solar panels parallel or in series

[galynd]

I recently purchase a Renogy solar kit with a pair of 100 watt Renogy solar panels and the MPPT 20A controller. I planned to connect them parallel, via the quick connections, but the cables from the panels are not long enough. However, I could connect them in series, as connecting the pos (+) of one panel to the neg(-) of the other panel is possible. Would this work? I realize I would be drawing 24v + and my boat is 12 v. Would the mppt controller convert back to 12 v with no loss?

Thanks, Greg.

 

[mitiempo]

Whether series will work or not depends on the controller. What do its specifications state as maximum input voltage?

On a boat series generally is a bad idea unless there is zero possibility of shading. If there is any shading the output will drop drastically.

 

[galynd]

Max input power: 260 w (12 volt), 520 w (24 volt)
So if the solar panels are actually putting out greater than 12 volts each would it work?
Also, I didn't think of the shading concern. good point. sounds like parallel is best option.

thanks,

 

[Chief RA]

I swapped the connectors/cables so I could parallel them as they came for series. Easy! Chief

 

[Ron20324]

I wired my 2 panels to 2 switches at the nav station, where the charge controller is. I can change from ser/par, or take either panel off line. As mentioned, because of possible shading series-wired, I leave mine in parallel. I believe when there is less sun, the combination in series keeps the charge to the batteries a little longer- like before sunset- and I'm trying to squeeze every bit of sunlight into the batteries..

 

[JK_Boston_Catalina310]

Get back on Amazon and get a set of Branch Connectors: http://www.amazon.com/RENOGY®-Branc...e=UTF8&qid=1432925436&sr=8-13&keywords=renogy

But to visit the parallel vs. series issue, as Maine Sail has said many times before the answer is: it depends.

Using the branch connectors you can try both parallel and series. Try each setup for a couple of days and record the data. See which actually works on your boat in your setup. Every setup is different and there is no universal answer.

As to the controller, what controller do you have?

I am running a Rogue MPT-3048. It will accommodate up to 100 volts.

 

[Bill Roosa]

the controller will have a "max input voltage". watts is pretty irreverent as it assumes the voltage (which you would have to guess at)
read the manual for the max voltage to the controller to determine if it will be less than the max voltage each panel will put out (added together of course).

 

[galynd]

Thanks for the info. I ordered the 5 ft extension cables.

 

[Chief RA]

Man! All you need to do is open the box on one of them and swap the cables! Parallel hookup with them and it is done, and connect to the box on one of them to your batteries! Whats going on here? Don't make it more complex than it needs to be! If you have a 12v system all this other stuff is not necessary! If you have 24v then hook 'em up as they are. Chief

 

[Maine Sail]

Man! All you need to do is open the box on one of them and swap the cables! Parallel hookup with them and it is done, and connect to the box on one of them to your batteries! Whats going on here? Don't make it more complex than it needs to be! If you have a 12v system all this other stuff is not necessary! If you have 24v then hook 'em up as they are. Chief

Many of these Renogy panels don't have a junction box that can be opened. It is factory sealed and filled with an epoxy like sealant.. MC4 factory installed pigtails are already attached.

 

[Chief RA]

Maine Sail: In all my years I have never seen such a mess! That is one cheap panel. I would then swap the connectors on the ends of the cables as these other approaches are not practical! If connectors are bonded then cut the suckers off and wire them in parallel! Chief

 

[walt]

Some time ago, I know at least Maine Sail had done some solar installs involving multiple panels and each individual panel had its own MPPT controller and the reason for this was to best handle the partial shading that is very likely to occur on a sailboat.

It’s interesting to look at why this makes any difference. Solar panels usually have some internal mix of serial and parallel diode connections. So when you connect two panels, you are just making them a little more either serial or parallel and its complicated for what actually happens with partial shading. Each manufactures panels may behave a little different.

This link is from a recent IJRET technical paper regarding MPPT and partial shading
http://esatjournals.org/Volumes/IJRET/2014V03/I19/IJRET20140319072.pdf

The above is from that paper and figure 7 shows the V/I panel characteristics for full sun and figure 9 shows the same panel with partial shading. Likely different panel architectures might behave different than what is in these figures but this likely does represent a fairly common V/I for a shaded panel.

If we assume we have two panels in our application, take a look at what power you would get with some different architectures for the panel in the figure.

With both panels getting full sun, both panels operate at the maximum power point in figure 7 (65 volts and .6 amps = 37 watts) so the total output power is 74 watts.

Now if we connect a full sun and a partial shaded panel in parallel with a MPPT controller, an interesting thing happens. One panel has is maximum power point at 65 volts but the shaded panel has its maximum power point at 47 volts.. however, since these are electrically connected, they both operate at the same voltage.

If you add the power of the two panels together (forced to operate at the same voltage), there is a maximum power peak that occurs at 47 volts where the total power is 28.2 +28.2 = 56.4 watts and a second maximum power peak at 65 volts where the total power is 39 + 3.9 = 42.9 watts.

A better search algorithm for the controller will find the best maximum power voltage of 47 volts and get the 56.4 watts. A not so good search may find the max of 42.9 watts at 65 volts.

Now consider the case of a MPPT controller on each panel. This now allows each panel to operate at its own best voltage so the unshaded panel operates at 65 volts = 39 watts and the shaded panel operates at 47 volts = 28.2 watts.

So the total power of the two panels with separate MPPT controllers is 39 + 28.2 = 67.2 watts.

