Featured Speculative 2-speed EV transmission

Here are some numbers I came up with with a 100% efficient motor with 50Nm torque limit and supplied with 50kW limit and I made it start limiting speed at 5,000RPM. (Rounded to the nearest whole number).

RPM Power Torque
100 5 50
200 10 50
300 16 50
400 21 50
500 26 50
600 31 50
700 37 50
800 42 50
900 47 50
1000 50 48
1100 50 43
1200 50 40
1300 50 37
1400 50 34
1500 50 32
1600 50 30
1700 50 28
1800 50 27
1900 50 25
2000 50 24
2100 50 23
2200 50 22
2300 50 21
2400 50 20
2500 50 19
2600 50 18
2700 50 18
2800 50 17
2900 50 16
3000 50 16
3100 50 15
3200 50 15
3300 50 14
3400 50 14
3500 50 14
3600 50 13
3700 50 13
3800 50 13
3900 50 12
4000 50 12
4100 50 12
4200 50 11
4300 50 11
4400 50 11
4500 50 11
4600 50 10
4700 50 10
4800 50 10
4900 50 10
5000 50 10
5100 25 5
5200 12 2
5300 6 1
5400 3 1
5500 0 0
5600 0 0
5700 0 0
5800 0 0
5900 0 0
6000 0 0

 

difference maker. iow, it has to substantially improve performance. of course, that is subjective.

 

While I wish there were a use case for a multi-speed MANUAL transmission in an EV, about the only use case I can see is to increase top speed beyond that of typical speed limits.

It could help reduce the cost of the motor by allowing for a much smaller motor, but I don't see motors as being the main cost right now, and most people would not want a 30-40hp electric motor coupled to a two speed transmission. It just makes so much more sense to put in a 150hp or better motor that has just as good low end torque as your common car in first gear and that does even the highest speed limits just perfectly fine.

 

My bad there.

Early applications seem to be in improving the sport performance. I believe wider range use will go to allowing the use of cheaper motors for non-sport models.

There is a lot of copper in an electric motor. Which is a more costly material than what is found in a LFP battery. Plus any rare earth minerals used for the car are in the motor's magnets. A smaller motor can mean a sizeable reduction in cost.

 

I guess maybe there's hope.

A reluctance motor runs a lot like an internal combustion engine and uses mostly iron. It has higher torque and efficiencies when not lugged. Add an automatic transmission, or an manual transmission with weird added sounds when it starts to get out of it's RPM band. Maybe even built right, the torque ripple when lugged could feel like lugging an ICE.

 

Personally I have no desire for going faster than about 60mph. If I could choose any car in the world just for the joy of driving this is what it would be:

 

The motor's Torque-RPM graph shouldn't change at all, with or without the second gear.

That is why we need to pick a second reference point, downstream from the gear change transmission. One possible choice is at the wheels. Or, pick the transmission's output shaft itself. Assuming the later, with 1st gear being a simple straight through, while 2nd gear is a 2X overdrive, I get this:



X-axis in RPM, Y-axis as torque in N*m.

Is this a more reasonable explanation? It cuts low speed torque in half, but enables a higher potential speed. But for any given speed that isn't either torque- or RPM-limited, the power output is the same.

 

That looks a bit more convincing.

Part of this is just about all that can be gleaned from a graph just by looking. The concave-up curvy part of the torque graph already has a lot to tell: the curve (extended to its asymptotes along both axes) is the set of all points whose torque and RPM coordinates multiply out to 50 kW.

Any point in the region above and to the right of that curve is a point with power > 50 kW. You can confirm it by doing the multiplication and scaling, but you already know just by looking.

 

^ What is it's output-shaft or wheel-axle torque vs RPM graph? Each gear, superimposed on the same chart?
Does it resemble my graph in post #29?

 

Thank you for fixing the graph.

The only thing that might make that graph a little different in reality is efficiency losses. I understand that that's why Porshe put a two speed on the Taycan.

 

And that lack of resemblance still seems paired with a lack of plausibility, because you have placed ink into the region above and to the right of the 50 kW constant-power curve, where no ink belongs unless you are getting more power from somewhere.

 

Let's see here.

• Say you have two gears, one is 2:1 and the other is 1:1 and we'll go with a 50kW motor. Note that it's pretty easy to make a motor that has a flat power curve.
  • Say the wheel is spinning at 2,000 RPM and you're in 1st gear. The motor is spinning at 4,000 RPM. Therefore max wheel torque is double motor torque.
    • If the max power is 50kW, then the max motor torque is about 12Nm and the wheel torque is times 2, so 24Nm.
  • Say the wheel is still spinning at 2,000 RPM but you shift into 2nd gear. The motor is now spinning at the same speed as the wheel.
    • If the max power is still 50kW, then the max torque is the same for motor and wheel, which is 24Nm. The same wheel torque as in the 1st gear.

Reasons why a motor would make sense:

1. You want to add more speed range to your power curve. If the motor has a max speed of 60mph, then doubling the gear would get you 120mph.
2. You want to reduce inefficiencies. The faster a motor turns the more friction. But to get the same power out of a motor at lower speeds means you have to feed more current through it at lower voltages, which translates into greater ohmic losses. Note that reducing ohmic losses would make it easier to make a motor with less copper in it.
3. You build a motor or controller that can't handle a flat power curve. If you have a voltage limit and you feed the same amount of voltage into a motor, most motors start with high power and lose power the faster they turn. So you could increase the voltage, or have a transmission that shifts up for more power. Oddly this would be kind of backwards from how a traditional ICE needs to be shifted.
4. You build a motor or controller that puposely follows a torque curve similar to an engine. A couple examples would be an induction motor with a constant A/C frequency or a reluctance motor. The main problem with these is that 1) with the induction motor it's easier to just increase and decrese the frequency as needed to maintain a constant output power and 2) the decrease in power in both of these motors is due to increased efficiency losses, which makes this method not at all ideal. 3) So that would bring us back to just programming it into the motor controller with a typical EV motor just so it would feel right with a transmission, which wouldn't be that much different than to just keep it a single gear and make it run pseudo gears.

 

Fair enough. Back in the day it was mainly motor limited. You had a fixed voltage, the battery, and the controller could put out no more than that voltage. So once the pedal was to the metal the motor would go from max current (max torque) to max voltages, and from there the current would decrease, as you said, from back EMF.

Fastforward to today and it's pretty easy to just increase the voltage to counter the back EMF. That way you're only concered with the maximum current. This is mainly the maximum current the battery can produce or that the controller is wired for. So once the motor's EMF reaches a point that the maximum battery/controller current would not be otherwise obtainable the motor controller simply steps up the voltage and keeps the total battery current flowing.

Because of this the battery (or perhaps controller) is kept at max current and at its acceptable battery voltage, which results in constant current and constant voltage and therefore constant power, at least on the battery side. On the motor side the current continues to drop and voltage continues to increase as the motor winds up higher and higher. The power is the same because current x voltage = power, but we also see the same thing happening with torque, speed and power. Torque x speed = power also. So less torque, more speed, same power.

 

This isn't a traditional DC motor.

Constant power curve added:

Both gears' torque vs RPM curves made more visible across entire range.