Electric Motorcycles

eMC
Here's my notes on building an electric motorcycle. I think I've decided by now that it won't be practical for me to build one right now, but it was fun finding out more about them.

Initial Parameters

My initial enquiry to the EV list ...
G'day all,

I just stumbled onto this list a couple of days ago, and I'm still reading back through the archives. Some of the things in the photo album look fantastic! Since the questions about the Z550 have lured me out of hiding I figure I should introduce myself.

I'm a 33yo software hacker from Melbourne, Australia, with a bit of an EE background and an obsession with motorcycles of all kinds. At the moment my main bike is a BMW R100GS, but it's a bit of a handful around town. I'm interested in seeing if an electric motorcycle would be feasible for me to build and use day-to-day.

So far I'm thinking:

Does this sound reasonable?

I've gone through the bikes in the photo album and going by the stated ranges and battery configs, the efficiency seems to vary enormously ... between 50 Wh/km and 400 Wh/km! Anyone know a good rule of thumb for a two-wheeled electric beastie?

I'd need to get a 'regular' range of maybe 40km (25mi) and a 'regular' speed of 60km/h (38mi/h) out of it to make it a practical bike, and maybe occasionally a 60km (38 mi) range or a 80km/h (50 mi/h) top speed.

Is this practical with the available tech, or will I just have to wait for fuel cells or unobtainium batteries or something?

Thanks!

-----Nick 'sharks' Moore

Some investigation

Wh/km

I found a whole heap of data on electric bikes on the Electric Motorcycles Photo Album, and tried to use this to work out what kind of Wh/km to expect. As you can see from the graph and table below, there's quite a lot of variation:

Battery Range Speed Drain Name Batteries Motor Final Drive
V Ah Wh km kmh Wh/km Type Ratio
48 22 1056 51-64 64 29 Jackal SLA ETek Chain
48 65 3120 32-51 77 76 EMB Lectra Optima D750S Etek Chain 3.375
48 55 2460 48 80 55 E-Scramler PowerSonic 55Ah Advanced DC Chain 2.53
48 40 1920 40 81 48 250XL SLA 40Ah Etek Chain
48 50 2400 30 86 80 Xineta Hawker 12V 50Ah Cupex (perm) Chain
48 44 2112 24-40 88 66 ReCycle 22AH SLA (paired) Etek Chain
36 115 4140 32 88 129 ElectroCycle trojan 12V TMH Prestolite Chain 6
72 41 2952 6-13 100+ 307 BEM Optima YT D51 Advanced DC Chain
60 55 3300 32-40 100 91 ElectricNinja 55AH Etek Chain
72 100 7200 27 104 265 Piggy Trojan 100AH Advanced DC Chain 4.6
96 42 4032 24-32 128+ 144 REVision Hawker Genesis Advanced DC Shaft ~3

Of course, this is total battery Wh, not actual Wh used to travel that range. There's a huge variation in claimed efficiency though ... perhaps different depth-of-discharge (DoD) strategies, different motors, different gearing, aerodynamics all add up.

Choosing Batteries / Optimizing Optimas

I'll probably be stuck with the lower lines on the graph, eg: Optima YTs or Federal Gel-Tech, because the deep-cycle types don't offer enough peak amps. Such is the way of engineering tradeoffs! Found some good numbers of Optima YTs discharge cycle, and it seems like about 40% DoD should give the best lifetime in km.

Practicalities

Okay, let's assume for the sake of the argument I can get 100Wh(cap)/km ... assuming a 50% DoD, that's really a consumption of 50Wh/km, at say 50km/h that's 2500W, which sounds about right.

So for a ~35km range I'd need 3.5kWh off battery capacity ...

# Type Capacity Weight Size ea Price Retail Value
V Ah(20hr) Wh(cap) CCA Peak kW kg mm AUD Wh / $
4 Optima YT D31 48 75 3600 900 43.2 108 325*165*238 $1840 1.9
4 Federal Gel-Tech 8G24 48 73.6 3530 410 19.7 97 276*171*251 $1200 2.9
5 Optima YT D34 60 55 3300 750 45.0 97.5 254*178*198 $1925 1.7
6 Federal Gel-Tech 8G22NF 72 50 3600 245 17.6 103 238*140*235 $1434 2.5
14 Thunder Sky LiIon TS-LMP80 50.4 80 4032 ?200? ?10? 42 220*145*61 ?$2000? ?2?
14 Thunder Sky LiIon TS-LMP200 50.4 200 10080 ?200? ?10? 95.2 285*182*100 ?$5000? ?2?

