it was all in and I tried to run it, but the motor controller almost immediately emitted smoke. Two mosfets needed replacement. (I have yet to check out the MC2133. But while I'm at it, I'll test several other MC2133s I've removed from previous boards just in case they were gone. It's a pricey chip - putting all the 'suspect' ones back into stock could save 100$.) I found a loose power connection and two wires that were unconnected at one end and should have been removed in the last modification quite a while ago. Any one of those could account for the fault, so it probably wasn't some unknown design problem now manifesting itself. (Whew!)

New Chemie Batteries

   I tried out a couple of other combos of battery chemistries to see if I could isolate the causes of self-discharge and poor conductivity in my NiMn-Mn in KCl ("MnMn" for short) cells to one electrode or the other. Earlier, an Fe electrode had undergone capacity decay with each cycle in slightly less than pH 14 solution and I gave up on that. MnNi (PP#4) didn't seem to work any better than MnMn including having high self discharge, so the suspect was the Mn posode. (Ni negodes don't work in pH 14 BTW - but 10-13 seems okay.) Then I tried NiNi in another cell with the same negode (PP#5), and that didn't work any better either despite its low voltage.
   Since PP#3 (MnMn) had been gradually improving with each cycle both in capacity and with gradually dropping self-discharge, further improving the construction and using MnMn chemistry seems like the best bet. If it had had about 10 times more capacity and if the self discharge had continued to drop (and also if it hadn't suddenly quit holding any charge, with electrode powder leakage evident) it could have been a practical cell - and over 2 volts. It might have even got there simply with enough cycling. While the NiNi also seemed to improve with cycling, it would seemingly need a substantially larger improvement, and even if that was obtained it's just over 1 volt and twice as many cells would be needed.

   Two things I meant to do in June and didn't were to finish the CAT standard 12 VDC plug & socket promo video, and to evacuate a pipe with ammonia as the fluid, to get lowered boiling and freezing points. You'd think I could have got to those rather short projects! But projects once started seem to have a habit of growing.



In Passing
Incidental news, editorial comments & opinionated rants

   Warnings from the spirit beings that oversee this planet have been coming over the ether then through e-mail lists (1111progress, TML, CCC, CWM...) for at least a couple of years now that we're headed for a very rough time. This month it was stated that when the main crisis hits, the disruption to global supply chains is likely to last about 3 years, and almost nobody has put away enough food to tide them over for so long. Plus, many people have left important things out of their supplies, like salt or whatever.
   It seems reminiscent of what I've heard of the Irish potato famine. When potatos came to Europe from the new world, they grew great, and they supported a large population increase in Ireland. Then the potato crops got a blight. It spread across the island and people went hungry. Nobody actually starved in the first year. But by the second and third potatoless year, everybody had run through everything they had and mass starvation ensued. There was a great exodus of Irish people desperate to go anywhere on any outbound ship. Many went to America. 2 million people, perhaps 25% of the population, died or fled. As with today's crisis, there were a host of political, ethnic, religious, social and economic factors exacerbating the problem.

   Accompanying the financial fraud that's impoverishing the world today, we also see increasing levels of natural(?) disasters occurring around the globe. And there are all those new, antibiotic resistant diseases as well as old diseases which will likely strike as diseases usually do - when populations are in hunger and want. It sounded like perhaps five billion people are likely to ascend prematurely to the mansion worlds in the coming years or perhaps in the next decade or two, reducing the global population to two billion. It's surprising how global it all is, with even China suddenly and unexpectedly facing a "credit crunch" similar to that of so many other lands.
   Although the crisis is essentially man-made, with the worst aspects being generated by the machinations of those in economic and political power, something fairly drastic has evidently become necessary in any event owing to an overpopulation of unthinking, uncaring and unspiritual materialistic people who are destroying the planet. The horrific crisis and its outworkings will however lead to tremendous quickening of individual thought and consciousness, and rapid planetary progress. The world will be a different place, just as pre- and post-famine Ireland were different places in just a few years.
   Things that will still have currency value after people lose confidence in fiat currency (which always collapses in a few decades if you look back into history) include silver and gold (which are now being bought up at record rates), and I expect bitcoin and other digital currencies. With internet transactions, digital currencies may be the only form of money that can be sent to another place quickly and without risk of theft in transit. Bitcoin is 'fiat', but it can't be inflated away to nothing: no more than 22 million bitcoins can ever be created because the program was written that way. If it catches on, people will be using milli- micro- and perhaps even nano-bitcoins for transactions. And even by itself it may be the beginning of the end for the financial system as we know it since most transactions will require no "middleman".
   We can see a crescendo of calamity rising around the world, but we don't know exactly how the collapse will unfold or the timing (everyone asks the economists on youtube "how?" and "when?", and the usual and honest answer is "I don't know."). One of the motivations for getting the electric Mazda is that it just might give me a way to get around in case extreme fuel prices or shortages develop before I've got my own 'ultra-efficient' vehicle drive on the street.

   As to those "natural" disasters, there's substantial evidence some of them are being generated by HAARP projects. When HAARP antennae aim millions of watts of radio wave power at an area of Earth's ionosphere, the ions heat up. They rise and are dispersed, being replaced by non-ionic air that wells up from the stratosphere. This causes low and high pressure zones directly and severely affecting weather, and also brings unfiltered sunlight with all its UV radiation down to ground level, which is a possible reason for otherwise unexplained big die-offs of plants and animals. Weather control was one of the ideas behind the system, as claimed in its 1990s patent. It's even being said HAARP effects can trigger earthquakes. I don't really follow that part. Fracking can almost doubtless create earthquakes. Then there's "seeded" jet "chemtrails" supposedly with stuff in them. I don't follow the rationale for those either and suspect the whole story is just vapor, but one theory says the alleged chemicals react with the HAARP radio waves to magnify the effect.

   Some are asserting that HAARP and or its sister antenna arrays are responsible for generating global warming (makes more sense than CO2, a so-far fairly small increase to a trace gas), for hurricane Katrina increasing in size and being steered directly onto New Orleans, Hurricane Sandy hitting New York, the Indian Ocean earthquake and tsunami, the Japan earthquake and tsunami and earlier recent earthquakes in Japan and China, dead dolphins with burned flesh washing up on beaches in the Persian Gulf after a bizarre "fire tornado" over the water, and the extreme US drought of summer 2012. Much of this stretches credulity. But then, so would some of these extreme weather and geologic events if we didn't have clear proof that they did happen.
   The celestials say the people running HAARP and fracking programs have no idea what effects they're causing and that the repercussions of these disturbances will take thousands of years to completely settle back to equilibrium.

   Aside from personal and family physical preparations to tide us over an extended supply failure, what can we do? Some may at some time find themselves in some position to make a big difference by a courageous personal act, like Julian Assange, Bradley Manning, Ed Snowden... or millions of unsung people doing what they think is right in spite of implicit or explicit negative consequences to their career or person. For the rest of us, the answer is to improve ourself. This seems too paltry, too insignificant, too mundane, to be the means for changing a world: but the whole world is made up of people like us, and only better people make a better world. Education is vital - learn what's going on, how things work in our screwy society, and think of better systems. We need to invent new and sustainable systemic tools and institutions to prevent people from relieving the citizenry of power by usurping it themselves and ruling instead of leading. Clear separation of executive and legislative powers and the choice ranking vote will help politically, and the citizenry also needs to control its monetary systems.
   And there's one thing the celestials repeat time and again - "hit us over the head" with, as it were: take 10 minutes a day out of your busy schedule and meditate, or take a quiet, meditative walk. Still your mind to make it receptive to the leadings of your own guiding spirit. These mentally quiet times are when you grow in spirit and advance in psychic status. Even if nothing seems apparent, it's happening.

