Solar panel diodes and series connection

South Africa
South Africa Solar Expert Posts: 295 ✭✭✭
Ok, my solar panel connection box burnt out, posted a while ago here. Appears that it is not such an uncommon thing.

The original diodes fitted was rated at 40v,12amps. Worked perfectly when the 2 x 200w panels where connected in series.
But 30 days after I changed the panels to series, for a MPPT controller, the box melted.

So my questions is:
When connecting the panels in series, volts hitting 59.6v with ease midday, are the diodes in the connector box supposed to match the voltage?
Or are the diodes meant per each string in the panel only, and not the full voltage that the systems generates?

Thank you in advance for your feedback.
5kVA Victron Multiplus II, 5.2kW array, 14kWh DIYLifepo4 bank, all grid-tied.


Comments

  • 706jim
    706jim Solar Expert Posts: 514 ✭✭✭✭
    Don't think anyone uses diodes anymore when using any charge controller.
    Island cottage solar system with 2500 watts of panels, 1kw facing southeast 1.3kw facing southwest 170watt ancient Arco's facing south. All panels in parallel for a 24 volt system. Trace DR1524 MSW inverter, Outback Flexmax 80 MPPT charge controller 8 Trojan L16's. Insignia 11.5 cubic foot electric fridge. My 30th year.
  • South Africa
    South Africa Solar Expert Posts: 295 ✭✭✭
    There must be diodes in the junction box, to protect the panel if it is partially shaded.
    5kVA Victron Multiplus II, 5.2kW array, 14kWh DIYLifepo4 bank, all grid-tied.


  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Today, we use series protection fuses to protect against a short circuit in one panel of a large solar array (typically 3 or more parallel strings of solar panels).

    In the olden days, they used series diodes (one per string) to prevent current back flow in the event of a short circuit (or, other reasons that I can only guess at).

    Diodes do work at replacing series fuses/circuit breakers... But diodes do have a voltage drop (0.2 to 1.0 volt or more)--And that is wasted energy plus excess heat in the "diode" array.

    Also, single diodes are not recognized by NEC/UL as "fuse equivalents" for safety unless you have at least 2 diodes (or one diode and one resistor) in series (primarily used in computer systems with rechargeable batteries for running the computer clock when the power is shutdown)... It is assumed that a diode can (and they do) fail shorted and are no longer a "protective devices".

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • South Africa
    South Africa Solar Expert Posts: 295 ✭✭✭
    Cool, so I will cut all the diodes out of the J-Box on the back of the panel, solder the wires to the connectors and add a fuse in place of the diodes.
    5kVA Victron Multiplus II, 5.2kW array, 14kWh DIYLifepo4 bank, all grid-tied.


  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    The diodes in the junction box are bypass diodes--They need to stay in the box to protect the cells in case one or more cells are shaded--Particularly if you have several panels in series.

    The other diodes I was typing about are series "blocking" diodes... They are not used much (if any) now in solar panels. And the use of series type solar charge controllers (the typical PWM and MPPT controllers that we have today) block battery voltage feeding current back into the panels over night. Older charge controllers where frequently "dump" controllers when used with solar arrays--Just like with wind turbines. Panels were permanently connected to the battery bank and needed blocking diodes to prevent leakage current from discharging the battery bank (particularly with 24 volt and above battery voltages).

    Just in case you weren't joking--And to prevent confusion with other users.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • South Africa
    South Africa Solar Expert Posts: 295 ✭✭✭
    Thank you Bill. ;)

    Now it is clear as mud.

    Learned long time ago, that when in doubt, ask a question in such a way that it can cause consternation. It is one sure way of getting down to the point of no interpretation. :-)

    Bypass Diodes will stay.

    http://www.digikey.com/en-US/article...nd-performance
    5kVA Victron Multiplus II, 5.2kW array, 14kWh DIYLifepo4 bank, all grid-tied.


  • South Africa
    South Africa Solar Expert Posts: 295 ✭✭✭
    But that goes back to the original question (improved version):
    When connecting 2 panels in series, volts hitting 59.6v with ease midday, are the bypass diodes in the J-box supposed to match the system voltage?
    Or are the J-Box bypass diodes specced to protect the cells in each panel only, and not the full voltage that the 2 panels in series can generate?
    5kVA Victron Multiplus II, 5.2kW array, 14kWh DIYLifepo4 bank, all grid-tied.


