zanydroid
Solar Wizard
Hmm, ok, so if disconnect is open effectively the resistance of the panels is added to the load induced by the multimeter. Though that should be a high impedance measurement already.
Voc on used panels, MPPT Voc limit etc, temperature coefficient
- Thread starter CaliSunHarvester
- Start date
zanydroid
Solar Wizard
Yeah at this point it's behaving a little too weirdly to be comfortable operating IMO.
CaliSunHarvester
Solar Enthusiast
I wonder if the weirdness has something to do with this:
Yeah, I paid for this; the guy in the blue shirt is on my payroll too.
CaliSunHarvester
Solar Enthusiast
Good news. I returned from a 3 day trip, had some time and everything is back to "normal", if you can call the lower voltage "normal". I started picking up panels and with the first one, I found the culprit. A poor MC4 connection. Redid this one and, to be safe, 3 others and both strings are delivering Watts now. And I found 2 missing tools under this panel!
Lessons:
1) don't make MC4 connections after sunset on a roof without a flashlight.
2) if the metal insert of the MC4 didn't make the click noise upon closing, do it again. Maybe use a different brand. But don't mix brands e.g. metal piece from one vendor with housing from another.
3) Don't try to stretch wire that is too short. Especially if they are connected with MC4s that were assembled in the dark and didn't click. This is not paint (which *can* be stretched, even and *especially* in the dark).
I did everything from this list.
Lessons:
1) don't make MC4 connections after sunset on a roof without a flashlight.
2) if the metal insert of the MC4 didn't make the click noise upon closing, do it again. Maybe use a different brand. But don't mix brands e.g. metal piece from one vendor with housing from another.
3) Don't try to stretch wire that is too short. Especially if they are connected with MC4s that were assembled in the dark and didn't click. This is not paint (which *can* be stretched, even and *especially* in the dark).
I did everything from this list.
NZ12VOLT
New Member
Gezz hope you dont pay him much (sorry) Laurel & Hardy clip,understand the reason for a break,someone else needs lessons too ?
CaliSunHarvester
Solar Enthusiast
So, everything is assembled and both strings have 12 panels. At times, one string puts out much higher voltage than the other one. Like right now:
Interesting enough, the current is the opposite way. The resulting power is virtually the same (both low due to overcast/ rain).
Should I leave it alone? The lower voltage is set by the MPPT? Sometimes they have very similar voltages. Here are more numbers:
More similar Volts:
Bigger difference, string 1 delivers MORE power:
zanydroid
Solar Wizard
Yes, the voltage is picked by the MPPT. Is the open circuit voltage the same?
The wattage looks really close. If there were multiple busted panels the gap would be bigger. It’s possible the slope is so gradual right now that the MPPT isn’t precise enough to find the true maximum, but it doesn’t really matter because a 10% difference in super overcast conditions is just a few watts.
Are they oriented the same way? Are there trees/other blockers not in the direct path? Those can impact scattered light.
I would expect them to have similar voltage if there’s no shading on the panels.
Maybe ask around on the SRNE 10K threads (I think that’s the same model you have) to see if there is any detailed MPPT graphs you can look at.
The wattage looks really close. If there were multiple busted panels the gap would be bigger. It’s possible the slope is so gradual right now that the MPPT isn’t precise enough to find the true maximum, but it doesn’t really matter because a 10% difference in super overcast conditions is just a few watts.
Are they oriented the same way? Are there trees/other blockers not in the direct path? Those can impact scattered light.
I would expect them to have similar voltage if there’s no shading on the panels.
Maybe ask around on the SRNE 10K threads (I think that’s the same model you have) to see if there is any detailed MPPT graphs you can look at.
CaliSunHarvester
Solar Enthusiast
I just went, turned off solar input at the disconnect and measured 365V for string 1 vs 440V for string 2. About 60F outside.
That's 30.4V vs 36.7V Voc (manufacturer's number is 37.8V)
But if 2 panels were bad (based on Voc), why is power output the same?
No shading- string 1 is far in the back and string 2 in the front:
zanydroid
Solar Wizard
So on an overcast day the shading is weird. Trees that do not block direct light will block indirect light, and on overcast days most of the light is indirect. That is, the power output is sort of proportional to the % of sky (in all directions) that the panels can see. This effect isn’t something talked about much in the forums here (I think people should arrange their series strings so a “circle” covering them is as small radius as possible, as opposed to making really long lines. The more compact layout gives them a better shot at seeing the same conditions. Not sure if this is a standard recommendation).
The panels closer to the trees will have more % of the sky covered.
I don’t have a good explanation for why the two panels being bad isn’t affecting the output more. One would be if they are already low output due to direct shading or indirect shading. This feels like too coincidental to be likely. Or something else is dragging the long string. Like there’s two bad panels that still have voltage. You would need 2 panels worth of damage spread across the array to match the loss from those panels not developing any voltage
Can you calculate the Vmpp expected based on spec sheet and compare that with the operating voltage? That would help identify panels that can’t keep up as the MPPT tries to draw more current.
If you think the panels see the same irradiance, then they should all be at the same output current. I think a cracked solar cell or other defects manifests as keeling over earlier than it should.
CaliSunHarvester
Solar Enthusiast
Hmmm.. those trees are like 100ft away!
Manufacturer has Vmpp at 30.7V
Will respond further later
zanydroid
Solar Wizard
LOL. Then the difference will be close to zero. All the panels will see the same level of blocking. It is as you say proportional to distance.
I have trees right on top and the panels that have trees on two sides are punished more than one side, on a day with overcast. (I’m using microinverters with module level monitoring so the data is very clean to analyze)
I see a structure to the top of your picture, that could block more scattered light on the closer panels.
Big picture: I have no clear ideas other than throwing out a bunch of ways you could be seeing a drop.
To get more intuition you might plot the power/voltage/current vs time, over the course of a day , for each string. Hopefully that will yield some anomalies, like discontinuous jumps, that can be dug into
CaliSunHarvester
Solar Enthusiast
The structure in the back/ upper right corner has another array that is near perfectly south facing.
These 2 new strings in question only get shade when the sun is setting. Maybe at 5pm these days.
I will do the previously suggested binary search, as I have 3 spare panels to replace damaged ones.
Last edited:
Similar threads
