One more time: Syncromesh or not. All forward gears are always in mesh in a multispeed box. Which is why I asked about the "syncromesh = sliding" comment. Which is where the gearbox tangent spiraled out of control.
The diff will generate some windage losses whether you are turning or not. I am trying to point out that opportunities for losses in the gearbox are oil seals, windage, bearings, hysteresis in the load carrying gears, side load in the load carrying gear, and losses from meshing the gears. Being as they will be involute teeth, sliding is pretty much a non issue. at least in a FWD application. So taking a few non load carrying gears out will drop some windage, a little bearing losses and some losses in meshing. You still have some meshing losses, hysteresis loss in the loaded gears, side loads in the load carrying gears, oil seal drag and the remainder of the bearing losses. AC motor efficiency is not always 90% or greater. At high torque at low rpm, you can be in the low 60s. How do you keep from doing that? By changing the ratio to increase shaft speed and reduce torque required at the same time. You will spend less time in a inefficient area of operation this way (quicker acceleration) and the losses will be less during this event. Which you will repeat at every stoplight. This more than makes up for the losses of having another gear pair. Now you also have two speeds of most efficient motor operation, not one. And you can determine which gear to use to get that efficiency. And you have a more regions of second best operation, and third best and so on. And finally, when the motor, controller and batteries are all pooped, you can still make it up the driveway. Another point or three: the car you keep referring to, the EV-1 does have a single speed gearbox. It is unclear whether or not it has electrical reconnection on the fly. The gearbox is double reduction, with all those extra parts. Unless it sounds like it is in reverse all the time, then the gears are helically cut. Not spur. Which I think you would find *very* objectionable in an EV. The EV-1 web site says 0-60 in 9.0 or less. I find 0-30 in 1.6 difficult to believe. It would require an average acceleration of 27.5 feet per second per second, or ~0.8G. If 80% of the mass of the car is over the front axle and there is no dynamic effect (unlikely) then the low rolling resistance tires would have to have a static coefficient of friction of 1.0. If it is more like a 60/40 F/R split which goes towards 50/50 under acceleration, then you are talking about a coefficient of friction of about 1.6. I don't think even DOT drag radials would do that. It looks like someone calculated that if you could put peak power in the EV-1 to the ground, you would get to 30 in 1.6 seconds. The math works out if you neglect air drag (it gets more preposterous if you factor in drag.) But AC drives don't do this. The power is limited by IGBT sizing, which limits current, which limits torque, which limits power until you hit base speed. Which is probably in the neighborhood of 2-3000 rpm. It is possible that it is lower, like near 1000 rpm, but that still means the first half of the accel run ramps from zero to full power, at near constant torque. So I find it really unlikely that the impact goes 0-30 in 1.6 seconds. From a power standpoint and the ability to deliver the torque required. It's my opinion that AC is only rarely flexible enough to do the job. And for now I am keeping that opinion. I have never driven an EV-1. I live in the northeast, so I probably never will. I sat in the Impact once, but that didn't tell me much. I don't think I can offer much more on this and keep it either informative or terribly EV-related. Seth