   And while making whatever preparations you feel are prudent and applicable, hopefully without unwarranted disruption to your life, fear not! The universes are evolving towards perfection according to the Universal Father's grand plan. The 'local' Lucifer rebellion is over (ca. 1986), and our world along with several others is now, somewhat swiftly and painfully, being yanked back into the regular circuits before we destroy ourselves with the leftover products of his ways of thinking, which has greatly delayed our evolution towards the eventual 'era of light and life'.



Electric Hubcap Motor Systems - Electric Transport

Chevy Sprint: Ultra-efficient variable "PGTC" Transmission... and the next blown motor controller

   The biggest draw to 'ultra-efficiency' is the substantial range improvement to be had with any given set of batteries. Another is with the Mazda burning 160 to almost 200 watt-hours per kilometer, or 1.6 to 2¢ [at 10¢/KWH], _assuming_ high charging efficiency (which I don't really think I've got), lower energy use would reduce the fuel cost, or the load on the solar collectors, notwithstanding that it's already so much cheaper than say 14¢/Km for gasoline.

 
  A while back, I came up with the idea that moving the whole motor and pulley shaft up and down might be easier, and would probably be better, than having an idler wheel to tension the belt.
   Now I considered that given the tiny amount of travel needed to tighten or slip the belt, perhaps the motor could be fixed in place after all, and only the other side, where the belt pulley was close to the end of the shaft, needed to pivot. That seemed much more doable. The slight twist of the pulley and shaft could be accepted.
   I also considered that if my plastic flat belt pulleys didn't hold up under the pressure or couldn't be tightened enough to prevent slipping when tight, I would swap them for metal V-belt pulleys and link belt.
   With these ideas I got back to the long-stalled project. On the evening of the 17th I put the shaft and pulley set back together with some adjustments, and figured out where and how to mount the spring to hold the belt tensioned, and how to position and apply the clutch cable from the pedal to compress that spring and slip the belt. On the 22nd I made the spring part and on the 23rd the clutch linkage.
   Not until the 30th did I put in the new battery box with a 36v, 30AH NiMH battery in it, and try it out. As soon as I tried reverse, the motor controller blew. A video from a camera showed that the belt had slipped off the pulley before that, so it not only blew, it did so with no load.
   This was quite unsettling, as this was the one controller with an impeccable record. It had run the Electric Caik outboard in February just fine. Did I have decent controllers, or were there still circuit problems? But as I disassembled it, I discovered a loose wire clamping screw, which held the output wires for phase "B", which was where the blown mosfets were. With any loose connection in the high power circuits, no further explanation of the failure is necessary... except for why it might have been loose and how it might be prevented in the future. The area was a bit cluttered with wires, and messes abet accidents - could it be made cleaner? A better question might be whether I should put in the TVS (transient voltage spike) back-to back protection diodes that I bought for the purpose quite a while ago.
   Further, when I took it apart and did the repair, I found two sense wires attached only at one end. They should have been removed long ago but went unnoticed in the wiring 'clutter'. If either bare end had touched the case or some other wrong place that too would probably have explained the failure. I don't think I'll look for some unknown design problem at this point.

   The drive belt will need a guide to keep it from going sideways. Another thing I found was that the clutch guide slot bolt had tightened itself up as the clutch went in and out a few times, so the mechanism no longer moved freely. I had already thought of tightening a nylock nut on the end to prevent it from turning. Evidently it's a necessity even for a little testing.

A Simple 2-Speed Transmission? - Morris Minor

   In thinking about the Sprint transmission, there are so many ways a transmission could be made that my mind goes off on tangents at times. This process was abetted by seeing a Morris Minor that was to be made electric, and hearing that another one had been done simply by putting an electric motor straight onto the drive shaft, and putting in a higher gear ratio rear differential. This got around many inefficiencies of a regular automotive transmission. But the motor was probably bigger than would have normally been necessary, and evidently it was sluggish getting rolling - not a surprise. It probably also ran at a rather high RPM on the highway.

   So I started thinking the Minor should work a lot better and need a smaller motor with even a two speed transmission. And, that any belt transmission would be far more efficient than a geared transmission oozing oil around. Also, a belt transmission with reducing pulleys would eliminate the need to change the differential.
   Then there's that clutch pedal idea. On pressing a clutch pedal, a belt-clutch system could loosen belt tension... or tighten it. ...Or, it could loosen one belt and tighten another. When the clutch pedal was pressed 1/2 way down, both belts could be loose - neutral. When it was fully pressed, the pressure of the foot could tension a belt (or pair of belts) providing a very large reduction ratio - a low 'gear'. Thus to start up, the right foot would press the electron pedal to start the motor turning, and the left foot, already part way down, would gradually press harder on the clutch pedal to engage low gear. Once the car had sufficient speed, the foot would be lifted, smoothly shifting the car into high gear to allow driving at higher speeds without undue motor RPM.
   Or we might conceive that this clutch pedal could be replaced by a lever similar to an automatic transmission lever, which would lock into place wherever the lever's button was released. Then the car could be kept in either gear effortlessly.
   And then we can conceive of having the lever tighten still another belt and pulley in the middle of its stroke and having a 3-speed transmission with two neutral positions, at 1/4 and 3/4 of the travel.

   Somehow, this seems simple compared to a variable torque converter, and still highly efficient. Perhaps the variable torque converter could be done with belts instead of a planetary gear, too... tho I can't see how to make something that functions like a planetary gear out of belts and pulleys.

A final note: The front wheel drive has an efficiency advantage. In a 'regular' rear wheel drive, the differential must always be turned by a 90 degree driveshaft gear, creating losses, whereas in the front differential Sprint, the entire differential and both drive shafts are (will be) turned in unison by the drive belt except when cornering: the gear teeth don't move, so it's essentially lossless.


The Electric Weel Motor & the 4-Runner Truck

   Of course, one way to avoid a transmission is to have a motor with stupendous torque that can turn vehicle wheels directly. One simple thought I've had for the still unmade Electric Weel motor is to make the stator in eight octagonal sections. What escaped me in that idea is that with 27 coils, 3-3/8 coils would go on each section. That might be awkward! If instead 3 coils went on each section, the motor would have 24 coils, spaced slightly farther apart. The rotor would have 32 magnets instead of 36.
   Then the easiest way to accommodate those coils would be to use two 48 volt motor controllers each running 12 coils (4 in series in each phase), instead of three 36 volt controllers each running 9 coils. It thus reduces the cost by one motor controller - 500$ at my present prices. I'll probably want to raise the voltage specs on a few of the controller components to have safe margins. However, I now consider that if 48 volts was plus 24 volts and minus 24 volts, or if the drive system was floating ground compared to the chassis, it would be virtually safe as far as electrical shocks. 200 amps times 2 controllers at 48 volts would be 19200 watts. 400 amps - 200 from each controller and four battery set - would provide around 240 foot-pounds of torque. (down from 270.)
   It is of course one thing to have a plan, but there are some other things that have higher construction priority for now. but I drew out the design for one octagon piece on a piece of legal size paper, using the actual metal center, a ruler and a triangle. It seemed surprisingly small and pretty simple. Simple and straightforward raises priority!
   A problem does arise in the inability of the CNC router... any router... to cut a sharp inside corner. Maybe I'll try finishing them with a chisel?

Homemade Segway Clones!

   On the 30th someone e-mailed me with a link to his homemade "Segway" pages after seeing one of my motors on a discussion list. Very cool to make your own if you like that form of transport! His write-up details the "Kalman Filter", microcontroller and gyro sensors used to give self-balance to the system, and his first, second and projected third "Segway" projects: Runner I, the improved Runner II, and Runner III - a project for some rainy weather.

Runner II

Die Plattform wurde dabei aus Alu-Riffelblech angefertigt. Die Motoren mit integriertem  Getriebe und direkt montierten Rädern sorgen für eine größere Stabilität.
Z.Z. ist noch keine Elektronik eingebaut (15.1.12).

Hurra, Runner II läuft.


