Turquoise Energy Ltd. News #76
  May 2014 (posted June 7th)
Victoria BC
by Craig Carmichael

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

Feature: Plan for complete torque converter transmission (see month in brief, Electric Transport)

Month In Brief (Project Summaries)

In Passing (Miscellaneous topics, editorial comments & opinionated rants)
* Hundreds of thousands of unsold gas cars fill hundreds of fields all over the world
* OMG: Nuclear War? USA considers a nuclear missile first strike? EVERYONE WILL LOSE! NO MORE WAR!

Electric Transport - Electric Hubcap Motor Systems
* "Centrifugal Torque Converter" is only a "Centrifugal Clutch"?
* A Bigger Plan: "CVT" transmission with planetary gear torque converter and centrifugal clutch
* Magnetic clutch torque converter: a new concept?
* Mushroom outboard: Leg from PP-epoxy composite?
* Individual Battery Monitor
* Mazda RX7 EV news: 18 V, 100 AH, 5 lithium-ion cells
* New concept in aluminum-air cells (can't recharge): install them in vehicle for when regular range runs out
* Caik Motor: Ready-made Rotors

Other "Green" Electric Equipment Projects
* Peltier Module Refrigeration
* Flat Panel LED Lights & Plant Grow Lights

Electricity Generating (no reports)

Electricity Storage - Turquoise (NiMn) Battery Project etc.
* Carbon Black (or graphite) from old dry cells (sigh!)
* Lithium Ion 100 AH Battery Sets

No Project Reports on: Lambda Ray Collector, Magnet motor, Pulsejet steel plate cutter (dropping this one until and unless further notice), CNC Gardening/Farming Machine (sigh, maybe summer... 2014... 2015?), Woodstove/Thermal Electricity Generator (may abandon), evacuated tube heat radiators.

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

Construction Manuals and information:

- Electric Hubcap Family Motors - Turquoise Motor Controllers - Ersatz 'powder coating' home process for protecting/painting metal
- Preliminary Ni-Mn Battery Making book

Products Catalog:
 - Electric Hubcap 4.6KW BLDC Pancake Motor Kit
 - Electric Caik 3KW BLDC Pancake Motor Kit
  - 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 craig@saers.com)

May in Brief

LED emitters soldered to copper sheets - good heat dissipation.

   I became rather enthused by the flat panel LED light design and spent some more days on it, mostly finding supplies and sources and ordering parts. Since the prototype alreay worked, I knew it was doable. It seemed like something relatively simple to make with inexpensive parts, an advance in ceiling lighting, not available in stores around here, and hence that might actually sell well, with the simple appeal of its low profile, inobtrusive design, and probably at a better price than most things I've been able to offer for sale so far. And plant growing lights, which are mostly fairly costly, could be made in the same design. I got the blue and red LED emitters for these along with the whites.
   When I found a thinner translucent acrylic plastic for the bottom diffuser, I made a new case for the same prototype circuit board. It let notably more light through. Then I abandoned the large circuit board in favor of strips of thin copper sheet, which would dissipate heat better, and only part of one strip would be the PCB. With this system I can make lights about 4" wide and any length from 4" to maybe 10", with 2 or 3 to maybe 8 or 9 series sets of three LED s for different brightness needs. Beyond 10" or so, the copper strips might be a little flimsy. A thicker gauge copper could extend the length even longer, but at a higher cost, and the current at 12 volts would head over 2 amps.

   At the same time I picked away at the 1/2 finished centrifugal torque converter, shaping new shafts for the drum and for the motor, which would butt together with a pin to hold them in line with each other. Then I made 8 plastic (UHMW-PE) pivoting shoes and mounted them on 4 threaded rods threaded through the input rotor in pairs, one on each side of the input disk rotor in each position.
   On the 21st I tried it out on the bench. In the forward direction there wasn't much torque. Going backward there was considerably more but it didn't seem like enough. But once again I had neglected to provide for a means of measuring the torque on the bench.
   Then I turned the motor up quite fast, and lost my grip on the drum. I soon found I could hold it from turning at high speeds, but just barely. In all this, the motor didn't seem to be working too hard. Evidently it wasn't being loaded down enough.
   There were 4 pairs of shoes, and room on the rotor to mount 8 pairs, which should presumably double the torque and motor loading. In addition if necessary I could add weights to the shoes, or springs, which should make them push harder on the drum. The results seemed promising, and if the car would only move one direction, that would be good enough for now. First I made a new drum/output shaft that could hold the torque wrench, and measure the present levels of pressure... which turned out to be only 5 foot-pounds. That needed to be multiplied by about 20 to put the Sprint car on the road.
   I went with adding weights to the shoes - 3/8" x 2.5" bolts. I put in larger (5/16") 'axles' for the now heavier shoes, and actually reduced the number to 3 pairs, 120° apart. The heft of the shoes now seemed more in keeping with the sizes and forces of other car-motive components.
   Bench testing commenced on the 31st. But I couldn't get more torque out than the motor had, and early in June I became convinced that my theory wasn't working and that I had only created an improved(?) centrifugal clutch. But if that clutch was combined with a planetary gear torque converter (PGTC) it would be a working system. I already know the PGTC works once it's turning, but it needs a clutch to allow it to turn before the car is moving.

