Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Solar Panels and MPPT
#1
Solar panels are the key ingredient for an electric boat.

They are conceptually quite simple, but in fact getting the most out of them isn't so simple.

Let's review some real basics of electricity, volts, current, and watts.

Watts are a unit of power, defined by volts * current. All the voltage in the world does nothing without current flowing. Similar to horsepower = torque * rpm, if a motor isn't moving, all the torque in the world produces no power. Voltage is like a electron force, and current is movement of those electrons.
The higher the voltage, the easier current flows. The more current (amps) that flow, the more power you get.
So consider a 100 watt solar panel. It can output 12 volts at 8.33 amps (12 * 8.33 = 100),
or it could output 24 volts at 4.16 amps. They are equivalent in power.

Solar panels are made up of solar cells.
A typical solar cell is about 3 watts. It outputs less than a single volt, usually .5 volts, and thus for 3 watts,
it would output 6 amps of current. One half a volt is not much, even the lowest voltage microprocessors need about 1 volt to operate. So cells are connected in series to get higher voltage. So just like three 12v batteries can be connected in series to get a 36v battery, solar panels connect many .5 volt cells in series to get say 12v, which would be 24 cells in series. A typical 100 watt solar panel connects 36 cells in series to output 18 volts with 6 amps of current (18v * 6a = 108 watts).

Then things start to get more complicated. Because the cells are connected in series, the output current from one cell is flowing through the next cell in series, which adds its voltage to the flow. This is all good until one of the 36 cells can't produce the same amount of current as the other 35 cells. This will occur if one cell is not getting the same sunlight as the others. The current is now limited to only what this one cell can pass through. So a 100 watt solar panel would be reduced to a 33 watt solar panel if a single cell can't output its 3 watts, but can only output 1 watt. If it is completely shaded, and can output .1 watt, the entire panel is limited to 3.6 watts. But it gets even worse, because all the power from the other cells that isn't going anywhere is generating heat that can destroy the panel. So what to do?
Reply
#2
The typical solution to series connected cells problem is to use bypass diodes.
A diode is a basic electronics device that only lets current pass in one direction.
When a diode is placed in parallel with a cell, if the cell can't pass current because it is shaded,
the current passes through the diode instead.  Adding bypass diodes around a cell adds cost and complication to the construction of a panel, so they are not used for each cell, but placed around a group of cells.  

   

Unlike a solar cell that increases the voltage as the current passes through, a diode DECREASES voltage as current passes through, which means it is consuming power.  This is called the forward voltage drop of a diode, which is .4 volts for low-cost diodes.   So when shading occurs, a bypass diode will bypass all those cells it is connected around, eliminating their contribution of power to the panel output, and also consuming power.  

When the bypass diodes are connected to columns as shown in the picture above,  it does not prevent shading on each row from stopping all the output of the panel.   This is demonstrated very clearly in this video

The lowest cost panels are desired for large solar-arrays and rooftops to make solar more affordable.
The installation of panels in these cases typically do not have issues with shading.  Certainly a large solar-array for grid power production would remove any possible sources of shading.  The only concern is variable cloud cover, and a small set of diodes is sufficient to prevent damage to a panel from reverse currents when shadow boundaries occur on a panel.   So when shopping for panels, the large lowest-cost-per-watt panels are made with all cells in series with minimal diodes.  The panel in the video is typical of large 'grid-array' panels.

Unfortunately, for installation on boats and other applications, shading is a major problem.
Using large series connected cells might look like you get 250 watts, but that peak performance is unlikely,
and a large 250 watt panel can often become a 25 watt panel, making the panel a very poor choice.

So what is the alternative?
Reply
#3
The alternative to connecting solar cells in series is to connect them in parallel.
Instead of each cell contributing its voltage to the combined output, each cell contributes its current.
If we have a 36 cell 100 watt panel where all cells are in parallel, we get .5 volt output, but have 216 amps output.
Now when a cell is shaded, it just doesn't contribute its current to the total, so we get 210 amps instead of 216,
which is still 97% of the rated output, whereas in the series case, we would only get, 66% output with 3 bypass diodes.  In the case of a 3 cells in a row, for a series connection, we get NO output, but in a parallel configuration, only these three don't contribute, so we still get 91% of our rated output.
This is a dramatically better design for solving the problem of shading.  
The only problem now, is we have only a .5 volt output which is not that useful, and we need large wires to carry 216 amps of current without losses.  

This gives us an engineering design decision of how many to put in series and how many to put in parallel.
When a series of cells is connected in parallel, shading of any one of those cells in the series would inhibit all those in the series from contributing its current.  So if we take our 36 cells, and connect 4 in series, and have 9 of these in parallel, we get 2 volts output and 50 amps.  One shaded cell would reduce the output by 1/9th or about 10%.
In the worse case, 8 of 9 are shaded, and the output is only one series connected set of cells.   Even if the 8 can produce current, because they have lower voltage than the non-shaded cells, they can't contribute any current.

