Turquoise Energy Ltd. News

I stopped working on the batteries about 3 weeks ago to focus on the motor and motor controller for a while, lest I discover later that the batteries were ready but I would have to order parts that might take a while to come for the other items. The motor is essentially complete but I do now have semiconductors - I.C. chips and MOSFETs - on order for the controller.

I should reiterate that this design of axial flux motor-generator is much less expensive than standard motor construction, and so simple any good handyman can make one at home. (Thousands of wind power enthusiasts have made quite similar ones, with propellers but lacking the nail strips for iron in the stators.)

In working on the motor, I had been checking the fit of the motor parts inside my car's spare tire in the living room. Everything was going as planned. Then I set the motor parts down on top of the tire and left the room. When I came back in I looked at the stack, and in just a moment I realized it was a better arrangement for vehicle retrofitting: the motor on the outside of the wheel instead of inside.

While it might seem odd, making a mounting bracket to reach around the wheel to the outside is more practical than extending the axle 2-1/2 inches to fit the motor inside, and there's more room to fit things. A stator mounting bracket is required anyway. This one is bigger, but easier to fit. More importantly, it means there are no changes to the car's existing components.

Having wheels sticking out a little farther might seem trivial, and probably is, but vehicles are safety approved as is, and making any such changes might raise "roadblock" issues: mechanical re-approval of the vehicle, and liability in case of any accident that might somehow be blamed, rightly or wrongly, on the extended wheels.

So, with almost no change to its main parts, the Turquoise in-wheel car motor became the Electric HubcapTM. The only work to the car to mount the assembly was to drill two holes through the back of the brake drum housing for bolts to attach parts of the mounting bracket to. Another hole will be required somewhere through the body to run the motor wiring in through. In the current version, the decorative ends of the wheel lug nuts will have to be removed for attaching the magnet rotor to the wheel, but I plan to simplify the next one for the other rear wheel. (One does of course want two for balanced drive.)

The main work of the motor is now done: I've wound and wired the motor stator coils (9 coils of 60 turns #14 wire each, 3 coils in series per fase, wired delta and intended for a 120 V battery, 0 - 1300 RPM for about 0 - 125 Km/hour), glued the magnets on the magnet rotor, and yesterday I cast the motor in polyester resin. There was some leakage of resin before it hardened, but I have "a cunning plan" for the next one. It remains to be mounted on the car and to have plugs soldered onto the cables. The cables include the power wires, an embedded tachometer sensor coil, and two embedded AD590 solid state temperature sensors. (Just a few days ago I thought it might be nice if an overheat warning light lit up on the dash before the motor actually caught fire, and I added this detail! The motors are air cooled, and I don't know how hot they'll get going up a mountain.)

I was very concerned about the safety of the NIB supermagnet rotors. If one suddenly clamped down on metal with a finger in the way, it would probably crush it, and it would take a crowbar to get it loose. Magnet incidents, which often seem to happen (in my very limited experience) when you're working on something near them and not even thinking of magnets, happen almost instantly. But I was surprised to discover that the back side of the rotor has no attraction at all to the steel wheel. Given that fortunate fact, the hubcap motor solves the problem: One ships the unit with the magnet rotor and the stator magnetically clamped to each other. Mounting the motor will consist of turning the car wheel to line it up with the mounting holes in the rotor, then bolting the stator assembly (rotor still clamped to it) to the bracket. Then put on the lug nuts to slowly pull the rotor away from the stator. A plastic feeler gauge can set the air gap. Removal is the reverse, allowing the rotor to clamp back onto the stator as the lug nuts are removed. So the rotor is never free to wreak its potential havoc!

Pursuant to being able to make the motors run a car, a 3-fase variable speed AC motor controller is required. I checked with Baldor motors, but they had nothing suitable, and what they did have was more than I wanted to pay. I searched on the web and instead of working controllers I found an I.C. chip, the Freescale MC3PHAC variable speed stand alone motor controller, and I started designing a circuit around it. In the course of designing the interface for that to power MOSFETs to drive the motor, I ran across a very simple sample circuit in an application note by International Rectifier (AN985) that generated the 3-fase signals with simply a 555 timer and 3 flip-flops of a '175 quad flip-flop!

Since the requirements are that simple, that ended my affair with the MC3PHAC and I've ordered the IR2130 3-fase motor MOS driver chip that the app note was about. That'll drive IRFP260 MOSFETs as the actual motor coil drivers. I'll use a 12V tap off the battery (or a separate 12V battery) to supply the electronics, and the old 4000 series CMOS standard logic chips at 12V, further simplifying things.

