Poor man's tracking - splitting arrays on one charge controller

f5f9 Registered Users Posts: 18 ✭✭
I may have some used solar panels lined up to combine with my other old panels (voltages are close).

Their is not enough room to place them on my pole mount and was going to mount them on some kind of stand next to it. BUT.... I was thinking of pointing them in a different directions than the pole mount panels to better track the sun.

Does pointing/splitting the arrays like I've mentioned into two different directions and having them both on one controller reduce power output?

i.e. one pointing SSW and the other SSE ---or a larger split... one pointing SW & the other SE

Do panels that are in the shade drag down the other panels that are fully exposed to the sun?

Hope my babbling made sense.



  • BB.
    BB. Super Moderators, Administrators Posts: 33,439 admin
    Re: Poor man's tracking - splitting arrays on one charge controller

    Actually, there has been a lot of discussion about that question... There is a Fronus White Paper saying the losses are very little:

    White Paper PDF file download

    But, Solar Guppy, who has the lab equipment and experience, believes otherwise:
    If you do go with GT PV, I would recommend two inverters, the often linked Fronious white paper is bogus in my professional opinion, its conclusions are not on measured arrays and did not take into consideration the thermal variations of the two arrays in different orientations, which is the single largest cause of vmp difference and why there is Mppt in the first place.

    If you assume that the drop in Vmp array voltage is around -0.4%/°C--And that you should match Vmp between different strings within 10%:
    • 10% difference / -0.4%/°C = -25°C or -45°F
    So the question appears to be, how far apart will the panel array temperatures be between the SE and SW arrays.

    Here is another thread where we had the discussion and I did a simple ASCII drawing to show the MPPT tracking problem:

    Two Strings in Parallel with Unequal String Voltages

    I think the recommendation is to use different MPPT inputs if you have different string orientations for best power production... And there can be times when differences in Vmp between the two arrays could cause and even worst than expected loss in power if the MPPT algorithm locks on the wrong peak...

    If your controller can manage the larger array--you could always try a few experiments and see if the single MPPT controller works for you or not... And if you are unhappy with the results, get a second controller when the funds permit.

    Regarding the shade question---When placing two strings in parallel where one or two panels in one string will be shaded--This will probably guarantee a large loss in combined output.

    Think of a three D Cell flashlight... Two 3xcell battery strings in parallel will work fine (both will share about 1/2 of the total current). However, with shading, one cell "goes away" and now you are left with one 3xcell string in parallel with another 2xcell string--The 3xcell string will carry darn near 100% of the current. And the shaded string will supply near zero useful power.

    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • f5f9
    f5f9 Registered Users Posts: 18 ✭✭
    Re: Poor man's tracking - splitting arrays on one charge controller

    Ok.... From what I think I understand now. <g>

    Sounds like to me that:
    • a good idea is to face all the panels going into a **single controller** in one direction (All panels being the same type/likeness going to that controller). . It's OK but not desirable to split the direction of each controller's panels (ie PWM controller panels all facing SE and all MPPT controller panels facing SW).
    • the better idea is to have all the panels for *both* type controllers facing the same direction even if one is a PWM type and the other a MPPT type.
    • an even better option (when there are two types of panels to deal with [old/new]) is to have two identical MPPT controllers. Have all the old panels going into one controller and all the new panels going into the other controller. Then have all the panels facing in the same direction.
    • the best idea is to just have one type of panel in your system that feeds into a single MPPT controller and have all panels going in the same direction.

    All-in-all keep the panels facing the same direction; pay more to get more in hardware and harvesting abilities, etc.

    I guess, like everyone, my goal is to get the most power harvested before the sun sets and back online as soon as possible after the sun rises. Without spreading out the panels your only option then is a real tracker.

    Where am I going wrong this time ;>

    What should I do or think differently to get the best real world results?

  • f5f9
    f5f9 Registered Users Posts: 18 ✭✭
    Re: Poor man's tracking - splitting arrays on one charge controller

    Hmmmmmmm....... After thinking about it.... I'm going to grab a bite to eat and reread your post and the links you mentioned.

    Until then, here's what I threw together while the site was offline (drawings attached). Also a little more background below for my situation.

    I currently have 4 old Photowatt panels (maybe 10-yrs old) and 6 Solarworld panels that are about 1.5-yrs old. I may have and additional 10 Photowatt-like panels coming that are about the same voltage, vintage, etc. I am also about to install 2 new Solarworld panels.

    When I hooked up the Photowatt panels a week or two ago into the PWM Morningstar controller I noticed that that my other panel's controller (Flexmax 60) goes into a Float or even non-charging mode when the old panels get too much sun; so, I decided to put the Photowatt panels in the other direction of the Solarworlds and tweaked the Solarworlds a little more westerly. This seemed to help allow both controllers to charge/play nicely together.

    Generically speaking....
    The Photowatts are facing SSE (400watts at the moment; possibly 1,400w) and the Solarworlds (1,050w now and 1,400w shortly) are facing SSW. It may even be more like SE and SW respectively.

    In the mornings I notice that the old Photowatt panels do not register any energy even though they are pointed in the general direction of the rising sun. ALTHOUGH.... the newer mono-crystalline SW175w panels register power going to the batteries... even though they are shaded by their own casings (backs to the sun). We're not catching the woods on fire by any means, but I'll see around 0.5 to 1.0 amps before the old panels even wake up. The power of better tech ---eh? :)

    When the old ones do come up on line, they'll collect about 1.2 amps and the new panels are still drawing around 1 amp... all while the sun is still favoring the Photowatt panels. Furthermore they seem to play better together when the sun is due south/splitting the two evenly with the new arrangement (direction of panels).

    Again, I was hoping to get some used panels that are similar to the old Photowatts, send that power to the Morningstar PWM controller and face them in due south/between the other panels that are here now to capture more light over the day. Continue facing the orig Photowatts more SE and the newer Solarworld panels SW on a different MPPT controller.

    Basically I wanted to track the sun over the day. Using my best panels last to grab as much energy before there is no more.

    What seems to work best in the real world?
  • n3qik
    n3qik Solar Expert Posts: 741 ✭✭
    Re: Poor man's tracking - splitting arrays on one charge controller

    I have a setup just like you are proposing. My two arrays, 512 and 434 watts, are 90 deg. of each other going into a PWM CC ( Xantrex C60 ).

    The key to this working is using a PWM an not a MTTP controller.

    The important thing to look at on your setup is early morning startup. That the old array does not feed into the new array.Then in early after noon, the new array does not feed into the old array.

    I am using Unisolar thin-film panels, they seem to resist being backfeed.