Slightly different voltage panels.

Wanderman
Wanderman Solar Expert Posts: 180 ✭✭✭
hi!
I am working on an update for my semi-flexible solar panel charging system on my RV. Te old one worked great until the anodized aluminum substrate panels failed. 15 months. Needless to say I will NOT be going that route again.

Question, the old panels were 36V~ for all three sizes. some had split modules to make that work.

What would happen if I used panels that had different voltages wired parallel (duplex home run wiring to a terminal block) then on to a morningstar ts-45-mppt controller.

I want to hear some opinions on this.

Thanks,

Rich

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Slightly different voltage panels.

    Basically, you can parallel panels (or strings of panels) with Vmp within ~10% of each other (35 volts, then 38.5 to 31.5 volt "new panels"would be fine). Also within 10% Imp for series connections of panels.

    The standard "12 volt" panels have Vmp~17.5-18.6 volts. So two in series would be fine.

    If you get large format >200 watt panels... You will find Vmp in the range of 30 volts. That is getting to be on the outside range for use with an MPPT charge controller finding the "optimum" Vmp*Imp=Pm peak (and even then, it still would not be "optimum" because the Vmp voltage are too far apart).

    You should look at >200 watt panels and see if you can fit them on your roof... Even if you have to pull up the "working" panels already there. These panels can be pretty cost effective these days.

    -Bill

    PS: I forgot to add--Get the Array Wattage first--then we can talk about MPPT controllers. But Midnite, MorningStar, & Rogue all offer some lower cost MPPT controllers that should work well (depending on ultimate array size/configuration).

    -BB
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Wanderman
    Wanderman Solar Expert Posts: 180 ✭✭✭
    Re: Slightly different voltage panels.

    Array watts will be around 450-500 watts.

    If you look at my old posts about the r&d...those specs worked well. the new panels will be made with full 1/2 V modules so based upon string config..maybe around 20V

    Rich
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Slightly different voltage panels.

    I believe the original thread is here:

    http://forum.solar-electric.com/show...ight=wanderman

    500 watts * 0.77 panel+controller derating * 1/14.5 volts = 26.6 amps "typical maximum"

    So, if you get a MPPT type charge controller with a minimum rating of 30 amps--You should be OK. The Rogue is 30 amps, and Midnite is coming out "soon" with some lower cost/current products too.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Wanderman
    Wanderman Solar Expert Posts: 180 ✭✭✭
    Re: Slightly different voltage panels.

    BB,
    The morningstar TS-MPPT-45 is already installed, as is the wiring, fusing, etc. The only thing being replaced are the three failed panels. Want to make sure I am doing the right thing!

    They are being custom built for me in the sta
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Slightly different voltage panels.

    You should be fine... As long as the three panels have Vmp in the range of 17.5 to 100 VDC (parallel, series, etc.), you should be fine. I would aim for ~30-80 volts Vmp for your array. 30-40 volts probably being the most efficient for the MPPT controller--But it does not matter much (at Vmp~17.5 volts, on hot summer days, the Vmp-hot for the array may be on the low side. For Vmp>>30-40 volts, the MPPT controller may be 1-2% or so less efficient. >~100 VDC Vmp, you could end up over voltaging the MPPT controller on very cold days (i.e., way below freezing).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Wanderman
    Wanderman Solar Expert Posts: 180 ✭✭✭
    Re: Slightly different voltage panels.

    according to the morningstar mppt docs. t appears the most efficient voltage is around 24V charging a 12v nominal system. That sound right? What if I have 3 panels around 20v? The issue has been figuring out how to assemble 1/2V full modules into strings that fit the dimensions. Looks like the largest could be 60 cells the next one down could be 40 cells and the last one around 36 cells. I am trying to figure out the max for the space...since they are building them. Probably 6" or so modules.

    Rich
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Slightly different voltage panels.

    In general, the optimum for an MPPT controller seems to be around 2x the battery voltage. But the difference is not that much (maybe 1-2%)... If you have a choice--I would go for 24-40 volts Vmp. Note that when a controller is 95-97% efficient and you lose 1-2% efficiency, that is 25% or more heat that has to be dissipated by the controller (3% vs 4% or 5% waste heat)--So it matters to the engineer (big enough heat sink/fan).

    The things to be careful about running Vmp~17.5 volt panels (which is perfectly within spec.) is that that Vmp drops on hot days (so you may not be able to quick/fully/equalize a cool battery bank on a hot day--And that you need very heavy cabling from the solar array to the MPPT charge controller (controller should always be within a yard or two from battery bank). Voltage drop (higher current with low voltage drop headroom) is a pain that you don't usually have to accept with MPPT controllers.