   His site has schematics, circuit board layouts, complete program source code for the ATMega in 'C', and even a PDF book about the operational concepts, Das Programm- und Filterkonzept für meinen selbstbalancierenden Runner. - von W. Schmidt. I must confess to not understanding much of it. (The language was beyond me!)
   Even more surprisingly, it turns out this German person had no idea where I live when he wrote me, but he's coming to see western Canada including Victoria this July.

A Superior EV Charging System... a superior battery monitoring system...
and an old "Derelectric" car: the 1982 Mazda RX7


J1772 EV Charging Plugs & Sockets

   It seemed galling that "EV charging Stations" were being installed all over town, but I couldn't take advantage of them - the Mazda has an ordinary 120V plug, and no ordinary socket was present at any of these charge stations. (Nor have I seen any EVs using the stations except those belonging to the owners of the station, eg, an Esquimalt municipal vehicle at the Esquimalt public library.)
   It's like they're really just being installed for show: "Look what we're doing for the environment - but hey, it's not for YOU - Get your low-life EV out of here or we'll have it towed!"
   But after getting a good look at one at Westburne Electric [picture], I began to hope that they maybe weren't very complex after all and could be used.
   Also I realized that probably there's no line voltage present on the main pins unless the plug and socket were mated, for electrical safety outdoors in rain, dew or frost. This is in fact the case: it's simple enough and the voltage is shut off until it's plugged in.
   There is however one huge problem: while the J1772 claims it allows for 120 volt charging, there is in fact no neutral pin to get 120 volts from. I'm trying to make EV's more practical with separate chargers on each battery. I have yet to see a 240 volt battery charger in any store, and yet this so-called standard doesn't allow for 120 volt charging. Evidently for the sake of one pin on the plug, it shuts out smaller and much more common vehicles like e-bikes and scooters, handicapped scooters, and converted EV's such as mine.

   Naturally Wikipedia had an article on them. There was a diagram of what circuits were needed for both the plug and the socket. (It doesn't make sense - surely pin 4 and 5 are reversed on one end... but which end?)



The article said:

J1772 signaling circuit

The signaling protocol has been designed so that:

The technical specification was described first in the 2001 version of SAE J1772 and subsequently the IEC 61851. The charging station puts 12 volts on the contact pilot CP and the proximity pilot PP (also "Plug Present") measuring the voltage differences. This protocol allows it to skip integrated circuit electronics as they are required for other charging protocols like the CAN Bus used with Chademo or – the SAE J1772 is considered robust enough for a range of −40 °C to +85 °C.

   The bare car side J1772 plugs cost over 100$ (ouch!) on the web, but I found a 3D printer design to print the body for one. In the absence of any other way to make it work, I'm going to make up a plug for the car that uses the ground pin as neutral. I can plug the front chargers into one phase and the rear chargers on the other to roughly balance the load - which is under 5 amps anyway. According to the spec, the ground wire has 'ground fault detection'. So the question is, will the breaker trip out with the slightest inevitable load imbalance, or will it tolerate a certain amount? I certainly intend to find out. And I'll try everywhere. Hopefully owners will disable or somewhat desensitize the ground fault circuit if it keeps getting tripped by people trying to charge. A "standard" that only works with 40,000$+ cars is of little use to many of the few EV'ers actually driving around. And the 40,000$ EV's generally have a long range and little need to plug in everywhere they go, whereas the 120 volters are the vehicles that really need to plug in everywhere.
   I'm somewhat hopeful that the 'ground fault' detectors won't be very sensitive, because the body of the car is supposed to be grounded to the wire, and that therefore the idea will work if the load is anywheres near balanced. Cars get dusty and dirty, and they would be almost bound to trigger a sensitive circuit in the rain. This would be a nuisance to the charging station owners. The above Wikipedia spec also says the car ('PEV') monitors the ground continuity, as well as the supply equipment. This leaves a lot open to interpretation about what the car is expected, or permitted, to do with the ground wire.

   Later I found the following diagram, which shows inside the charging unit, with neutral and ground tied together... and on the car socket, pin 3 is labeled neutral instead of ground. If this isn't a mistake, current flow is permitted in a neutral wire, and one can get 120V. That would seem more reasonable than excluding 99.9% of battery chargers in North America, but I'm not sure how the ground fault breaker would work, and it can't be both 'neutral' and 'ground' in conformity with normal North American electrical practices. But it's also unusual that neutral and ground are tied together inside the charging unit. Who can really tell - maybe using 120V chargers will work fine, maybe it won't. It's certainly worth a try, since these stations are popping up all over the place.


   On a final note, regular 120VAC sockets out in the rain are dangerous electrocution hazards. The intent of this charging system to leave the power disconnected by a contactor until activated when a vehicle is sensed is good. A similar system could be made for regular outlets by having a socket with a pushbutton on it, which is pressed by the plug on the cord as it's pushed home. However, the cord would become 'live' whether or not the cord was plugged in at the other end, and users would have to remember to plug the car end in first, then the cord into the charging plug. This would still be dangerous and people would have accidents. The button could be on a special charging cord that plugs directly into the car, with the note "no extension cords!" or somesuch on it.
   With somewhat more effort (walking back and forth), I can operate as safely by plugging one end of the cord in indoors where it's dry, and plug the car end in first and unplug it last.
   A "trick" that could make a safer 120 volt charger from regular parts would be to have no actual ground wire, but instead tie neutral to 'ground' in the car. A contactor in the charger unit would have its coil activated when the "ground" wire in the cord was pulled to neutral by plugging in the car. But this would be an unorthodox use of plugs and cords that might lead to unanticipated problems, obviously no ground fault breaker could be employed... and like regular sockets, they probably won't be found at EV charging places.

More Mazda Miscellany

   On the 4th I put in a both-sides carpeted plywood cover over the battery well in the rear of the Madza RX7, with a few more pieces of matching carpet surround. In addition to looking nice and providing a cargo space, this muffled the noise coming through the battery well from the rear axle and wheels. I was a little disappointed by the sonic results. The rear was audibly muffled, much quieter. But I seemed to be hearing all the same unmuffled sounds now coming from the front, as if they had simply moved forward to compensate. But no doubt the sound level is down a few dB.
   Later on I noticed that it was quieter in third gear than in second, and even quieter out of gear. That makes it transmission noise, probably entering the car from both the front and rear end of the driveshaft. Not much to be done about that besides insulate. People outside the car say it's quiet. The silence when stopped at red lights is great!

   Sometime I'd like to get the car up on a hoist, or over a pit, to grease the suspension and steering (it's 31 years old!). But Tom Sawyer and I managed to crawl underneath and spray zinc "galvanizing compound" and then textured black paint on the rusting angle iron battery well supports at the rear, on flat pavement. Then Tom sprayed the bumper black as well, which had various white scrapes on it and looked pretty ugly. Another day we sprayed over the flaking black paint at the sides of the windshield. I've done the mechanical and electrical stuff to make it work. Tom has supplied a couple of batteries to up the range and has done a lot of what's making it look presentable! I presume I'll be helping him out when he gets an electric car or a car to convert to electric, which he very much wants.

   I bought an 11th power adapter, for the 11th battery, and on the 7th I wired it up as a float charger. It was only 3.3 amps instead of 5, and 25$ + taxes - including a discount. I'm now paying for having bought only 10 of the 3.95$ ones at XS Cargo instead of 30 or 40! But one can do better (than 25$) pricewise and also get more ideal 14 volt, 6 amp adapters, by shopping on line - worth it for several for a car charging system.
   This adapter was 15 volts, and instead of opening it, I simply cut the plug off the end and soldered on two diodes on the end of the "+" wire to give a 1.2 volt drop to the 13.8V float voltage (another 8% charging efficiency loss right there), and then a .82 ohm resistor to limit the current when the battery was low (more loss). I used the larger resistor since it was only a 3.3 amp charger.