   Around mid month I put the MnO2 plus carbon black substance from a dry cell in a jar, and rinsed it with 3 fills of filtered water to dilute out the NH4Cl electrolyte, then dried it a few days. Then I was too busy to get back to it - Rats!
   But I wanted to move ahead with the torque converter too, and there seemed to be no end of other things that needed doing. I finally saw an accountant about my corporate tax filing, now over 2 months late, that I was having trouble figuring out. He said it was too hard for him. But he gave me the name of another accountant - who was on holidays. But I finally saw him and he got me started again by getting me to call a number at Canada Revenue for statements about my 'payroll account'. (It took a whole day to get through on the phone. I could have had this straightened out a couple of years ago if Canada Revenue had an office in BC's capital, where one could talk to someone when confused.)

   On the 30th I was informed that the U-Vic Ecosat satellite, with the diamagnetics & laser navigation experiments, in which Jim Harington of AGO Environmental Electronics has played the leading role, was a go for funding to be launched. Until that day this was still under consideration and nothing was certain.
   My role in the project, at least so far, has been trivial yet apparently crucial. In looking up some diamagnetics info on Wikipedia, I passed on a link to a Japanese researcher mentioned in in an article's references. It looked to me like his work was closely related. He was contacted and indeed he became involved in and important to the project.
   I'm fairly certain that magnetic drive is the key to practical travel across the solar system and between the stars. In the more credible stories of UFO s, awesomely powerful magnetic fields often seem to be a key feature, causing odd electrical phenomena, and compasses to go wild. Room temperature superconductors have recently been discovered here on earth. (For the first time ever, the researchers had to raise the temperature, to 30°C IIRC, to find the critical temperature.) If and when these are perfected, it will become practical to produce such magnetic fields. But exactly how the drive would work I'm very hazy on.

   So as I write this newsletter the LED lights are 'in process' for when I find time to do the new circuit board, I know in principle how to do a working torque converter car transmission, and I think (as I have for 2 months) that the NiMn batteries will work well if I leave out the doubtless impure 'artist supply' graphite. It all boils down to finding time to do these things along with other projects (including aquaponics) and several aspects of arranging financial affairs -- partly to plan ahead for possible financial and economic collapse.

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

Where Unsold Cars Go To Die

   Someone sent me the link below to an incredible article. It shows photo after photo of acre after acre of brand new unsold cars, in various locations in various countries all over the world. There are hundreds of thousands of them, if not millions, just parked there and deteriorating. It also says there are 10 billion cars in the world - more than there are people.

Evidently, everywhere cars continue to be produced as if there were buyers, in order to hide the lack of demand from the public. This in turn hides the fact that economies worldwide are imploding: Relatively few people can afford a new car. As John F Kennedy said when I was young, "We all drink the same water, we all breathe the same air." And today all our financial and economic systems are, if not one, at least intricately linked. No land is exempt from what's happening... and filling with unwanted cars.

   I like to think that there are also people waiting for electric or plug-in hybrid cars to improve and come down in price, who are putting off purchasing a new gas car just as they're (hopefully) becoming obsolete. I wonder how these vehicles would be selling if they were electric, with a similar or lower price tag than gas? (Ideally with efficient variable torque converter transmissions, NiMn batteries, and some sort of 'free energy' charging system.)
   I anticipate it's likely that soon, if there's fuel to be had, it will be pretty much unaffordable. The few electric cars on the road now will become priceless.

Where the World's Unsold Cars Go To Die | Zero Hedge

Oh my God, the news Gets Worse!
Anyone for Thermonuclear War and Extinction of All Life?

   The ongoing financial, economic and environmental turmoil which must soon lead to a colossal crash is bad enough, but at the start of June it developed that the Hypocrite in Chief and the military-financial complex in Washington are talking about a first strike nuclear attack on Russia and maybe China. Why? Simply because they don't want any power to rise to challenge American hegemony over the world. They believe like Nazis that they are "over-men" and all others are "under-men", and they care not who or how many they kill to maintain the hegemony and the illusion of "greatness" - a quality which has been under continual attack for a over century and which they have now completely eradicated from their institutions of national life. Hopefully nuclear war is only talk and there is in fact some sanity to keep such ludicrous ideas in check. But on the present administration, no bounds of reason can be set.

   On Greg Hunter's UsaWatchdog.com (can be found on youtube), Dr. Paul Craig Roberts, a long connected and well informed figure, says the hubris of the insane people in Washington is so great they actually believe they can win it. Talk around governing circles is "What good are nuclear weapons if you can't use them?" They're building up an ABM (Anti-ICBM) "missile shield" in eastern Europe. They are surrounding Russia and China with these and other "military assets" in various countries and bases. And the situation is all the more perilous because, being widely known, it may tempt the target countries to launch their own nuclear first strike before they are themselves hit.

   But little has changed in the figures of "Mutual Assured Destruction" ("MAD") since the 1960s. Roberts reminds us of studies including a recent one that show that if even 1% of Russia's and USA's nuclear arsenals are used, the death toll will be in the billions, not millions. And if either side lets off less than half its 'nukes', there will probably be no more higher life on this planet. Heavy strikes anywhere would unleash deadly radioactivity to drift around the globe, and also cause a nuclear winter, with daily temperatures in temperate zones below freezing for about three years: there would be no food for anyone and most or all food animals will also become extinct. He also says the "missile shield" probably won't stop 5% of the return fire, but that that's irrelevant for the above reasons. Everyone on all sides will lose everything including their lives, and those responsible will be found blameworthy before the judges of the Universe.

   Here's a notice to anyone and everyone in control of our affairs everywhere that may chance to read these words: You have no more right to lie, steal, defraud, cheat, extort, intimidate and murder than anyone else. And especially:


Electric Hubcap Motor Systems - Electric Transport

Torque Converter: the Whole Transmission?

