Turquoise Energy Ltd. News #95
  covering December 2015 (posted  January 3rd 2016)
Victoria BC
by Craig Carmichael


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

Happy New Year, Everyone!

A Highlight:
- Self-turning Magnet Motors are Powered by VHE/Lambda Rays?
(see Month in Brief, Electricity Generation)

Month In Brief (Project Summaries)
- Unipolar Motor Controller - VHE Lambda Ray Converter - and - Magnet motors appear to be lambda ray converters! - Chevy Sprint & variable transmission - Miles ZX40 Electric Truck - Mazda RX7 EV & battery longevity notes - LED/Aquaponics & Space Lighting Note & Peltier Fridge

In Passing (Miscellaneous topics, editorial comments & opinionated rants)
 - Lifting the Veil: The Pathocracy (book) - Bail-Ins Coming? - Techniques for Suppression of Clean Energy - Peace and Forgiveness - Pluto images

- In Depth Project Reports -

Electric Transport - Electric Hubcap Motor Systems

* Unipolar Motor Controller: pulse speed control input to improve performance.
* Hubcap Motors: Rotor Magnet Attachment

Other "Green" Electric Equipment Projects (no reports)

Electricity Generation
* VHE/Lambda rays - ideas
* "self turning" magnet motors are apparently powered by lambda rays!
* microcontroller triple timer for the converter

Electricity Storage - Turquoise Battery Project (NiMn, NiNi), etc. (no reports)

No Project Reports on: CNC gardening/farming machine, Electric Weel, reluctance motors, battery making, aquaponics.



December in Brief

   As I think back over 8 years of renewable energy developments, I can't take a whole lot of satisfaction in my successes. There are components and aspects to it all that have been very successful, but I still haven't put an 'ultra-efficient' EV on the road, and I haven't so far made any really satisfactory electrical generator or a practical battery. Some things, as far as development goes, seem to be on the brink of happening, but they've often seemed that way before, then deficiencies in the ideas, designs, or my physical constructions show up, and they haven't yet. A major problem has been that as I develop one thing, I discover something better. It seems unsatisfying to work on the old thing once I see a better way, so I start on the new thing instead of perfecting the old. One might also say I divert too readily into new projects which are only marginally related to what I've been doing. Sometimes I expect they'll be a short, minor diversion, but they tend to grow and metamorphose.

   The new battery chemistries, NiMn and NiNi, seem to almost work, but have never quite performed in a manner such as to be practical. I now have a couple of things I want to try next on them, but I haven't had time to look at the project in months.
   My BLDC motors are great in theory and highly efficient but so far they've broken down somewhere or other under stress. When they do I improve the design of whatever piece broke, and they are now much stronger. But presently the magnets still fly off the rotors from over-revving. Theoretically I have an improved design for attaching them, but I haven't made a proper new rotor in quite a long time. The new unipolar types - the unipolar controller, BLDC4-3 and reluctance motors, promise great results but still need more work, and now I can't say I'm proceeding on them at any great speed. And of course if the reluctance type is better, why finish perfecting the BLDC ones?, which I might otherwise have done by now.
   The variable torque converter has been a bugaboo all along. It took me years to find a workable design, and there were always motor and controller problems to cope with while trying to work on it. And now that I have a theoretically practical design with the slipping planetary gear (proven to work in Sptember 2012) and large centrifugal clutch (can hardly fail... if it's adequately robust and the relatively low RPM is high enough to lock it), and again, now why worry about the torque converter if the high-RPM capable reluctance motor shouldn't even need one?
   The off-wheel add-on car wheel drive setup (with belt or chain drive to the wheel - TE News #91 & #86) looks like a real winner, solving the problems of ruining vehicle handling with extra unsprung "wheel motor" weight, but I can't make it without a truly working pancake-shape reluctance motor - or perhaps a BLDC motor with torque converter.
   And now having gained a practical understanding of the whole process, especially as reported in TE News #94, the VHE/lambda ray energy converter seems like such an important development as to be worth pushing everything else aside for. Imagine being able to generate the electricity one needs 'on the spot'. Then imagine also using such energy to desalinate seawater and turn vast deserts into gardens!

   As the years go on and various things come up, I find myself somehow with less and less time, and less energy, that I can devote to all these projects. I can't help but think that things would go much faster, and projects would get completed, if I had a helper or two or three. People who I could give the ideas and designs to who could then turn them into pieces for assembly into products. Maybe younger people with more energy. I envy Edison for having had such resources, a rare state for an inventor.
   So far somehow it hasn't seemed to work out. I'm certainly not much of a manager and I seem to have trouble delegating work. I may think of doing so initially, but once a design is in my head it usually seems easier to do it than to explain it to someone else. And for CNC parts, doing the design up so it can be made is usually the only step. The machine makes the part.

Unipolar Motor Controller

   The new (purchased) energy return choke coil now ran cool, but the energy return/flyback diodes were quickly getting hot. I hooked up the oscilloscope to various points to check voltage patterns and their relationships with other voltage patterns, and tried a few different connections like double choke coil and no choke coil. Then I thought: the diodes conduct all the flyback current spikes regardless of whether the power is going back to the supply or merely being shorted across the motor coils. It should be all the same to them. Eventually I came to the conclusion that things were actually working as they should, that the diode heat was probably what it should be with all the high current spikes. They just needed a good heatsink, and preferably active rectification to reduce the losses. Perhaps I had a conceptually better unipolar motor controller with half the active components after all!

   I could have it wrong, but an odd "extra voltage" spike right at turn-on looked to me like the full-current pulsing drive was imparting some "DC kick" to the motor coils. From Bedini and then other more commercial motor makers comes evidence of getting something extra out of motors from pulsed drives. It's explained as being 'the coil's magnetic field energy is being returned'... but this energy return has always been the norm. The 'extra' energy may some "DC kick" VHE/lambda ray radiant energy being imparted with each pulse.
   Later on, with the ground clip in a different place, there was no sign of the extra voltage pulse. There was a slight drop there instead. And yet the pulse really wasn't explainable by where the ground point had been. Main conclusion: looking at and interpreting and meaning of transient waveforms is complicated and it's easy to jump to unwarranted conclusions. I didn't know if the pulse was real or not. And I started looking at how to get more power - increasing peak current and shortening the OFF times.

   I decided to give it a rest, do some software development, and wire up the lambda ray collector. Later with Christmas concerts and various festivities, I didn't get back to it.
   But I did start thinking about beefing up the rotors even more - with a solid cover shell of tough plastic that fits over the magnets and bolts onto the steel rotor with flat-head machine screws, clamping each magnet inside a rectangular pocket.

VHE Lambda Ray Converter - and - Magnet motors appear to be lambda ray converters

   On the evening of the 6th I started in on code to read the serial pushbuttons in the display-controller. I got it going after a while, after some seemingly inexplicable results that really had simple causes. (like serial shifting after reading the first bit instead of before). Over the next few days, I found the multiply, divide and analog value reading programs I'd done in 2013 and tidied up the code some, along with reading and learning how to use some complex features of "timer A" that looked just right for timing the resonant pulses. But I didn't get things beyond that point.