    You run a 35-40 Vmp solar array, you can have 1/2 to 1/4 the diameter cable from array to controller (to 2-4x farther away on smaller diameter cable).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: Slightly different voltage panels.

    actually 2x would be a misnomer if talking about battery base voltages other than the standard 12v battery voltage. it generally goes that one would take it one pv higher than that which would be nominal. nominal is 36 cells for every 12v battery output desired. so if one has 2 36 cell pvs it's best to put them in series. now if one has 3 36 cell pvs and desired the same 12v output it does not configure optimally with the 2 pvs in series as one pv would be left out of it unless it is placed in series. it still works, but is not at the highest efficiency. to gain the efficiency of the 2 36 cell arrangements in series one would have to put in multiples of series strings of 2 36 cell pvs and parallel those strings. aka 2, 4, 6, 8, etc so as to continue the voltages seen by 2 36 cell pvs in series.

    for a 48v battery output using 36 cell pvs one would optimally go with 5 in series. nominal would've been 4 36 celled pvs in series so you wind up as 1 36 cell pv over the nominal 4 x 36 celled pvs for 48v battery outputs.

    now this is a rough rule of thumb and the optimal point was easier to figure before the advent of 60 cell pvs. at this point it really boils down to the first viable number of cells that would be higher than nominal and that number is governed by the pv makeups. it is possible to have a 36 cell pv in series with a 60 cell pv to attain a better voltage as long as the currents are within a certain percentage of each other.

    for instance say we have a 36 cell pv at 18.1v vmp and 7.6a imp in addition to a 60 cell pv at 29.4v vmp and 7.5a imp. this would be good for charging a 24v battery bank as the number of cells exceeds the nominal for 24v charging which would've been 72 cell, but is in fact 96 cells and quite viable. now if in keeping the pvs identical you would have need of doubling the pv cell count for 60 celled pvs because the 60 cell pv does not provide enough voltage even to meet the nominal voltage requirements as that would represent 72 cells in series. the 2 60 celled pvs would give 120 cells in series and is the closest one can get to the nominal and still be higher than that nominal.

    i do hope my ramblings clarified this some rather than confuse. if one were to have a 40 celled pv it would suffice quite well for mppt operations for a 12v battery in most locations. i would be leery of some locations that exceed 110 degrees f often and higher cell counts in the pvs would be desired over the 40 celled pv then to overcome voltage losses from the high heat.
  • Wanderman
    Wanderman Solar Expert Posts: 180 ✭✭✭
    Re: Slightly different voltage panels.

    I was doing quite well until your post Niel! I have to figure out what number of modules to place on each of these dimensionally challenged panels to get 3 duplex home runs back to the charge controller (actually to the common negative and 3 fuses to a common positive THEN to the controller) The question is what is the VARIANCE in module/cell numbers that will give me the best efficiency even at slightly differing voltages.

    Rich
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: Slightly different voltage panels.

    sorry if i confused you. in your original question of paralleling differing voltages, bill likes to state within 10% of each other, but i like a tighter tolerance of 5%. the farther away from each other there will be some increases in the losses in efficiency.

    what i said was more comment/continuation from bill's last post and may have strayed too far off from your original concerns. you may just ignore my last post if it confused you.
  • Wanderman
    Wanderman Solar Expert Posts: 180 ✭✭✭
    Re: Slightly different voltage panels.

    Niel,
    No worries. I've been a bit brain absent this past week. I was teaching a class in the field and it was eating up most of my bandwidth :)

    Assuming a 12V battery bank, and a ~24v MPPT "sweet spot. I have figured out that I can build (in the space available) a 36 module panel, a 40 module panel and a 60 module panel. The trick here is to maximize watts/amps out while minimizing the differences (and therefore efficiencies) of each panel's voltage.

    I have 3 spaces to fill:

    #1 26 1/2" x 100 1/2" (1956x680mm)

    #2 26 1/2" x 64 1/2" (1066x680mm) This could be extended out to 34" wide.

    #3 41.5" x 39" (1066x990mm) this can have the 39" extended to 45" if needed/wanted.

    That's all I have. This would be with 6" or so modules.

    Rich
  • Wanderman
    Wanderman Solar Expert Posts: 180 ✭✭✭
    Re: Slightly different voltage panels.

