jackatfsi
10 W
Still a newbie at this w/ biases for hub motors and 12v battery packs......
Anyway the idea is to get as much useful energy out of the battery pack and into bicycle motion as possible.........
The most efficient transfer of power occurs when the source impedance of the battery matches the "sink" impedance (load) of the motor. The impedance of the battery is (usually) DC (non-complex..in the engineering sense) while that of the motor is both complex (ie. contains inductive and capacitive components......note here that the "filter cap in the controller is part of the load)
Since the motor impedance (load) is usually higher than that of the battery (source) the oldest method of matching was to put a variable resistor (rheostat) in series w/ the battery to raise the source impedance (and just throw away the energy lost as I^2*R in the rheostat as heat)
The early transisorized regulators worked the same way...the heat went into the transistor "heat sink".
Then someone realized that if a "good" transistor was operated as a switch, then negligable power would be lost when the switch was open (I = 0) and also when the "switch" was closed (R = 0)......
The simple-minded way to think about this setup is to ignore the AC component of the switched source and to ignore the complex impedance of the sink (motor)....namely the RLC circuit composed of the motor brush and winding resistance, the motor inductance, and the "filter" capacitor.
Oh yes, and if you don't just ignore them, the AC component will either match or not match w/ the RLC circuit. Mismatch = some loss, match = some additional power transfer to the motor (circulating currents and all that)......what "safe" calls the "PWM Effect", I believe (for the moment anyway)
To continue....newer "inverter" type controllers used in welders and LED drivers etc etc......use a high quality inductor and capacitor circuit w/ an integrated "digital signal processor" (DSP) controlling both the frequency and the "on time fraction" of the switch to optimized the DC output impedance of the source. I said "DC" because now, the brushed DC motor a least, can now be represented only by it's non-complex impedance (unless the load is varying at an unusually high frequency..."corduroy road" ???) This makes the spreadsheet formulas a lot easier.........
The things I like best about this type of controller are:
I can get rid of the relays I'm using between the batteries for series parallel switching. (principally for ease of charging)
The "inverter" can continuously change the voltage down or up (within the capabilities of the components, of course) That is, the "PWM Effect" can always be being used at its maximum.....
The parts count goes down......
At the moment I'm chasing down the actual parts required for a homebrew unit and will report back.......
Anyway the idea is to get as much useful energy out of the battery pack and into bicycle motion as possible.........
The most efficient transfer of power occurs when the source impedance of the battery matches the "sink" impedance (load) of the motor. The impedance of the battery is (usually) DC (non-complex..in the engineering sense) while that of the motor is both complex (ie. contains inductive and capacitive components......note here that the "filter cap in the controller is part of the load)
Since the motor impedance (load) is usually higher than that of the battery (source) the oldest method of matching was to put a variable resistor (rheostat) in series w/ the battery to raise the source impedance (and just throw away the energy lost as I^2*R in the rheostat as heat)
The early transisorized regulators worked the same way...the heat went into the transistor "heat sink".
Then someone realized that if a "good" transistor was operated as a switch, then negligable power would be lost when the switch was open (I = 0) and also when the "switch" was closed (R = 0)......
The simple-minded way to think about this setup is to ignore the AC component of the switched source and to ignore the complex impedance of the sink (motor)....namely the RLC circuit composed of the motor brush and winding resistance, the motor inductance, and the "filter" capacitor.
Oh yes, and if you don't just ignore them, the AC component will either match or not match w/ the RLC circuit. Mismatch = some loss, match = some additional power transfer to the motor (circulating currents and all that)......what "safe" calls the "PWM Effect", I believe (for the moment anyway)
To continue....newer "inverter" type controllers used in welders and LED drivers etc etc......use a high quality inductor and capacitor circuit w/ an integrated "digital signal processor" (DSP) controlling both the frequency and the "on time fraction" of the switch to optimized the DC output impedance of the source. I said "DC" because now, the brushed DC motor a least, can now be represented only by it's non-complex impedance (unless the load is varying at an unusually high frequency..."corduroy road" ???) This makes the spreadsheet formulas a lot easier.........
The things I like best about this type of controller are:
I can get rid of the relays I'm using between the batteries for series parallel switching. (principally for ease of charging)
The "inverter" can continuously change the voltage down or up (within the capabilities of the components, of course) That is, the "PWM Effect" can always be being used at its maximum.....
The parts count goes down......
At the moment I'm chasing down the actual parts required for a homebrew unit and will report back.......
