E-Busa

[Frank06]

Early on in the process I contacted Remy and got some good information on things to watch out for.  These motors are designed for continuous operation at higher loads, so coolant temperature is an important parameter.  Output is derated depending on temperature.  In the worst case, the magnets can become demagnetized and there are sensors inside the motor to tell the controller to reduce output if temperature gets too high.

I ended up communicating with some motorsport enthusiasts who gave me some good tips.  The motor uses ATF (Dexron VI is recommended) for cooling and lube purposes.  It doesn't take that much flow for light loads which are typical for street use.  We all love our high-hp machines but the reality is that it doesn't take much power at all to go down the road at 60 mph.  Remy recommends use of a dry-sump system for higher loads i.e. one pump to ensure the oil sump is evacuated (it's on the small side) and another to supply oil back into the system.

One take-away for me was the suggestion that I "just throw an oil pan under it" which I did.  The first step was to disassemble the motor and mark the area of interest on the oil sump.  I used a hold saw to make the initial cuts.  A milling machine would have been nice but I decided this could be done with hand tools.  The rectangular area was drawn when the motor was in the frame and showed me the space I had to work with between the frame rails.



I don't know what grade aluminum this thing is made from but it is not easy to machine.  The guy who milled the top of the case down said the same thing.  Regardless, I got through it then cleaned the holes up a bit, making sure there was minimal opportunity for oil to remain trapped inside the OEM sump.



A smallish oil pan was fabbed up, mounting holes were drilled/tapped once everything was in position (being careful to keep chips out of the sump) then the sump was installed with a cork gasket and gobs of ATF-specific silicone.

[55chevr]

I have new understanding for this.   It appears that nothing is off the shelf for an electric vehicle.
Joe

[Frank06]

Joe, the air cooled motors are a lot easier to install.  My drag bike was a piece of cake compared to this one.

[Jessechop]

So with no transmission you are limited to the RPM the motor will put out. You mentioned peak hp @ 2750 and you were double that @ 140 mph. So is a regear in order? Or is over reving a electric motor ok? I know playing around with brushless RC cars the motors can free rev to the point of exploding a rotor. With a load there is no concern for this.

[Frank06]

Right, with a series-DC motor (armature and field in series) they will rev no-load until they break (unless the voltage is kept very low.)  In fact the standard way to wear in the brushes with a series motor is to run it at 12V for a long period of time.  A small motor like on my dragbike runs at about 2K rpm and draws 15-20 amps during this process.

The permanent magnet motor is more sophisticated.  It has a red-line of 10K rpm and my controller has it set to 8K in case a chain breaks or something similar happens.  Because of friction (I'm assuming it's friction) the "constant power" curve isn't really constant and it decreases with speed.  If I regeared so my target speed occurred at peak power I would presumably go faster.  But this is a street bike so I chose not to do that.  The gearing on it now puts the max power speed at around 70 mph.

Two upgrade options are: (1) change the motor core to the parallel wound version.  Torque is lower but goes on longer.  This is attractive because I can use the existing charger and batteries and partially compensate for lower torque with gearing.  Developed power is higher at the higher rpm but I'd need to upgrade my existing controller.

(2) run at higher voltage.  Torque is the same but goes on longer.  In this scenario the motor core stays the same but batteries have to be reconfigured, I'd need a new charger and controller.  Plus, running at 600V is at the limit most components are rated at (wiring, fuses, contactors, etc.)  That's getting into some very serious design considerations... Option (1) is cheaper and safer and I'm considering it.  This is the approach Lightning takes. 

[Frank06]

I wanted to keep stock bodywork as much as possible.  It made sense to mount the small electric oil pump for motor coolant underneath and behind the sump.  Oil is introduced into the motor at the appropriate fitting and gravity drains into the oil pan.  It's pumped via a check-valve through a small radiator and filter mounted aft then back to the motor.



The radiator for motor cooling is on the right hand side of the bike, the controller radiator is on the left.  It's bigger as the controller has less mass and rejects more heat than the motor.  Both items only run 5-10 F-degrees above ambient.



Also shown is a bunch of 12-volt Control wiring being installed.

[Jessechop]

3 questions

#1. Have you monitored oil temps leaving the motor vs leaving the cooler? Its just me being curious if temps are 125 or 225 sort of deal.

#2. Back when we built an EV is school our wiring was much, much less. Granted that was almost 20 years ago and the technology was basic at best back then. So is that much wiring common place on newer EV's?

#3. Is that a home made blast cabinet?

[Frank06]

3 questions

#1. Have you monitored oil temps leaving the motor vs leaving the cooler? Its just me being curious if temps are 125 or 225 sort of deal.

