Shading on Solar Panels

[sailinglegend]

Originally Posted by noelex 77

...In practice, Vmp will be lower than 18v unless you are cruising artic areas...

Another pointless pedantic argument from a moderator who should know better.

I said a 32 cell under shading was "about" 16v, with 36 cells it's about 18v. It's the principle that's important not the absolute value.

[Celestialsailor]

I wonder why all my 36 cell panels have been over 20v open circuit

[noelex 77]

Let's have a look and see if we can find why in the video when one panel was shaded the output from both panels dropped to very little.

Unfortunately, this requires an understanding of solar panel voltages. Assuming the Vmp of a 36 cell panel will be 18v or more, is simply not accurate in the real world. If the Vmp was 18v the bypass diodes should have allowed the charging rate to be much higher.

There is very little technical information on the first panels, Go Power, used on the test. Fortunately Kyocera also tested is a reputable manufacturer and publishes comprehensive specifications. For the KD series of panels the specs are:
Voc 22.1v
Vmp 17.9v
Bypass diodes 8
Temperature coefficient is 9.26x10-2

If you want to understand how solar panels work you need to understand these numbers.

Voc is largely irrelevant. At the open circuit voltage, the panel is by definition producing zero power. Vmp is the important voltage as indicates the voltage where the panels will be producing the most power. However, 17.9v is measured under very artificial conditions. The solar panel manufacturers use very bright flashes of light to measure this number. These brief flashes don't allow the solar celłs to heat up, as they usually would in very bright conditions in the real world. Fortunately, all reputable manufacturers produce a temperature coefficient so you can work out the voltages in real world conditions.

The solar cell temperature was not measured in the video, but in conditions shown the cell temperature would be around 45°C. Thus the theoretical Vmp in very bright conditions is 17.9-0.0926x20= 16.0v

The 16v only applies to very bright conditions and a new perfectly clean solar panel. It does not take into account the voltage drop in the wiring, connections and the regulator. The voltage drop in wiring alone is around 3% so another 0.4v even in an well installed system.

Nevertheless, with an unshaded panel the voltage at the controller should always be above the maximum charging voltage, which might be 14.6v for lead acid battery.

Deeply shaded solar cells will transmit very little current. A deeply shaded solar cell acts like a break in the wiring. Solar panels have bypass diodes that allow the current to travel around these deeply shaded cells. If panels are connected in series and one panel is shaded, the current from the unshaded panel must pass through these bypass diodes before it reaches the battery. Each one of the bypass diodes introduces some voltage drop. The Kyocera panels have 8 bypass diodes (the number is not specified for the Go Power panels, but is likely to be less). The Kyocera panels use the best Schottky diodes with a voltage loss of about 0.35v per diode. Unfortunately, with a totally shaded solar panel the current from the unshaded solar panel must pass through each one of these diodes (they did not install any additional bypass diodes).

This means the voltage reaching the solar controller when the panel is at Vmp is 16-0.35x8=13.2v. This is without taking into account voltage drop in the wiring connections etc. We don't know the voltage of battery in the video. Another complicating factor is that it was a lithium rather than a lead acid battery, but it is likely the unshaded solar panel could not operate at close to its Vmp hence the very poor output.
These calculations explain why the solar panels did so poorly when wired in series with one panel shaded.

Originally Posted by sailinglegend

Another pointless pedantic argument from a moderator who should know better.

As you can see, assuming the Vmp of 36 cell panels is "always over 18v" is not accurate. In fact, it is a couple volts out. To explain the behaviour of the solar panels seen in the video and in other real world situations the correct numbers need to be used.

[thomm225]

Originally Posted by noelex 77

Let's have a look and see if we can find why in the video when one panel was shaded the output from both panels dropped to very little.

Unfortunately, this requires an understanding of solar panel voltages. Assuming the Vmp of a 36 cell panel will be 18v or more, is simply not accurate in the real world. If the Vmp was 18v the bypass diodes should have allowed the charging rate to be much higher.

There is very little technical information on the first panels, Go Power, used on the test. Fortunately Kyocera is a reputable manufacturer and publishes comprehensive specifications. For the KD series of panels the specs are:
Voc 22.1v
Vmp 17.9v
Bypass diodes 8
Temperature coefficient is 9.26x10-2

If you want to understand how solar panels work you need to understand these numbers.

Voc is largely irrelevant. At the open circuit voltage, the panel is by definition producing zero power.

It's nice to know the Voc though because it can tell you if your batteries are nearing full charge.

