Interesting. Does that mean that performance in a multi-string system could be enhanced by measuring each panel's current output and putting panels with closely matched current output in the same string?

Yes, the issue though is you can't pick what the factory sends you and without the factory flash tester data you don't know what the panel performance is.

You can't just stick it in the sun as everything is changing (tempature, sun angle) and you would need an electronic load or curve tracker to even attempt this which most home owner don't own

Sorry for joining the discussion late and not letting this subject die. Solar Guppy, I have read your previous posts on this subject and thought I understood them but now I am not so sure.

If you have multiple strings in parallel, and one string is subject to shading, most likely some bypass diodes have kicked in on that string resulting in a lower voltage on that string relative to the unshaded string. As the strings are physically connected at the inverter, only 1 voltage can exist there. Does the unshaded string "pull up" the shaded string or does the shaded string "drag down" the unshaded string or a combination of both?

If the voltage is "dragged down" to any extent in parallel strings with shading, the best design practice would seem to be to stick with single strings, select a inverter with a large MPPT voltage range and operate unshaded near the high end of the range. If there is any shading, the output would only see to suffer by the percent shaded (down to the lower end of the MPPT), not disproportionately. This would favor the smaller string inverters (like the SMA 2500 and Xantrex GT3.0) and not all these larger inverters where paralleling strings is necessary to get near their full output. It would also mean that the Multi-string capability of the Sharp Inverter can be a benefit in sites subject to shading.

Any comments are appreciated.

The string that has a shaded panel(s) will just operate closer to the VOC of the shaded strings combined panels left generating voltage.

If the voltage of the shaded string ( at voc ) is less than the string in sunlight , then very little power back-flows into the shaded string. What most poeple don't know is a dark string wont draw much power, I have run tests for this to verify things for designs and it comes down to the fact that solar panels are basically diodes. In total darkness , applying a VOC voltage to the array which is back-feeding with bench supplys and measuring with lab quailty power analyzers to the one of my 3000 watt arrays, it has a draw of 15 watts. Not allot and in any type of day-light its zero.



Yes, a single string is always better than parralled strings due to mis-matches between the strings. Every single solar panel is a little bit different. When you have a single string , even with panels that have mis-matchs in current of a few percent, the panels with the higher current ability will be limited by the lowest current panel BUT , they just run at a slighty higher voltage and hence slighty lower current and end up producing near optimum power.

Having said that, ALL inverters perform much worse at the higher voltages , this is due to high switching losses in the electronics. Running string inverters near there bottom end will from the inverters prospective yeild the best efficiencys



I think the large 5kw+ inverters are to meet marking demand, it is not the best solution from a panel harvest, cost (they are more expensive than 2 smaller units) and they are ALL fan cooled which I don't like. A pair of GT3.0's will out harvest Any 5kW+ system , for both the panel losses will be lower and the operational range be more optimal for the inverter



The Sharp inverter is the worst performing unit of ALL the CEC certified units, 91%+ efficiency. Mppt is dominated by tempature which for arrays on the same roof, regardless of orientation, will be similar and hench the mppt. I've run intentional miss-matched strings and the differences are 2-3% due to the mis-matching (running a 7 panel string with an 8 panel string) so while its a hit on harvest single panel shading is not really that big of a deal for daily harvest


I don't bother responding to the news-group questions on this, there is some-much dis-information out there. Hope this info helps

Solar Guppy, Thanks for the help understanding the effects of shading on parallel strings.

You mentioned all inverters perform better at a lower voltage (close to the output voltage I believe). What are the ball park numbers? For example, what is the approximate efficiency difference between a GT3.0 running at 220 Vdc and one running at 400 Vdc. Unless the difference is huge, it would seem to be better to run near the high end of the MPPT to give you headroom for some panels being shaded, as well as voltage degradation over time, and panels being in the -5% of their specification.

I am aware that the Sharp Inverter has a much lower efficiency than some other inverters. In spite of this, I have seen some pretty good reported kWh harvest from it though and I was trying to understand this. Most residential installs are subject to some shading, and I speculated that because the Sharp Inverter has 3 smaller independent strings, it has more shading tolerance, which for some sites, may help overcome its lower efficiency.

Another possibility is the kWh reporting. To what accuracy level, can we believe the reported kWh numbers from any inverter (be it Sharp, SMA, PV Powered, Fronius, or Xantrex)? Is this specified in the new CEC inverter tests or will it become the next area of contention between the inverter companies?

You can look at the actual test data at the CEC site for the differences on efficiency vs input voltage. It different for each inverter and is dependent on the design typology of the inverter, some it will effect more than others. On the GT, its about 1.5% I beleive.

As for the Sharp, its the bottom of the heap, the key is as you figured out is the phantom "reported" khw by the unit.

SMA was 2-3% high, PVPowered and Fronius were 5% high, I never tested a sharp so I can't say what it does, but the CEC data speaks for itself. The GT is better than 1%, and I HOUNDED them to make this right, benifit of being part of the design process. I've had discussions with the top levels of some of these firms, they claim the customers "are happy" with the rose colored kWh reported.

While not strickly Grid-tie, the MX60/FX inverter combo is the worst, its off by 25-30%, the benifit of no input current dectect, that's why MX60 owners are always boasting about how great there arrays work with the MX-60, it really aint so!

Partial shading IMHO is a non-issue. What are the chances of getting only one or two panels shaded and the rest of the array in full sunlight?. If your talking about a tree, its going to take out the whole array or enough to make the array basically non-functional from a harvest point, and no, its not that common from what I've seen to have a couple of panels shaded.

I'm not suggesting you design for the minimum, but target your low end voltage using the vendors sizing tools to be within a panel or two for worst case (high heat). In typical condidtions, your be 100 volts higher, only in desert heat will the minimum voltage be an issue

Anyways, thats my take and you get what you pay for advise wise