  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Think of a flashlight with 10 akaline cells in series. Batteries are fully charged, each provides 1.5 volts at @ 1 amp (or whatever).

    However, if one cell goes high resistance... Now, you have 13.5 volts from the other batteries trying to get current flow through the one battery outputting little voltage/current. That one battery may see upwards of 13.5 volts in reverse voltage and can a) kill any current through the string or b) can over heat.

    Solar cells are sort of like this... With equal sun on each cell--Each cell wants to output 0.5 volts @ 1 amp (for example). If one cell is shaded, that cell goes "high resistance" (no voltage generate and no current flow).

    When that happens, the cell gets "reversed biased" (goes from +0.5 volts across the cell to negative array voltage across the "dark cell"). Remember the solar cells are just "giant diodes", and when they are reverse biased, it putting voltage across the diode backwards. For Solar panels, roughly this maximum reverse voltage is ~10-12 VDC. If the reverse voltage goes over that rated reverse voltage, it will damage the cells and even cause current to flow through them--And remember that Power=I2R=V*I -- So that the dark cell can overheat and catch fire.

    The Bypass diodes are setup to "short" solar array current around any "dark cells". Preventing damage to the dark cells.

    If you have only a 12 VDC panel--You may not even need bypass diodes. But when you have a high Vmp panel in a 100-600 volt solar array string, the bypass diodes are very much needed to prevent damage/reduce the safety risks of a dark or failed cell.

    The Bypass diodes do not add or recover any voltage from a dark cell--The panel with a single dark cell can lose upwards of 10 volts of output voltage (depending on the exact panel wiring and number of bypass diodes).

    This is the same as if we put a bypass diode around each 1.5 volt battery in a flash light. If one cell goes dead, the battery stack will still output current, just at a reduce voltage around the "bypassed" bad cell.

    If you have two parallel battery strings, the "failed cell" string will not output very much (if any current) because its output voltage is reduced vs the other parallel string with all "good cells".

    Yes, if you can guarantee that a solar panel never has shade (when the sun is up)--You could cut out all bypass diodes and the panels will run just fine... But if you have a bird perch on the array (or leave a present behind), then you run the risk of cell damage/panel fire.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    Cool, so I will cut all the diodes out of the J-Box on the back of the panel, solder the wires to the connectors and add a fuse in place of the diodes.

    NO !! Those are bypass diodes, for when part of the panel is shaded !
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • South Africa
    South Africa Solar Expert Posts: 295 ✭✭✭
    Got it.

    Thank you Bill.

    Ps. Not a chance I will put up a panel without the bypass diodes in place.
    I have seen what happens if they are damaged.
    Have felt the heat of the panel due to damage diode ... and I have touched the panel connection with panel in full sunlight. Finger is still smarting ... it was 'smoken' hot.
    5kVA Victron Multiplus II, 5.2kW array, 14kWh DIYLifepo4 bank, all grid-tied.


  • RCinFLA
    RCinFLA Solar Expert Posts: 1,484 ✭✭✭✭
    Protect diodes are incorporated into panels to prevent cells with low resistance shunt leakage from getting very hot during periods of partial shading. Usually a panel will put a parallel reverse bias diode across a down and back pair of cells within a panel. This makes it convenient to connect at junction box.

    A little bit of explanation on cell shunt defects is in order. A 1 ohm shunt defect resistance (on the extreme end of defects) means little during normal forward cell operation when only about 0.55 v is across the shunt cell defect. It amounts to about a half amp loss in overall panel current production. When a section of the series string is shaded, the illuminated cells still produce there open circuit voltage. This will reverse bias the shaded cell. The reverse bias can be high voltage depending on how many cells are encompassed by a single bypass diode. If 20 cell series is encompassed by a single protection diode then the maximum reverse bias on a single shaded cell would be about 19 x 0.55v or 10.5 vdc. The shaded cell that happens to have a spot shunt defect of 1 ohm now get hit with 10.5 vdc / 1 ohm or up to 10.5 amps flowing through the 1 ohm spot defect. If the panel has cell size that produces 8.5 amps max then maximum would be 8.5 amps through defect site. This is still 10.5v x 8.5 amps through the defect or 89 watts pumped through a small defect area on the cell. It can get blistering hot. Image what would happen with no shunt protection diodes with even a higher reverse bias across the defect spot. So that is the reason for the protection diodes. Typically two or three diode are in the junction box, usually equal to the number of up-down row pairs within the panel. Six cell width panels will have three diodes. No more then 24 series cells (13v) covered by a single protection diode.