We can improve this performance significantly by using active electronics, which leads us to a MPPT design.
Reply
#4
Let's review how a solar cell works.

When extracting current from a solar cell, it causes the voltage to drop.  
Because power = current * voltage, to get the maximum power from a cell, if the current increases but the voltage drops too much, you would have less power, and if not enough current and too much voltage, that is also less power.  There is an optimal combination of voltage and current that yields the maximum power called the maximum power point or MPP. If you plot the current vs voltage on a graph, and compute the power, you can see the point where peak power occurs.  This point changes with changes in sunlight.

   

MPP Tracking or MPPT works by adjusting how much current is extracted from a cell to find the amount that produces the most power.  
While the improvement varies, MPPT control can yield a 10% improvement in output over a static design.
Reply
#5
So we now have the technical background to understand how to create a solar panel that will achieve maximum performance under the challenging conditions in a world with shade.

In the ideal design, we have a MPPT controller that is connected to a single cell, and extracts its maximum output under all shading conditions, and boosts its voltage from .5 volts to a constant higher voltage, such as 6v.
Now we combine all cells optimal output together in parallel and get the maximum current at 6v.

While this ideal is possible, it would increase the costs significantly. The first problem is getting electronics to run at .5v, which is difficult. The lowest cost electronic components often need 2v or 3v to operate.
ST Micro does have a microMPPT controller that will operate at .3v, however it is limited to just 1.8 amps,
so it can't be used with a full 3 watt cell, it would require 1/3 size cell, and thus increases the cost by 3 times as well. The issue of the cost becomes significant when operating at a per-cell level. The alternative is to increase the 'grain size' of the microMPPT design to use more cells in series. This both increases the voltage to a more manageable level,
and also reduces the effective cost of adding the active electronics.

ST Micro has another chip that is closer to the ideal design, called the SPV1020. This chip will operate down to 6.5v, and can handle 9 amps of current. So we can use a solar panel with 14 cells in series as input for the SPV1020.
It can also operate up to 30v of input.

If we want to use 'off-the-shelf' solar panels, these panels are designed for 12v battery charging, and thus output approximately 16v volts. So a simple solution that can implemented quickly and easily is to use a circuit board with the SPV1020 chip that is designed to handle these off-the-shelf panels.

The SPV1020 can handle a maximum of 100 watt 12v panel, so if one wants 400 watts of solar power, the can purchase four 12v 100 watt panels and 4 of the SPV1020 boards.
The smaller the panel size, the better performance you can achieve. So one could use 8 50 watts panels,
or even 16 25 watt panels that are all connected in parallel. It comes down to a matter of cost, as panels less than 100 watts are usually more expensive. The smaller wattage cells are more expensive because they don't use full-size cells, the cells are cut into smaller pieces in order to create more in series to get the 16v output.

So we have designed a circuit board that uses the SPV1020. We are doing a kickstarter project to get the quantity of boards produced to at least 100. At quantity 100, the costs are significantly less, 500 would be even better, but at minimum is 100. At his quantity, we will be able to sell them at only $25 each.

At $25 each, it adds minimal cost to a 100 watt panel. It becomes more significant for a 50 watt panel, but is still not unreasonable given the significant increase in performance of the solar output in the shaded world of a sailboat.
Reply
#6
I got the circuit boards back from fabrication, and have done some initial tests to confirm they work.
Here are some pics that show a 12v 50 watt panel connected directly to a 12v battery is charging
the battery at 2 amps.  Then I connect it to the miniMppt which outputs '24v', so I use two 12v batteries in series, and it increased the current draw from the panel to 2.2 amps.  So it works.
The 97% efficiency boosting the voltage, and looks like perhaps 10% improvement itself from the MPPT.

   

   
Reply
#7
Kickstarter for the microMPPT circuit boards is now live.

microMPPT controller kickstarter
Reply
#8
(04-21-2015, 07:44 AM)jackb Wrote:  Kickstarter for the microMPPT circuit boards is now live.

microMPPT controller kickstarter

It seems I am late to discover this, but this is fascinating idea and has lots of potential. Now that kickstarter is over, what next? Any chance you could post scematics and maybe even board layout for reference for others?

Next boat I want to build would almost need the whole 100pcs lot alone Smile Well maybe not quite, but tens of boards anyway.
Reply
#9
Welcome to the forum mr. monk. The schematics are well documented for the SPV1020 eval board.
I just don't have the battery charger circuit. If you want dozens of them we can try another kickstarter.
I will need a lot when I get to that point on my solar boat. Are you planning a solar powered boat also?
Reply


Forum Jump:


Users browsing this thread: 2 Guest(s)