 Captain's log

   On the 10th I put columns on a piece of paper for date, mileage and trip meter, amp-hours, watt-hours and comments, and I started keeping track of my driving. Short as the trips were (the longest one being 13 Km), I went 171 Km by the end of June, using about 32 KW of electricity - 3.20$ worth if charging was 100% efficient, but also if none of it had been done 'for free' from the solar PV panels.

   On the 12th, Rick Pitts, who had nicely converted an S10 pickup truck came and saw the RX7. The two conversions had fairly similar parts, so it was rather to my surprise that he was impressed by the performance and handling. The big difference is the batteries. He lives out of town and a 20 Km range would be inadequate. His S10 has 24 six volt golf cart batteries, which would weigh around 1500 pounds - 3/4 of a ton. He said it's like driving a concrete block down the road, and the brakes are only just adequate. The RX7 had only about 450 pounds of batteries, the four NiMH'es being only about 115 pounds or so.

   The weakest lead-acid battery seemed to be getting better inch by inch - I think with cycling, but perhaps just because of warmer weather, as suggested by Jim Harrington. But I wanted kilometers rather than inches, and on the 13th I changed it for a new one that Tom brought me to use. I put the old one outside under the wiring closet, to use with the solar collector system (see under 'Electric Equipment Projects').

   After I got the windshield wipers working last month, I thought they'd loosen up with use. Instead they went lethargically, had to be pulled by hand a couple of times to get them moving again, and then seized up again. After driving home in the rain with no wipers on the evening of the 17th, I took the passenger's side pivot apart, with difficulty twisted the rod out of the pipe with vise-grips (it should have been *loose*), filed off some rust from the outside of the rod, cleaned and greased it, and reassembled it all. Now they work without hesitation.

   I don't get around a lot, but I drove the Mazda almost every day and with all the electric driving, I didn't buy gas for the Tercel from April 29th until June 22nd. Unfortunately all the money saved on gas and more is going for extra car insurance. It is amazing how artificial monetary factors often act to keep us from making the best real investments, what with having to pay extra in order to do the right thing, or obtaining savings, eg, energy savings, that are trivial in monetary terms compared to gigantic 'imposed' expenses that we may have little control over such as a mortgage and property taxes.

Longest drive: 13Km

   On the 26th I decided I needed the 6-tray NiMH battery from the Mazda for the Chevy Sprint experiments. I decided to first see how much "ultimate" range the Mazda had with the present batteries. I drove it 8.1 miles (13Km), the last mile going around streets near home. Voltages under load went from 130's to 120's to 110's and finally dipped down into the 100's, the lowest I've driven them to. The CycleAnalyst showed 17.5 amp-hours - 2310 watt-hours at 132 volts. That's the most I've used in one charge. 17.5 amp-hours doesn't seem like a lot of juice to get from batteries rated from 60 to 95 amp-hours. Part of the problem is doubtless that continuing currents of 50 to 150 and even 200 amps are very high for regular marine-RV type batteries. This is a less obvious reason for using golf cart batteries, which are made to handle higher continuous currents. Battery performance improvement with reduced currents is also a reason an "ultra-efficient" car drive system will give a much improved driving range.

   After the trip two of the NiMH batteries were below 12 volts, eg 11.8. They surely had little life left. The other two were just over 12v, eg 12.1. The lead-acids were all above 12.2 volts. If I beefed up the charging of the NiMHs to get them up to 14.0 volts would it be good for 10 miles/16Km? Then again running lead-acids down too far with no sodium sulfate in them is worse than running down the NiMHs. (Now 3 of the 7 PbPb's have sodium sulfate added when new, but none presently installed are ones 'renewed' with sodium sulfate.)
   Three or four batteries were still at 12.4 volts and could have gone a good further distance. The best battery appeared to be the "reconditioned" one I had purchased at Battery Doctor. It was still at 12.6 volts and so apparently could well have driven as far again to a total of at least 16 miles/26Km.

Need for individual battery monitor

   Of course, while the ever decreasing voltages can be seen by glancing at the CycleAnalyst while driving, there's no indication of which batteries are dropping by how much. A manual measurement of each battery afterwards is tedious and can only give 'after the fact' readings. Is it mostly the NiMHs are at the end of their charge and dropping way down under load?, or are they doing okay while the weakest lead-acids are going down to 7 or 8 volts?, or are they all dropping more or less equally? These are the sort of questions my planned individual battery monitor is intended to answer, pretty much automatically, while driving.

   The point was acutely illustrated after a drive on the 29th. The 120 volts DC was low enough that I checked all the batteries. I found one at 10 volts. It would obviously have been pushed much lower while driving - into battery damage territory. In removing a battery 3 days previously, I had accidentally left a charger unplugged. (Wouldn't I make a good surgeon?) The monitor would have shown this in good time. When I turned the car on, probably on the previous trip, I would have seen that one battery didn't seem to be charged, and if I'd missed that, at some point it would have gone yellow and beeped when I stepped on the electron pedal.
   After the same drive one of the NiMHs was down farther than the others, too. I might try to drive too far for it, too some time -- without the monitor.

Ultrasonic Noise!

   The evening after driving 13 Km, my tinnitus was loud and piercing. Yet I had played no music nor worked with noisy tools. It seemed to me I had noticed this before after driving the Mazda. I looked up the PWM frequency of the Curtis 1231C motor controller. It was 15KHz -- the same frequency as the TV flyback transformer whine that originally caused my tinnitus when I was 6 or 7! There it was: the car apparently makes a piercing ultrasonic sound (15KHz is ultrasonic to me these days) that is apparently damaging my already mediocre hearing. That squeal would be coming off the controller, but more so as vibration of the coils of the motor. I decided to carry earplugs in the Mazda and use them. Another brand of controller said it was 16KHz.
   In my controllers, the 'CRM' frequencies may go as high as maybe 30 KHz. But the frequency is variable. It drops lower and lower with motor speed and as the load decreases, and is quickly into the audio range where its loudness can at least be heard and assessed. At higher speeds or low loads the CRM signal may be absent. In the car the CRM is gated on and off by PWM at around 300Hz, so it's only there part of each second. 300Hz is audible (providing some warning to pedestrians), but it has far less sonic energy than 15,000.
   So my controllers aren't 'innocent', and in fact the motor coil humming will probably be heard, but they'll probably be much less of a problem, and a less hidden problem, than fixed frequency units in the worst "near ultra-sonic" frequency range for human ears, often more felt as a discomfort rather than heard. With my units, no one is likely to suffer hearing loss - without even knowing why.

120 volts better than 132?!?

   The next day, the 27th, I took out the NiMH 'stack of trays' battery to use it in the Chevy Sprint experiments. I put a size 24 battery from the back in its place. In removing one battery of eleven, 10%, and dropping the voltage from 132 to 120 volts, one would expect a proportional rise in the current needed to move the car. Instead I got a surprise. When I left home, the car seemed to have more pep, or at least no less, rather than less - if anything at slightly lower currents. I drove 4.4 miles using just 8.1 amp-hours - my first recorded time of using under 2 amp-hours per mile. With the lower voltage, this translated to even more watt-hour savings:
 8.1 AH * 120 V / 4.4 miles = 221 WH/mile or 137 WH/Km.
   It's hardly been below 270/170 before, and sometimes above 300/185. True, I neglected to reset the trip meter on the Cycle Analyst before starting and the amp-hours estimate might just be slightly off, the streets were quiet and the battery made about 30 pounds gone from the car... but the route had the usual hills and I wasn't being super-light with my foot.
   It appeared the car had more range with 10 batteries than with 11! Could it be that the motor or controller just runs more efficiently at a lower voltage?
   If this was verified by further 120 volt drives, I certainly wouldn't be bothering to add another battery. Instead I might be trying out 108 volts again to see how the statistics stacked up. But in the next couple of drives I'd left a charger unplugged and couldn't trust the rather high figures because of the low battery. The third drive was 4.2 miles using 9.06 AH. That's 260 WH/mile, 165 WH/Km, with a very light foot on the pedal, and the next one was 270/172. I'm guessing I made some mistake in the first 120 volt trip's readings. Any slight drop in watt-hours is probably explained by the shedding of the 30 pounds of the battery.