Centrifugal Torque Converter - or is it just a Centrifugal Clutch?

   Let me start this by saying that while my intent has been to create a centrifugal torque converter, I've never got more torque out than the motor itself had. At the start of June, I tentatively concluded that what I may have made instead is just a new (... and hopefully improved?) design of centrifugal clutch. Unless I have some further inspiration, I'll go back to the planetary gear torque converter (PGTC) that is known to propel the car, since it did so in September 2012 -- but with this unit as the clutch, the previously missing component to the PGTC transmission. That would provide all the necessary pieces for a complete transmission unit that is known in theory to work -- once I redesign the housing, shafts & bearings et al again to fit everything in.
   So the drive train would be: the motor driving one of the planetary's gears, the slipping gear of the planetary attached to the tension rope & lever, the planetary's output gear driving the centrifugal clutch input disk rotor, the shoes hitting the clutch drum and driving it around (with high torque from planetary gear converter), the chain from that shaft driving the differential (car's original parts from here), and the differential to the CV drive shafts driving both front wheels. I'm not sure which element of the planetary gear will go where. The centrifugal clutch might work best with a higher RPM than the motor itself. Or, the slipping gear might somehow be combined with the centrifugal clutch if I use the right combination, eliminating the tensioning rope, manual lever and slip pulley.

   I pecked away lethargically on the half done torque converter. On the 5th I drilled holes for pivot pins for 4 pairs of shoes in the input rotor and cut and mounted the first set.
   I decided to put a pin between the motor shaft and the output shaft where they butted up against each other to keep them turning in line. On the 11th I drilled a hole in the end of drum shaft. The next day I bought a 3/8 x 1/2 x 1" bronze bushing. On the 13th I threaded the shaft hole for a 3/8" bolt, screwed a bolt in and cut off its top, leaving a smooth 3/8" concentric pin sticking out.
   Then I drilled a 1/2" hole in one end of the motor shaft and inserted the bushing. Theoretically the hole, drilled on the lathe, should have lined up on center, but I could have done it better freehand. I took the shaft to AGO and had it - the other end - done there. AGO not only had a big enough lathe to simplify things, but something I didn't know about called a 'center drill' (sets of them)... and more know-how. "Of course your drill wandered off, because you didn't square off the end of the shaft in the lathe first." That seemed obvious once stated, but I don't remember it from my electronics courses at BCIT.

   Then I made 6 more plastic (UHMW-PE) pivoting shoes and mounted them: 4 sets on 4 threaded rods threaded through the input rotor in pairs, one on each side of the input disk rotor in each position. I used no springs.
   Finally I mounted the motor to fit it, took it off and mounted the rotor disk on the shaft, and finally re-mounted it complete.

   On the 21st I tried it out on the bench. In the forward direction there wasn't much torque at all. Going backward there was considerably more but it didn't seem like enough. (Actually I don't understand why it didn't just jam going backwards. Later it did occasionally.) I had put the coupling pin in an existing shaft I found for the drum that was just right... except not long enough to grind a hex shape on one end for attaching a torque wrench. Once again I had no means of measuring the torque on the bench.
   Then I turned the motor up quite fast. It didn't seem to be working very hard. Evidently it wasn't being loaded down enough. (And in the other direction, even less.)
   There were 4 pairs of shoes, and room to mount 8 pairs, which should presumably double the torque and motor loading. In addition I could add weights to the shoes, which should make them push harder on the drum. The results seemed promising enough to make more shoes and plan on that basis - and if the car would only move one direction, that would be good enough for now. But the first step would be to make a new drum/output shaft that could hold the torque wrench, and measure the present levels of pressure.

Setup, with torque wrench to try to divine the torque figures

   I had that finished on the 24th, again with a bit of help from AGO to drill the center hole. It turned out the torque was only about 5 foot-pounds - about the same as the stalled motor. (This low figure for stall was because I was running it off 18 volts for the tests instead of 36.) In the other direction the needle barely moved, maybe 2 foot-pounds.
   Obviously the setup had problems. Doubling the number of shoes to 8 and upping the voltage would still only produce maybe 12 to 15 foot-pounds. With the 4 to 1 reduction following, that might move the car on level pavement, but it certainly wouldn't put it on the road. But I was pretty sure the approach, a disk rotor on the motor turning within a drum on the output side, with slippery plastic bumping into things at specific points, had potential. The size of the slots - or should they be protrusions? - and the makeup of the plastic 'shoe' parts in between, remained wild variables.

   I decided to try weights on the shoes. The light plastic pieces would have much more impact with a hefty bolt bolted through each one, oriented for the weight to centrifugally push the plastic outward. While it was apart, I would smooth off the edges of the slits in the drum to a 45° angle. (As usual the raw aluminum edges were cutting the plastic.)
   The four 1/4" threaded rods seemed too light for heavy 'boots', so I cut holes for three 5/16" axles. I thus reduced the number of shoes from four to three pairs, through which I put six 3/8" x 2.5" weight bolts. They now had a promising heft to them that seemed much more in keeping with a vehicle transmission part. I finished two on the 27th, then had two hectic days and got nothing more done.

