kin> "So, as I believe it, you are suggesting to connecting both LiPO4 and LiPo ground together, and that goes to the black of the controller batt input. Then, a schottky diode goes facing "forward" out of the LiPo battery, at which point it connects to the LiPO4 battery positive, and the line continues to the controller.
So the operation will be that the LiPo battery, which starts charged at a lower voltage, will lose .2v across the schottky but otherwise contribute to the bike. Meanwhile, the LiFePO4 battery can't dump into the LiPo battery because although things can go "out" of the Lipo battery, they can't go into the Lipo battery.
I'm thinking this through in my head, but I still can't quite understand what's going on where the LiPo and LiFePO4 is connected. So...uhh...does the LiPo just not do anything until the LiFePO4 is of the same voltage (actually, .2v lower, technically)? It feels odd, but I guess diodes feel odd."
the part of the diode connected to the top of the battery is called the cathode. the anode of the diode is the arrow tip thingy in the diagram. current will normally flow out of the diode in the direction of the anode, and when you try to make current go backwards through the diode it is blocked until you raise the voltage pushing it back into the diode up to the 'breakdown' voltage.
for diodes, the way they are made to increase the breakdown voltage causes them to have different forward bias or voltage drop as the current flows through it in the direction of the arrow. for a diode with a high breakdown voltage, the forward bias is higher than it is for a diode with a low breakdown voltage. this is important because the diode creates heat when the current flows through it in the forward direction.
this heat is expressed as power or watts. W=I x V, current times voltage. the voltage here is the forward bias voltage drop as the current flows through the diode. so to reduce heating and power loss in the diode, you use a diode with the smallest forward bias possible that meets the requirements of the breakdown voltage needed in the circuit.
in this case that is the difference in voltage between the lifepo4 pack and the lipo pack. there is no need to use a 100-200V diode since the pack voltages are not more than 20V different so a 20V diode would do the job and have the lowest forward bias. the one i mentioned on miro's thread was the 20V25A axial diode with a .45V forward bias instead of the 1.2V forward bias of a 100V diode. so for the same current it would only consume about 1/3 of the power and produce only about 1/3 of the heat.
the 16S lifepo4 can discharge down to about 34V and the 12S lipo would end up discharged at about the same voltage, say 36V. so the lipo would not push current until the voltage of the lifepo4 pack had dropped to the 50V level. and yes this 50V would include the .45V forward bias of the diode on top of the lipo pack so the pack itself would be 50.45V and the diode would eat up .45V of that potential so that it was identical to the 50V of the lifepo4 when the lipo would finally be able to push current. kinda arbitrary numbers since it will depend on the actual lipo pack voltage, which could be all the way down around 48V even.