The Voc on my 100 watt panel is around 22.1 volts also

When I first turn off all the lights, instruments etc the voltage on the panel side will be about the same as the battery voltage. (12 something)

As the battery charges, on a sunny day, the voltage on the panel side of the controller will increase.

After a while, I usually check it the next day, the voltage on the panel side will be nearing 22 volts (on a bright sunny day) which indicates the battery in almost fully charged due to the decreased current flow

[longjonsilver]

If shading of the solar panels is such a big deal, then like so many do and put the panels aft on the bimini or arch or davits is a good idea. However, for panels on the coachroof like some do, can we not reduce the shading of these panels by lifting the boom with the topping lift so as it is right next to the mast?

Secondly, with our limited financial resources, perhaps the best use of our money is more efficient charging by LiFePO4 batteries, so as to capture a greater percentage of what the sun gives us, rather than to add more panels in locations where there is a greater chance of shade?

just food for thot

jon

[a64pilot]

If I understand correctly then if all other things are equal, your better off with a 60 cell panel in shading situations?

[TeddyDiver]

Originally Posted by noelex 77

Let's have a look and see if we can find why in the video when one panel was shaded the output from both panels dropped to very little.

Unfortunately, this requires an understanding of solar panel voltages. Assuming the Vmp of a 36 cell panel will be 18v or more, is simply not accurate in the real world. If the Vmp was 18v the bypass diodes should have allowed the charging rate to be much higher.

There is very little technical information on the first panels, Go Power, used on the test. Fortunately Kyocera is a reputable manufacturer and publishes comprehensive specifications. For the KD series of panels the specs are:
Voc 22.1v
Vmp 17.9v
Bypass diodes 8
Temperature coefficient is 9.26x10-2
....

I think neither one of those panels on the video had any bypass diode whatsoever..

[noelex 77]

Originally Posted by TeddyDiver

I think neither one of those panels on the video had any bypass diode whatsoever..

Bypass diodes are necessary to prevent local heating of individual cells, so it is very rare to see reasonable sized solar panels constructed without any fitted. The lifespan of the panel would suffer. Usually even the cheapest "no name" Chinese panels have some installed. They are buried in the junction box so they are not always obvious.

The specifications should list the number of bypass diodes. Kyocera "12v panels" commonly have 8 fitted, but most other brands use a lower number.

There is a convincing argument for fitting an additional bypass diode around each panel if you use series connection. If this had been done for series connection in the video, the performance would have been much better. The voltage loss over the shaded panel would have been less. Importantly the reduced voltage loss would have allowed the solar controller to maintain the unshaded panel at its Vmp.

However, I still think parallel connection would have been shown as the best option, so the overall winner would have been unchanged.

[Paul Elliott]

Originally Posted by noelex 77

Bypass diodes are necessary to prevent local heating of individual cells, so it is very rare to see reasonable sized solar panels constructed without any fitted. The lifespan of the panel would suffer. Usually even the cheapest "no name" Chinese panels have some installed. They are buried in the junction box so they are not always obvious.
[...]

Is the bypass diode a recent thing? My 15-year old BP/Shell Solar 100W "12V" panels have two diodes in the junction box, but these are wired as series diodes, not bypass. I recall that my panel can also be configured as a "24V" panel, so perhaps these diodes can be arranged to bypass each of the "12V" cell strings?

In any case, I'm changing my setup so rather than have the three "12V" panels all wired in parallel and going to one controller, each will have their own MPPT controller. I'm going to remove the diodes entirely, unless someone can show me how the two no-diode "12V" cell strings in each panel can become damaged in a partial-shading situation.

I'm also putting in per-panel voltage / amperage monitors, and a V/A/Wh monitor between the combined controller outputs and the battery. This part is more of a science experiment, but I will be curious to see how everything is working. Here's a photo of my work in progress (there will be three Genasun controllers):

[noelex 77]

Originally Posted by Paul Elliott

Is the bypass diode a recent thing? My 15-year old BP/Shell Solar 100W "12V" panels have two diodes in the junction box, but these are wired as series diodes, not bypass.

Bypass diodes are not recent. Two bypass diodes was a common configuration on early panels. I remember an early BP 75w fitted on previous yacht wired in this way. You need to trace the wiring but I would be surprised if they are not both bypass diodes, rather than blocking diodes. The other clue is two blocking diodes does not make any sense if it is a 36 cell panel.

Here is a photo of typical solar junction box with two bypass diodes fitted. Note the orientation of the diodes. I suspect this configuration has been fitted to your panel:

Originally Posted by Paul Elliott

In any case, I'm changing my setup so rather than have the three "12V" panels all wired in parallel and going to one controller, each will have their own MPPT controller. I'm going to remove the diodes entirely, unless someone can show me how the two no-diode "12V" cell strings in each panel can become damaged in a partial-shading situation.