   Hold onto your hat: It would seem that self-turning magnet motors - some types if not all of them - get "bursts" of magnetic energy from VHE/lambda rays. A sudden change in magnetic field as one magnet flies by another generates a burst of electricity, which converts lambda rays into another burst of electricity, which makes a burst of extra magnetism. That's their power.
   They are thus a special form of lambda ray converter, converting the rays into magnetic force leading to mechanical rotation rather than directly into electricity. From the amount of energy they seem to make available, they don't seem like terribly effective converters, but their power would seem to come from the atomic interactions between magnetism, electron motions and lambda rays. I've written more on this idea, along with several more ideas derived from the web about lambda rays, in the detailed project report.
    Since the electrical voltage switching types of converters seem to be much more effective, I'm ending any thought of trying to further develop magnet motors myself. (Here at least is one small simplification of goals!)

Chevy Sprint & variable transmission

   This project has been repeatedly pushed aside by competing attractions that are also taking longer than hoped. When I got a 300 amp, 36 v Kelly Brushless DC Motor Controller, I thought everything would work but the slightest touch of the control over-revved the motor and the magnets broke off the rotor - all 12 at once. Then I wanted to put it back together with the new magnet strapping system to be good for higher revs, but I had just come up with the unipolar idea and wanted to orient them NSNSNSNSNSNS for the unipolar controller, instead of NNSSNNSSNNSS for the regular bipolar one. Why put it back together the same way if I was then going to change it? Then I found out about reluctance motors, came up with what seemed like a good design for one and was also given the basis for a second good design, and wondered why I would be using permanent magnet motors at all. Why not just go for the new type? But the motor controller energy return system needed work, and so far the whole thing seems to have the power of about 3 hamsters. As above I was doing some testing this month partly to see what needed improvement to get more power.

   On the 18th I went to some little used shelves in the shop and realized... Wait a minute! There's a second Electric Hubcap motor rotor sitting on the shelf! To run the motor in the Sprint all I had to do was reassemble the motor with this spare rotor. It's still the old style magnet strapping, so would be just as prone to over-revving as the first one, but I could connect up the transmission linkages so the motor has a load and won't over-rev without some serious power going into it... and run it at 24 volts instead of 36. That would also help limit the RPM to a lower value.

   That would make it easy to get going on the project again. I'd be leery of taking it on the street without beefing up the rotor and using 36+ volts, but it would at least test out the planetary gear torque converter with the centrifugal clutch transmission, hopefully to drive it on the lawn and pinpoint any weaknesses or flaws in the system. After all, the theory of the transmission is good: The planetary gear torque converter part drove the car along and upslope on the lawn with only about 1 KW of power in 2012. It's my attempt at implementation I have my doubts about. Especially the centrifugal clutch seems inadequate and just might fail somehow under load.

   At this point, with so much else to do, I'll probably await nice spring weather before continuing.

Miles ZX40 Electric Truck (and MAZDA RX7 EV FOR SALE?)

   I had been thinking that since I hadn't got the Sprint on the road, and several factors have kept me from going ahead, and since the Mazda RX7 EV had such limited range, I perhaps ought to buy a decent electric vehicle to replace both it and the gas Tercel - if necessary renting a car for my very occasional longer highway trips.
   I now have pretty much all 100 AH batteries in the RX7, but what it really seems to need is something more like 150 AH or better, in order to supply sufficient current with less voltage drop and stress as well as to have more actual storage. I figure that might nearly double the driving range, perhaps to 25 or 30 Km (without running it into the ground). But unless one ordered the special 8 volt golf cart batteries (180 AH), which would probably cost substantially more than buying a common type, one couldn't fit 144 volts worth of higher amp-hour batteries except with very pricey batteries (like Li-ion, NiMH).
   It crossed my mind that there is one common size, 'frame 31', a little bigger than the '27' size I used. I figure one could fit 6 in the back, and just barely squeeze 5 in the front, for 132 volts instead of 144. At 145$ each those would cost maybe 1800$ - at least considerably less than most lead-acid EV refills. But it seemed that size 27 were 105 amp-hours, while size 31 was 115 amp-hours. Apparently it would be a very modest improvement. There were some other 12v sizes up to 150AH, but again they were probably less common and more costly. A last option would be eighteen 6v golf cart batteries, which would be 240 amp-hours but just 108 volts. The 26 KWH (for 3000 $) would no doubt give it decent driving range, but performance would no doubt suck.
   Anyway, batteries for the RX7 seemed like stop-gap measures. It's still a leaky beat-up old car with a several things that don't work, with the motor connected to a lossy standard automotive transmission.

   Then a friend sent me a link to a 2009 low speed electric cargo truck at "Burnaby Repo" (repo.com) I decided to buy it. I figured it would doubtless do 50 Km/Hr around town, which, assuming the practical range was even 40 Km, would actually take me most places where I commonly go. It really whined away at higher speeds but went up to 34 MPH on level ground. The noise seems to be in the transmission gears rather than the motor.

   It wasn't quite what I wanted, but it would surely be more practical than the RX7. And it could carry lots of cargo. And I figured I could surely sell it for as much as or more than I bought it for, electric trucks still being almost unobtainable.
   It was a Miles ZX40, which Chinese company I had never heard of. It apparently operated from 2003 to 2013 when it went out of business, making gas and electric vehicles that 'meet vehicle safety standards'. Several types used a single frame that could be a flat deck, pickup, open deck passenger carrier, van or cargo van. The front end of all of them looked the same. The transmission for the electric was the same as for gas but stuck in 2nd gear, with no shift lever and no clutch. (I can't help but feel that would be a significant source of inefficiency, as usual.) It had a 72 volt AC motor with a variable frequency drive (VFD).
   It was on sale for around 6000 $C... 10000 $ less than a used iMiev I saw (cheapest one yet for 16000 $). Similar new models on their web site (vans or pick-up trucks - I didn't see another cargo-box style) were around 21000 $US. (I can't help but think that for such a price, if a regular car would do instead of a truck, an Mitsubishi iMiev or Nissan Leaf would be better value. At 6000$ it's a different story.)
   It has 12 golf cart batteries under the floor of the cargo cube - half as many as some vehicles. Those would give it 240 AH * 72V = 17280 WH of energy, weigh somewhere around 62.5 * 12 = 750 pounds, and cost 145 * 12 = 2000$C (with taxes). Golf cart batteries (having sodium sulfate in them) are apparently the longest lasting of the lead-acid types, and if I also used 6 of the pulse chargers for charging them, they should last quite a long time. Maybe, with luck, headed for 700 to 1000 charges?
   At "Burnaby Repo" it was stuffed into a warehouse with no room to move. I had no reason to suspect it had problems. But when it arrived it didn't work right - it worked intermittently, and let me down on the street a couple of times. But I found that if I unplug the 35 pin cable to the motor controller for a moment it starts working again. Going by the symptoms it seems to be either the motor rotation/speed encoder or the wiring to it. This may be one reason it was more affordable. But I wanted to buy a vehicle, not another project! I'll cross my fingers and order another encoder (60$) from Canadian Electric Vehicles (CanEV.com).

