Mixing panel types

McGuyver
McGuyver Registered Users Posts: 14
Good day friends, I have 6 x 68W Unisolar panels in parallel controlled by an Outback Flexmax 80. I am replacing my FLA batteries with 800Ah LiFePO4. The panels are rated at 16.5v Vpm but according to the controller, I get an average of 22V in from the panels. I have a space issue on my RV and would like to add another 300/400watts. Can I add mono or poly panels into the array? What Vpm should I get - 16.5v or 22v?:confused:

Comments

  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Mixing panel types
    McGuyver wrote: »
    Good day friends, I have 6 x 68W Unisolar panels in parallel controlled by an Outback Flexmax 80. I am replacing my FLA batteries with 800Ah LiFePO4. The panels are rated at 16.5v Vpm but according to the controller, I get an average of 22V in from the panels. I have a space issue on my RV and would like to add another 300/400watts. Can I add mono or poly panels into the array? What Vpm should I get - 16.5v or 22v?:confused:

    You'll either need a separate controller or be willing to suffer some performance loss as mono and poly panels won't have the same Vmp. A "12 Volt" panel typically has a Vmp around 17.5 - 18.5 and a "24 Volt" double that. There are some "oddball" Vmp's like this Kyocera at 23.6 http://www.solar-electric.com/kyocerakc170gt.html. However, don't expect your Unisolars to stay at that level; the flexible panels tend to drop off a bit over time.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Mixing panel types

    I am not sure what and average of 22 volts means with the FM controller... Normally solar panels will output near Vmp (16.5 volts on a very cool day) when under load and near Voc (21+ volts cool) when not under load (batteries fully charged).

    And as Marc says, the panels have a "wear in" period:
    1 During the first 8 to 10 weeks of operation, electrical output exceeds specified ratings. Power output may be higher by 15%, operation voltage may be higher by 11% and operating current may be higher by 4%

    But--in any case, with a MPPT controller (like the Outback FM family), you can put three of your panels in series, then put the two strings in parallel (3x2). That will increase the Vmp-array to ~49.5 Vmp (rated)... More than enough for a 12 or 24 volt battery bank.

    You should be able to put all six in series for 99 volt Vmp (double check Voc-cold for your area to make sure you do not exceed 150 VDC maximum of the controller).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • McGuyver
    McGuyver Registered Users Posts: 14
    Re: Mixing panel types
    BB. wrote: »
    I am not sure what and average of 22 volts means with the FM controller... Normally solar panels will output near Vmp (16.5 volts on a very cool day) when under load and near Voc (21+ volts cool) when not under load (batteries fully charged).

    Well, thanks, that answers it for me. What the Outback shows is the "no load" volts at 22v.
    If I have this right I should check the input voltage on the BULK cycle to see the "under load" volts which should be in the region of the listed Vmp- correct?
    And as Marc says, the panels have a "wear in" period:

    My panels have done 2½ years so that should be settled by now....
    But--in any case, with a MPPT controller (like the Outback FM family), you can put three of your panels in series, then put the two strings in parallel (3x2). That will increase the Vmp-array to ~49.5 Vmp (rated)... More than enough for a 12 or 24 volt battery bank.

    You should be able to put all six in series for 99 volt Vmp (double check Voc-cold for your area to make sure you do not exceed 150 VDC maximum of the controller).

    -Bill

    I'm not exactly sure why you suggest the series connection Bill? Please give some more info...will it be better at all? The way I have it is that in parallel I may lose a little charge at low volts (early morning, late aft, very yuk weather?) but I will gain when there is some shading on only one of the panels or it is less efficient for some reason due to bad angle or whatever.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Mixing panel types
    McGuyver wrote: »
    Well, thanks, that answers it for me. What the Outback shows is the "no load" volts at 22v.
    If I have this right I should check the input voltage on the BULK cycle to see the "under load" volts which should be in the region of the listed Vmp- correct?

    Yes. Even then an MPPT controller may load the panels "differently" than the panel would expect. But it ought to be close.
    My panels have done 2½ years so that should be settled by now....

    Agreed. Should be working nicely now. :D
    I'm not exactly sure why you suggest the series connection Bill? Please give some more info...will it be better at all? The way I have it is that in parallel I may lose a little charge at low volts (early morning, late aft, very yuk weather?) but I will gain when there is some shading on only one of the panels or it is less efficient for some reason due to bad angle or whatever.

    The main reason for running serial connections is to increase array Voltage resulting in less loss through wiring. With an MPPT controller there is some gain to be had under less-than-ideal conditions, but it's not easily quantified. Shadows are usually better dealt with by parallel configuration. I think Bill is suggesting upping the Vmp of the original array so that it might more closely match a new panel configuration. That gets to be a game of "hunt-and-peck" sometimes.