Using a single MPPT controller for each panel is fine, but I would not recommend removing the diodes if they are bypass diodes. In certain conditions individual solar cells can overheat, which is primarly why the diodes were fitted in the first place. Removing the bypass diodes will not cause any improvment in the output.

A blocking diode could be removed benifically from your system, but I would be surprised if either diode is a wired in this way.

Cool display BTW

[Paul Elliott]

Originally Posted by noelex 77

Bypass diodes are not recent. Two bypass diodes was a common configuration on early panels. You need to trace the wiring but I would be surprised if they are not both bypass diodes,rather than blocking diodes. The other clue is two blocking diodes does not any sense if it is a 36 cell panel.

I just checked the panel specs and you're right, they are bypass diodes. Yes, there's no reason to remove them since they don't create a voltage drop in regular service, and do provide some protection. Funny, since these panels have two parallel 36-cell strings, I had assumed that these were series blocking diodes to prevent back-feeding a shaded string.

I will have to study the connections. I wish I could find a schematic of the cell and diode connections. There is a diagram (like the one you showed) with the junction box connections, but no schematic showing how these things are actually connected. With only two bypass diodes, I don't see how we can protect a 2x36 panel used in a parallel system.

[Paul Elliott]

(About bypass diodes in my solar panels)

Originally Posted by Paul Elliott

Yes, there's no reason to remove them since they [...] do provide some protection.

The more I look at this, the less I am convinced that these bypass diodes I have in my 2x36 panels could possibly provide any "hot spot" protection in a parallel panel (or single panel) configuration. Sure, if I were running my panels in series, the diodes would bypass a panel with shaded cells, and protect a the shaded cells, but for parallel, there is no possible connection with just two diodes and two parallel strings that serves any purpose. If the 36-cell strings were bypassed at the 18-cell midpoint that would require four diodes for my panel. There are only two diodes, and I don't think there is a midpoint connection available.

My diodes are doing no harm, but providing no protection.

I think these diodes are there to be used if the panel is configured for 1x72 cell "24V" operation, using the junction box jumpers. In this case each diode would bypass a 36-cell string.

What am I not seeing?

[TeddyDiver]

Originally Posted by noelex 77

Bypass diodes are necessary to prevent local heating of individual cells, so it is very rare to see reasonable sized solar panels constructed without any fitted. The lifespan of the panel would suffer. Usually even the cheapest "no name" Chinese panels have some installed. They are buried in the junction box so they are not always obvious.

The specifications should list the number of bypass diodes. Kyocera "12v panels" commonly have 8 fitted, but most other brands use a lower number.

There is a convincing argument for fitting an additional bypass diode around each panel if you use series connection. If this had been done for series connection in the video, the performance would have been much better. The voltage loss over the shaded panel would have been less. Importantly the reduced voltage loss would have allowed the solar controller to maintain the unshaded panel at its Vmp.

However, I still think parallel connection would have been shown as the best option, so the overall winner would have been unchanged.

If there were a bypass diode present the output of two panels wouldn't drop near zero in series.. ( with one shaded cell)

[CornelisBent]

Originally Posted by Celestialsailor

Someone posted this link awhile ago here and I found it very useful and it may answer all of your questions. It's an experiment with shading.

Guys in the video didn't seem to have bypass diodes in the cabling, one per panel. This is a pretty standard thing to do in series. With bypass diodes, the output would have dropped to only half (instead of close to 0) with one of the two panel badly shaded. Knowledge Db – SUNBEAMsystem.com

It's a very good video and I learned a lot, but it's sad they didn't read up on the use of external diodes.

[Franck Bavaria]

Dear sailors, I have plenty of space on the roof, so between mast and dodger, to install pannels. Furling main sail, so I do not need to walk there.

This "shading" thread confuses me a lot. So it seems that the best would be to install 4 pannels in parallel, in the hope that 3 of them do not get knocked off by shading. A serial mounting, or even 2+2, makes no sense, since Volts do not add up in shading conditions, it seems.

I was also condisering buying thin film CIGS pannel (like the MiaSole FLEX-03N), less power but less sensitive to shading, but it seems it won't work any better (or does it?)

So another headache upgrade. I though I'd put two 150W flexible pannels in serial and a Victron MPPT 75V 15A, but I now understand this won't work well. And also I need to figure out where to put fuses and diodes (if needed ?).

I may end up buying a Honda portable generator and a set of ear plugs. Please, dear eco-friendly sailors, help !