Solar Panel Mounting on Bimini

[noelex 77]

Originally Posted by Pelagic

Great guidance, Thanks again Nolex
For the Voc 41v limit, should I be using STC or NOCT (about 4V less) when making my shortlist?
I am assuming the reason behind a high Voc is that in poor solar conditions it will still flow at above 24.8Volts? (Just not sure how much leeway I have for parallel)

Has anyone found a consumer rating list using NOTC, as some of the manufacturers try their best to hide it!!

If the the Vmp is below the battery voltage the charging will be reduced quite considerably. The STC Vmp is measured with a cell temperature of 25c and in practice will be more like 40 c which reduces the Vmp.
A good general rule of thumb is the solar panel need 36 cells for 12v and 72 cells for 24 v. If has 72 cells it should be fine for your 24 v system. Panels with this number of cells generally have a Voc of of about 44v or a Vmp of about 35v under STC. Not many solar companies list outages for NOCT.

The reason for caution is there are a number of panels with 60 cells. This is a popular type of panel for the domestic solar industry. These panels are often called "24v" panels, but in fact their voltage is too low to efficiently charge 24v panels on their own or in parallel. (They are fine for a 12v system with a MPPT controller)
Typically these 60 cell panels have a Voc of 36.5v and a Vmp of 29v.
36.5v sounds like enough to charge a 24v battery, but particularly at the higher cell temperatures the Vmp is reduced to below battery voltage which is is likely to above 29v during the absorption phase.

Higher voltages and number of cells are fine with an MPPT controller which you want anyway. The good MPPT generally have a limit of 150v. Some of the cheap controllers have a low limit so check if you want to use a very high voltage panel, but for your sized array you really need a good controller anyway.

[mcarling]

Originally Posted by mrm

No theory needed, rather practical understanding of how panels are assembled, what is the effect of shadowing a single cell in a series-arranged panel, and what bypass diodes are for.

I wonder if you might have misunderstood me. Please allow me to clarify my remarks, which may have been less than clear.

I hypothesize that revolving a square solar panel under a given partially shading (with a typical arrangement of 36 cells for a 12V panel, 72 cells for a 24V panel, 144 cells for a 48V panel, etc.) 90 degrees about the vertical axis will not change the power output.

[noelex 77]

The bypass diodes divide the panel different sub panels. On a 12v panel there are typically between 3 and 8 bypass diodes.
These diodes are mainly to protect the cells from overheating, but they do effect power production.

The thing to rember is a deeply shaded cell in a solar panel conducts very little current. Think of a deeply shaded cell like a break in the wiring.

Without bypass diodes one deeply shaded cell will stop production from the other 35 cells. Bypass diodes allow the conduction around the shaded cell. If other cells, covered by the same bypass diode, are shaded it will make no difference to the output. If a cell covered by another bypass diode, is shaded, then another section of the solar panel is shut down and the output of the panel will fall further.

The above is a simplified explanation, ( shading is rarely dence enough so the cell conducts nothing. The voltage may not be sufficient when a portion of the panel is cut out by the bypass diodes) but it illustrates that the pattern of shading in relation to arrangement of bypass diodes effects the output.

Therefore rotating a solar panel about the vertical axis will generally effect the output depending how the shading falls in relation to the bypass diodes.

( not the above applies to rigid mono and poly panels. Amorphous panels work differently)

[mcarling]

Thanks noelex 77! That was a very clear explanation.

[Pelagic]

Originally Posted by noelex 77

A good general rule of thumb is the solar panel need 36 cells for 12v and 72 cells for 24 v. If has 72 cells it should be fine for your 24 v system. Panels with this number of cells generally have a Voc of of about 44v or a Vmp of about 35v under STC. Not many solar companies list outages for NOCT.

Thanks again for taking the magic out of it!!
Measured and confirmed that what is most efficient for me is if I can find fairly square Solar Panels as I have max 4m beam on new soft canopy tapering to 3.2m.

Tested the largest size I found online (4 x Sunpower E20/327Watt) and they would install nicely, with some minor modifications.

Decided to fit panels along outboard edges of both sides of Bimini stainless support as a watertight hard top, removing the canvas underneath to maximize cooling. Then modify fitted center piece (yellow) to flange up to panels to keep things dry.

Shipping everything by Sea Freight to Subic Bay is a given for me as there are no dependable suppliers in the Philippines.

Should be available from HK, Taiwan or Singapore as many Tier One/Two types are made in Asia.