... fitting them into a bike-sized vehicle won't be easy. Cardboard models seem the way to go. I've added in LiIon battery options because they might be a good option ... they seem to be about the same $/Wh but half the kg/Wh. On the down-side, they require a per-cell Battery Management System. See Victor's CRX under 'Links:' below ...

Motor and Drivetrain

The simplest setup is likely to be the permanent-magnet Etek M7 from Briggs & Stratton. They're pretty widely used in the electric motorcycle world. There's quite a bit of info on the Etek at robotcombat.com and currietech.com.au (scroll down) and it looks like it'd be easy enough to mount in a motorcycle application.

The Etek will spin at 72RPM/volt, so at 48V that's a top speed of 3456RPM. A motorcycle wheel&tyre is about (hand wave) 2m in circumference. Front sprockets go down to about 13 teeth, and rears up to about 57, so we'd have a max speed of 3456RPM * 13/57 * 2 m/R = 26 m/s, or about 94 kmh flat chat. Sounds about right, perhaps a little high, but to get a higher final drive ratio would mean adding an idler shaft or similar, and I'd rather not. On the other hand, if you were going to add an idler shaft, you could add it coaxial to the swingarm pivot, to keep the chain/belt lengths constant.

Motorcycle sprockets are generally splined to fit a shaft, but it's probably easier to machine a sleeve keyed on the inside to fit the Etek output shaft and stepped on the outside to make it easy to braze/weld the front sprocket to the sleeve.

Regeneration!

One of the disappointing things about a permanent magnet setup is that it makes it hard to do regenerative braking. Since motor voltage is proportional to motor RPM, the only time you'll get a voltage high enough to recharge the batteries is revving the motor beyond its normal maximum RPM ... and on a direct-drive machine, that'd only be when rolling down a hill quicker than the bikes top speed ... not likely to be a common occurance ...

One way around this is to use a "Contactor controller", or series-parallel switch. Put simply: split the pack in half, and wire these halves up so that they can be connected in series or parallel. This provides two 'gears' ... in parallel, maximum current is increased for more torque, and in series maximum voltage is increased for more top speed.

Serial vs. Parallel

Pretty obviously, fuses would be needed to prevent the batteries exploding if a contactor stuck or was closed at the wrong time.

And, of course, dropping down into Parallel reduces the battery voltage, so if you've sped up to top speed in Series and then shift into Parallel, there's sufficient voltage to regeneratively charge the batteries. It's not exactly sophisticated, but it's a start. The idea can be extended to more battery packs as well: eg with six packs, you can have 1,2,3 or 6 batteries in series, or to put it another way 6,3,2 or 1 packs in parallel.

Just as the selection of series vs. parallel is a bit like the selection of gears in a gearbox, the two likely strategies are manual or automatic switching. For manual switching, a control would select which contactors to close, giving a selection "Parallel-Neutral-Series" just like 1-N-2 on a motorcycle gearbox. The throttle setting would provide PWM control over power and regen. For automatic switching, a controller would monitor throttle position, battery voltage and motor voltage, and select the appropriate 'gear' and PWM setting.

Old Shopping List

Kawasaki GPX250 or similar rolling chassis
inc. minor repairs
trading post? $500
Chain & Sprockets (motorcycle) bike shop $250
KIT: Etek M7, Alltrax controller, charger, throttle currietech.com.au $2400
4 * Optima YT D31 ? $2000
Motor mount plate & front sprocket carrier & battery cages custom $300
Sundries: (lights, paint, shrinkage)   $500
TOTAL: $5950

... which, sadly, probably puts this project out of my league. Compared to buying a decent GPX250 for $3000 and $3000 worth of fuel, it'd take perhaps 40,000km to break even ... and given the range (40km or so), that's 1000 discharge/recharge cycles (perhaps 3 years) and by then the batteries are probably half shagged. Oh well. It'd be a nice toy :-).

