Simple MPPT for solar, am I missing the point?

Discussion in 'General Electronics Chat' started by Dyslexicbloke, Jul 24, 2012.

  1. Dyslexicbloke

    Thread Starter Well-Known Member

    Sep 4, 2010
    I have solar panels, with an open circuit voltage of about 19.5V in full sun.
    My batteries, depending on temperature, could require as much as 14V to fully charge although 13.8 would be more usual.

    I appreciate that if the panels are delivering less power that the batteries can accept, at any given point, then the resultant terminal voltage will be less than ideal because the charge current will be lower, I am not expecting magic or over-unity but I am looking for a simple efficient way to get the maximum power from the panels.

    I don't, at this point, want a circuit suggestion I am looking for comments on my understanding, or lack of it for that matter, with regard to power conversion.

    I may be making assumptions here that are simply wrong but then that is why I am asking the questions.

    Firstly I am assuming that capacitors store energy and that the voltage, assuming there is enough to work with, is of little importance.

    For simplicity lets assume that the batteries are not fully charged and effectively capable of holding the panel voltage down at 13V when connected directly.

    If I charge capacitors with the panels and use PWM and a choke to control the battery charging current which is modulated to maintain the optimal capacitor, and hence, panel voltage will I gain anything as a result of operating the panels at a higher voltage?

    Taking this a step further, if the panels are arranged to output say 30-50V, dependent on irradiation, can I use the same system to generate a charging current that would result in a much lower battery terminal voltage without inuring huge losses?
    The reason I ask about this extreme situation is that useful power may be available at low light levels if some of the panels are connected in series.

    I will be using a small PLC, or possibly a PIC, to manage things and I appreciate that I have not addressed a charging strategy but for simplicity I would like these questions treated in their simplest form.

    In short … Using PWM and a choke to generate a current into some load is the following possible:-
    Panels at 18V delivering 5A into the capacitor bank (90W)

    Current drawn from the bank, via PWM switch and a choke, to hold the bank at 18V
    I need to take out 90W right?

    Assuming 5% losses in the choke and switch my target load must be recurving circa 85W, which, if we work on a terminal voltage of 13V, would be about 5.6A, far more than would flow if the panel were connected directly to the battery.

    Will that work or am I just missing the point somewhere?
    Last edited: Jul 24, 2012
  2. shortbus

    AAC Fanatic!

    Sep 30, 2009
  3. Dyslexicbloke

    Thread Starter Well-Known Member

    Sep 4, 2010
    Interesting links, thanks. I hadn't appreciated the history although the basic principle I do understand.
    I suppose what I should have asked is can I build a simple, capacitor based, buck converter and expect it to work efficiently for a few volts drop.
    Secondly if the voltage conversion is 50% or more will it still work?
  4. nsaspook

    AAC Fanatic!

    Aug 27, 2009
    MPPT when used with panels under a few hundred watts to charge a properly sized 12VDC battery bank is close to a wash unless you run the panels at max power all day. You will see gains in charging current when the batteries are low and need the max possible current over plain PWM but when the charge controller starts to limit voltage/current in absorption and float stages most of the gains will be lost. With panel power of 90 watts input the possible gain vs PWM might be 5-10 watts max during bulk mode. I see little bang for buck with a MPPT converter at 90W vs just adding more panels with plain PWM at such low power levels.
  5. Dyslexicbloke

    Thread Starter Well-Known Member

    Sep 4, 2010
    I sort of guessed that which is why I am not considering an expensive commercial unit.

    We are off grid so power is expensive anything we get from solar is offset against the generator which changes the maths RE ROI.

    Batteries were circa 430Ah when new and are designed for very beep cycle operation. Although they are very long in the tooth now they seem to be doing well.

    I have 280W of amorphous units and 95W crystalline, panel rating.
    The amorphous is split between two banks of dissimilar units, 120W of 1 type and 180W of another.
    The plan is to add a further 200W of crystalline and mount 295W on a tracker.

    Given the context I think a few amps is probably worth chasing especially as my loads are modest.
    circa 80W @ 24 hr
    and perhaps another 100W, average, @ 6Hr

    I just bought a new inverter, 600W Pure sine, that will eat even my intermittent loads.
    Anything big water pump, standby battery charger, runs directly from the genny.

    Last edited: Jul 25, 2012
  6. cork_ie

    Active Member

    Oct 8, 2011
    MPPT is Maximum Power Point Tracking and it's main advantage is to extract the maximum power at any particular time depending on the level of sunlight. I had to do some extensive research on this lately, while designing a solar system for a larg'ish Catamaran. The best explanation I found on the net was the Wikipedia article
    The main advantage of MPPT controllers is that they can precisely deliver optimal 3 step battery charging irrespective of solar levels and panel ouput voltage and utilise the maximum power available from the panels irrespective of the output voltage, until such time as the batteries are fully charged. Battery temp sensors are often added to reduce the charge voltage in warm conditions
    MPPT is sophisticated and not easily replicated by us amateurs. All the true MPPT controllers run on software. To be honest unless the system is
    A) large, & B) has constantly varying solar levels. it is unlikely that the cost benefit of MPPT would be worth while . PWM charge control is also widely used and quite efficient and I am sure with a bit of effort it should be easy enough to design a homemade pwm based controller.
    By the way I eventually settled on 2 BLUE SKY ENERGY Solar Boost 50L MPPT controllers. They cost a lot of $$$ but for my application were the best choice. if you check their website you will get a lot of useful info.
    Last edited: Jul 25, 2012
    PackratKing likes this.
  7. Kermit2

    AAC Fanatic!

    Feb 5, 2010

    Yep! The 'simple' part.

    1 - a single diode to prevent reverse current flow in shadows or low light.

    2 - A method to monitor voltage and disconnect the panel when a set voltage is reached. (op amp comparator and associated relay perhaps??)

    A few tests to 'dial it in' and you would have a simple system and with minimal components to lose power in.

    Or not. Just saying :p

    (edit added later)
    If you are using amorphous panels(mostly) then you will have more power available at low light levels than with crystalline types
    Last edited: Jul 25, 2012
  8. Dyslexicbloke

    Thread Starter Well-Known Member

    Sep 4, 2010
    Amorphous ... I would have gone that way for exactly that reason but they are difficult to find these days.

    Space isn't an issue for me but I can see why it is critical in most installations.

    Simple with minimal component count is what I have now, almost exactly as you describe.

    This thread has convinced me that buying a controller is not going to be cost effective but I think I will still have a go at building one, if only for the experience.

    I think I would like to change tack a little with respect to the focus of this thread ...
    Should I start a new one specifically about buck converters or carry on here?

    I am only interested in the power conversion circuitry, control will be from my PLC which will manage either PV voltage, for max power, or battery voltage, if the panels are providing more than the batteries can absorb at any particular phase of the charge cycle.