New update on the pulse motor.
Nice work Tom! Looking pretty good.
On the hall sensor, I'm guessing you have the Arduino looking at the signal
trailing edge instead of leading edge due to the polarity coming out of the
comparators. At least that's what it looks like from my angle. If that's the
case, I can see why the length of the magnet is determining your timing.
I had a crazy idea about collecting back EMF when the coil is triggered
bipolar. You did a lot of work in this area, so are probably two steps
ahead of me, but I'll try to describe it so others can chime-in if they want
to.
Suppose you connect a full-wave bridge rectifier to your coil and then
take that output to a cap bank. When the caps are empty initially, they
will suck a lot of juice from the coil that is trying to actually run the
motor. I got that much and would expect that behavior. But what happens
when the cap bank gets charged up to the voltage level you are powering
the motor at? Seems to me like the motor would no longer even notice the
cap bank anymore, like it's not even connected. But the cap bank would
still collect back EMF, but what you have to do different is not allow the
cap bank to get discharged below the voltage you're powering the motor
with. Let me toss some numbers out as an example:
Let's say the motor is being driven with 12 volts. The cap bank starts
at 0 volts. We switch on and the motor lugs until the cap bank gets
to 12 volts. From there the motor starts to run normally and the cap
bank juices up to say 24 volts. Let's keep the ground for the cap bank
completely separate from the ground of the motor supply. And let's
bring in another 12 volt power supply (or battery) and connect its
ground to the cap bank ground. Now between the separate power
supply positive and the cap bank positive we can attach a dump
load. Because of that separate power source, it's impossible to drop
the voltage on the cap bank below 12 volts. If we leave the dump
load and separate power supply connected, by way of the dump load
the cap bank will charge up to 12 volts and we can start the motor
without it even lugging initially.
Efficiency-wise this may be a bad implementation, but the concept is:
we can collect back EMF from a bipolar fed coil, we just have to be
careful not to dump the storage of it below the voltage driving the coil.
I have a hunch since the cap bank voltage will be higher than the separate
battery, it will actually end up charging the separate battery every
time we dump it. That could be a good thing, since we can power
a load and charge a battery at the same time. We just have to pay
initially to energize the coil. Maybe with some kind of battery swap
relay system this thing could run for quite a long time.
Anyway, just a crazy thought I had. Maybe you can make something
of it.