I too think it’s downright absurd that it’s so difficult to fast charge LiFePO4 from perfectly capable alternators.
I’ll offer a few ideas that have come to my mind to generate some more conversation:
1. I’ll call this approach, cheap and big
2. I call this one alternating alternators. This idea I’ll credit to my algorithms instructor for teaching me to think outside the box. This approach assumes that overloading the alternator is safe until it reaches a maximum temperature, at which point it only needs to cool down. Is this a safe assumption? That would be useful to know.
- Add a second high output alternator to your engine so you have two alternators, a1 and a2
- Create circuit c1 that charges the LiFePO4 and circuit c2 that charges the vehicle starter battery
- Configure c1a1 and c2a2
- Once a1 gets too hot, swap c1a2 and c2a1 using temperature sensors, relays, and either hardware or an arduino to control when the circuits get swapped. I’m not aware of a DPDT relay large enough for this application, so I think you’ll need 4 of the above 500amp relays to swap the circuits, but there may be a way to simplify the circuit I’m thinking of.
- Note: If the alternator heats up too quickly, this will become a non-solution.
3. Let’s return to first principals for a second.
How does a battery charge anyway? It seems to me it would require a voltage to be applied that is greater than the battery’s internal, open circuit voltage - otherwise how could it charge if the net circuit voltage is 0v? After all, isn’t this the exact same thing as hooking up another 12v battery at an equal charge in parallel? No charge is transferred if the batteries are at equal voltage - they merely sit there maintaining their capacity. Therefore, I conclude by intuition that the battery charger by definition
must apply voltage greater than the battery open circuit voltage.
From there, Ohms law seems to indicate the battery charges at a rate of (chargerVoltage - batteryVoltage) / (batteryInternalResistance). The issue with LiFePO4 is allegedly that its internal resistance is lower than lead acid and therefore it needs a charger to “walk” it up the charge curve maintaining a voltage lead on the battery such that the amps it consumes does not exceed acceptable levels. Lead acid allegedly doesn’t need this since it’s internal resistance is high enough to limit the charge current given a relatively high constant charge voltage of 14.4ish volts from the alternator.
Therefore, it seems to me that if we can “simulate” the correct internal resistance for LiFePO4, we can hook it to a constant 14.4v alternator charge voltage and expect it to charge at the amps we desire, at least momentarily. One academic paper I read measured its test 100Ah LiFePO4 battery at 0.0018 Ohms internal resistance. So given a 14.4v alternator current, we can charge a LiFePO4 at 10% capacity (12v) at a current of 100A if we simulate a battery internal resistance of R = V/I = (14.4v - 12v) / (100A) = 0.024 Ohms. As battery voltage rises to approach our alternator voltage, we need to decrease the simulated internal resistance to maintain the same 100A charge current. I’m guessing at about 70% - 80% we could stop simulating additional resistance and let the battery’s internal resistance take it to full charge.
So now the only question is how? How can we dynamically simulate with resistors the internal resistance of the battery such that the battery charges at 100A until full? Do they make variable 12v 100A resistors in the 0.024 Ohms - 0.0018 Ohms range? Did I just ask how you design a DCDC charge controller? Idk, I’ve reached the end of my expertise now. I’ll let others chime in.