So we're on the same page here, are you talking about switching in
an up-converter when the batteries are low or are you talking about
leaving it in the whole time?
Both have drawbacks, but different ones.
Leaving the upconverter in will waste your smaller 12V source at a
greater rate than the 36V series pack. This is because there will
be internal resistance from the series pack that won't be found in
the electronic up-converter (depending on how it's designed of course.)
That unit should supply 36V gladly at whatever current you need to
drive the load up to the rated power of the unit, till the pack is exhausted.
This could lead to high current and shortened battery life on your 12V unit.
If, however, you are talking about switching it in only when the 36V pack is low,
then you need to make sure to disconnect the 36V pack or you waste energy
dumping power into the battery chemistry rather than driving your bike.
So now we are talking about double switching the power in and out, more hardware
more weight to drag around.
It appears to me that, no matter how you look at it, the answer is to have a bigger pack
at the voltage you need. YMMV, of course.
Interestingly with batteries once they are flat you wouldn’t want to draw more from them anyway. Could use it as a speed booster on the highways at the cost of wasting more current but the losses would be smaller during peak cruise on flats.
The killer application for a intermittent booster circuit to have switch in is for ultracaps.
I have an ultracap LED torch, unfortunately once the energy drains the voltage drops below the LED threshold and its useless but still about a good two volts sitting their to be had. Higher energy caps could be exploited in such a fashion to tap deep into higher voltages.
The perfect capacitor ESU when below LVC should have a booster to get above LVC.
Capacitors don't suffer from cell death like batteries so if you can find a way top suck them down to 0v go for it. At 0v, they are almost a dead short to a charger. Even a cap that could give 60v@100 farad would unvail some awesome stuff.
Enough about caps.
A booster uses inductors to produce the voltage rise, a motor is an inductor. You can achieve the same result as a booster by altering the inductance of the motor too. It works on the same pricipal but its direct and more efficient inside the motor. DC motors, PWM can do amazing things with inductance.
A circuit board, and electronic switches, PIC, multap motor coil windings. This would be the cream of the crop and ready to play with. The coil tappings could interface with the controller, PWM, sensorless phase angle logic and all sorts. We already have a voltage booster inside the electric motor.
