| running directly from the battery. Output voltage is independent of variations in battery voltage, and efficiency is usually pretty high, so not too much heat generated.
You can design and build one yourself, or order ready-made small PCB or even potted component like this (except this particular one will not tolerate your input voltage). I've made a buck converter for 68V input using the LM7757 in the past. To be honest, these kind of circuits need some care when designing the PCB, especially the ones using higher switching frequencies (>50kHz, I'd say). So you may be better of by buying a ready made design, or possibly an evaluation kit from the manufacturer of the chip.
As for the battery voltage and driving the motor:
- Was the 45V measured under load? Measuring an unloaded battery will usually yield a noticeably higher voltage than when under moderate load.
- Regardless, overcharging a battery is bad for the battery, mkay? So you want to look into a way of preventing that.
If the battery really does deliver higher voltages under load than the 36V, I see basically 2 options:
1. Use another DC/DC converter to bring it down to 36V. You could even opt for a buck/boost type converter, so you still get 36V when the battery drops below 36V (using more current, of course). The thing is, with the currents I expect in an EV, losses in the whole converter may still be pretty high and generate a fair amount of heat. Not quite as much as a linear regulator, but still.
2. Use the raw battery output to the motor controller, and accept that it may sometimes be higher than 36V. Compensate for this by making the "on"-time of each phase shorter, without changing the overall frequency.If you measure the battery voltage, you can even change this on the fly, and so keep motor power constant. (This is essentially what kahuana suggested).
|