New Shopping List

On the other hand, if the batteries and controller could be gotten cheaper, perhaps by sponsorship on the batteries and by importing parts directly from the States (I've seen Etek+controller for US$700), perhaps it'd break even ... perhaps this explains why most of the E-vehicles seem to be from the States, even though fuel is cheap over there.

The constant torque of a DC motor, as compared to the jerkiness of a ICE, probably means I could get away with a light chain. Belt drive is another option, but finding matching ratios may be a problem. Actually Cloud Electric have #40 sprockets down to 11 tooth (on a 7/8" shaft) and up to 60 teeth on a four-hole rear sprocket. So ratios up to about 5.5:1 might be possible.

Kawasaki GPX250 or similar rolling chassis
inc. minor repairs
trading post? $ 500
Chain & Sprockets (#40) Cloud Electric $ 100
KIT: Etek M7, Alltrax 300A, etc Cloud Electric $1200
4 * Gel-Tech 8G24 Solar Online $1200
2 * 24V 6A Smart Chargers Jaycar $ 200
Motor mount plate & battery cages custom $ 200
Sundries: (rattlecan paint)   $ 100
TOTAL: $3500
... which makes it look a lot more tempting ... hmmmm ...


Two Crazy Hybrid Ideas I Had

I posted these to the EV list in July 2005 ...
One idea I was thinking about ages ago was to machine off two cylinders of an inline four and attach a big electric motor in their place, straight onto the crank. You'd end up with a small (225 in your case) parallel twin. With a regenerative controller, the motor could do a lot of the work when taking off, and the engine could then take over to recharge the motor ... you'd probably need fly-by-wire EFI to hide this behaviour from the user though.
(eg: you could take a CBX750 engine and do something terrible) to it.

Here's another crazy idea [for turning a bike engine into a hybrid]: the BMW airhead motor uses a dry clutch, so there should be very little friction with the clutch disengaged. And the clutch spins at engine RPM ... up to about 8000rpm. If you made a spacer to fit between clutch and gearbox, and connected your electric motor in there (mount it off to one side with a belt drive or something?), you could just pull the clutch in to disengage the engine, and your motor could drive through the gearbox. And since it's already a shaftie, you don't need to muck around with final drives for your [tadpole trike], just use the BMW single-sided swingarm and bevel final drive for your back wheel.

The 'airhead' motors vary in size between R45 (450cc) and R100 (1000cc). They're aircooled, opposed-twin pushrod engines, although the bigger ones have oil coolers too. For a hybrid, you could make the motor a good deal smaller by removing the alternator and starter motor ... after all, your main motor will do both those jobs.

(I don't know as much about BMW K-series motors, but it's possible they'd be even better for your purpose, as they're watercooled and more fuel efficient than the old airheads.)

(it'd look something like this)

Much Simpler ... but not terribly dignified ...

On the much simpler end of the scale, I'm very tempted to build a scooter around a Golden Motor Powered 12" wheel ... basically just that and a plank (and probably a pivoting front wheel on a T-bar). And a whole bunch of (30+, 9Ah) NiMH D-cells or a LiIon pack or something similar underneath. If built around a HSB-36, top speed would be about:
317mm * π * 333rpm = 6.00 m/s = 21.5 km/h
and max power with torque/speed:
17.68 Nm / .158m = 112 N = 11.4 kgf
317mm * π * 263rpm = 4.36 m/s = 15.7 km/h
At which point it'd be drawing something like 20A, so to make it useful the power pack would have to be biggish. Might make a nice "first EV" though ... and you could always carry it in the boot as a spare!

It's interesting to think about range: at max speed (21.5 km/h) it's drawing about 5A, so assuming a 10Ah supply would go about 43km. At max torque, it's drawing about 20A and only going 15 km/h, so its 10Ah supply would go about 7.5 km ... about 1/6 as far!

This 10Ah NiMH battery pack has 30 D-cells and is 162 * 97 * 45 mm and 5kg.

(If I put skateboard trucks or similar on the front, it'd technically be a leaning tadpole trike :-). Yes, I know there's already plenty of these kind of things on the market ... but it'd still suck less than a Segway.)


Links

Other People's EVs:

EV Mailing List(s):

Australian Suppliers:

International Suppliers:

High-tech doodads:



homepage