   I'm not fussy about the Chinese construction. Both the batteries under the body and the electronics are open to the road beneath, and there's no undercoating on the body. It already has some light surface rust in various lower places. Randy at CanEV didn't think much of it either. I may take some anti-rust zinc spray paint to it.

   Anyway, if anyone wants the Mazda RX7 EV I'll sell it for 2000$... but without my pricey NiMH and lithium batteries. Battery option ideas for the buyer I've stated above. (The electric drive parts are doubtless worth more out of the car than in it.)

Tercel & RX7 EV battery notes

   After about 5 years of use and abuse, the 3 x 10 D cell (30AH, 12V) NiMH starter battery in the Tercel is now failing to start the car if it sits for 3 or 4 days or more. That would be after many thousand starts, and many hours of waiting at red traffic lights, and coasting down grades or while slowing down, with the engine idling much too slowly to keep the charge up... thus saving the maximum possible amount of fuel, which at one point (before the NiMH battery) I calculated from fuel purchase records to be about 10% savings in mostly city driving. (When I first turned the idle down, with "reconditioned" lead-acid batteries, they were lasting months, not years.) Hauling around a 12 pound NiMH battery instead of a 28 pound PbPb has also doubtless saved a bit of fuel.

   My records of battery replacements in the RX7 EV somehow are lacking - from the time the first piece of paper in the glove box ran out of room. That's rather pathetic record keeping for a car I partly have for battery longevity experiments! One of the four remaining PbPb batteries appears to be giving up the ghost, limiting the range to under 5 miles in spite of pulse charging it for the last 2 or 3 months. I'll hazard a guess it's been in there about a year and a half. I'll have to keep better records!

LED - Aquaponics & Space Lighting Note & Peltier Fridge

   I watched a video on youtube of an indoor farming project in an 'abandoned' multi-floor Chicago building - a parkade or something. LED lights had been donated by GE. It seemed like 90% of them were red, and only 10% were blue. Using mostly the cheap red LED.s and only a few of the costly blue or violet ones improves the economics of making grow lights. Perhaps 3 rows of reds and one of blues would be better than two of each. But really I need to find a design that's easier, that doesn't take hours to assemble. And it needs either better cooling or to be derated from 15 watts to maybe 10. Or perhaps an efficient switching current supply would sufficiently reduce the heat generated, because much of the heat comes from the current regulator pass transistor, especially during the day with the solar panels raising the supply to 14 volts. It's only (so far) the emitters near the circuit board that start failing.

   I also note that for all the claims of longevity of LED lights, the first LED house lights I made in 2012 don't seem to be as bright as they were when I put them in. The emitter makers did say that organic outgassing or vapors from nearby organic materials would contaminate the plastic and reduce the brightness. If that includes carbon dioxide, the bit of CO2 in the air probably accounts for it. That would tend to indicate that the emitters should be enclosed in a sealed space, or at least one with very restricted air circulation. On the other hand, some of them might be running too hot, which demands that they be better heatsinked to the open air they shouldn't be exposed to.

Peltier Module Fridge

   The peltier module fridge is cooling more effectively with the big 15 amp emitter replacing the 8.5 amp unit. It probably actually draws around 8-10 amps, the fan inside circulates air through the fridge, and the outside heatsink fan needs to run at a higher speed to cool the heatsink. It's been maintaining about 3° near the ice tray, which always has ice, and 7° near the top and the far end. I am no longer concerned that a big tub of margarine or whatever is too warm for the long time it takes to use it all. My take is that around 100 watts, on most of the time (I occasionally shut it off manually for a few hours or overnight) is a lot of juice for an "apartment size" fridge. But if improved COP, long lived peltier modules ever make it to the market they'll be a game changer. Even 50% higher COP would be most useful, and with 200+% effective COP they'd definitely start replacing compressor based fridges.



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

   Somehow, three topics, written separately and at different times, fit together like a single narrative. Okay, once I realized it, I changed a couple of lines to enhance the effect. It was still completely unplanned. "Bail-ins Coming?" was inserted on January 2nd after I suddenly realized what is likely to happen with the banks - perhaps this year.

The Pathocracy

   Not too long ago, I noted how people with various personality defects or disorders had come to be in control of governments and the economy almost everywhere. This has been referred to as corporatocracy, oligopoly, and cleptocracy.

   Now a book called Lifting the Veil, available on line, has coined a new term, The Pathocracy, explaining how and why psychopaths, just 1 to 4% of the population, have come to dominate the positions of economic and political power.
   (Personally, I have a hard time believing that actual, genetic psychopaths with no soul who only look like normal human beings, could manage to function in society or that they are such a large percentage of the population. But the leanings and behavior of those who have come to be in charge certainly can certainly be broadly described as psychopathic.)

   According to the last chapters, the road toward eliminating the power of this group is education (what else!!), and non-participation rather than confrontation. Learn what goes on 'behind the scenes.' Don't keep your money in the big banks. Choose local over big corporate. Etcetera. (And I might add, buy as little petroleum as is workable. I'm sure that would be the best help for the middle east especially, where wars are being fought for it and corrupt powers are being sustained by the revenues. Great argument for EV.s!)

   For anyone who wishes to delve into this rabbit hole of power lust, greed, satanism and pedophilia, here's the URL. (Myself, I couldn't bear to read even half of it.)

Lifting the Veil - An Investigative History of the United States Pathocracy.

Bail-Ins Coming?

   Legislation allowing "bail-ins"* has been passed in most western lands including Canada. I understand that six European countries that have "failed to enact" the demanded laws by 2016 are being taken to court by the European Central Bank (ECB). Why even in the surreal culture of the early 21st century should banks have such gall, and such clout over elected governments... and why are they so concerned about a few holdouts?
   I've said ever since Cypress that if a second country implements "bail-ins", people everywhere will start pulling their money out of banks, and the global financial pyramid system will quickly collapse. The banks must realize this. Suddenly I glimpsed the obvious purpose of them having demanded that all countries enact the legislation: when their "Lehman Brothers moment" arrives and they can no longer carry on, bail each other out or save themselves any other way, bail-ins* and capital controls** are to be implemented throughout the entire western world, all nations, all on the same day. That way, no one anywhere will have the chance to pull their savings out.

   But how, really, can that save the financial/banking system? Confidence will still be shattered and all economic activity disrupted. The gargantuan, ever-growing debts foisted upon everyone over decades will still be utterly unpayable. The advice for several years now of many of the same economists who predicted the 2008 collapse of the US housing market at least by 2005-2007 can only be reiterated: Get your wealth out of the financial/banking system. Buy any real assets you think will be useful, have some physical cash, silver and maybe gold. And put aside some good food and water. No one dares to predict which straw will break the camel's back or when it will happen, but whenever it does, it will be Earth-shattering.