    PV cells will have shunt defects. Ones with lower shunt resistance, weeded out by panel manufacturers, end up for sale on eBay.

    Now for need for reverse bias rating on diode. In normal full illumination operation, theoretically the maximum reverse bias for a diode covering 20 cells in series would be 10.5 v plus some extra voltage for cold weather where the cold would have higher voltage per cell. Due to minor mismatching of components there may be higher voltage on some diodes. It is like connecting ten diodes in series and putting a high reverse bias across the series stack. If they all match, having same leakage, then the high voltage is equally distributed across each diode. If reverse bias leakage on the diodes are not exactly the same the higher leakage (lower resistance) diodes get less voltage and the other diodes get higher voltage. If they exceed the reverse breakdown voltage then they have to carry high current at high voltage meaning they get very hot and blow out.

    Most panels have too small, axial lead diodes. When are they are needed during shading there is up to 8.5 amps x 0.9v or 7.7 watts of power being dissipated by a small axial leaded diode. The axial lead diode has insufficient heat sinking to handle that much power and it gets very hot.

    There is no need for low voltage drop hot carrier / Shottky diodes for protection diodes. These type of diodes usually only come with lower breakdown voltages. I recommend TO-220 packaged regular silicon diodes that can readily have high breakdown voltage rating and can better handle the heat of 8.5 or so amps when called upon during shaded periods. Electrically insulate the TO220 tabs and connect them to a heat sink metal plate or case of the junction box. This will result in a near bullet proof solution..
  • South Africa
    South Africa Solar Expert Posts: 295 ✭✭✭
    RC, thank you. You answered my last question: What diodes to put back in?

    Panel was never shaded, I connected it wrong for less than an hour, fixed that, then 1 month later the problem surfaced.
    So I am hoping that it was a case of the one diode weakened (as a result of my weak state of mind at the time), and therefor failed.

    By replacing all of them, I hope to resurrect the panel, but as BB also said, I will not put it back in production but I can still use it for other projects where the panel can be checked frequently.

    And if not then I will learn something new (always a silver lining): Where does one recycle solar panels in Cape Town? :-)
    5kVA Victron Multiplus II, 5.2kW array, 14kWh DIYLifepo4 bank, all grid-tied.


  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    If one or more of the bypass diodes fails open, then the panel will continue to be usable but will be vulnerable to shade damage.
    If one or more of the bypass diodes fails shorted, then the corresponding group of cells will not add any voltage to the panel output.
    Either way, I would recommend replacing all of the diodes, even those that test good, with new higher power rating diodes. There may have been a design flaw in the original diode specification that allowed them to burn out when stressed over a long period of time.
    If one panel is in circuit and completely shaded, all its bypass diodes might conduct at the same time, causing more combined heat than they were designed for.
    The maximum voltage across a bypass diode will be the output of that group of cells only, so they do not even need to be rated for the full panel voltage.
    If you put one panel backwards in a series string of three or more panels, then the bypass diodes could be subject to a reverse voltage equal to the Voc of the rest of the string. That could indeed cause them to break down in a way that creates a medium resistance short which could then burn out the diodes completely later in either reverse or normal configuration.
    SMA SB 3000, old BP panels.
  • South Africa
    South Africa Solar Expert Posts: 295 ✭✭✭
    Thank you inetdog.

    All will be replaced with TO-220 packaged regular silicon diodes, if I can find them, 40v, 12amps range.
    5kVA Victron Multiplus II, 5.2kW array, 14kWh DIYLifepo4 bank, all grid-tied.