I like the Sunpower E20 Series:
SunPower E20 Series Solar Panels
Does anyone have any experience with them?
Mono Crystal x 96/ Vmp-Volts-54.7/ Imp-Amps 5.98 /
Voc-Volts 64.9/ Isc-Amps 6.46

Found it strange that all the large output panels only have 3 bi-pass Diodes

[noelex 77]

The Sunpower panels are some of the most efficient, but they are unusual. They active some of this efficiency by polarising the surface of the solar cells. To do this they require a positive grounding system which is not practical on a yacht. The sun power information is here:
http://www.ftp.tognews.com/GoogleFil...owerPanels.pdf
There was some discussion on CF here about this:

http://www.cruisersforum.com/forums/...ing-90082.html

These are the only panels that do this. So having said its very simple, I guess there are some traps.

Sun power do make some slightly less efficient panels that are suitable for negative grounding.
There is a representative that has posted on CF who will be able advise you much better on what sunpower panels need possitive grounding.

If you get the sunpower panels that do not require positive grounding your plan sounds great.

[mrm]

Originally Posted by mcarling

I wonder if you might have misunderstood me. Please allow me to clarify my remarks, which may have been less than clear.

I hypothesize that revolving a square solar panel under a given partially shading (with a typical arrangement of 36 cells for a 12V panel, 72 cells for a 24V panel, 144 cells for a 48V panel, etc.) 90 degrees about the vertical axis will not change the power output.

I think I understood you correctly. Gettinthere mentioned a relationship between panel output and its orientation relative to a typical shading pattern in mounting location and called it '(panel's) grain' or '(panel's) direction'.

You responded to that observation with a negation, which you later confirmed:

Originally Posted by mcarling

"I hypothesize that revolving a square solar panel under a given partially shading (with a typical arrangement of 36 cells for a 12V panel, 72 cells for a 24V panel, 144 cells for a 48V panel, etc.) 90 degrees about the vertical axis will not change the power output. "

I responded to your first post by giving you pointers to subjects for further self-study rather than a complete answer, as I deemed it more beneficial long-term. Please bear with me, part of my profession is teaching students

Later noelex 77 gave you a full answer. I will offer a simple analogy too. Imagine a garden hose carrying water. Now you pinch the hose someplace - water stops flowing. Now imagine the same garden hose divided into sections and arranged in such a way, that each section can be bypassed with a pressure relief valve should it become blocked. Again, you pinch the hose someplace. Water continues to flow albeit with some losses.
Think of shadowing of a single cell as analogous to pinching.

Now, this analogy is only an analogy and does not hold water (pun intended) on closer inspection, but I think it is good enough

Going back to gettinthere's observation: if you pinch just one section (think boom shadow along a single string of cells) some water will flow. If you pinch all sections (shadow across all strings of cells) no water will get through (panel will be useless). That is why his observation is valid.

Hope it was helpful.

[Roy M]

I hope this isn't a non sequitar: A new company, SOITEC, with French roots, has opened in San Diego, producing a different type of PV panel. It's described as a concentrated photovoltaic solar technology that uses a lens to focus light on a circular, multi-junction cell no bigger than a ladybug. At present, they are manufacturing arrays that are garage-door sized, and for the utility market. Apparently, they provide a better dawn-to-dusk performance than conventional units. Has anyone heard of these? I'm asking because PV panels will be the LAST things I mount on the boat before sailing over the horizon, as they are continuously improving in design and output. And thanks, to all of those who have shared their experiences, especially with shading and placement of panels. On further investigation, it appears these things are much heavier and bulkier than conventional units, oh well.... And, they require a tracking device to maintain the focus. Rats!

[Pelagic]

Originally Posted by noelex 77

The Sunpower panels are some of the most efficient, but they are unusual. They active some of this efficiency by polarising the surface of the solar cells. To do this they require a positive grounding system which is not practical on a yacht.....//......These are the only panels that do this. So having said its very simple, I guess there are some traps.

Somehow I knew there had to be a catch; it was just too good to be true!

From my Panel research the Mitsubishi concept is different.
“Low Loss Half-cut Cells”
The 2565 HD, has 120 x mono-crystalline Cells with 4 bus bars.

Given the shading issues on sailboats… is this one I should consider even if the panel Voc is at 38.2V for my 24v Battery Bank?

Solar Power - MITSUBISHI ELECTRIC
Solar Power - MITSUBISHI ELECTRIC

[Pelagic]

Everyone has advised me to Max out on Wattage for the open space available.
What I found out is that all of the larger (300W) high output panels are configured long and narrow (1.97m x 0.99m) which is too long for my 4 panel set up!

One exception is the Kyocera 320 Watt (1.662m x1.32m) that fits ok, but only leaves 37cm between the 2 aft panels at end. (See below)
My boom is 12cm wide, so can fit on a centerline crutch in between panels, but from experience is the possibility of added shading still worth using this one?

My other option is to use 250-265Watt panels like Canadian Solar…or the Mitsubishi that x 4 would deliver about 200 Watts less than the Kyocera.