* "Bail-ins": recapitalization of failing banks by taking (a very substantial percentage of) depositors' savings, in return for (plainly worthless) bank stock.
** "Capital Controls": you can only take a little capital - a little of your own money - out of the bank each day. I think (don't quote me) it's 300 Euros in Cypress and 400 in Greece, and in Greece there are weekly maximums.


Techniques for Suppression of Clean Energy

Moray B. King, electrical & systems engineer who had worked around the frontiers of physics for decades listed 8 ways that revolutionary energy inventions have been blocked before they started on the road to the marketplace.

* Academic Suppression [ridicule by 'experts', etc.]
* Blocking of Funding
* Blocking of Patents
* Litigation
* Threats to the Inventor
* Property Destruction
* Framing the Inventor with a Crime [...but hardly necessary when trying to patent certain types of
                                                       inventions is in effect already legally considered to be a crime!]
* Assassination

These tactics come from academia, business, and black ops.
(Per: Breakthrough Power book by Jeanne Manning & Joel Garbon)

   It is easy of course to become cynical, to give up and simply accept what's doled out by "the establishment". But in the coming times, the establishment as we have known it will wither and be replaced by one populated with people having the will to the three core values of quality of life, growth and equality, and having empathy, compassion and love (values and qualities psychopaths entirely lack). For the prepared, the lucky, and for coming generations, hope for the future is brighter than seems warranted today.

Peace and Forgiveness

   On Christmas day in 1914 after 5 months of appalling slaughter, along 2/3 of the western front the German, French and British soldiers came out of their trenches into a no-man's land strewn with corpses and fraternized, singing Christmas carols, exchanging greetings and photos, cigarettes and chocolates, and even playing soccer. The high commands on both sides declared this to be treasonous and subject to court martial. It took until March to get the killing machine back into full operation. (I'm not quite sure of the timing - at one point it was dark, so I assume it must have started Christmas eve, but playing soccer games indicates it stayed well into Christmas day, perhaps all day. And okay, I'm a year late for the centennial of that event!)

   In August 1914 youth went enthusiastically off to war to try to vanquish the nation's hated enemies. After all, they were attacking inferior, strange foreigners who spoke other languages - and it was high adventure! Everyone thought it would be quick, but everyone was thoroughly sick of war by the time it was over, and war was never contemplated so lightheartedly again. (The Guns of August by Barbara W. Tuchman is a good read.)
   Today we speak of the 'international community' and the whole world is gradually becoming well connected. With the advent of electronic communication we are coming to know, sometimes personally, other members of our human family in other lands and cultures. We recognize much commonality of purposes and desires with them. And a single language has become the "exchange language" that everyone wants to learn to be able to communicate with others everywhere.
   There are, and probably always have been, far more humans of good will who desire peace than those few who are hell-bent on destruction. We are fed up with war after war. But in good-natured or apathetic tolerance, we have allowed the few to usurp the positions of power and influence. And these few ever manipulate and arrange awful situations and "false flag" crisis events to arouse our inner fears and prejudices and to have us grant them ever more power to "protect" us - by commandeering everyones' resources to fill their war coffers and removing our rights and liberties. Much of the west is on the verge of becoming dictatorships. (Again it's a good argument for having to have all legislation ratified by the whole population with on-line voting.)

   Sometimes such twisted people stop and examine their own motives, methods and goals. Many eventually revolt. When they do they cease their machinations and start to recover their sanity. Some start trying to expose and explain what's going on. When we hear them we feel sympathy for them. We forgive them. In spite of their past victimizing, they were victims of a system that asked them to do immoral things in return for employment. Some who feel powerless in spite of their position, through various mixed motives, wait until they retire to speak out against what they've been contributing to. (This seems paradoxical: Why now, when you have no more influence? Where were you when you were in charge of so much?)
   We should help any we are in contact with to start thinking and examining what they're doing. And forgive them in advance. A peaceful, forgiving environment will help them to recover their life instead of resisting change for fear of people who seem to be hostile to them, against whom in their insecurity they should build mental and physical bunkers of defense.

   Our thoughts and intentions for peace and forgiveness are more powerful than those for war and conflict. The angels magnify them and help them to bear fruit and spread amongst open minded people. The overall consciousness of the world has been rising, and this will accelerate into the future. The present distressing and depressing world conditions are the dying manifestations of shadowy greed, avarice and lust for power, on which the light of day is at last being shone, exposing them for what they are. The more we have optimism and hope and concentrate on being peaceable and forgiving, having peace and forgiveness, and initiating progressive processes and developments which promote peace and forgiveness, the sooner they will be vanquished, even in the hearts of many or most of those presently working harm.

Pluto: Color Images

   I keep hoping for close-up color images of Ceres and comet 67P. Instead, some lovely color images of Pluto are being put together. This is a reduced, cropped version. See the original for better detail and coverage.

http://pluto.jhuapl.edu/Multimedia/Science-Photos/image.php?gallery_id=2ℑ_id=389


Surface features on Pluto in color (reduced, cropped)

   The things that look like mountains are indeed of mountain size scale. They may have been (at the time of tidal disruption) a huge melted slurry of "ice" blocks floating in liquid, which then froze. This is judging by their similarity to broken up ice blocks on Europa, and because they appear to spill over onto the "cantaloupe" terrain with the wrinkly dune patterns at the far right. Is that similar to Triton's cantaloupe terrain? Per my tidal disruption theory on the "Origin of Pluto and Triton" [Somewhere on here: http://www.saers.com/recorder/craig/] they were once one much larger body, and so it may be the same terrain - a bit like Africa & South America, where one landscape was separated by continental drift.)
   The white material (whatever it is - frozen nitrogen?) appears to be blown from the north to northwest, preferentially covering faces sloped that direction in the center mountains and in the terrain at the right, where dunes reminiscent of those on Mars appear to go right across the polygon edges. (they're clearer in the full size image.)
   All this amazing and windswept geography occurs on an intensely cold sphere 1/3 the size [volume] of Earth's exceptionally dull and uninteresting moon.

   In the last part of December, the Dawn spacecraft should have entered a lower orbit around Ceres (still not very close!). I haven't had time yet to go see if there are any interesting findings.



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

Construction Manuals and information:

- Electric Hubcap Family Motors - Turquoise Motor Controllers
- Preliminary Ni-Mn, Ni-Ni Battery Making book

Products Catalog

(Will accept BITCOIN digital currency)

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



Daily Log
(time accounting, mainly for CRA - SR & ED assessment purposes)


Dec 1-3: Writing previous newsletter
4: Unipolar Motor Controller tests - oscilloscope readings
5: Study & interpretation of the readings.
6: More readings, more study. Evening/night: Programming code for MSP430G2553 to read display-controller serial pushbuttons.
7: Continued above. Working (finally!) by early afternoon. Started studying "timer A" in the MSP - to be used for pulse drive timing interrupts. Initial write of Timer A interrupt handler.
9: Mor Study: Timer A has mor complex features/modes than most timers. Looks like the unusual features might be perfect for the lambda ray converter. I'll have to redo that to take full advantage of them.
11: Mor programming.
15: Traveled to mainland to purchase an electric truck.
18: Bought toy quadcopter EV. HD camera, wi-fi. Wow! It flies - didn't try camera/wifi.
19: MILES ZX40 electric truck arrived. It didn't run properly. SIGH! Started troubleshooting it.
20: More truckbleshooting. More figuring out MSP430 TimerA - this time how the interrupts are handled. (not well explained in datasheets.)
22: Got truck running and drove it a couple of short trips.
23:
24: Truck proved to be intermittent.
25: Wiggled wiring cable and got truck home.
26: Removed a couple of connectors in a wire and soldered it together.
27: Truck quit again. Rats!
28-31: Hey, it's Christmas Holidays!