Pipe Battery Problem (It's the chargers!... mostly)

   The two newest 6-pipe NiMH 60 D cell batteries are a mix of older batteries, some of which were in soldered together cases that I took apart. They seemed very good and after a drive were down to around 12.5 volts along with the lead-acids.
   But the one in the stacked cases, and the older bottom one in the box, made of twelve 6-volt tubes, previously used in the Sprint tests and then in the boat with Electric Caik outboard, were getting almost as discharged as the weakest PbPb after 5 miles. The worst one was the twelve 6-volt tubes in the bottom of the box. I suspected that one of the sets of tubes wasn't making contact properly.
   On the 7th I took the battery box apart. When I pulled out the bottom battery I found it had not one, not two, not three, but four end caps broken loose (two of them on the same series pair - and they were all 'plus' ends) and only three tubes were working. Yikes! And yet I seem to remember testing every tube before I put the sets of 6 together, and fixing, at least, 2 or 3 of the new ones that I'd just made. (This was the morning after thinking the voltages were getting low unexpectedly soon near the end of a drive, so one or two of the breaks may have happened just within a day or so.)
   I glued the caps back on and put the box together again the other way up, with this suspicious set of 12 six volt tubes on top where I could keep an eye on them. Some thoughts are:
 - that the caps are ABS while the pipes are PVC. (There's supposed to be a "transition glue" for both materials in the hardware stores.)
 - That it's possible but not easy to twist an end cap off when glued with the present methylene chloride solvent, which seems about right in case access to the cells is needed. Having them come off by themselves, of course, is no good.
 - that this battery had the pipes oriented vertically in its original box. The means the weight of the batteries was bouncing up and down, eg with waves on the sea or being handled roughly or dropped, stressing the end caps.
 - These particular tubes have a "U" spring of nickel-brass on the plus end, a supposedly advantageous feature to help hold the cells under a bit of pressure. This may have had some odd effect - maybe there's more pressure on the caps than with other tubes.
 - I think it's the first time I've had even one cap accidentally come loose, so I think I'll just see how they all hold out now.

   The next day, the 8th, this battery seemed to do okay, while the other low one, the one in the 6-stack of 3D printed cases, was about empty (reading 11.99 volts) on return from a too long drive. Without knowing what it's problem was, I decided just to replace part of it at a time and see what happened. Before dark, I pulled it from the car and replaced the bottom tray, and the cells in the second from bottom one. The "-" terminal of the 2nd tray seemed quite loose when I was removing that tray. Perhaps it may not have been connecting, but the cells in it had the exact same voltage as the others. Then I put it back in the car and got it on charge for the night. Ideally I'd have replaced 1/2 the battery instead of 1/3, but I now only have 20 free NiMH D cells.

   On the 10th I took a an 8 Km drive. The same two batteries, in spite of it all, were at the end of their charge while the other two had a few Km left in them - about the same as before. I started thinking about upping the low ones to 70 amp-hours to compensate.
   It would at least be simple to add one more tray to the stack of 6 - I had made it with that in mind. But I persevered and swapped out the middle two trays with the ones previously removed from the bottom two which had no apparent effect. Most of the trays had a lower voltage following the drive, but one seemed to be more fully charged. It also seemed to have a bad connection. I suspected the bad connection was the problem.
   Naturally one bad connection drops the range to less than 5/6 - maybe to 3/4 or worse - owing to the higher currents being drawn from the remaining five banks.
   The tube batteries showed no obvious problem. With the end cap problems, probably from the push and pull on individual tubes from the copper "springs" in between them, I might want to break up the 6v tubes and make 6 new 12v tubes with the same cells. Those would have 1/2 as many end caps, with no special stresses on them.
   Finally I put in a 7th tray, and a 7th pipe. In spite of the theoretical added capacity, at the end of an 8.5 Km drive, they were down to 12.0 & 12.1 volts while the other two were still at 13.3 or so.

   After all that, it finally occurred to me to check the voltages when the car was charged, instead of only after a long drive. On the morning of the 13th the lower ones proved to be only about 13.7 volts while the higher ones were at 13.8. That seemed to explain the differences, but evidently none of them were being fully charged - at least, not in 8, 12 or 20 hours. The chargers were putting out about 14.0 volts, but into no load. Going through the .62 ohm resistors, the charging current drops more and more as 14 volts is approached. At about 13.8 volts, only 300mA is flowing, or 50mA per 10 amp-hour bank, and at 13.9 volts, it's half of that, 25mA. 50mA may not even exceed the 'idle' current once charged, and so they won't charge up to the desired 14.0 volts at all.
   I changed the resistors that set the regulated voltage in the low pair of chargers and got them up to 14.11 and 14.07 as measured in the lab. (It's slightly pot luck with no adjustment potentiometer.) As wrote about this, I considered that I should probably raise them all ideally to about 14.15. Without the current limiting resistor, 14.2 would be too high, but with it, the batteries probably won't get above 14.1 or so. The resistor will prevent excessive currents. Or, anywhere from 14.05 to 14.15, a smaller series resistor would bring up the voltage more quickly and closer to the set voltage. But that would need a charger either with more amps capacity, or one that would do constant current at its maximum capacity until it reached its set voltage.
   I decided to do more measurements on the NiMH batteries, charged and discharged, over a few days before doing anything more. With the weak lead-acid now out, if the NiMH's would get up to the optimum 14.0 volts, I figure the car should have perhaps another 3 to 6 Km of range. 12 to 15 Km will take me a few more places than 8 or 9. 20, 25 or 30 would start making a big difference.

   Then I started thinking about the 13.8 volt float chargers for the lead-acids... maybe for the same reason those 13.8 volt charges should be upped to 13.9 or 14.0? Some of them already were 13.9... were those batteries the ones that were staying highest after a long drive?

   At the end of all this, the NiMHs were still only charging to about 13.8 volts, and the lower ones to 13.75.

Lead-Acid versus NiMH Float Charging


   The float charging turned out to be the opposite of what I originally expected. The PbPb batteries charge and attain full charge at a lower voltage, then rise up to near the 13.8 volt float charge voltage when they're full. Since the chargers put out their highest current, 5 amps, at around 11 to 11.5 volts, and less and less as the voltage rises towards 13.8, the lower voltages mean they charge more quickly, in relative terms.
   The NiMHs, on the other hand, rise quickly to higher voltages at lower states of charge, and thus they end up charging much more slowly. 8 or 10 hours wasn't enough - they were needing a whole day, and they only seemed to hit about 13.75 to 13.85. At that the NiMH batteries would be only partly charged and were being drained to 12 volts while the lead-acids still had plenty of range left. Thus with the weak lead-acids having been replaced by strong ones, the NiMH batteries, which should have been great, were definitely limiting the range of the car.
   Partial workaround solutions would be to set the chargers to a voltage bordering on "too high" like 14.15 instead of 14.0, and to reduce the charging resistor to .4 ohms (at 10 watts). That would mean that at 12 volts, the chargers would be being pushed to their 5 amp limit, but would source somewhat higher currents than before as voltages rose.

   As a first step to seeing the difference, I took out my regular 6 amp battery charger that's set to 14.2 volt cutoff for
NiMHs, and started more quickly charging up the 4 NiMHs, one at a time. Surely having all good, full batteries would provide a 20Km range! It helped, but it too cut off without getting them fully charged - and the voltages actually dropped again once the cells were back on 'regular' float charge.
   13.8 volts seemed to be about all I was getting. Nevertheless on the 19th I drove 6.9 miles/11.1 Km on a single well charged charge. 3 of the 4 NiMHs were down to 12.0 volts as measured afterward. They may have had another kilometer or two in them above 10 volts, but no more. This didn't get put to the ultimate test. The lead-acids were all over 12.2 volts and mostly over 12.4, indicating a few more kilometers were available from them.