   On the 30th I dug out an old 'right angle blade' scraping tool I've had for ages and scratched 45° bevels into tops of the slots to smooth the edges. This gave them a "Y" appearance. They were then much easier on the plastic. Probably a 45° angle "V" would be the ideal slot shape.
   As I did it it occurred to me that where I was worried the shoes would jam in one direction, I could simply shallow out the slope on that side, ideally a shallow italic sort of "V" shape, worked down until the shoes wouldn't catch on the shallowed slopes in that direction. Then I made the 3rd set of shoes.
   The next day, the 31st, I fired it up. In the weak direction it now had about 5 foot-pounds. In the stronger it started getting more. It was hard to read the dial on the torque wrench. It vibrated back and forth, since the force was on-off pulsing - as intended. It seemed to vibrate from about -20 to +30 foot-pounds. I had the impression of double force or better - 10 or more foot-pounds, and the motor was turning slower to get it. It was better loaded down and didn't readily speed up. This time, with 24 pulses per revolution instead of 5, the pulses were too rapid for the motor to change speed much, so a flywheel would seem to be unnecessary. 10 foot-pounds still wasn't enough to get the car moving. 15 or 20 probably would be on smooth, level ground, but it still wouldn't put it on the road. 100 would be ideal. The peak torque at the moment of impact of the shoes with the grooves was probably higher than the average. Would the car see the peak force or the average? - was it worth trying out on the car? I decided to try a couple more things on the bench first.
   The first thing to try out was higher voltage. Half voltage is only 1/4 power, and consequently the RPM wasn't getting all that high. I had a 24 volt NiMH set up and I put it on to charge, then connected it. (I've blown up too many motor controllers at 36 volts. I need to improve them.) After trying it out and still watching the needle bounce all over, I decided to mount it on the car. Not surprisingly, the car didn't move.

   The next idea was more weight. I only had to remove the motor to modify the shoes - the transmission box stayed in the car. The weight bolts were fairly long, and I got the idea to put coupling nuts on them. These extended past the ends of the bolts and probably doubled the weight or better. I remounted the motor. It didn't seem to work very well and I had the impression the battery was low. I put on a 50 amp shunt and a plug to plug in the voltage and current monitor. Sure enough, the battery was very low, and I put a couple of 12V chargers on the two 12V sides. But with the meter I could now see what was happening. 1/2 hour later I went out and tried again, but the power was still dopping off. According to the meter a lot more current was being used than I had expected - up to about 80 amps, which dragged the battery voltage way down. But the 50 amp breaker didn't blow, neither did the two 30 amp (total 60 amps) fuses.

Testing in the Sprint car

   I had to conclude in multiple runnings that the car was seeing the average torque, which was of course no higher than the motor's torque. The 'torque pulses' were too brief and too small to transfer through all the slack in all the linkages and the rubber tires and start the car moving bit by bit, more with each pulse. At this point, I've decided to again examine the planetary gear type of converter, but combining this centrifugal clutch component with it as the required clutch.

Electro-magnetic Clutch Torque Converter?: a new concept

   On May 2nd someone thought I should look at a magnetic clutch - a broken one from a lawn tractor was being replaced. We picked up the old unit and I wondered how it might apply. Obviously a working clutch is what the planetary gear converter was lacking, but I'm presently trying to make the centrifugal converter, which doesn't need a clutch. On the other hand, the centrifugal converter is similar to a centrifugal clutch.

Electric clutch opened. The stationary electromagnet (left side) is energized
to spring the two clutch pieces (right) together and connect the input to the output.

   The electromagnet contrives to pull the two rotating elements of the clutch together without itself having to spin, which greatly simplifies the wiring. The input has fittings for a one inch shaft such as those I'm using. However, the output has a V-belt pulley and no means for attaching a chain drive or engaging a shaft. The whole unit is also probably too small for a vehicle. But there may be larger units more suitably configured.
   The output has a springy mounting and is pulled slightly sideways by the electromagnet to engage with the input. When not engaged the input element rotates freely, but the output is pulled to the other side, into contact with two non-rotating permanent magnets that act as a brake. (It's a clutch for the cutting blades, so they should stop rapidly when not wanted. For a vehicle motion clutch that feature would be removed, or selectively engaged as a "Park" drive lock.)

   On the 4th it struck me that I had conceived various ideas for things that would start a car moving, but wouldn't work because they would "wind up" or engage, and there was no way to "unwind" or disengage the system to create a repeating cycle, or else it would randomly stop in the 'torque stroke' position, from which the car couldn't start moving. But a magnetic clutch could engage and disengage rapidly and repeatedly, resolving the impasse.
   In fact, it can do much the same thing I'm trying to accomplish with the centrifugal converter. Used as a pulsing element, within mechanical limits the electric clutch can engage for any desired length of time via its pulse width, as often as desired via its pulse frequency. These times would probably be in terms of fractions of a second, say 1/20th to 1/2. These two factors, pulse width and pulse frequency, could be controlled more or less manually at least for testing, or automatically by a microcontroller monitoring vehicle speed, motor speed, and probably motor current. And there are other parameters that could be adjusted, like putting in slots and patterns like on the centrifugal converter, or using one plastic element to smooth things and reduce noise. It might or might not need a flywheel on the motor, depending on how the configuration and pulses work out.
   It soon also struck me that this smaller clutch with the V-belt pulley, if repaired and used this way, looks potentially almost ideal as a torque converter for the motorcycle. Perhaps I'll continue making the centrifugal converter for the car, and try this hopefully pretty simple idea for the bike. No shortage of projects!

   After finding the centrifugal converter as made didn't seem to be more than a clutch, my enthusiasm for this approach has waned some. It might still be worth trying.