Which is the better trade off that you would choose?

[noelex 77]

I would not dismiss the sunpower panels. I think they make some without the polarisation. These are only slightly less efficient.
I am afraid the Mitsubishi panels are no good. They have 120 cells, but they have wired two lots of 60 cells in parallel. This makes them an effective 60 cell panel.

Look at the Vmp this is the voltage that the panel produces maximum power. The MPPT controllers can reduce the solar panel voltage, but they cannot boost it. So the solar controler will operate the panel at the Vmp ( which varies with conditions), but this has to above the battery voltage.
The quoted Vmp of 31.4v of the Mitsubishi panels is this is under ideal conditions with very low panel temperature. In practice with higher cell temperature, lower light and some shadowing the operating Vmp be lower by several volts than 31.4 and therefore will be lower than the battery voltage during charging.
The controler will have to keep the panel at the battery voltage so the power production will be less. Often much less.
You could wire two of these panels in series, but that is the only way they will work.

[Pelagic]

Thanks Nolex…. Now last question before Xmas…..promise!!
Would you install the Kyocera 320W or the Canadian Solar 265??

[noelex 77]

Don't worry about asking questions, that's what the forum is for.

The only specs for Canadian solar panels were 260w not 265w.

The larger wattage of the Kyocera panels will give you more power. The extra wattage will be greater than the extra loss to shading. Even with one of the 4 Kyocera panels shaded and putting out nothing you still have almost as much. Under NOCT 3x230=690w compared to 4x188= 752w if all 4 canadian solar ,panels were working.

The Kyocera panels are also well made and regarded, the Canadian solar are more of an unknown.

[tartansail]

Originally Posted by noelex 77

I would not dismiss the sunpower panels. I think they make some without the polarisation. These are only slightly less efficient.
I am afraid the Mitsubishi panels are no good. They have 120 cells, but they have wired two lots of 60 cells in parallel. This makes them an effective 60 cell panel.

Look at the Vmp this is the voltage that the panel produces maximum power. The MPPT controllers can reduce the solar panel voltage, but they cannot boost it. So the solar controler will operate the panel at the Vmp ( which varies with conditions), but this has to above the battery voltage.
The quoted Vmp of 31.4v of the Mitsubishi panels is this is under ideal conditions with very low panel temperature. In practice with higher cell temperature, lower light and some shadowing the operating Vmp be lower by several volts than 31.4 and therefore will be lower than the battery voltage during charging.
The controler will have to keep the panel at the battery voltage so the power production will be less. Often much less.
You could wire two of these panels in series, but that is the only way they will work.

What is your opinion of controllers like the Genasun that claim to boost the voltage? If that works, it would seem to address this issue. The early Solbian panels were smaller and put out marginal voltage for 12 volt systems. The Genasun's were recommended to bump the voltage. The current larger Solbian's don't need the voltage bump under nominal conditions, but it still seemed like a useful feature to mitigate the impact of shading.

[noelex 77]

Originally Posted by tartansail

What is your opinion of controllers like the Genasun that claim to boost the voltage? If that works, it would seem to address this issue. The early Solbian panels were smaller and put out marginal voltage for 12 volt systems. The Genasun's were recommended to bump the voltage. The current larger Solbian's don't need the voltage bump under nominal conditions, but it still seemed like a useful feature to mitigate the impact of shading.

The voltage increase is not normally needed. Even under low light solar panels will still produce lots of volts just no amps. It is possible sometimes when some cells are under shade to get more power by boosting voltage, but in practice the panels will be all be at different voltages.

So voltage boosting is only used under very specific circumstances such as tring to use a 12v panel in a 24v system.

Solbian have manged to produce some very efficient panels. It's cheaper (and possibly more efficient) for them to use the same cell in different panel sizes. This means the smaller panels have fewer large cells instead of the conventional approach of leaving the number of cells constant and decreasing the size of the cells for a smaller panel.
Because the small panels have fewer than 36cells the voltage needs to be boosted. There are very few solar controllers that will do this. Gensun make one of these.

The effeciency loss in the controler in boosting the voltage needs to be traded against the improved effeciency of the panel. They can, if enough panels are available be hooked up in series, but this does introduce some new advantages and drawbacks.
The low voltage panels with a controller boosting the voltage are a new and insteresting development I have not personally used them and it would need some controlled tests to match the output against "conventional panels". At the end of day features like light weight and flexibility are probably more important than a small difference in effeciency.
They are unfortunatly very expensive at the moment.

The only caution I would suggest is if using a low voltage panel that the output may be a bit less once the voltage boosting has taken place.

If using conventional voltage panels the parasitic losses of driving a controller that will boost voltage are likely to outweigh the rare occasions where the voltage boost is helpful.