More Amps & Motor Controllers

   The Mazda uses around 25 to 95 amps cruising along the street, depending on the slope of the road. It can easily use 150 to 250 amps when accelerating or going up a steeper hill. And that's at over 100 volts. The Electric Hubcap seems to have pretty high torque per amp, but noting the RX7 performance, I'm considering that pumping 200 or 300 amps into a 36 volt Electric Hubcap motor at times will be a minimal requirement, even with an efficient transmission and optimum motor speeds from the torque converter. I'm putting temperature sensors into the motors, but I'm not half so concerned about the motors overheating (the Electric Caik Outboard hardly got more than cold at 50-60 amps) as I am about my motor controllers blowing up. Theoretically they'll take 240 amps continuous, and higher surges for short periods... Theoretically!
   Related to this, someone looked at my controllers and pointed out that a problem with TO220 and TO247 transistor packages is that the screw-down hole is at one end while the actual semiconductor inside is at the other. The active end isn't being pressed against the heatsink very well, and if contact isn't good, the heat has to travel across the metal tab to the area of the screw before it's dissipated into the heatsink, so the transistor is running hotter. And as someone keeps pointing out on the motor controller discussion list, "Heat, not current, is the enemy." The TO247 seems better than the TO220, but still at very high amps, thermal runaway is possible. ...and what if that plastic around the edges isn't exactly flush with the metal? Potentially it could actually hold the metal off the heatsink. My early controllers from 2008, which weren't very well current limited and might well have pushed out a few hundred amps, often burned up, seemingly, or so I thought, from high currents causing sudden thermal runaway.
   The thermal transfer aspect of the design may bear some investigation and measurement - or just redesign to ensure the best attainable heat flow from that spot on the lower right where the silicon actually is, to the heatsink. Maybe even a copper heatsink bar under them instead of aluminum (which is inevitably aluminum alloy, which is less heat conductive than hard-to-get pure aluminum).



Electric Equipment Projects

Renewed Lead-Acids for Solar Batteries

   With having the 'renewed' PbPb batteries that didn't seem to have enough capacity to use in the car, and which nevertheless should have hundreds of cycles in them and might well improve with use, I decided I should use them with the solar PV system to increase the storage capacity, which was quite small all along and had shrunk with the loss of most of the NiMHs, now used for cars.
   I decided that rather than bring the batteries into the house and upstairs, and having to vent the closet to the outdoors in case of gasses, I'd run a long wire to the ground and just put them there, outside below the 12V wiring closet. I used some #14 house wire. Rather thin, oh well - the hole in the wall was small. The bare ground wire would be ground, the white wire (taped red) would be to the main house supply (which is 14.0 volts in the sun, dropping to 12 or less at night, via a circuit breaker), and the black wire (also taped red) would be the DC-DC converter output, also through a breaker, about 14.35 volts when the sun is out.
   Each battery would supply the main house power, but isolated through a diode. And similarly to the car, each would charge through a diode and a small resistor (1/4 ohm?) from the DC to DC converter. The diode would drop the 14.35 volts to about 13.7 for the constant voltage float charge, and the resistor would limit the charging current for gentle charging.
   It occurred to me that the place to get low voltage, high current diodes for batteries was an auto electric store rather than the electronics store - they have just the right ones, for car alternators. The clerk couldn't understand wanting any number of them besides six (the number in an alternator), so I bought six instead of four, with the negative side being the case. That way they could all go into one heatsink without being isolated from each other. The plus side would come from each battery.
   At ground level would be a high current diode from each battery to the house supply cable. If the system voltage slipped below the battery voltage less the .6 volt diode drop, the battery would supply into the system.

The equipment of the Solar PV System is improving.
However, so far it's all store-bought stuff and not original creations,
except for the 10" plate to mount circuit breakers on.

The NiMH batteries are disappearing into car projects.
The white wire through the wall is for lead-acid batteries on the ground below.

MSP430 Microcontrollers - Interface & Software Development

   You'd think in a whole month I'd have had time to do something on this, but all I got done was to solder a few parts onto a proto board for the fridge control. I didn't even get it wired up.



Electricity Storage - Turquoise Battery Project (etc.)

Production Prototype Cell #4 ("PP#4")


  
NiMn "+"ode on flexible graphite current collector, separator paper

 
'extra' nylon separator cloth, vertical bar grille 'basket' for "-"ode.

 

Upper nylon & paper separators

 
Ni "-"ode, copper foil "-" current collector
 
 
Rubbery stuff to press everything together


Cell, opened later.


End of Month: PP#6 with carbon fiber cloth folded under osmium doped graphite current collector sheet.
To be folded down, over electrode briquette and under separator sheets, allowing the electrolyte to pass through
while making electrical contact points all across the front face of the electrode to improve conductivity.

   Last month battery research by others showed use of insoluble Barium Ferrate. This seemed promising - what about barium manganate? ("Slightly soluble" potassium permanganate is somewhat more soluble than I remember reading, at 6g per 100g of water. Where did my mind get 2g from?) By changing the electrolyte from potassium chloride to barium chloride, perhaps the manganese dioxide (MnO2, valence 4) would charge to [probably] insoluble barium manganate (BaMnO4, valence 6 - 2 electrons per reaction) instead of potassium permanganate (KMnO4, valence 7 - 3 electrons per reaction). This would seem to make the whole idea of charging to a higher manganese oxide and moving more than one electron per reaction definitely workable even by "regular" battery chemisty standards without chelation.
   Further reading however discloses that there's a valence 7 'barium dipermanganate' (Ba(MnO4)2), which is very soluble. If it charges to this instead as seems likely, the purpose of using barium is negated.

   On Sunday the 9th, after once again seeing everyone looking for better batteries with nearly identical discussions on "Electric Boats" and "Battery Conversions" discussion lists, plus having been presented with some new ideas at the end of May, I decided to throw together another test cell.

   This time, I decided that if my cells had poor conductivity and high self discharge for reasons unknown, I'd make a completely different negative electrode and see what changed. I decided to use nickel, which is only -.7 volts and will yield a cell well under 2 volts. But at such a low voltage there should be no hydrogen outgassing even if graphite is added to improve conductivity, and so it should have little or no self discharge. I used monel powder (Ni:Cu alloy (or 'solid solution'?), ~65:35). With the copper content, conductivity should be high. (Reminder: nickel can only be used as a negode in a moderately alkaline cell: it's somehow immune to oxidation at pH 14. It's the only metal that is.)

   If either or both problems were still there, it would seem they originate in the nickel/permanganate posode. If both were gone, the Mn negode is to blame. At that point I could either try things to improve the Mn negodes, or abandon them in favor of the monel. I'd rather not do that if it can be avoided, since the manganese adds about .7 volts more to the cell.
   If the self discharge was being caused by the posode, perhaps it might still be worth a try to replace the potassium chloride, or some of it, with barium chloride. I can also try some ferric chloride, a strong oxidant, but I discovered that has to be mixed into the electrode, not added to the electrolyte later.

   Another point to the new cell is that "PP#3" suddenly 'went bad'. Up until that point, its performance had been gradually improving with each cycle. In "PP#2" there didn't seem to be much left of the separator paper when it was disassembled. Apparently #3 had the same problem, because when it went bad, electrode powder material came out the filler hole when the electrolyte was drained. (Yes, I really ought to rip it apart and look.)
   My remedy is to add a layer of thin 100% nylon cloth in front of each each paper. The paper won't let even minute particles through, and the cloth will support the paper at all points even if it has more or less turned to mush.