Mushroom Outboard Motor

   It occurred to me one might make an electric outboard leg out of two pieces of UHMW polyethylene, drilling and milling out the spaces for the shafts, bearings and the double U-joint where the direction changes for the prop. It seemed neat and tidy until I thought about how to shape the outside to a streamlined teardrop profile.
   Then I thought of carving a leg out of wood, making a mold of it, and forming the leg out of polypropylene-epoxy composite. Light, strong... But I have no plans for actually going ahead with this idea at this time. No shortage of projects!

Boat Speed Indicator?

   On the 25th, thinking about going fishing, I suddenly had a new idea for a boat speed indicator for slow boat speeds. Instead of a pitot tube, this one uses a weight on a pivoting stick. It would be less susceptible to error from the depth of the transom changing in the water with shifting weight or waves. But perhaps somewhat susceptible to error due to the angle of the boat changing (bow to stern). A disadvantage would be that the calibration would be individual depending on the weight, length and cross section, and would have to be determined, from some other speed indication. For somewhat higher speeds, a heavier weight and a smaller cross section would be used, but the whole design wouldn't be good for the sort of high speeds where the typical commercial pitot sort of speedometer is used.
   With the pitot tube and vertical plastic speed viewing tube, the amount the water will rise (above sea level) for any given speed is a known constant.

EV Individual Battery Monitor

   On the 10th I drove the EV Mazda a fair distance, and the voltages seemed to me to be suspiciously low after too short a distance. Again I couldn't tell if it was all the batteries getting a little low together or one problem battery getting very low. I measured them after I stopped, but nothing seemed amiss. Then, having left it unplugged for the tests, I forgot to plug the car in and it sat uncharged all night.
   Again a monitor showing the state of each battery would have eliminated the need for the time consuming checks afterwards and would have immediately disclosed any problem during the drive. The next morning I got enthused and decided it was high time to make it.

   First I itemized the chief challenges:

1. There aren't 12+ analog inputs on most microcontrollers, and I didn't want to do some .5mm spaced chip with dozens of pins. I decided to make the monitor do only 6, 7 or 8 batteries - however many pins were available. One can install two units to measure 12 batteries, or else measure 24 volt steps instead of 12, which should at least narrow down problems to one of two batteries.
2. The voltages are high and floating. Best to insulate everything well and have nothing metal the user can touch.
3. Each battery is at a different voltage than any other compared to the circuit. This means different levels of attenuation required and different voltage ADC rez for each one.
4. The little color display (128x128 pixels) would be a challenge both to wire and to program. The leed spacing, 12 fingers in less than a centimeter, will be a real challenge to my coarse toner transfer circuit board techniques. One would expect there to be a connector made to plug it into, but I haven't found it.
5. Control pushbuttons. I decided these would be the capacitive type - just little interleaved fingers of circuit board that are touch sensitive under software control.

   Just writing these out caused much of my zeal to evaporate. Nothing was especially hard (except wiring up the display), but it adds up to a considerable project. As I started in on the PCB design and it started getting more complex than I'd hoped, it became apparent that it was going to take considerably more time than I was bargaining for... and I quit. On the 13th I decided to just buy a bunch of those tiny, cheap 3-30 volt 3 digit LED voltage displays, mount 12 of them in one panel for the dash, and hook one to each battery. The colored bar graphs would certainly be superior, but the voltmeters should fulfill the essential requirement, at the cost of distracting the driver substantially more to read all those numbers than to glance at bars and make sure they're all green and even. (Really I barely have time to glance at the overall volts, amps and amp-hours now and then.)

   Anyway, I've defined the color display for the circuit board layout program, which I also want for the fridge, heat pumping and solar controls, and I got a start on the circuit board layout if I change my mind. It would be an excellent product. I may work on it occasionally. Oh, to have people I could farm out some of the work to!

The little three digit LED voltmeter displays arrived on June 2nd. I decided to mount just 4 or 5 in a small plastic panel with CAT plug pins on the back to plug them into batteries, and connect them to whatever batteries seem most vulnerable or suspect at the time, and maybe use the 24 volt idea to cover a couple more than the number of displays. Of course, that won't show unexpected problems in the other batteries, like unnoticed capacity deterioration, or unplugged or non-functional chargers.
   Hopefully, by monitoring the (presumed) weaker batteries, I'll be able to extend the driving range of the RX7 (also see next - lithiums in RX7) by knowing whether it's time to quit as I normally do now just in case, or whether there's some more safe life left without possibly damaging one. More on this hopefully small project, I expect, next issue.