Electric Hubcap Motor Systems - Electric Transport

Unipolar Motor Controller

   The energy return system in the motor controller seemed to be the overriding concern. I reconnected the ARM (axial reluctance) motor. I hooked up the oscilloscope to various points to check voltage patterns and their relationships with other voltage patterns. As I did these measurements and initially copied them in here, I was mystified by a few things. After a day to reflect, some of them started to clear up, and I've written both of the initial confusion and the later answers that came to me.

First, what was the supply voltage before and after the "choke" - the energy return coil?

     
Supply Voltages before (Yw) and after (Bu) energy return choke
at Motor coil turn-on (L), Motor coil turn-off (R)


   Since the voltage for the motor coil is drawn after the choke, one would expect the blue line to drop at turn-on and yellow to follow. Why the blue should rise, and somehow the energy seem to be drawn straight from the supply, is puzzling. Hmm... what's the "reverse recovery time" of the mosfet body diodes? If they conducted while reverse biased, that would drag the supply down. Okay, typically under 40 nSec. The drop comes on more slowly than that, and lasts for about 400 nSec (the fall time of the motor coil drive), so that can't be it. And the capacitance across the mosfets/diodes should be trivial.
   Finally I remembered that the main current starts to build from zero once the motor coil is turned on. It draws no current at first, so why should the voltage drop? On the other hand... why should it rise? And what could in the first instant be drawing current from the battery side of the choke besides nothing?

   I hesitate to mention the "DC kick" of a suddenly switched voltage here in connection with the "wrong project". I would think it would only explain the blue line rise, not the yellow line fall. On the other hand, Bedini was claiming results beyond the expected with his pulse driven motors and now others are claiming extra energy from certain new types of motors... and I've chosen to drive this controller with pulses. Now I read that Bedini actually says the extra power of his motors comes from radiant energy. (Per: Breakthrough Power - book by Jeanne Manning & Joel Garbon.) That would doubtless be correct.
   Later I figured out that the apparent dropping of the main supply (above, yellow) was really the ground reference side rising (mentioned again below), so if the yellow line were flat at the beginning of the pulse, then the spike of the blue after-choke voltage is really its rise above the yellow, making it a considerable 6 or 7 volt spike. (Maybe this is why the battery seemed to last longer than I expected? It seems somewhere along the road to the reputed Bedini results.)
   But later again, with the ground clip on the circuit board, there appeared to be no voltage spike at all - in fact, a slight dip was seen (below, yellow). Is anything there at all besides measurement vagarities?


   Perhaps it should be noted that there are filter capacitors on both sides of the choke (>600 & <1000 µF), and that the scope probe grounds are connected at the supply side filter cap ground. The choked side capacitors' ground goes through a fat 8" wire to the same point. So the two signals are referenced to the same ground point, and unless the scope isn't sampling both simultaneously, the traces shouldn't be misleading. (If they aren't sampled simultaneously, why did I spend so much on a new, modern oscilloscope?) The ringing - apparently back and forth across the choke - might indicate that it needs more filter capacitors or that the 8" wire should be eliminated. (What a lot of ringing there is, and especially in the coil drives!)

   At turn-off, the rise of the supply line with flyback current coming from the coils, compared to the after-choke voltage, is what's expected and desired.
   Of course, all these are very rapid transient effects. That can be a whole study in itself.

Here are some traces of one of the motor coils turning on and off, first with the after choke supply voltage - the whole pulse, then the on and off areas expanded. In all the following traces, zero volts is one dotted line up from the bottom one.
   

Then with the battery side supply voltage, showing one pulse, then longer time to show the spacing between pulses. The energy return flyback diodes prevent the coil from rising much above the supply voltage.
 

Perhaps there's a clue to what's happening in the nearly identical voltage drop of the supply and the motor coil in the first 1/2 a microsecond after turnon? In fact, I think that's it... both voltages are referenced to "ground". As the mosfet turns on, it draws current from the ground line. So the apparent identical drop in voltage of both the power supply and the coil at turn-on would actually be a transient rise in the ground reference voltage.

Finally here are on and off with the choke removed - the one supply line to the motor coil.
 

There's no separate "after choke" line to get a pulse of voltage into as the coil comes on. And when the diodes conduct current to return the energy, they merely short out the motor coil flyback, heating the coil instead of returning it to the supply.

   On the 5th I put just one choke coil back in, 750 or 800 µH, and again checked the waveforms on the scope. The differences from using the two parallel chokes making 400 µH were fairly subtle. Here's the supply before the choke and one coil (colors are reversed from above).
 

One significant apparent change that I note is that there seems to be lower diode drop voltage from the motor coil to the supply at turn-off.

And after the choke, same coil.
 

(Hmm... the 500-700 nS delays at turn-on and turn-off - the switching speed - may change with the gate resistor value, now that I think about it. At least, it's supposed to. I should probably go over the old IR app notes and see whether I should reduce them, increase them, or they're about right.)

   And why in all of these traces does the bottom end of the coil seem to sit about a volt above its supply voltage when idle? Perhaps it's one .7 volt forward diode drop, and over a longer timescale the coil voltage would drift down to its supply voltage.

   As to the immediate concern, it would appear that the energy return system actually is doing its job. The first fix would be good heatsinks for the energy return diodes, and the second would be active rectification to reduce that heat.


   I note in these drive pulses that they are only 5 microseconds long, with 25 microseconds between pulses. With a good 25 volt supply the average current might hit 20 amps. Thus a continuous drive then be 100 amps. It would seem the peak currents will have to be permitted to rise much higher, or the 20 microsecond OFF time shortened, or both, in order to obtain an average current of 200 amps to run a motor at 25v*200a=5KW.
   If the pulses build up to 200 amps in just 5 microseconds, they should decay in an similarly short time, so shortening the 20 µS OFF time to under 10 µS, if not to 5, should be feasible. But maybe I should check the BLDC4-3 motor too and see what it does?

   With the ARM motor connected (and turning no great speed, hence little back EMF) the coil pulses are only 5 microseconds long, and the rest period between them is 20 microseconds. The 15 amps max would be 75 amps if there wre no rest periods. That's at least headed towards the 200 battery amps that (at 24 volts) would indicate the desired 4800 watts. That was right in the middle of the desired rest period duration. The surprise is how fast the current builds up and shuts off the drive pulses. Could the current sensing be faulty?