    OK, after some additional research these are some conclusions. Assuming 6" cells, I can have a 6x7, a 5x9 and a 4x11 in the spaces available.
    Those are:

    #1 26 1/2" x 100 1/2" (1956x680mm)
    #2 26 1/2" x 64 1/2" (1066x680mm), this can extend to 30.5" instead of 26 1/2"
    #3 43" x 39" (1066x990mm) to go larger here would involve "short" strings as the space gets narrower (on the 39" side) as it gets longer.

    That works out to 42 (6x7), 45(9x5) and 44 (4x11) cell modules (panels) assuming around 1 volt (maybe a tiny bit more) that gives me ~21v, 22.5v
    and 22 v respectively. I believe with the fractions above 1v I will have very close to 23v, 24.5v and 25v. I could be wrong as I do not have the
    exact cell specs.

    I can add an extra string to each of the long ones (not the 43" x 39") to be at: 50, 48 giving volts at 25v, and 24v (again a bit more based upon the
    extra fraction per cell)

    This will feed a Morningstar TS-45-MPPT controller with all three duplex (plus AND minus) wires going to a fuse panel. Common neutral and separately fused positives then feeding into the controller with 1 Negative and 1 Positive.

    It appears that the Morningstar is most efficient, charging a nominal 12V battery system right at 24V input.

    Will this cause me issues with efficiency? Which voltages would you suggest? Any other ideas?

    I really only get one shot here and want to maximize output and efficiency.

    Thanks!

    Rich
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Slightly different voltage panels.

    I would not worry too much about the "exact" Vmp-array vs the MPPT controller's optimum efficiency... You are probably only talking about 1-2% maximum for best case vs worst case efficiencies.

    Another question--You are using ~1.0 volts per cell? Usually it is 1/2 volt per cell and strings are ~32-36 series cells for a "12 volt" panel...

    My suggestion is to try for a minimum of 36 cells in series. You can go for ~72 if that works out for you (i.e., a "24 volt" panel or Vmp~36 volts).

    Another reason for "standardizing" with either "battery rated" panels, or GT panels (Vmp~30 volts for many)--Is it allows you to mix/match off the shelf panels down the road with your custom panels (say you buy a couple panels you can set beside the RV when camping for extended periods of time).

    I would avoid doing "custom for you" Vmp voltages--You will be stuck in the custom panel market if you add/replace panels later (unless you purchase a 2nd MPPT controller for those "additional" off the shelf panels).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Wanderman
    Wanderman Solar Expert Posts: 180 ✭✭✭
    Re: Slightly different voltage panels.

    Bill,
    First off, I have no choice as there are no standard sized panels that will fit nor thin semi-flexible versions available. The chinese ones I had made a couple of years ago (three years really) died in 15 months. These will not be on an anodized aluminium substrate so should not die the same way.

    The cells are all .5~ volts so a 44 cell module will give us around 22+V.

    The numbers I have specified is the current and in some cases maximum cell count for the specific spaces I have to work with.

    There is no adding panels later. This is all the space I have to work with.

    Rich
  • Wanderman
    Wanderman Solar Expert Posts: 180 ✭✭✭
    Re: Slightly different voltage panels.

    OK, so I have temporarily installed the new prototype panels.

    The main oddity I have is Morningstar Tristar MPPT 45 related.

    These panels should output around 21Voc and ~8amps. oc is fine, checked with am meter.

    I know the tristar remote meter displays are weird, but what should the main display read?

    I was seeing 14.7v (that's what it's set to) at around 10-11amps. The watts calculate out to around 150 or so.

    This all seems OK, but shouldn't the panels be around 300 watts total into discharged batteries?

    What am I missing?

    Rich
  • bill von novak
    bill von novak Solar Expert Posts: 891 ✭✭✭✭
    Re: Slightly different voltage panels.
    Wanderman wrote: »
    I was seeing 14.7v (that's what it's set to) at around 10-11amps. The watts calculate out to around 150 or so. This all seems OK, but shouldn't the panels be around 300 watts total into discharged batteries? What am I missing?

    If you are at your setpoint then you are not getting full power. It is regulating to keep voltage from going over the setpoint. Either:

    1) your batteries are more charged than you think they are
    2) the bank is too small
    3) the bank is too old so the ESR is high
    4) your wiring is too thin (also adds to ESR)
  • Wanderman
    Wanderman Solar Expert Posts: 180 ✭✭✭
    Re: Slightly different voltage panels.

    OK, I will drain batteries again and re-test. Sounds Reasonable!

    Rich