Lithium-Ion Cells in Mazda RX7 EV

   On the 28th, having used an 18 volt bank of lithiums in torque converter tests, and then having received a voltage-current dual panel meter, I decided to try a bank of them in the RX7, where it looked like one in its tub would just fit where a battery was missing. This upped the nominal voltage from 120 to 138, the highest yet, with the cells: Li-ion 18V, NiMH 36V, PbPb (size 27) 60V, PbPb (size 24) 24 volts. Most of these are about 100 amp-hours rated, except the size 24s are maybe 75-85 with lower current drive, and are surely now the limiting factor in the car's range, but larger ones won't fit where they are. If it had 144 volts of size 27s (in good condition) it should theoretically have 35-45 Km range according to Canadian Electric Vehicles, which is much more than I've actually had to this point. Going up to 138 volts should help.
   First I (at last) put together the charger. It wasn't a good afternoon to try it out, because I had to drive about 3 miles to an evening appointment. I drove around the block. Performance was great, but I found that the lithiums had dropped in that short distance from 18.7 volts to 16.7. They aren't supposed to be driven below 3.2 volts per cell, or 16 volts. They should have had at least as much reserve capacity as any other battery in the car. Why they should be down so far so fast was puzzling, and bode ill for a longer trip. I decided to leave them installed and I checked the overall voltage on the meter frequently. It's hard to tell what 18 volts of 138 volts is doing, but it seemed good enough all the way that I didn't stop. At my destination (where I charged it almost 2 hours), and then at home again, I measured the voltage on arrival and found it was about 16.4 volts. Evidently it drops to around that level rather soon and then drops little for quite a while. After a trip or two, all 5 lithium cells read exactly the same voltage (3.31v).
   With the higher voltage (and warm weather), the car uses less amp-hours per mile - typically 1.8 to 2.3 instead of 2.2-2.5. It's as many watt-hours, but the additional voltage/cells will take it farther before the batteries are depleted.
   On the 31st I dared a trip to Hillside mall, the farthest destination yet with nowhere to plug in while there. On the return trip the voltages were still high enough that I took a detour to another store, and the whole trip was the farthest between charges yet, 7.8 miles (12.5 Km). On June 4th I made almost the same trip, but took a detour to specifically extend it, to 9.0 miles (14.5Km). The lowest batteries (the two size 24 PbPb s and one of the 90 AH NiMH s) were getting down, and without an individual battery monitor it's dicey to test the limits too far - you can't be certain when the weakest one(s) starts being driven too low.

  It should also be pointed out that Vancouver Island is pretty hilly and Victoria city traffic is frequent stop and go. With flat ground, or fewer stops and starts, or both, the car would consume substantially less energy.

New Concept: Lightweight Air-Aluminum Batteries for when regular cells are used up

   Alcoa (aluminum company, of course!) has a new plan for adding air-aluminum cells to EV s. Aluminum has an enticing amount of energy as a negatrode, a light atom whose alkaline reaction; Al + 3 OH-  =>  Al(OH)3 + 3 e-; moves 3 electrons, at -2.33 volts. (Compare with my 'super energy' rechargeable manganese reaction; Mn + 2 OH-  =>  Mn(OH)2 + 2 e-; just 2 electrons at -1.5 volts.) I don't know how they prevent the reaction from occurring spontaneously at such a high voltage, but one assumes they have worked out some means.
   The idea of non-rechargeable cells in an EV has held little appeal to me, but Alcoa's plan seems very good: Rather than having these lower cost but non-rechargeable cells used for regular driving, they would be the "spare tank" reserve, taking the car as much as 1200 Km farther than the regular rechargeable batteries will take it. When the aluminum is running low the cells are replaced (or the aluminum in them is), but that wouldn't happen except after some accumulated mileage of long trips - trips that would otherwise have been impractical or at least less practical in the EV. It replaces the need for a gas engine (plug-in hybrid), stopping too often to charge, or having to make the trip in a gas burning car instead.

Electric Caik Motor: Ready-made Rotors!

   At Princess Auto I found a '7.8" brake rotor disk' fitted for a 1" shaft for 35$. It's machined nicely and it fits nicely in the Electric Caik motor housing, and the center takes less space than the "H" bushing type. In fact, I think I could make the Caik motors 1/2 to 3/4" thinner, under 4". (This should also be true for the smaller "AJ" size bushings that I recently ran across. Why do the stores keep these things in the back so you don't find out they exist?) The textured surface should give the epoxy a better grip than smooth steel to hold the magnets on. The price contrasts well with an "H" taper lock shaft bushing for 10-15$ plus having a 7.5" steel rotor cut by abrasive waterjet for 35-40$.
   One possibly negative attribute is that the disk plate is thinner, about 3/16" instead of 1/4". But I don't really know what thickness is required to complete the magnetic circuits of the supermagnets nicely anyway, and just picked 1/4" owing to some gut sense of proportionality. I doubt any difference will be noticable. Also for some reason, there's a key slot, but no set screw hole(s) at the shaft to lock it in place. That can be remedied.

Other Green Electric Equipment Projects

Peltier Module Heat Pumping/Refrigeration

   I really hoped to bring this project to some sort of conclusion for a while early in May, with some new '24v' peltier modules I'd ordered in April. If I had to work on it, I wanted it to be on the still unmade solar control. It was not to be, and the fridge is barely staying cool.

   In the warming spring weather, the TEC12706 peltier module I had put in temporarily continued to barely keep the fridge cool to 7°c, making only a little ice in bottom of the water tray.
   On the 7th I got the "24 volt" peltier modules, numbered C2406. Unfortunately I could find no specs on these units - only the same text as at DX.com, repeated on several other web pages. It can't be a 24 volt unit at all, as it draws 6 amps at 12 volts. Most peltier modules say how many thermocouples are in them and their rating, eg, a TEC12708 has 127 thermocouples (in series) and is rated at 8 or 8.5 amps. If this one has only 24 thermocouples either they're a completely different material or it should be rated for about 2.8 volts. If on the other hand it has 240, that would make it a 28 volt unit, pretty much as advertised... but it doesn't behave that way. It acts as a 12 volt unit, despite the text.
   Installed in the fridge, it drew 5.25 amps (almost 70 watts at 13 volts) and soon had the heatsink up to 37°c, but seemed to be cooling about well as the previous unit. It became apparent I had merely replaced one 15v, 6a module with another. I tested all three units and they were the same. When I momentarily put 24 volts across one, it drew at least 12 amps.
   Furthermore, I then tried the three 4 volt units from the same order, the C0405 s, thinking to make a 15+4 volt double unit. They seemed to be 4 volts, but they only drew 2 amps instead of 5, which wouldn't pump enough heat or match any other modules I have. If one pumped 5 amps through them, both sides got warmer, tho unevenly.
   This is the first time I've got goods from dx.com that aren't as advertised, but needless to say I'm not very happy with it.
   So I was left trying different things to get decent cooling, when I wanted to be doing other things entirely. I tried mixing a lower voltage TEG module with a TEC12708. But it only dropped it 2.5 volts, and it was thicker, which made it very difficult to get good thermal contact with the other, main, module. There wasn't room on the heatsink for 3, 4 or 5 TEG modules to get to a good 12 volt working voltage.