   At 15 volts with the Honda battery, the pulses lengthen to 6.4 µS, and the total period to 28 µS. I cut the current sense shunt in half (electrically speaking). The pulses doubled to 13 uS and the period lengthened to 35 uS - which is still 22 uS dead time.
   I put one probe on the voltage sense resistor. There was more ringing than anything else. I moved it to the R-C filtered side, where the motor controller chip senses it. Still lots of noise (and straight across a .1uF capacitor!), but one can see the sense pin voltage rise from 0 to .1 volts (blue, .1v/div) and then the pulse shuts off. One sees the sense transiently go negative with return energy. (I should have followed the trace farther along.)


I tried a few positions across the current shunt, using it as a variable resistor. Lower resistances seemed to perk things up a bit, but at that point I was running off the 10 amp power supply, so the effect wasn't large. But definitely if one expects to anything like 200 amps of average current for the intended 4800 watt motors, allowing a much higher peak current is necessary, and the current pulses are going to have to be relatively closely spaced.

   But early in the month, I turned my attention to microcontroller programming for VHE lambda ray conversion, and later Christmas and other year-end events intervened.


BLDC Motor Rotor Magnet Holder


   I started thinking about beefing up the rotors even more than better PP strapping. How about a solid cover shell of tough plastic that fits over the magnets and bolts onto the steel rotor with flat-head machine screws, clamping each magnet inside a rectangular pocket? It could be up to 1/8" thick over the magnets and still theoretically leave a 1/8" or better gap between the 1/4" center wall and the shell with a 1/2" (or wider) flux gap. Somewhere it would still need 'fins' to blow cooling air outward, but I can think of a couple of potential arrangements. One is to simply make the plastic thinner than the gap. Then the magnets will still be 'fan blades'. On the other hand, it would be strongest if the plastic pockets are the same depth as the magnets. The more I think of the way the magnets broke off the rotor in the Sprint the more I think I like strong.
   Then, what type of plastic? I occurs to me that kitchen countertop laminates are strong and rigid. Some of them are fairly thick - not thick enough for the whole thing, but perhaps thick enough to be the cover over the magnets (and under 1/8"). Then the main piece would be perhaps PVC or lexan (something glue will adhere to) the same thickness as the magnets, with the square magnet holes going right through it, and the top, solid laminate, covering the holes.

   A short term advantage to doing it this way would be that the magnets don't need to be glued to the steel rotor. For the first one, I could arrange them in traditional BLDC configuration for the Kelly controller, and then assuming I get the unipolar controller working well, I can take the rotor apart and change them.



Electricity Generation

VHE/Lambda Ray Energy Converter


Some Thoughts, Info, on Lambda (λ) or "VHE" Rays

   First... the name? Calling rays orders of magnitude above gamma rays "gamma rays" is misleading, even qualified as "VHE gamma rays". It distorts thought and study on the subject. It was once thought that the gamma ray bands were the top end the electromagnetic spectrum. Hence any rays found above those bands could only be higher frequency "gamma rays". But the rays are qualitatively different, like infra-red rays (eg 3µm) are qualitatively different from ultra-violet rays (eg 300nm). In fact, the lambda and gamma bands are much farther apart in frequency than that. This is why I unilaterally retitled them "lambda rays".

   It seems to me now that in case one doesn't like that name for terra-electron-volt and higher (TeV+) rays, one could drop the misleading term "gamma" (γ - MeV to GeV range rays) from "VHE gamma rays" and just call them "VHE rays". Or perhaps "VHE lambda rays" would trigger recognition of the "VHE" term in physics and astronomy circles without applying the "gamma" misnomer.

   (A similar distortion of sorts occurs in the metric system. It was intended to be scalable. This seemed to work down to the smallest sizes of nanometers and even picometers - sub-atom sizes. No one speaks of "15 millionths of a millimeter". But when scaling up, beyond kilometers, even space scientists have subconsciously yielded to familiarity or "force of habit" and speak of distances in the solar system as (eg) "millions of kilometers", thus splitting up the scaling factor into two places, instead of either as "billions of meters" or as "gigameters". The moon orbits 400 megameters from the Earth. Even the spelling checker doesn't know these words! But I digress.)

---

   Next is a thought on detection: Different parts of the electromagnetic spectrum have qualitatively different radiative effects. For example, a radio tuned to the 30 meter band won't tune into or detect visible light rays at 500 nanometers, and a photosensor won't react to radio rays. Likewise, with lambda rays at perhaps 1000-100000 times the frequency and energy of gamma rays and again having qualitatively different radiative effects, it seems likely that typical gamma ray detectors aren't going to detect lambda rays. As Dr. T.H. Moray said, the most powerful radiative energy has been unknown because we lack the "reagents" needed for its detection.
   Today that statement needs revision: terra-electron-volt rays were detected in 1989, and rays at up to 2.9*10^27 Hz have been detected within the last decade. (There's more on such developments below.) But how much do we know about their intensity - the strength of the "CLBR" ("cosmic lambda background radiation"), or about its strength coming from our own star? I'm sure it's very little.
   Each photon is of course far more powerful than those in any other band, but it's mainly (or only) from the fact that people over the decades have obtained powerful electrical outputs we can infer that it is very strong. Scientifically, however, AFAIK it hasn't been measured and there is probably no instrument currently existing capable of making such measurements.
   Perhaps related, I expect that Dr. Gustav LeBon's theories that radioactivity in matter is likewise caused by such high energy rays will someday be proven correct - that atoms (below atomic number 101) don't just blow up on their own at specific half-life rates. An external energy source is disrupting them.

---

I found this article: Probing the Sources of VHE CR Electrons (Cosmic Gamma Radiation)
http://what-when-how.com/cosmic-gamma-radiation/probing-the-sources-of-vhe-cr-electrons-cosmic-gamma-radiation/

   Well, I wont try to explain things I don't understand, but this had some interesting (if poorly reproduced) graphs. Considering their all-inclusive title "VHE CR Electrons (Cosmic Gamma Radiation)", the relationship between "electrons", "cosmic rays", "VHE" and "gamma" rays seems hazy. (Electromagnetic rays are of course photons, which are not cosmic "rays" (charged particles) nor electrons, but which may release electron-positron pairs.) This ambiguity may stem from the still only vague realization that the electromagnetic spectrum doesn't end with gamma rays.

Distribution of high energy "electrons"(?) with different assumptions of how they diffuse in the interstellar medium.
Note the concentration of energy in the lambda ray band, perhaps around 10^12 to 10^14 or 1 to 100 TeV
(1-100 trillion electron volts and perhaps centered on 10 trillion) in the right graph.
As people have obtained very substantial electric power from such radiant energy,
perhaps we may assume that the "slow diffusion" graph is more the actual case.


Another "fast diffusion" graph would see major attenuation of ten trillion electron
volt and up photons, yet the highest energy levels of all around one trillion.