#2. Back when we built an EV is school our wiring was much, much less. Granted that was almost 20 years ago and the technology was basic at best back then. So is that much wiring common place on newer EV's?

#3. Is that a home made blast cabinet?

(1) I'm monitoring temps at the oil pan.  The BMS (battery monitoring system) has 4 temperature probes.  The highest I remember was when ambient was about 80*, oil pan was 85 or so.  Electric motors are extremely efficient compared to ICE's.  Nothing in the system has ever been hot, just warm to the touch.  The only heat-related issue I've had so far is the DC/DC converter overheating (this is what keeps the 12V aux. battery charged.)  It's running the lights, horn, two pumps and keeps the main contactor pulled in.  I'm dealing with that now.  It's fine for about 25 miles or so (30-45 minutes around here) than craps out and has to sit and cool down.

(2) The controller is pretty sophisticated and has a whole bunch of inputs, although not everything is used at this time but I decided to terminate them all onto terminal strips in case I ever need the inputs.  Series DC motors (RC style) are a whole lot simpler.  Plus, doing the wiring harness for all the street gear entailed a bunch of wiring.  The BMS uses a wireless OBDII connector wired into a CAN-bus line to shoot data to an Android tablet I use to monitor various parameters.  There's wires for forward/reverse (reverse is cool!) and I can disable regen when I just want to coast.

(3) Yes it's a homemade blast cabinet.  I used a kit from TP Tools and keep modifying it for various projects.   

[Frank06]

I forgot to mention, the other big deal with wiring is inputs into the BMS.  For safety's sake, every single cell has an input into the BMS so there's 96 wires coming from various places, all heading forward.  The BMS sets an alarm if a cell gets too low, too high, too weak, etc.  It's necessary wiring but a real PIA.  The BMS wiring all goes through plugs and harnesses but some of the street wiring is still a rats nest.  I've got it all documented but want to clean it up as much as I can.  (another winter project, lol) 

To answer the other part of your question, you won't find an OEM EV that doesn't use a permanent magnet or AC motor.  Their development cycle lets them figure everything out and design for minimum cost/maximum simplicity so their systems will all be pretty cleaned up by the time they're in production.

[Frank06]

The controller's job is to send current to the motor in each of the three phases.  It's really a variable frequency inverter.  It functions by "trading" voltage for current when you initially call for torque.  As the motor starts turning it acts as a generator and creates a voltage that opposes current flow, so the controller has to increase voltage in order to keep things spinning.  They are pretty sophisticated little guys, packing a lot of silicon and engineering into a pretty small space.  It's desirable to keep wire leads as short as possible; I chose to mount mine directly over the motor.  The three connectors are where the leads to the motor exit.



Eva uses two of these controllers on KillaJoule, although I'm assuming hers are "hot rodded" a bit to increase output.  The controller has to be matched with the motor pretty well or you can get unpredictable performance, never mind the risk of demagnetizing and ruining the motor.  Battery input is via a fuse and contactor (relay) into the back of the controller.



Mounted in place:



Braced underneath (to the motor face):



The space underneath is not wasted: there is a small sump mounted there (unpainted in the photo above) the coolant pump attaches directly to it.  Somewhere I have photos of the coolant loop but it's been over a year since I did the build and they must be hidden inside a sub-sub-directory or something.

[Frank06]

Batteries.  It's all about the batteries...     Actually everything has to work together.  There's no point in having a very sophisticated motor/controller system then try to feed it with flashlight batteries.  Likewise, there's no point in choosing batteries that are a lot more capable than needed for the rest of the system.  The analogy I make is that the batteries are like the fuel supply system in an internal combustion engine, the motor is like the engine and the controller is like the carburetor.  Any one of them can limit performance depending on their individual characteristics.

I went through a lot of investigation before choosing Enerdel, an Indiana-based company.  They make a pretty sophisticated lithium nickel-manganese-cobalt (NMC) cell that is a good compromise between power, energy and safety.  They're used in a lot of commercial applications like hybrid-buses.  They're also the battery used by Lightning Motorcycles.  They come in two flavors: "energy" or "power" modules.