   I found another source of peltier modules at aliexpress.com, which seems to be an alliance of east Asian companies - apparently even factories - who ship small parts directly through the post office, with a central payment system. Searching hundreds of modules shows only a few that aren't 127 thermocouples for 15 volt ratings. Then, the promising looking ones were all sold in lots of 10 for over 100$, so one can't try out a couple of each of 2, 3 or 4 types and find the most satisfactory - if it exists - without spending a lot of money. I couldn't find any like the CP85338-7108 - 8 volt one I used in the 15+8 volt dual module system that worked so well until the 8 volt module got cracked. (71 thermocouples, I think.) All the "cheap" peltiers at Digikey have gone up from 17$ to 22, when one can order similar from China for around 5$ - and the Cui ones at Digikey are probably made in China anyway.
   The frustrating ubiquity of the 127 element/15v peltier modules was such that I started thinking that using them at about 7-10 volts via a DC to DC converter may be the only way this is going to work well, and that the converter losses would have to be accepted. The converter could be simplified, and it's output voltage made programmable, as a programmable part of the microcontroller based fridge control circuit.
   I ordered a very large 62x62mm, 15 amp module. (20$ each, package of two) That should pump plenty of heat even at half voltage. (Unfortunately I can't fit two to put them in series - unless I put one on the underside of the copper bar and use two heatsinks and two fans.) And for solar use with variable voltage, with extra cooling power one might adopt a strategy of aggressive cooling to freeze all the ice during the day when the sun is providing lots of power (COP be dammed!) and much less if the supply voltage is down a bit and (hopefully) none at night.

Flat Panel LED Light

   I continued looking for suitable supplies for this project -- and found them. Supplies seem to often be much the biggest issue as soon as you're looking for something not commonly used by many other people and you don't have a budget to order from some pricey specialty supplier without giving it serious consideration first. Consider that I was paying 10$ plus for LED emitters for the globe lights, with two emitters per light. Things became much more economical in principle when I ran across smaller emitters for 20¢, even though 9-15 would be required per light.
   This time around I found "Aliexpress.com", an umbrella sales and marketing agency for many Chinese/Asian export dealers. White LED emitters similar to those for 20¢ at dx.com were to be found at outlets here as low as 10¢ - and with better brightness specs, according to the numbers given.

   There were also several types of emitters for plant growing lights. They cost a lot more - 30¢ to over a dollar - but were still affordable since people will pay extra for plant growing lights. Of course, my lights would be pretty low output for plant growth, but the larger ones appear to be quite pricey, so there may be a market for small, lower cost ones. (Seedlings? Shade-loving plants?) It seems Earth plant photosynthesis largely uses two very specific wavebands of light: near ultra-violet to blue in the band 400-450 or (other sources) 410-470 nanometers (foliage), and red to deep red in the band 640-675nm (promotes flowering). Thus it seems green light is pretty much useless, and white including sunlight is good only insofar as it contains the blue and red. The balance between these two bands, it seems, varies with the time of day and with the season - which may be how plants 'know' when to flower, each at its 'preferred' time. (We tend to think the mechanisms of life are simple, and it's postulated they all came about by random chance rather than by design - here's just one more amazingly sophisticated one!) For basic growing, it's probably enough to have a mix of the two colors, and for just leaf growth, just blue.
   The blue emitters seemed to all be about 450nm or '445-455nm'. I'd have preferred something nearer the middle of the 400-450nm band, but I didn't see any. (After I'd ordered I found a couple of 430nm emitters.) There were 1 watt and 3 watt blue emitters... which seemed in my quick scan to have almost the same lumen figures... so I ordered some 1 watt 450nm s. In the red were 'high efficiency' 630nm, and 660nm. I chose to get some of the latter as that's at the center of the band for plants. 630 may perhaps be close enough, but the 660 s specifically said "plant growing".

   Then, the 1/8" translucent acrylic plastic was said to absorb 1/2 the light. But LED s need diffusion. Those tiny, intense emitters, doubtless harmful to vision, leave sharply defined spots in front of your eyes, as is typical of too many LED lights. I was given a sample of a thinner piece at Plexi-Klass locally, and it seemed to let more light through. Making a new case for the prototype with it, the light seemed substantially brighter. Maybe it blocks 20 or 30% of the light. One can see where the emitters are, yet they're diffused into 'blobs' of light, not the intense sharp spots. That's success! It's probably better transmission than my globe lights (for which I had never given the issue any thought), or at least as good. Lamp shades for incandescent bulbs absorb light too, some of them a high percentage.
   Getting an even glow from the whole light surface would be tricky with my design and these somewhat 'aimed' emitters. I don't see it as a real problem. The cure might be worse than the disease. Other than a thick diffuser blocking much of the light, one might put in a thin sheet of something to start the diffusion before the bottom cover - two or multiple layers spaced apart - but the fixture would probably have to be 2 or 3 or 4 inches thick. (Optimum distance for the single thin cover seems to be around an inch.)
   I disassembled a Chinese flat panel unit that had an even glow. It had the 60 emitters in strips of 30 along two opposite edges, shining inward into a 1/4" thick clear acrylic sheet, with a thin translucent sheet under it. Held up to a light, their translucent sheet looked just a little brighter than my sample and had similar diffusion. (Rats, where'd they get it? I suspect they had it made, optimized for lights.) The aluminum outer edges act as the heatsink for the emitters.