   I'm not sure this article was intended for the layman. Of course we all realize the implications of the following statement which specifically seems to specify VHE lambda ray energy levels:

"Above several GeV, Compton and synchrotron losses dominate over bremsstrahlung and ionization losses."

...don't we?

---

High Energy Gamma Ray Sources (Cosmic Gamma Radiation)
http://what-when-how.com/cosmic-gamma-radiation/high-energy-gamma-ray-sources-cosmic-gamma-radiation/

   This extensive treatise gets into measuring techniques, specific astronomic sources of "MeV" and "GeV" (mega- and giga-electron-volt) gamma radiation, specific space missions, etc. While undated on this web page, it is rather dated as it mentions nothing beyond 2003. But it looks like a good overview of the whole subject area, at least as it was known in 2003 or 2004. One "TeV" (terra-electron-volt) is as high as they consider - as I see it, the bottom end of the lambda/VHE ray band.

   (I note with some disgust that scam ads for "Magnetic Generator? $49 - Eliminate Your Power Bill Easy Do It Yourself. Great Discount Now!" appear within that article's space. Note the "?" after "generator". All they give you is links to stuff you probably already found for yourself if you were interested, and instructions, plans or drawings that leave out vital details... because they won't work anyway. Yes I fell for one of those once. Why is nothing done to stop these con artists, who could easily be located and shut down? Because the "powers that be" approve of anything that causes confusion about "free energy" and tends to discredit the entire subject, so they choose not to investigate. They probably collect tax on the proceeds.)

---

   Below is info from a presentation for a talk from 2005 says that [one?] terra-electron-volt rays were seen in 1989: "TeV Gamma rays from Crab Nebula: 1989", and says that a new generation of instruments is needed for observing VHE rays. 1989 is much earlier (if only 1/10th the frequency/energy) than the 2007 date I found with the 2.9*10^27 Hz (~10 TeV) observed figure. The presentation predicts the dawn of a new golden age of TeV "gamma" ray astronomy using Cherenkov telescopes. It seems to figure the dividing line between "HE" gamma rays and "VHE" rays is at 100 GeV, 1/10th of a TeV. (How come my search results are hitting these older links and not newer ones? Anyway, this is much little known info even if it may be a little out of date.)

VHE gamma-ray astronomy: observations

http://www.unizar.es/taup2005/Charlas%20Speakers/05%20-%20Miercoles%2014%20Septiembre/T20050914%20-01-%20Martinez.pdf



   It also sheds light on a possible reason VHE rays seem to come from all directions: scattering by Earth's atmosphere.

"shower Detection of TeVgamma [lambda] rays using Cherenkov telescopes"

   And yet, doesn't the ray have to reach the ground intact, and its shower of particles occur inside the wire, in order to be utilized? Perhaps rays 1 TeV and lower are scattered by the atmosphere but the 10 and maybe 100 TeV rays go right through? Or perhaps only the occasional ray interacts with the atmosphere?
   If it is assumed that most rays are absorbed in the atmosphere and they aren't, the amount of energy available at Earth's surface in a little-diminished band could be grossly underestimated. I suspect that the density of energy in the VHE bands is unsuspected by the astronomic community.




   There is also an interesting chart of sources of VHE rays, here defined as being 100 GeV and up. If these sources are where the energy comes from, it would seem to explain why VHE/lambda ray converters evidently work day or night. (Or does most of what we get come from our own sun? Maybe I should look up TH Moray's Sea of Energy book and see what his ideas were?)


---

   A cable TV technician came to the house one day. They deal with long 'transmission' wires... so I asked him if he had heard of the "DC Kick". He had. He also knew nothing about 'free energy', so that wasn't the connection. I still have found nothing definitive about the kick, but it certainly sounds like it's real enough. It sounds like one of those things no one writes much about because they don't really understand it, and it doesn't seem like anything really important, so it gets mentioned casually to new workers when they enter a field where it needs to be known. Then again, one of Mark's e-mails said there were trade or scientific journals with articles on the subject, that he had cited to his correspondent. I wonder how to find those? If only it had a more distinctive word to search on in its name.

---

Magnet Motors are VHE/lambda ray converters?

Finally, here's what I think is a "game-changing" idea. In TE News #92 I mentioned a 'youtuber', "MotionMagnetics", who had posted a lot of info on self-turning permanent magnet motors and (as I recall) said that there were three essential types. Especially for one type I remember him saying that in a successful motor, there was a sudden burst of extra magnetism that helped propel the rotor on its way. This seemed surprising. Where would that come from?
   But here's a connection: moving magnetic forces induce voltages into nearby metal... not only into copper wires to make generators, but eddy currents into any solid metals including into other magnets. A powerful enough, sharp enough flux change (such as passing by a magnetic shield to suddenly enter into full supermagnet strength flux) could cause a sharp enough change in voltage, in electron pressure within a piece of metal, that the "DC Kick" conversion of lambda ray photons to a shower of electrons and positrons would take place. These particles would generate electric eddy currents, which would create more magnetism, most likely an "extra" counter-magnetic field: the magnetic burst that makes the magnet motor work.

   Ergo, it would seem that magnet motors - at least those making use of the 'magnetic burst' effect... which just might be all of them... are in fact getting their rotational energy from lambda rays and are thus a special form of VHE lambda ray radiant energy converter rather than a wholly separate type of energy source. Even ones like the V-gate, which seem to work because a cam pushes the stator magnet out of the way at the critical point when reverse flux would otherwise stall the rotation, may actually work because the suddenness of the motion and change of flux when the cam drops the magnet, causes a magnetic burst at that point.

   So, those who say repetitive work can't be done by magnets without an energy input are probably vindicated. But people have made magnet motors that run. They just don't work for the reasons the proponents and builders think: there actually is an unsuspected external energy input.
   Having seen the difficulties of getting a self-turning permanent magnet motor to work and the low power outputs available from them, versus the high powers from the oscillating coil types of VHE ray converters, I think I'll continue to concentrate on the latter and forget about self-powered magnet motors.

   OTOH, an understanding of how such magnet motors are actually powered could lead to design improvements that make them much more powerful and practical. This could have great potential. It seems to me for example that an outboard motor that turns itself with no electrical supply or motor controller, its speed governed simply by mechanically moving something near the rotor to adjust the field strength, could have great practical potential.

   Then there are "Bedini" and other "hybrid" sorts of motors that seem to get extra radiant lambda ray energy as they switch their coils on and off. One of them seems to greatly improve the battery range of certain e-motorcycles over what is normally expected. (the "Zero" motorbike, IIRC?) And why not? Fast switching of coils is the best technique to convert lambda rays. (I need to do more testing to see if my 'full strength pulse' unipolar motor controller and reluctance motor combination is actually getting this 'extra kick', or if it's just some trick of the high frequency measurements. And I could try it with the BLDC4-3 motor, too.)

---

Meanwhile, back at the project...