    To return to the original premise, if it were me I'd pick a standard "12 Volt" panel with a Vmp of 17.5 - one Volt off from your existing array and thus less than 10% difference. You usually can do well with that.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Mixing panel types
    McGuyver wrote: »
    Well, thanks, that answers it for me. What the Outback shows is the "no load" volts at 22v.
    If I have this right I should check the input voltage on the BULK cycle to see the "under load" volts which should be in the region of the listed Vmp- correct?
    Exactly--Under "Bulk" the MPPT converter is trying for 100% power transfer. In "Absorb" it may run from 100% down to 1% power transfer--so the Vmp reading (Vpanel) will be "variable". Float is going to be in the 1-5% or so current range.
    My panels have done 2½ years so that should be settled by now....
    Sounds good... I tend to be a "Gloomy Gus"... Engineers typically want the stuff to meet specifications at the end of the product's life--so we throw in a few more "fudge factors" towards that end.
    I'm not exactly sure why you suggest the series connection Bill? Please give some more info...will it be better at all? The way I have it is that in parallel I may lose a little charge at low volts (early morning, late aft, very yuk weather?) but I will gain when there is some shading on only one of the panels or it is less efficient for some reason due to bad angle or whatever.

    Nothing fancy--Just like your 12 volt battery is 6x 2 volt cells in series... You put a number of solar panels in series to raise the Vmp-array voltage...

    Normally, Vmp~17.5 volts for a 12 volt battery bank... A "hot array" with a cool battery plus some voltage drop in the wiring and charge controller makes sure that Vmp-array>Vbatt-charging.

    With PWM controllers (pulse width modulation--really just a "simple" on/off switch or transistor), the Vmp:Vbatt ratio is fairly critical... Too low of Vmp and the battery will not charge well on hot days because the solar array Vmp voltage is too low.

    With MPPT controllers (maximum power point tracking), they are typically Buck Mode down converters (switch power supplies). They can take Vmp-array>>Vbattery-charging and efficiently (95% or better) take the "high voltage/low current" from the solar array and down convert to "low voltage/high current" required for the battery bank.

    The marketing literature "sells" the fact that MPPT controllers will generate "up to" 30% more power (sub freezing sunny days in winter--they don't tell you that)... And usually 10-15% more power.

    However, the advantage is you can run a Vmp-array>>Vbattery and use smaller copper cable, and longer copper cable from the Array to the Battery Shack/Charge Controller.

    Two systems--one that would require "OO" copper cable and long wiring run, with Vmp=17.5 volt array and 12 volt battery bank... Run Vmp-array > 70 volts, and you can do the same thing with 14-12 awg wire.

    So--for you, if Vmp gets to close to Vbatt-charging--You can just put 2/3 or even possibly 6x solar panels in series to raise the Vmp-array voltage. This can help you batteries charge faster/better on very hot days (with cool batteries).

    6 panels can be connected as:
    • 1 panel in series, 6 parallel strings Vmp = 16.5 volts
    • 2 panels in series, 3 parallel strings Vmp = 33 volts
    • 3 panels in series, 2 parallel strings Vmp = 49.5 volts
    • 6 panels in series, 1 parallel string Vmp = 99 volts [tried without calculator--dumb]
    Or, if you wanted move your array farther away, or add more panels later--the higher array voltage will allow you to use smaller diameter cable.

    There are other reason "higher voltage" series strings can make things easier...

    If you have three or more panels/strings in parallel, each panel should have a fuse or breaker at the Rated Series Fuse rating... If you have 3 or more panels in parallel, then the two+ good panels can feed a short in the bad wiring/panel and cause to overheat/start a fire.

    This is why people tend to use "combiner boxes" with larger arrays and 3 or more parallel connected strings.

    If you are happy with your system, there is no reason to change to series/parallel wiring.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • bluewickedburner
    bluewickedburner Solar Expert Posts: 78 ✭✭✭✭
    Re: Mixing panel types

    Think carefully before using series connections for panels on an RV.

    Those of us using RVs and solar have a few things to consider our land affixed friends do not, one important one are the various things on the roof such as A/C units, sat dishes, anemometers for awnings, tank vents, air vents and their covers and the very limited space available on the roof.

    In series, you get shading from any of those things are the game is pretty much over for that series. If you can avoid the shadows then series can make some sense.

    If you are stationary most of the time then figuring out the shaded areas is pretty easy. Here is a tip I used and it works very well:

    Get on the roof at sun up and using a dry erase or grease pencil, mark all shadows. Do that every hour or so until sun down and you'll know exactly where all your shadows fall and at what times. If you can avoid those areas in your panel placement, series is good to go. If not, go with parallel. That beats laying all the panels up there and watching meters.

    Since the lines drawn around the shadows are a roof level, any panel going on will be elevated slightly allowing for mounts so you'll be safe. This works for a place you are most of the time since when traveling all bets are off.

    One thing you can do is note the direction your RV is facing and then you have some reference for later travels. Not foolproof but it takes little effort and is easy to do.
  • McGuyver
    McGuyver Registered Users Posts: 14
    Re: Mixing panel types

    I am thanking all of you for these absolutely valuable inputs.
    However, I am in New Zealand which is known as "The land of the great white cloud" for good reason. So looking for shadows on the roof is a joke in this country. That is partly the reason for getting PVs instead of crystal panels. This place also hardly ever gets much hotter than 24°C :cool: and the winters are long. Luckily I don't have any funny stuff on the roof causing more shadows apart from the flue So it sounds like parallel may be the best for me according to all your feedback. I may try the 2P3S sometime just to try and see the difference although it may be negligibly small and difficult to determine. I do not have any fuses currently in the 6P array which I recon I should look at. Will the fuses cause a drop in efficiency at all?

    Here is a pic of the array: I had to stagger and turn the panels to fit between the huge rivets on the roof - therefore the strange layout.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Mixing panel types

    The fuses will not have any noticeable effect on panel output (unlike blocking diodes).