I was looking at ways to get volts-per-hertz, but I noted there are individual "undervoltage protection circuits" on the high side drivers, and if you use small capacitors, the drive outputs will shut down after a given time at low RPM's. Presto! Why bother with high speed PWM, which just creates RF noise and works the driver transistors harder? Pick the right size caps for the desired single pulse width. (Not what they had intended the protection circuits for, I'm sure - but convenient!)

There'll be at least a couple more chips on the board - a 4053 for reversing the 'D' inputs (and hence fase generation) on the flip-flops for backing up the car, and an LM339 to turn on the temperature warning lights at set points.

My next door neighbor just gave me a big dead power inverter. It has two fans and the whole case is a giant heat sink, lots of great mountings and hardware (even very scucum battery terminals), and in fact 4 good IRFP250N power MOSFETs that I'll use (along with four dead ones). Well now, that'll be something like $65 in parts and some design and fabrication work, for a custom made motor controller in a fine case. Not bad - I heard figures in the several $1000's of dollars go by for the mythical AC car motor controllers that I can't seem to find on the web. (though, I'm sure they do exist!)

Back on the subject of vehicle safety, I don't know how other electric car controllers work, but the way this one does should provide more safety than standard gas engine controls. To go, eg, 60 Km/H, hold the 'gas' pedal 1/2 way down. To go 120, press it all the way. The car will do everything in the motors' power to get you to your selected speed including accelerating, idling and decelerating (which recharges the batteries). It won't dangerously pick up speed as you go downhill or lose speed going up. Take your foot off the 'gas' and it will immediately start to decelerate to your selected speed of 0 Km/H. Vehicle accidents happen very quickly. How many could be avoided or mitigated in severity if the car was braking as soon as the driver began to lift his foot, well before can get it to the brake pedal? I think back to my own serious car accident in 1993 and I remember how I covered most of the distance between me and the other vehicle while my foot was going from the gas to the brake. It wouldn't have prevented that accident, but it surely would have mitigated the severity of the impact and reduced my injuries.

When I set the battery project largely aside for three weeks, I felt I would have working batteries but for leaks. I didn't originally understand the batteries operate under a certain amount of pressure, called the hydrogen plateau pressure, which can be anywhere from .1 bars in some new alloys formulations to 20 bars with mitsch metals. A typical Ni9La2Co alloy is around .5 bars. I don't know what mine with monel and lanthanum hydroxide will be, but I do know that when I try to charge the batteries, they soon push the window glazing tape seams apart. With the seams glued, they force the tape open somewhere else. The electrolyte/electrode paste (and doubtless the hydrogen) oozes out. Edge leaks are why battery makers haven't used the otherwise superior bipolar flat plate design, preferring solid cans with just one opening for a terminal post.

b) each cell must be separately sealed in solid ABS plastic so that if there are any leaks through the edges, there is nowhere for anything to go. The function of the tape seals is then just to prevent leakage between the two adjacent cells under the ABS seal, and as they should both be under similar pressures, it seems workable. Only for the two outermost cells does the tape go through to the outside air, and even there it may be necessary to seal the ends with a flat ABS plate, leaving only the terminal post sticking through with a tight friction fit to the plastic.

To satisfy these conditions, the cells should interlock, the bottom of one fitting precisely over the top of the next and the two sealed by methylene chloride. Having already found pinhole leaks in the seams of the ABS cells I've done, I concluded that it was a case of injection molding or having a lot of batteries with a bad cells or two. When I went to an injection molder, he was quite discouraging and basically told me I could make the mold without a milling machine and that it would cost me $10,000 plus to have one done. The visit was however quite valuable: In an offhanded way he suggested I could make the ABS into a paste with M.C. and work it into some hand mold.

I'm not sure if he was serious, but I dissolved some ABS in MC overnight in a glass spice jar with a polyethylene lid. It makes a syrupy liquid which I hope can simply be poured into a simple mold. Now I'm making a metal battery cell mold. How it will work I'll find out in the next few days. At least it's all room temperature and there'll be no thermal shrinkage!

Naturally, I can hardly wait to see all these things rolling down conveyer belts in assembly lines here in BC, even in Victoria, and barbaric combusting vehicles being converted to plug-in hybrids en masse! If any and especially all of these prototypes work out well, the principle that they can be done and that it's practical to retrofit existing vehicles to electric hybrids will have been established (and on a peanuts budget), and I'll have had the honor of doing that.

But I can also safely say that, while I would love to help set up such assembly lines, determine the most economical production techniques, see the first products out the door, and suggest areas of future production or product improvements and even potential markets, I am not the person to run such an enterprise on a day to day basis once it is established.