Pulse Timing Made Simple? Triple timers in the MSP430 microcontrollers

   Looking at the "Timer A" datasheets - one of the several complex devices integrated within the MSP430G2553 microcontroller chip - it had "hidden" within it two "extra" interval timers along with the main one. For a while this seemed confusing and complex, but really it's like having three timers in one. The more I explored it, the more I began to realize the brilliant design behind a system where multiple output timings are performed in hardware, but right after each change in output, there's one CPU interrupt to set up the next time period under program control. Somebody really knew what was needed and figured out in fine detail how to get it. Someone tells me this system was first developed at Motorola, the same place that came up with the fabulous MC6809 and MC68000 microprocessors in the late 1970s -- and might we suspect, probably by the same brilliant designer?

   And in fact, the chip has two such triple timers built in! This seemed perfect for the triple-transformer/coil converter, where each one is driven by a frequency in resonance (or nearly so) with the other two, forming a "chord". With a single counter, I'd have had to have a "quantum" time when the micro was interrupted and did all the counting for each frequency, incrementing three numbers until each overflowed, at which time a transformer coil would be switched on or off by software and the next time interval would be loaded.
   With the triple timer being also a hardware counter, each of the three intervals would be added to the current timer count, then stored in its end count register in the timer, almost like before. But the timer/counter would then count up automatically. No CPU interrupt is needed until each interval ends, at which time the appropriate coil/transformer would be turned on or off and the next interval length would again be added to the current count to be the new end count. Thus the timer/counter features allow there to be just one interrupt per each pulse length interval, instead of so many I feared the MSP430 just might not be able to handle the pace.

   Then as I was trying to write the interrupt service routines, I started thinking: there was something in the datasheets about outputs from these timer compare registers, and different mode settings for them. If these came out to pins on the chip, they could be used in place of simple port outputs and could toggle the output in hardware at exactly the right time, without a short but uncertain length of delay while the CPU answered the interrupt. (The interrupt service routine (ISR) should easily manage to put the next time value into CACCRx long before the next switch time.) This might be a bit nit-picky, but it would be the "proper" and more exact way to do it. I looked at the pinout of the MSP430G2553. Some of the many bizarre names on some of the pins said things like:

Pin 4: P1.2/TA0.1/UCA0RXD/UCA0SIMO/A2/CA2

   So if I set everything up right, pin 4 could be TA0.1 -- presumably TimerA0's TACCR1 output. Then I looked at the pins I had actually chosen to drive the coils: 11, 12 and 13. In additon to being P2.3, P2.4 and P2.5 (as stated on the part in the Eagle PCB layout files), they could be TA1.0, TA1.2 and TA1.2 (wasn't shown there, but it was in the datasheet. Lack of the full pin designations on the Eagle part fooled me into thinking that the pins I was choosing couldn't be used for anything but general purpose I/O port pins, so there was no point delving deeply to make sure I wasn't selecting pins that would be more valuable as something else. ...but probably academic since I didn't know what I was doing with the timers anyway.)
   Too bad the middle one wasn't TA1.1. Not quite ideal, but if I set it up right I could pick the second timer (TA1) and have two of the three outputs driven by the hardware, eg:

11. TA1.0
12. TA1.2
13. P2.4 (software switched)

   Considering I didn't know what I was doing, I did well to get two out of three. And just how close do the frequencies have to be, anyway? Switching one in software certainly should be manageable - a new circuit board is not yet! But it will be at some point and I'll shift pin usage to make it all hardware switching. When I was half my present age, I used to delve into every minute detail of the specs and operation of a piece of digital logic and be sure I had every minute parameter of the whole circuit within specs and sure to work. Now I gloss over the details like everybody else, and cludj my way through. Of course back then, the functions contained on this one microcontroller chip would have been spread across maybe a dozen chips. Each pin would have one exact function, and each component/chip would have seemed individually much more important and worthy of scrutiny - which is of course not the case.
   And in fact, one of the lines is one of the two traces on the board connected with a jumper. Removing the jumper and soldering in a somewhat longer (still short) length of fine wire can change it from pin 12 to any unused pin. Pin 14 is unused and can be configured to be TA0.1. That uses up both timers to get all exact timed hardware switching... is there some other use for the other timer? Or, I could cut a trace to the display-controller and liberate a pin that can be TA1.1. That would be more ideal.
   Hmm... now a trace cut and two jumper wires. Still faster than doing a new PC board, and (so far) not too messy! Looks like the way to go. I removed the jumper, cut the trace, and simply swapped pins 10 and 13 by soldering on two fine wires.

The timer now connects:
10. TA1.1
11. TA1.0
12. TA1.2

The Display-controller now (SPI) connects:
4. P1.2 - Data In (from DC)
8. P2.0 - Data Out (to DC)
9. P2.1 - Serial Shift Clock
13. P2.5 - Shift/Load Enable

I fixed it on the schematic, and in theory on the PCB layout too, but it's hard to route them single sided without more jumpers, so I may revise it again before I make another board. (Using pin 7 [P1.5] instead of 13 [P2.5] for the DC would help.)
   On the 30th I jumpered across the voltage regulator so it could run with a 5 volt supply. I hope that's enough for the power circuits and the VHE rays. I also found and (in Eagle for the next board) corrected a mistake - the Display-Controller was powered from the wrong side of the regulator. The 12 volts would have blown it. It all comes of changing the design as you go for "version 1.0".

Next

   (29th) The next step will be to try out the PC board, with a 5.0 volt supply. I'll modify the display-controller drivers to use the I/O pins selected on the converter board rather than on the "launchpad" development board, program the MSP430 and plug it into the socket on the converter board. Then I'll see what surprises await. Hmm... where does the chip get its clock from? (When I was designing the board was probably a better time to ask that question!)

   Once the board is running properly, probably the biggest task will be to sweep through a frequency band with only one coil running and find resonances, frequencies where the DC kick creates the highest voltages. Presumably the first coil can then be set to the maximum point, then the other coils can then be set to harmonically related resonances to that. ...Or, might each coil have its own set of resonances? I had best sweep the frequency band with each coil separately and find out.
   Now, how to go about that? With non-symmetrical pulses, there's a host of possibilities for setting the ON and OFF period even for one frequency. And then the frequencies are to be varied ad infinitum. With a 'real' computer I'd just put the result for each frequency into a table for later analysis. This microcontroller however has a very limited RAM memory space - a ridiculous 512 bytes instead of billions, which will quickly be filled. (Even in 1986 I put 32K byte SRAM memory chips into my 8/16 bit control systems! Not to mention 64K DRAM.s!) Some sort of processing and checking results is going to have to be done, repeatedly, each time the mini-table is full during the frequency sweep. Only the high and low points can be recorded.
   Then, I think, when each sweep is completed, the peaks and dips should be reported to the user via the display-controller. Perhaps I'll use the top 3 buttons to initiate a sweep on each of the 3 coils. Maybe one of the bottom ones plus a top one to have the results display repeated.
   This is going to take some doing, but the resonant pulses are the heart of the system and I want to explore them relatively thoroughly. If it can be made to "learn" to configure itself to various configurations of coils and wires of different lengths, it will make it much easier for anyone to make up a converter.



http://www.TurquoiseEnergy.com
Victoria BC Canada