There's a lot of different types of lithium batteries out there today, generically known as "lithium ion".  Lithium Iron Phosphate (LiFePO4) is used in a lot of automobile conversions.  They are a pretty safe chemistry and typical format is prismatic (rectangular) which simplifies installation, although they are not as energy or power dense as other types.  Lithium Polymer (LiPo's) are commonly used in RC vehicles because they are very power dense but they are more susceptible to damage if over or under charged.  LiFePO4's have a nominal voltage of about 3.3 volts per cell and stay fairly flat when discharged i.e. the discharge curve is pretty flat as they give up current, then dives at the end.  Most LiPo's have a nominal voltage of 3.7 vpc and their discharge curve is not as flat.  The Enerdel NMC's have a discharge curve that mimics lead-acid in many respects.  On this bike I "lose" about one volt for each mile traveled.  I generally charge to 4.1 volts per cell for a starting voltage of about 393.  After 50 miles pack voltage will generally be around 345 or so.  It's a pretty convenient way to keep track of how much further I might go.  The trick is to avoid the ends of the discharge curve.  Don't overcharge and don't suck them down too far.

Enerdels are shipped in 12s2p modules.  Two cells are hooked together in parallel then twelve of these are hooked in series in one block.  I purchased eight such blocks or modules.



The aluminum bits at the end are connected to sheets of aluminum separating each cell.  This conducts heat away from the cell for passive dissipation.  Enerdels are pretty unique in that the positive tab (connection point) is at one end of the cell with the negative tab at the other.  Supposedly this keeps heat generation uniform.

With the plastic finger guards removed you can see the busbars which make the series connection.



The cells in each module are clamped together by four insulated aluminum bolts with a nut on each end.  Things are pretty tight in front of the motor.  This is where I want to install the first subpack.

[Frank06]

Because the cells are bolted together I was able to make my own modules using stainless steel threaded rod (1/4") inside carbon tubes to sleeve the rod and protect the cells in case of rubbing or other accidental contact.



Ultimately I ended up with 24 cells directly in front of the motor. The aluminum angle is part of the frame for the next 26 cells.  The cells are connected through insulators to aluminum sheet on the end of the pack.  This is bolted to the subframe.



The next group of 26 cells is fit on top of that.  Not shown: these are secured by a piece of steel bar bolted to the 'Busa frame.



I had to spend some time thinking about where the "most negative" and "most positive" cells would be, and what the optimum wire runs would be.

[Koncretekid]

Frank,
Thanks for the great detailed description of your bike and the power system.  My mind is too full of internal combustion things to comprehend all of what you've described, but maybe in my next life......!  I hope that both Ed (my pit b*tch who is planning to build an electric bike after he finishes drag racing his Triumph triple) and I will get to witness your build next July at Loring for a better understanding.
Tom

[Frank06]

Hi Tom,

I know what you mean!  When I semi-retired here in Maine it gave me a chance to explore EV's a bit more and they've given me an outlet to get my "tech fix" but it's hard to get past ICE's.  There's just so much energy stored in fossil fuels, a shame they won't last forever and that burning them isn't good.  My H2 is on the stand right now for race-ported cylinders, new carbs, etc. and I'll probably redo the crank, the seals are at least 12 years old.  I'm thinking to bring the E-Busa in July and the H2 at the Harvest Event (maybe my RD125 also, another open class, lol.)  It's all fun, life is good.

[Frank06]

We've had a pretty abrupt weather change here so it's time to think about firewood and winter projects.  I decided to up the ante and spring for the motor upgrade and I thought you guys might be interested in output curves.  The motor I'm running now is Series wound and has a ton of torque.  It's wonderful for the street but doesn't have as much power as the Parallel wound.  The parallel motor's torque is smaller but goes out a lot further.  Here's some curves to show what I mean.  There's a lot of lines on them but I don't know how to get rid of what is non-essential.  The torque output is solid red (top-most line) and power starts at zero (dashed red) and climbs linearly upward.

Series:



Parallel:




The series motor has constant torque of about 420 N-m or ~300 ft-lb out to about 2500 rpm, then drops off.  Power climbs linearly to ~100kW or 130hp.  I went through the traps at 140 mph (~5500 rpm), well off peak-power speed.  This chart says power there was only 85-90 hp so hopefully this means my aero is working pretty well.

The parallel motor on the other hand has constant torque of about 340 N-m or 250 ft-lb out to about 5500 rpm.  Power at this point is about 180 kW or 240 hp.  I know that's enough power for 200+ mph but I don't know if I'm that interested in those kind of speeds or not.  I might be able to run a smaller i.e. lighter battery pack for "sufficient" output to keep me happy.

Another advantage of the parallel motor is that the casing is slightly smaller which means I can raise the motor in the frame.  This will raise the output shaft in relation to the swingarm pivot which will unload the idler wheel I'm running.

From a street perspective, the torque numbers are at the countershaft sprocket.  Present gearing is 17/48 so I could make that even steeper to restore rear wheel torque, but I probably won't.  Even at its present "porkey" weight this bike is plenty quick.