   Again no one uses such thin 'translucent' acrylic much - possibly it's good for almost no other purpose - and to buy any I had to order a whole 4' x 8' sheet. I was lucky to find a sample of acceptable material and get it locally, and it's doubtless much cheaper per square foot in whole sheets.

   There were also 15cm x 20cm single sided circuit boards at aliexpress.com, again for a low cost, and I ordered a pack of 10 as they were about the right size for the lights. Luckily I didn't order more, because later I thought of simply using thin sheets (.016"?) of copper in strips, with the LED emitters soldered between them. (When they eventually arrived, except for the top one, they also appeared to be cheap phenolic boards, where I'm pretty sure the site said epoxy.) It'll cost a bit more, but the first prototype ran too warm, and copper sheets will be better heatsinks for the emitters, absorbing heat better and radiating it well from both faces. The gaps between strips will improve airflow. And it solves the problems with etching the large boards with large copper areas. Just the end strip will be circuit board, for the control circuits and connectors.
   I'll also improve the cooling airflow, with a ventilation gap between the top and the ceiling, formed by the mounting tabs.

   There I left it on the 9th, awaiting parts, intending to get on with the torque converter. But I sent someone who was going past the rather distant store anyway, to get 2 thicknesses of copper to try out, .010" and .016". I shouldn't have asked for 3 square feet of each, as the bill came to 85$. But I got to try it out immediately. Next the new circuit board "strip" is to be designed and made. But I switched to working on the torque converter.

Electricity Storage

Turquoise Battery Project

Carbon Black powder... from old dry cells

   Having found "aliexpress.com" with a multitude of Chinese and East Asian businesses listing their products under an overall sales system umbrella, I checked for "conductive carbon black powder". A company had what appeared to be the perfect 1.5Kg bag of high purity powder for about 80$c. But when I went to buy it, the web page said "can't ship to the selected country" (Canada). They were the only company selling carbon black, and it seemed none of their products could be shipped to Canada. No reason was given, nor any alternative suggestion. Was this some strange delivery issue? Did the president of the company hate Canada? Or was it some stupid Canadian restriction passed by those who are elected to represent our interests, to solve some little problem someone once had, or even might potentially have, and now in company with thousands of other stupid restrictions arbitrarily impeding the ability of all Canadians to do anything but sit and watch TV all day? Or even to abet some corrupt corporate interest(s) in relieving us of our money?
   Unfortunately, Canadian government interference seems the most likely. Churchill said something to the effect that if 10,000 laws were enacted, people would lose their respect for the law. I suspect there are far more laws than that today, and I feel more and more that the bulk of them simply limit our freedom to do useful things, without having any proportionate - or even discernible - benefit to society. But I digress.
   I finally decided that the path of least resistance might be to get old dry cells and use the MnO2/carbon black (or graphite) mix in place of the suspect graphite and some of the KMnO4. It wouldn't be as concentrated as desirable and the whole un-optimum mix propotions can't be exactly known, but at least whatever carbon and manganese is used in them is pure, battery quality, and won't cause self discharge.

   So guessing the actual percentages, apparently the amount of carbon used is a little over 50% by volume, but carbon being un-dense, that's more like 30% by weight. If I take 40 grams and add other ingredients it should make some semblance of an appropriate mix:

28g - MnO2 (or Mn2O3 or MnOOH or Mn3O4 or Mn(OH)2 in various proportions depending on discharge state)
12g - carbon black (substantially less than I'd like)
18g - KMnO4
18g - Monel
18g - Ni(OH)2
6g   - Sm2O3

(Rats, I didn't get to it!)

Lithium Ion Battery Sets: 2 odd cells out of 11

"18v" - really about 16v in operation - lithium battery

   When I got the lithiums, I was a bit taken aback to hear that they had self discharge and had to be kept topped up. I thought that was a big argument against alkaline cells. Then, of the eleven 100 amp-hour lithium cells, which I got for the boat & Electric Caik outboard, one seemed to discharge itself much more quickly than the others, and one wouldn't charge up to 3.8 volts, refusing to rise above 3.34 and soon dropping back to 3.31 after charging for any length of time. The 'lossy' one had to be charged to full separately after the other four in the 18v set had finished.
   This probably shows poor quality control, although the 'almost new' cells might possibly have been subjected unknowingly to overdischarge or overcharge at some point. It also shows that the individual battery monitor could be put to advantageous use even with supposedly identical and ideal cells - that anything was amiss would have been obvious instead of undetected.
   Whatever the cause, I set the low voltage one aside. I had thought to use that last cell as a token lithium cell in the Mazda just to be trying out three distinct types of batteries in it: PbPb, NiMH and Li-ion. That's now out. But I installed the best of the two sets of 5 temporarily as an 18 volt battery (really about 16.5v), making 120+18=138 nominal volts. That can be read above in the Electric Transport section. It improved range and performance, making the car more useful. After the second or third trip all 5 cells read exactly the same voltage, 3.31v.

Victoria BC