   Another and rather risky change I made was to use PLA plastic for the cell body instead of ABS... and without even verifying first that methylene chloride ("MC") would glue it! The intent here was to use a plastic that hopefully wouldn't wick the water up the sides to the filler hole. The PLA proved to be MC proof and the face cover easily popped off. I waited until Monday to go to Industrial Plastics and get some MEK to try instead. It worked, but the bond wasn't very strong.

   The posode was made with the same mix as last time (was that from March?). That ran out without filling the compactor, and to finish I used an older previous mix having MnO2 instead of KMnO4, to which I added some monel powder to get a bit of granularity, and some KMnO4. These were wetted with Diesel Kleen, compacted (about 1/2 of each), and the electrode placed in the cell. Its resistance was moderate, in the lower 100's of ohms from any surface point to the graphite current collector.

   For the minus I used 40g of fine monel powder (still sort of coarse & granular for battery work) to which I added 3.5g graphite powder to hopefully improve conductivity (which should be good anyway), and 5g of nickel hydroxide to get some fine nickel in the voids between the granules. There was no need to add transition metal compounds to raise the hydrogen overvoltage since the reaction voltage of the nickel, -.5 volts or so, is lower than the base hydrogen generation voltage. So common things like graphite and a copper foil current collector should work fine. (A sheet zinc current collector here would oxidize since it has a higher reaction voltage than nickel.) The electrode used all of the mix - it was the right amount.
   But I should have used more graphite - resistance was in the K ohms. And the effect of compaction was apparent. There were about 5 fills and compactions making 5 visible 'bands', and the different bands had different meter probe contact resistances from the lower 1K ohms to over 20K ohms. This was all high, and varied doubtless from uneven compaction.

   The cell didn't perform well. On the 14th I opened it up. The MEK didn't glue the face on very tightly, and I slit it open without damaging the insides. Also I had realized that with the new design there was no reason to glue the internal basket in, making it removable.
   Underneath, the soft, tough watercolor separator paper of the posode appeared to have become a very brittle sheet with yellow colors, probably sulfates. That wasn't an expected part of the plan. But it probably explained where the yellow and orange crystals on some of the [leaking?] terminals came from - the paper, and probably the sulfates in the Sunlight dishsoap.
   On the 17th I removed the negode from the tray, ground it back to powder (it broke up easily), and added 10g of graphite to its 46g. The graphite seemed to be more bulk than the electrode material, but after I compacted it back into the square, only 11g of material was left, and it didn't look like much bulk. This time, negode resistance readings were in the mid tens of ohms between any two points. That was more like it!
   I also poured a little ferric chloride into the posode. It bubbled fiercely for a brief time. Unlike rather inert ferric oxide (rust), ferric chloride is a strong oxidizer. And I rinsed off the separator bars area. It was surprisingly dirty for a zone that was supposed to be clean with no powders leaking into it. It's possible that crud between electrodes could explain all the self discharge problems - if this has been regularly occurring. Note to self: once a battery is assembled, don't squeeze it, even if discharge currents seem to increase a little!

   This time, I decided to assemble the battery simply by clamping it between two flat plates of metal, with a piece of rubber for the face instead of plastic, and simply leaving the four C-clamps on - it's a test cell after all. No glue except some modeling clay to attempt to seal the electrode terminals. Then it would be simple to open, and anything can be inspected, cleaned or replaced. It proved impossible to stop the leaks, but I just kept pouring in bits of salt solution to keep it from getting too dry.
   On the morning of the 18th the smell of diesel kleen was pretty much gone, and I put it together, with a new piece of nylon cloth in the positive separator side. It started out apparently discharged at .6 volts and it would supply virtually no current. With a 35mA charge, it charged up to 1.7 volts or so quickly, then gradually rose and fell, hitting a peak of 1.800 volts. Then it started falling to around 1.73 volts. That this was owing to electrode conductivity increasing as it charged and not some increasing self discharge seemed indicated by the voltage when momentarily disconnected staying higher and falling more slowly with time. However, currents were only somewhat higher, and the cell still didn't perform much like a 'real' battery. It completely discharged itself overnight. But the next day I took it apart and rinsed it, charged it a few hours, and gave it a discharge test with a 50 ohm load. It looked pretty hopeless with quickly dropping voltages, but it started to level out at around 1.25 volts, and then started dropping much more slowly. This sort of voltage didn't make much sense for KMnO4 discharging to 2 MnO2 (with the Ni negative discharging to Ni(OH)2)... but it was about right for MnO2 discharging to Mn2O3. It hit 1.0 volts in 1/2 hour, and continued dropping about 10mV per minute down into the .7 volt range for an hour. That would probably be the negative continuing to make Ni(OH)2, with the positive going from Mn2O3 to Mn3O4 and finally to Mn(OH)2. But it was all milliamp-hours instead of amp-hours.

Self Discharge Changes With pH!

   It finally occurred to me to check the pH. It was 6. At that acidity, many things that are insoluble above pH 8 become soluble, so crud saturating the separator paper and self discharge would be expected. Aha! I dumped in some calcium oxide to raise the pH to 12 or 13 before trying to charge the cell again.
   Low pH in general, even 8, 9 or 10, might explain a lot. There's the "predominant species" - hydroxides - then there's what else might be floating around in small quantities that might have untoward effects. Perhaps the pH really needed to be kept up to 11, 12 or 13... or (perish the thought!) 14 to prevent the self discharge I've been having all along. Having thought of it, I checked out alkalinity versus self-discharge.
   Sure enough, 3 hours later, when the charge was disconnected the voltage dropped from 2.3 volts to 2.08, and there came almost to a standstill as far as immediate observation. Considering only trace additives or trapped hydrogen could cause such high voltage, it was a dramatic reduction. When I tried a load test after this relatively short charge period, voltages started out higher and stayed higher until about the 16 minute mark.
   Why have I never thought of or noticed a correlation between pH and self discharge before? Well, after all the standard dry cell (non alkaline) is low pH, and it holds a charge very well.

   However, staring at a voltmeter and leaving the cell overnight were two different things and as usual most of the charge was gone by morning.

   Also to check out self discharge factors, I decided to make a new posode with entirely traditional stuff:
* nickel hydroxide (15g),
* with graphite for conductivity (7.5g),
* samarium oxide to raise oxygen overvoltage(2.1g),
* and 1% CMC gum for binder(.5g).
Nickel hydroxide electrode with the paper peeled back.
Far left is the negode with a copper grille on the back.
   This made an electrode with a bit to spare. The only non-traditional part was wetting the mix with toluene to dissolve the graphite to have it reform into more conductive forms - nano-tubes and lamilae - as the liquid evaporated. (I used toluene [methyl benzene] just in case the diesel-kleen [with tri-methyl benzene] has been leaving anything behind after evaporating, causing the self discharge.) I also omitted the Sunlight dishsoap with its sulfates and sulfonates. Theoretically - my theory - not chelating the metal ions is what gives electrodes their limited cycle life and chelation makes them permanent. But after all my poor results, this time I want to try things more in accordance with things that have worked in the past. The new nickel hydroxide electrode (June 19th) had the silvery color of reformed graphite and was around 250-400 ohms between various points if the meter probes were dug in a bit.

   I don't know why, but I opened the cell, hoping to put the new electrode into it. It would have been better to first simply add enough potassium hydroxide to the electrolyte to raise the same cell pH to 14, and see if that helped. Could my whole idea of a "moderately alkaline" electrolyte be flawed? Anyway, it doesn't take much KOH added to salt solution to raise the pH to 14, and I still think the main electrolyte can be KCl, potassium chloride, which will still be much safer than pure KOH owing to the much lower concentration of hydroxide ions. And of course NH4Cl doesn't discharge standard dry cells even at low pH. (But I wonder why they use NH4Cl instead of KCl?)
   Trying to get the posode out broke up the graphite current collector. Oh well! I took another case and put in a new graphite sheet, separator sheets, and the basket still holding the nickel negode. ...In case nickel will work at pH 14 with the chloride in the electrolyte. According to Hoyle, at pH 14, it should be just a 1.1 (nominal) volt cell. It definitely didn't want to charge up to the sort of voltages of my permanganate cells, sitting at under 1.4 volts on charge.

