jonescg
100 MW
Many folks across several forums have expressed concerns about the high voltage system I plan on using for my electric race bike. Your concerns are not unfounded, as anything over about 300 V DC is dangerous territory. But there are a few good reasons why I have gone this way, and rather than reply to every post around the shop, I hope to explain to you why I made the choices I did.
Motor Power.
I want lots of power for two reasons - more power means faster acceleration and more fun, and secondly, it means even if I only use a fraction of the available power, I won't risk incinerating a rather expensive component. The most power dense motors out there today are permanent magnet AC synchronous motors. Induction motors are bigger, heavier and bulkier, and they all tend to be radial flux which limits how they fit in a bike. So the options were EMRAX, Evo, Carbon, YASA and UQM. All have their upsides and downsides, but I settled on the Evo because it was cheaper than a YASA, more powerful than an EMRAX, smaller than the Carbon. UQM was another option but they effectively stopped selling one-offs to people like me.
The Evo came in two sizes, the AFM130 and the AFM140. Both of these came in two viable winding configurations. Given peak power from the AFM130 wasn't quite enough to sustain the racing duty cycle, I opted for the AFM140 which is 40 kg in weight and 380 mm in diameter.
So with the motor size selected, I needed the right windings. Malte from Evo suggested the #4 winding would be better as it would have more inductance, use lower currents (due to the higher voltages) and generate less heat:
OK, so the inverter becomes the big issue - the motor is useless without an inverter supplying it with power. Lots of inverters are out there, but I had to settle for one which delivered enough volts to the motor. In the case of the Evo AFM140-4, it needs 565 V AC rms to get to top speed (5000 rpm). Thanks to field weakening, it can achieve the same thing at lower voltages, so the "base" speed is about 4000 rpm, or 450 V AC rms.
How do you produce 450 V AC rms? From an inverter running a DC bus of sqrt(2) times the AC voltage, or about 678 V DC. The Rineheart Motion Systems inverter, PM150D- series can operate from either 360 V DC Bus maximum (PM150DX), or 720 V DC (PM150DZ). In the case of the AFM140-4 motor, the high voltage inverter is needed to achieve a respectable speed.
So how about a motor which can be spun up using a lower DC Bus? Well the PM150DX would be able to produce a maximum motor voltage of 240 V AC rms from a maximum battery voltage of 360 VDC. The #3 wound Evo still needs 424 V AC rms to achieve top speed, but with field weakening it could be achieved at ~240 V AC, but at what current? The PM150DX is limited to 450 A peaks, or 109 kW. Wasn't this inverter supposed to be able to deliver 150 kW? Where did the power go? How about a Tritium Wavescupltor 200? Would it work with an Evo motor? Well yes, but it can't seem to employ field weakening like the RMS inverter, to the top speed and ultimately motor power would be restricted to about 100 kW max.
In the end, I needed a motor which could deliver 80-100 kW continuous. In racing, continuous power is what matters, since you are on the throttle all the time. Getting the battery, inverter and motor package right was my biggest conundrum. Either it couldn't handle the DC bus required, the correct winding wasn't suitable, or the motor wasn't powerful enough.
The AFM140-4 was the best motor option, as it could be driven by an RMS inverter running the maximum DC voltage limit allowed in the rules - 700 V top of charge. Oh, and trying to find a voltage which relies on a number of cells divisible by 4 with the resulting product being an odd number to satisfy the terminations... The combination of all of these factors lead me to believe that the option I went for gave the best power for the least compromise, at the expense of a high voltage.
Motor Power.
I want lots of power for two reasons - more power means faster acceleration and more fun, and secondly, it means even if I only use a fraction of the available power, I won't risk incinerating a rather expensive component. The most power dense motors out there today are permanent magnet AC synchronous motors. Induction motors are bigger, heavier and bulkier, and they all tend to be radial flux which limits how they fit in a bike. So the options were EMRAX, Evo, Carbon, YASA and UQM. All have their upsides and downsides, but I settled on the Evo because it was cheaper than a YASA, more powerful than an EMRAX, smaller than the Carbon. UQM was another option but they effectively stopped selling one-offs to people like me.
The Evo came in two sizes, the AFM130 and the AFM140. Both of these came in two viable winding configurations. Given peak power from the AFM130 wasn't quite enough to sustain the racing duty cycle, I opted for the AFM140 which is 40 kg in weight and 380 mm in diameter.
So with the motor size selected, I needed the right windings. Malte from Evo suggested the #4 winding would be better as it would have more inductance, use lower currents (due to the higher voltages) and generate less heat:
Malte from Evo said:You are also right that you could potentially use a AF-140-3 machine instead. This would mitigate the need for field weakening but would have two big drawbacks: Firstly, the inductance of the machine is much lower, which translates into a choppier ac waveform and, consequently, magnets which will overheat when operated at higher speeds for a longer period of time. This is difficult to quantify but the risk is very real. Once demagnetised, you will see an irreversible loss in performance, i.e. torque and power will drop significantly. Second, since you have a lower voltage constant on the AF-140-3, you also have a lower torque constant, which means that you need more current to produce torque. For example, the AF-140-4 needs about 350Arms to produce 600Nm whereas the AF-140-3 needs 465Arms to achieve 600Nm.
OK, so the inverter becomes the big issue - the motor is useless without an inverter supplying it with power. Lots of inverters are out there, but I had to settle for one which delivered enough volts to the motor. In the case of the Evo AFM140-4, it needs 565 V AC rms to get to top speed (5000 rpm). Thanks to field weakening, it can achieve the same thing at lower voltages, so the "base" speed is about 4000 rpm, or 450 V AC rms.
How do you produce 450 V AC rms? From an inverter running a DC bus of sqrt(2) times the AC voltage, or about 678 V DC. The Rineheart Motion Systems inverter, PM150D- series can operate from either 360 V DC Bus maximum (PM150DX), or 720 V DC (PM150DZ). In the case of the AFM140-4 motor, the high voltage inverter is needed to achieve a respectable speed.
So how about a motor which can be spun up using a lower DC Bus? Well the PM150DX would be able to produce a maximum motor voltage of 240 V AC rms from a maximum battery voltage of 360 VDC. The #3 wound Evo still needs 424 V AC rms to achieve top speed, but with field weakening it could be achieved at ~240 V AC, but at what current? The PM150DX is limited to 450 A peaks, or 109 kW. Wasn't this inverter supposed to be able to deliver 150 kW? Where did the power go? How about a Tritium Wavescupltor 200? Would it work with an Evo motor? Well yes, but it can't seem to employ field weakening like the RMS inverter, to the top speed and ultimately motor power would be restricted to about 100 kW max.
In the end, I needed a motor which could deliver 80-100 kW continuous. In racing, continuous power is what matters, since you are on the throttle all the time. Getting the battery, inverter and motor package right was my biggest conundrum. Either it couldn't handle the DC bus required, the correct winding wasn't suitable, or the motor wasn't powerful enough.
The AFM140-4 was the best motor option, as it could be driven by an RMS inverter running the maximum DC voltage limit allowed in the rules - 700 V top of charge. Oh, and trying to find a voltage which relies on a number of cells divisible by 4 with the resulting product being an odd number to satisfy the terminations... The combination of all of these factors lead me to believe that the option I went for gave the best power for the least compromise, at the expense of a high voltage.