Compare the best power you get with a single controller of 56.4 watts to the separate controller on each panel of 67.2 watts. The separate controller per panel outputs 19% more power. Now if the MPPT controller for both panels connected together found the lower maximum of 42.9 watts, the controller per panel of 67.2 watts is a whopping 56% higher output.

The other really interesting thing for the particular curves we are looking at is that both the non shaded and shaded panels still operate at about the same current. So if these were operated in series and with an MPPT controller, the current through the panels would be .6 amps and the unshaded panel would operate at its maximum power voltage of 65 volts and the shaded panel would operate at its maximum of power voltage of 47 volts resulting in a total power of 39 + 28.2 = 67.2 watts.

So for this case, you get the worst partial shading performance connecting the two panels in parallel with one controller. And, both the series connection and the controller per panel each netted the same maximum power available. The series connection doing as well as the controller per panel likely is not always the case.

 

[Maine Sail]

Yep when I first started doing this everyone told me I was crazy. I suggested they: A) Do the real world data collection as I had, or B) Do the math...

All in all these gains are not huge but when you are out of real estate they can certainly be worthwhile...

 

[Chief RA]

Good presentation Walt. It is easy to grasp the simple fact that 2 separate solar systems have advantages and especially with regard to shading. With such a technical presentation just be aware that few people will comprehend the detailed narrative or graphs. Chief

 

[walt]

I didn’t really understand the controller per panel myself until recently as I had the opportunity to play with a grid tie inverter for a project at work. We didn’t exactly have a partial shading problem but something related.

Also a little interesting.. The grid tie MPPT controller I was looking at was one of those "cheap Chinese" ones that some people are so fond of ( LOL..) and it did something interesting. When a "panel" source was applied to the controller, it would ramp up to the correct maximum power point VI fairly rapidly similar to the video Maine Sail posted maybe a couple years ago with the Genasun controller.

But.. once it found this maximum power point, it would just sit there operating at this point for exactly 90 seconds. Then every 90 seconds, we would observe it doing a sweep in the vicinity of the where it had previously picked a power point. The sweep only lasts maybe 1.5 seconds. This was interesting to me as I had assumed that the controllers were continuously looking for max power. For our application, this was no big issue but in a solar application, you might lose a little power by only updating the mpp every 90 seconds. I don’t know what some of these controllers that get recommended here do. The video Maine sail had shown only showed the start up response.

 

[Maine Sail]

I didn’t really understand the controller per panel myself until recently as I had the opportunity to play with a grid tie inverter for a project at work. We didn’t exactly have a partial shading problem but something related.

Also a little interesting.. The grid tie MPPT controller I was looking at was one of those "cheap Chinese" ones that some people are so fond of ( LOL..) and it did something interesting. When a "panel" source was applied to the controller, it would ramp up to the correct maximum power point VI fairly rapidly similar to the video Maine Sail posted maybe a couple years ago with the Genasun controller.

But.. once it found this maximum power point, it would just sit there operating at this point for exactly 90 seconds. Then every 90 seconds, we would observe it doing a sweep in the vicinity of the where it had previously picked a power point. The sweep only lasts maybe 1.5 seconds. This was interesting to me as I had assumed that the controllers were continuously looking for max power. For our application, this was no big issue but in a solar application, you might lose a little power by only updating the mpp every 90 seconds. I don’t know what some of these controllers that get recommended here do. The video Maine sail had shown only showed the start up response.

There are vast differences in MPPT tracking speed. The cheap Chinese controllers are using 20 year old processes if they are even MPPT at all controllers like the Genasun, Midnite, Outback etc. are tracking multiple times times every second.

 

[walt]

Just to point out.. for the case I showed in the figures, the two power peaks were at 47 volt and 65 volts. Some controllers will actually do a large enough sweep to find the best peak at 47 volts. But in order to do a large sweep, the controller has to actually vary the load so that the operating voltage covers the range of concern. During this time of the sweep, there is only one instant where the controller operates at the peak power point. The rest of the time during the search, it would be operating at a non optimal spot and lose some power conversion.

So.. I have some doubts about full sweeps multiple times per second (would need to see some backup info). What they may do is small sweeps around the current maximum power point very often but the larger sweeps not so often. This is all just how the software is written for the controller so can be done a lot off ways.

 

[Rick D]

Well, That's The Pitts!

I have seven panels all ganged together into one controller. I can only imagine the inefficiencies of that layout!

 

[Chief RA]

Rick: what size panels? Input panel or panels current per controller is of greater concern rather than one controller per panel. This is especially true in your case with 7 panels. Use a controller per pair and hook the last panel directly to your stern DC beer cooler! Works for me!
Chief

 

[walt]

Probably no big deal.. but another thing those graphs show is that if you have more than a couple panels and one of them gets partial shading - and with MPPT, the panel that gets the partial shading doesnt just have reduced power, the power of that panel can actually go almost to zero.

Say we have seven panels all in full sunlight. All operate at the same voltage and the MPPT controller finds the optimum power point - easy since they all have the same power point. In our figure, that would be at 65 volts.

Now if one panel gets shading, the rest of the panels/MPPT controller still hold the maximum power point voltage at 65 volts as this is true peak power voltage for the whole set.

But if you look at the graph for the panel that has partial shading, it only is producing 3.9 watts at 65 volts. It is biased so that it is ONLY putting out 1/10 of full sunlight. I.e., the rest of the panels picked a power point voltage that pretty much biased the one with the shade OFF.

If you just avoid partial shade, all this junk goes away.