   This cell now, PP5, had the usual low conductivity on the first try... and it got worse after that. And it still lost most of its charge overnight. Alright then, what about Ni-Fe at pH 14?. If that didn't work...

Back to PP #3?

   In some e-mails with someone else who's having trouble making batteries that perform acceptably - NiFe in his case - I started thinking that I should try the same formula as cell PP3 (TE News #62). That one seemed to be working and gradually improving for a couple of weeks until suddenly it failed. Probably one of the separator papers gave way and allowed electrode material to leak between electrodes. Three things I'll change: (1) add some cloth separators (nylon?) to the paper ones to help hold things together. (2) Shorten the torching of the negode. I think 7 or 8 seconds of torching oxidized it too much, reducing the conductivity and performance, resulting in the poor initial performance. I don't like reading 200 ohms before torching and 200,000 afterward. (3) Put a carbon fiber cloth next to the posode, under the separator sheets - see below.

   I mentioned to him, and should do so here in case I haven't mentioned it in some previous newsletter: the idea of chelating the slightly soluble KMnO4 with sulfates and sulfonates (in Sunlight dishsoap) to prevent the MnO4- ion from migrating appears to work. The evidence for this is that the electrolyte initially turns purple with permanganate (obviously a little gets away), but then it clears up, indicating no more permanganate in the electrolyte. But if the posode is later broken apart and wetted, purple permanganate comes out into the water, indicating that it's still present in the electrode. (The initial dissolved permanganate probably hits the negode and goes through the whole charge sequence MnO4- -> MnO2 -> Mn2O3 -> Mn3O4 -> Mn(OH)2 -> Mn metal, adding a minute bit of Mn substance permanently to the negode - which is of little consequence if it's Mn anyway.)

   On the 28th I put in the collector sheet with an edge of a carbon fiber mat tucked under it, and made the posode. I used nearly the same mix of chemicals as for #3 (TE News #62), compacted as hard as I could. The resistance point to point was just 20 to 30 ohms. I put it in, folded the carbon fiber on top of it, put in watercolor paper and nylon separator sheets, and pressed the upper basket down on top. But I didn't finish the cell in June.

Carbon Fiber separator sheet to improve electrode conductivity

   What happens if the posode separator sheets are three: (1) carbon fiber, (2) paper, and (3) nylon? The carbon fiber would help connect any low conductivity areas in the electrode by connecting the whole face area together. I decided to immediately try this in the present cell with its low conductivity nickel posode. I took it apart, and found that the paper was in good shape. It must be the Sunlight dishsoap I usually add that suffuses them and makes them brittle. It also was only stuck to the electrode in a couple of spots, so I got it up pretty easily. That doesn't say much for holding the electrode compressed, which probably explains its decreasing conductivity. I didn't think to check electrode surface resistances, added the carbon fiber matt on top of the posode, and put it together again. It would doubtless be ideal to wrap the cloth over the electrode edge on one side and have it press on the current collector sheet, but I couldn't do that without pulling up the electrode and probably wrecking it.
   The cell worked better again, whether from the conductivity of the carbon fiber or simply because its thickness pressed the electrode together better, I'm not sure. Probably both. I looked for a thin carbon fiber cloth or non-woven fiber to continue using this little trick, but the plastic place only had more of the same as I had, 50mm, 100mm, 150mm or meter+ width. After all, low conductivity is one of my main problems. I wouldn't trust a fine carbon fiber cloth not to leak powder or short to the other electrode across the vertical grill space. Otherwise it would be a theoretical possibility to have no other separator sheet.

NiNi, continued

   To continue with this NiNi cell: this time it didn't seem to leak. (yay!) pH seemed to sit around 12-13. It seemed to charge up to about 1.45 volts, and with 50 ohm loads it ran down to 1.0 volts in: 50", 70", 90", 4'30", 11', 17', 16', 15' on seven tries. As usual: hours of low rate charge, milliamp minutes of discharge. After about the 5th load test, it seemed it would now go up or down in voltage only depending on how long and how hard it was charged.
   On July 1st, after charging all night at ~30mA, it started at over 1.7 volts open circuit and lasted 22':45". This time I left it going, and found that opposite to most Ni-Xx cells, the steady 30mV/minute voltage drop had started slowing somewhere around 1.15 volts, and it slowed dramatically under a volt to 10, 6 and finally 4mV/minute. The cell ran another 50 minutes before it was down to .75 volts. I ran it for 2 hours total, down to .636v, where it was dropping at just 2mV/minute, still putting out over 12.7 mA into the 50 ohm load: total about 30mA hours. It ran even longer down to 5.3 volts - 4 hours and around 55 or 60mAH, still working (if 1/2 a volt is useful). The voltages being so far below the theoretical voltage of Ni-Ni (1.2v no load) is probably attributable to high internal resistance, but then it didn't recover very far after the discharges. The question seemed to be not why it worked well at low voltages, but why Ni-Ni should charge up above about 1.3 volts or so in the first place.
   The initial high voltage was probably hydrogen ions or gas trapped in the negode, forming a nickel-hydrogen cell. This is supported by bubbles seen on the surface and a lowered electrolyte level noted later. Once the hydrogen was exhausted and as it fell under 1.2 volts or so, the Ni-Ni chemistry became active.

   Ni-Ni chemistry is probably 'perpetual'. Whereas metal hydrides gradually corrode, iron gradually over-oxidizes and turns to non-conductive rust, and cadmium and zinc gradually dissolve or form crystals that short out the cell, discharged and even overdischarged nickel can be charged back up to metallic state and there would seem to be no soluble forms. And it's just .1 volts lower voltage than the other Ni-Xx types.
   However, a Ni/Mn - Mn cell, PP#3, gave 100mA-hours at double the voltage of Ni-Ni in March. It was on the borderline of being a useful cell. Furthermore, until it suddenly 'went bad', it appeared to be improving gradually with each cycle. I think I'll go back to that chemistry and see what improvements I can tweak out of it. With a few times improvement it would make a good cell, where this NiNi cell would need at least 20x and twice as many cells would be required to get a desired voltage. I may first try Ni-Mn using the Ni posode of and in this cell with an Mn negode. When I take it apart, I'll try to check its surface resistance (carbon fiber cloth permitting) and see if it seems worth it.

   And I really should find out for sure if graphite and graphite powder cause hydrogen bubbling and self discharge of Mn negodes. If not, why am I using a zinc current collector sheet and zinc powder as a conductivity additive? If the carbon fiber separator sheet could be used on that side, too, it would doubtless make for a good improvement in conductivity. If that doesn't work, I could perhaps make a perforated zinc sheet - see next.

Making holes in metal for perforated metal electrodes - a technique

   An e-mail correspondent suggested a technique that sounds great for perforating metal: silk screen a mask with tiny holes onto the metal and etch it in acid, much like making a printed circuit board. Naturally I would do the laser printer toner transfer for the pattern. Both sides would of course need screening and there may be a complication or two, but without having a special punch to do it, it just might be the practical way. I should think zinc would erode pretty fast in acid.
   Pursuant to that, someone said a faster way to etch circuit boards is to soak a sponge in the etchant and repeatedly wipe the board with the sponge, wearing rubber gloves (and eye protection) of course. The trouble with all the bath methods is that the waste product builds up on or at the copper and stops the reaction until it's washed away by agitation of the bath. The sponge continually wipes off the byproduct. I'll hazard a guess that it only takes 4 or 5 minutes (for a circuit board) instead of 15 , 20, 30 or more.



http://www.TurquoiseEnergy.com
Victoria BC