Solar MPPT Charger for Li-Ion Battery

[forgoreth]

We are building a PV system on our electric car and we decided to mount three 30W panels that are placed in different tilt angles due to car geometry.
I thought that since each panel will produce a different amount of power, it would be better to build three MPPT controllers and connect their outputs via diodes so that the current will flow only to the battery. Our MPPT controllers are boost converters and they are designed, simulated with an MPPT algorithm. Moreover, they are built and tested by manually changing the duty cycle. Everything looks PERFECT! But...
I have concerns about the output voltage of the MPPT controllers. The connection is simply like Solar Panel->MPPT Circuit->(diode)-> Li-Ion Battery
I always thought that two cascaded DC-DC converters (One for MPPT and the other one for battery charge regulator) will serve a better solution for efficiency for this kind of application. But battery is usually connected to the output of the MPPT in the examples that I have checked.(this is cheaper and more simple.)

My first question is: What happens if the output voltage of the converter is different than nominal Li-Ion battery voltage at the maximum power point?
If it is greater than the battery voltage, current will flow through the diode between the mppt and the battery. And set the output voltage directly to the battery voltage. But I cannot envision that simple power conversion there... Please somebody help me understand it!
For instance, the output voltage at the MPP is 70 Volts if a resistor was connected to the output of the mppt circuit. But when it is a battery, it will equalize the MPPT output voltage to its own voltage. Is the difference between MPPT output voltage and battery voltage directly affects the charging current? If it happens like this, there is no need to worry about and no need a define a "maximum output voltage" in the control algorithm (as long as the output current does not exceed the nominal charging current).
If output voltage @MPP is less than or equal to the battery voltage, the current will not flow because of the diode. So, should I add a part to algorithm like "if the output voltage @MPP is less than the measured battery voltage; forget about maximum power and apply nominal charging voltage"?

I am a little bit afraid to use those Li-Ion batteries. Should I forget about MPP tracking and make a constant voltage/constant current charger?

P.S.: Vmpp of the panel is 17 Volts and the nominal battery voltage is 48 Volts (54.6 Volts when fully charged). By the way, I need suggestion for those protection diodes. Which diode model should I use? Any suggestions will be appriciated. Thanks in advance!

 

[Johansen]

with a boost converter you should not need a protection diode, as the boost diode will prevent current from flowing backwards into the panels, or through the converter.

the only problem you have is making sure that the mppt circuit cannot overcharge the battery, and making sure that it does not push too much current into the battery at end of charge.

you do need 3 mppt converters, but since you are using a boost topology, there is no need for extra diodes. btw: use the voltage of the battery multiplied by the current going into the battery to calculate the power flow, and keep all three panels on an isolated ground (put the current sensor on the negative line). that is the cheapest and fastest way i can think of at the moment.

 

[forgoreth]

with a boost converter you should not need a protection diode, as the boost diode will prevent current from flowing backwards into the panels, or through the converter.

the only problem you have is making sure that the mppt circuit cannot overcharge the battery, and making sure that it does not push too much current into the battery at end of charge.

you do need 3 mppt converters, but since you are using a boost topology, there is no need for extra diodes. btw: use the voltage of the battery multiplied by the current going into the battery to calculate the power flow, and keep all three panels on an isolated ground (put the current sensor on the negative line). that is the cheapest and fastest way i can think of at the moment.

Thank you for your suggestions. If I do not use an extra diode for protection, does this mean that MPP voltage automatically equalized to the battery voltage and I should not fear about where the MPP is? I am still not sure if the current can flow through the diode when the battery voltage is higher than the boost converter output voltage. Don't I need to stop tracking if the calculated output voltage is less than battery voltage?

Thanks again

 

[Johansen]

you can design your algorithm so it does not care what the battery voltage is.
simply calculate the power flow from the panel voltage, and the panel current.
then adjust the pwm duty cycle, calculate again.

if your boost converter output voltage is less than the battery then there won't be any power flowing, and your mppt algorithm will adjust the voltage higher to push power into the battery. at least.. that's what should happen.