AC upgrade
From RC-Autopilot Wiki
>= Designing an AC drive for the Microcar =
Contents |
Motor
Together with a large German motor manufacturer, I'm working on a 15kW async 3-phase motor, wound for 33V RMS (for 48V battery voltage) with nominal torque of 24Nm upto 6000rpm (max torque 50Nm upto 3000rpm), max speed 12000rpm. See the datasheet in the graphic to the right (it's in German, but surely you can see whats what). The aluminium motor weighs 35kg and is 35cm long by 21cm diameter. I plan to connect it directly (shaft to shaft) to the Microcar's stock gearbox, providing 400Nm torque to the wheels.
Update: After several weeks of negotiations, I've put aside those plans due to a sudden change in the tone of the supplier (from "no problem, we can do it" to "we have to do lots of research first, pay us 8000 EUR upfront before we even try"). Instead I now (29.01.08) got a cheap 200 EUR 4kW (1400rpm) Async motor and I'll try disassembling and re-winding it myself. According to the very interesting (German) page http://www.energie.ch/at/asm/index.htm a 3kW motor with round squirrelcage bars and low flux distortion ("Streuarm") can deliver 100Nm at 1500rpm (and more) when run with a variable-frequency inverter, which is exactly what I'm planning to do. So ideally I can:
- re-wind this motor from 400V to 48V by reducing the winding turns by about 1:10 (and increasing the wire diameter by 10:1)
- replace the squirrel-cage bars, which are probably waterdrop-shaped or rectangle, with round ones
- somehow make the magnetic field more concentrated and less strewn (although I have my doubts that will be possible with a pre-fabricated motor)
According to the formula P = 100 Nm * 1500 U/min * 1 min/60 s * 6,28 = 15700 Nm/s = 15,7 kW --> meaning the 3kW motor can output 5 times the rated power!!! At 90% efficiency that would be 1.57kW heat loss, so the motor must be cooled well.
Photos of inside:
And during removal of the existing coils:
Performance
According to my calculations (see graphic to the right), the new drive will make the car reach 30km/h in 3 seconds, 60km/h in 7 seconds, and 100km/h in 17 seconds. The top speed (limited mainly by air friction) will be just under 120 km/h. Since I dont have the exact numbers, I'm estimating the Microcar wind values at 0,35 cW and 1,5m² front projection.
If my friction calcs are correct, cruising at 80km/h uses about 5,5kW, or just under 10kW at 100km/h. That would mean you could drive at 100km/h for an hour, and have a range of 100km with 10kWh/100km (about 1,50 EUR) milage. Or drive at 80km/h, using 5.5kWh/80km or 6,9kWh/100km (1 EUR/100km), almost 150km range. Or go a super-saving 60km/h using only 3kWh/60km for 200km range. If you like to floor it, at max speed (about 115 km/h) you'll still get 75km range, plenty for most daily uses.
As for hill climbing, it will climb a 10% incline with over 70km/h, reaching 70km/h (from standstill) after only 15 seconds. Even up 20% it can still do almost 50km/h. (But only when lightly loaded with just one driver, a total weight of 500kg). Loaded up with 700kg? Then you can still climb 15% at 45km/h, or 10% at 60km/h, which is better than many large trucks.
400V design
Another approach could be to use a standard industrial AC motor controller (Frequenzumrichter) which can cost as little as 500 EUR for a 15kW version, which is made for a 3-phase 400V input but can also take a 560V DC (400V*1.41) input. So the battery voltage must be 560VDC (due to different loads it will vary between 500 and 600V). The advantage is the lower currents and lower losses along the way, thinner cables, and theoretically we could use standard 15kW motors off the shelf, without any modifications. These would not run at the highest possible efficiency, but they are cheap (<1000 EUR). But on the downside, the BMS for a 560V battery pack would need to monitor up to 175 cells (for example using 3.2V LiFePo cells) or 151 3.7V LiCo cells. This and the mere effort of wiring up 175 cells in a battery pack will raise the price significantly. Probably only feasible when using passive cell protectors that bypass current on their own when charging.
