Tesla Roadster's 98% efficient transmission nominated for innovation award
 

Found this interesting tidbit:

http://www.blogcdn.com/green.autoblo...rdrive-630.jpg

(Image source: Borg Warner announces new single-speed eGearDrive transmission for Coda electric sedan — Autoblog Green)

BorgWarner, the designer of the gearbox for Tesla's roadster has been nominated for the "prestigious 2010 Automotive News PACE Awards, recognized around the world as the industry symbol of innovation."

(Here's a link to past PACE Award winners & finalists.)

98%!

Something to be said for a single speed transmission, for sure.

(Makes me want to open up the ForkenSwift's gearbox and get rid of all those extra wasteful gears that are uselessly meshed together all the time.)

Source: BorgWarner's eGearDrive(TM) Transmission Named 2010 Automotive News PACE Award Finalist | Reuters

I'd call it a gearbox rather than a transmission, since tranmissions in automotive use generally imply multiple speeds. For a single set of helical cut gears, 98% is good, but not tremendously higher than average.

The gearboxes we use in my industry tend to be about 80% efficient and that is good. But those are worm gears.

See? I've been advocating for fewer gears instead of more for a long time now...

That's good context. Any idea what average would be?

The Solectria Force Geo cars also all used custom single speed gearboxen.

I couldn't really get away with it in the low power ForkenSwift... well maybe I could (the Citicars were all single speed 48 volts). But I like to have the ability to downshift to keep amps low when cruising (or climbing very steep stuff at low speed), and upshifting when I want more power.

I actually do use all 5 gears - but mostly 2 & 3 with a bit of 4 once in a while. I could probably get away with just 2, 3 & reverse. Wonder what kind of difference it would make...

You'd like the Golf twin-drive hybrid then. One forward gear. Electric up to about 30 mph, then diesel power (and/or blended ... I forget) from there up to top speed.

I guess ForkenSwift and E-Sandra are like apples and oranges: big motor and 48v vs. itty-bitty motor and 72v, but I find myself using gears MORE than I do with an ICE! I go from 1st to 4th to get to about 25 mph. I could certainly stay in one gear, but it would be much slower to get there. I attribute that mostly to my throttle only going to 1970 Ohms, but just the throttle limit (I would guess anyway) wouldn't change the need to shift, would it?

Oh, I think it definitely would. You need amps to accelerate, and need to be able to open up the pot to get the amps.

The fastest way for me to accelerate from a stop to 30 mph is to put it in 3rd gear and floor it, then switch to 4th when it starts to run out of steam around 25 mph. More or less. Pulls very battery-unfriendly amps when I do that though (200-300).

Well I didn't get my amp gauge from China yet, so it's all conjecture for me:p, but my guess is I am attempting to overcome lack of amps by upshifting to draw more amps. Accessing more throttle would then make it so I would need to shift much less. Makes sense:)
Also, when I get to 4th is when the voltage drops the most (down to about 58 volts) which I'm guessing equates to more amps drawn? I really should know more about this stuff before I converted the car:p

I think your conjecture is probably right.

Don't worry - I didn't know anything either (lots I still don't know). It was an "immersion" learning experience.

I'd agree. For the gearboxes we make (helical gearing), we use a 1.5% loss per gear mesh. That is our average.

1.5% loss per mesh. I really should think about ditching 1st and 5th then: 3% improved trans efficiency... And if I go ahead with the taller 5th for the Flea... guess what... I'll have an extra shaft I can remove gears from (though in terms of wear, it'll have unmatched gears on it).

I believe the gears must be engaged to actually be loosing that much. Obviously, if they are just churning oil you have additional losses.

When I had my 65 GTO I mostly only used first and fourth gear. I'd take first up to 3500 rpms and shift into fourth. It would drop down to 1,600-1,700 rpms. Got to love a close ratio four speed and a big V8 with lots of torque. :D

First gear at 5,500 rpms would take me to 55 mph. Fourth at 55 mph I would be running 2,750 rpms. Accelerating up to speed on an interstate on ramp there was no need for second or third gear. I would just hit 3,500 in first and skip to fourth and continue up to 55 mph. :p

That is the advantage to an electric motor also instant torque at lower rpms.
There is no need for a multi gear transmission or at least go back to a three or even two speed transmission. With a two or three speed transmission you can optimize the power to speed ratio so your not running the electric motor too slowly forcing it to pull more power than is really needed.

Wasn't Tesla trying to get a transmission with 2 forward gears to work? Did they completely give up on that?

telsa had a 2 speed to begin with, but had issues with it breaking the gear box. Thus the one speed design came second; although they wanted a one speed, implementing it by production time wasn't happening, so the "simpler" 2 speed was used as an interim.

Multi geared boxes usually make up for the loss in efficiency by having better acceleration. 1 gear boxes are rare. With electric and reversible drive 1 speed is all you need for 55mph or less so long as the motor isn't too small. But even electric motors have their peaks. The 2 speed Tesla box was supposed to give them a better top speed. Electric motors produce a lot of torque for their horsepower. Torque is the force that blows gears through transmissions and snaps axles.

Run-of-the-mill 5 Speed manual trannies with OD are 85%-88% efficient. Borg-Warner is probably the largest manufacturer of transmissions. My personal favorite is their T-56 :D. One speed transmissions are there just to give a better ratio than the final drive can provide.

I noticed the default on the EM performance calculator is 95%.

I've been under the impression it's more like 85%, what with all those meshes and all.

Haven't found a really solid number though because this sort of thing is harder to pin down than it would seem.

If I'm understanding what I'm reading, with an EV, it's not about the transmission, but about the controller. The EV1 had a 1 speed transmission, and maintained full torque through the power band because of the controller. Do any of the DIY conversions have this kind of controller technology available, or is that the magic bullet that they don't want to let out of the box?

I don't think I posted in this thread, I think I missed it altogether, but when I was into dirt track racing, one of the first things we'd do with the Borg/Warner T5's was remove all but 2/3. Didn't need the rest, only needed 2 to start, and 3 after the first corner.

I know 85% is what most RWD guys use on the dynamometer numbers and it should be proven many times over by now. With 1.5% per gear it does sound about right too, 7.5% loss at the transmission for a 5 spd and 5% at the differential. That leaves 2.5% to other things like axles, wheels, etc. I've heard comments that suggest FWD only lose 12.5% but I haven't seen anything proving or even discussing it. A driveshaft for a RWD car can weigh 20lbs for Steel and 11lbs for Aluminum, which is roughly the weight of a forged wheel with no tire. I also know with AWD you get two more differentials and the loss is roughly 25% to all wheels.

Well, a trans doesn't HAVE to engage all the gears at once. I recall making something like this from my lego days, might need computer shift assist to not break teeth, but more efficient design anyway. If it is on a bike (or car) with chain reduction then some way to have direct drive in high gear with no meshing would also be beneficial.
http://ecomodder.com/forum/attachmen...1&d=1261335074
(found this messing around with google sketch) envision the output on a sliding splined shaft.