New to solar - finding the MPP point

GrahamG
GrahamG Registered Users Posts: 4
I understand the concept of Vmp and Imp but, using a microcontroller based charger design, does one accomplish finding this point ?

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,433 admin
    Welcome to the Forum Graham!

    Are you asking as somebody that wants to install a solar arrary, an MPPT (or PWM) solar charge controller, battery bank, etc...

    Or are you asking on how to design a MPPT controller from an electronics/engineering point of view?

    Very basically... Solar panels are (for the most part) constant current devices... I.e., under full (100%) sun, the panels will output Imp (current maximum power) from Vpanel = zero volts to Vmp (voltage maximum power). Above Vmp, the current starts to fall and ends at Voc (voltage open circuit).

    This is very different how we usually use power sources. Batteries, AC mains, etc. are all constant voltage devices. This hold ~12 VDC from zero current to maximum current (trip breaker, over load AC inverter, genset, etc.).

    An MPPT charge controller is really a "switch mode Buck down converter":

    Buck converter

    More or less, think of it as the DC equivalent of a Variable AC transformer. A buck converter can take high voltage/low current and efficiently down convert it to low voltage/high current for charging the battery bank (or other needs).

    The MPPT charge controller (one common method) needs to do is figure out the Vmp of the solar panel/array. That point where we get "peak power" from the solar panel array. It can do this by varying the current it draws from the array and creating a table--Remembering P=V*I

    P=V*I
    P=10 volts * 5 amps = 50 watts
    P=15 volts * 4.9 amps = 73.5 W
    P=18 volts * 4.8 amps = 86.4 W
    P=19 volts * 2.4 amps = 45.6 W

    So for the (made up numbers) above, Vmp~18 volts.

    Note that Vmp is relatively stable over a wide range of sun energy and panel output current. However, Vmp does fall as the solar cells get hot (or on very cold days, Vmp can rise over the Vmp-standard test conditions).

    The MPPT controller can run a "sweep" every 10 minutes and get a "new Vmp-measured" and use that for the next 10 minutes. Note that the controller dumps "full available current" into the battery bank during these sweeps (and less than "optimum current" too as it searches for the optimum Vmp*Imp measured).

    There are other methods, some work a little better, other work a little less well.

    Anyway--Is that sort of what you are looking for?

    Of course, we have to look at the solar panel specifications, the input Vpanel range of the charge controller, etc... To ensure that we have a "match" between the Vmp-array and the Imp-array, the controller capabilities, and the battery bank voltage.

    Many higher end MPPT charge controller companies will even make a website (or download to a PC) to help you design the series/parallel/wattage of the solar array to work with the MPPT controller and connected battery bank--Here is one such tool for Midnite brand MPPT controllers:

    MidNite Solar Classic Sizing Tool

    The first time through a "sizing" question (solar array, panel Vmp/Imp, what MPPT charge controller) can be very confusing... After you do it a couple times, it will make more sense.

    If you have some starting information (like you need a 1,000 watt array, or have 4x panels with Vmp=18v and Imp=5a, a 24 volt battery bank, and want to use a Midnite Classic 150 Charge controller)--We can help you size the array/controller/etc. to work. It is usually much easier to do the design based on your needs--Not some theoretical system.

    Solar power systems and their components are very much based on the over all size of the array/battery bank/loads/etc. A small system will use different components vs a large system (and you will have to make different design choices/decisions based on size of system).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • GrahamG
    GrahamG Registered Users Posts: 4
    BB - this is JUST the information I was going to want to start out with. Yes - I want to design a small 3A MPPT unit and to incorporate on it a larger PCB that is going to monitor and log temperature data. Solar water in/out, pyronometer value and some other parameters etc.

    The current solar panel is a 10watt and the charger module is a "Chinese" 3A unit based on a PIC16F716, it works but for the quality is not worth the US$15.00 that the local electronics supplier charges for it.

    I will want to write my own C code for the PIC (hence the reason for my question) whether it is incorporated on the main PIC24FJ256GC006 or a smaller separate PIC like the one in the Chinese unit.

    This is the start as I want to expand on the MPPT idea and produce a solar system for myself, especially as we now have daily load-shedding problems. !

    Graham
    Durban
    South Africa (with load shedding :)
  • jonr
    jonr Solar Expert Posts: 1,386 ✭✭✭✭
    Note that DC-DC conversion (buck) and tracking MPP are separate features and you could (but usually don't) have one without the other.

    I am available for custom hardware/firmware development

  • GrahamG
    GrahamG Registered Users Posts: 4
    When you say DC-DC buck and MPPT are separate features surely they are found together as (a) you need to obtain the MPP and then (b) use the solar panels, at the right voltage, to charge the battery(s) ?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,433 admin
    More or less, as I understand, you need to adjust the buck converter duty cycle such that you keep Vmp-measured at its optimum point. Of course, you need to monitor battery bank voltage too--So once the output power (P=Vbatt*Ibatt) is less than Array Wattage, then you no longer really need to track/control MPPT.

    By the way, there are some chips which can be used as MPPT charge controller for smaller systems:

    http://www.ti.com/product/BQ24650
    http://www.ti.com/tool/bq24650evm-639

    This thread has some more information and a chip from Linear Tech too:

    http://forum.solar-electric.com/forum/solar-electric-power-wind-power-balance-of-system/solar-beginners-corner/20941-diy-solar-usb-charger-mppt?22175-DIY-solar-USB-charger-MPPT=

    Since most common MPPT designs use "buck" converters... They can only take higher voltage and down convert to a lower voltage. Buck cannot do 14.0 to 14.0 volt regulation (need a buck-boost converter) or up convert to a higher battery voltage (boost converter).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • GrahamG
    GrahamG Registered Users Posts: 4
    Bill - thanks for that.

    I never thought that there would be a dedicated IC.

    I have an R&D account with TI and so I have ordered some samples to see what they perform like.

    Normally they are $5.18 each plus shipping but at least a dedicated and stand-alone IC.

    In the meantime I am going to build a test rig and evaluate the use of the adjustable PWM and buck DC-DC. It will be fun if I can get the algorithm right as I can incorporate the concept into other designs that use a PIC micro for "other" functions.

    I have been watchg Julian Ilett's YouTube videos on MPPT design since you replied to my post. Quite an interesting subject all round.

    Thanks for your input again.

    Cheers and enjoy the weekend.

    Graham
    Durban
    South Africa