RichardS
N/A
This is the second installment of thread http://www.ybw.com/forums/showthrea...ments-with-4-solar-panels-and-MPPT-controller
The setup was the same but this time with an 7A LED array as a load on the car battery:
The other big difference is that this time the tests were done in sunlight starting at 1:00pm. At this time of the year the trees around my garden cast long shadows so I set everything up and waited for the shade of the first tree shadow to move out of the way. I knew I would have about an hour before shadow of the next tree came across. The photo was taken with the righthand two panels in sun and the two lefthand still in shade. It was interesting to watch the meter as the sun moved across .... in full shade the voltage was still around 40V and this only increased to 44V when all the panels were in sun. However, the current steadily increased from around 1A in full shade to 4A when all were in sun.
Perhaps this excellent voltage performance is why MPPT controllers are highly regarded and why Victron recommend using a high a voltage as possible within the controllers capability?
Anyway, for those who are interested, the tests were exactly the same as last time but this time starting with a Parallel/Parallel - Series - Parallel/Parallel configuration and then the S/S-P-S/S configuration used last time.
Once again the pieces of cardboard were used to cover 8 cells at a time and for each shading condition I used every possible permutation of the outermost 8 cells of every panel and averaged the results. once again, the results were highly consistent with a maximum variation of only 0.2V between each set of readings, although the currents were much higher this time, of course. I've not used Vic's method for the tables as I have copied and pasted all this from a text document, so apologies for all the dots.
Once again, the first two set of results were taken with the panels on open circuit connected directly into the multimeter i.e. no regulator or battery in circuit.
PARALLEL (LEFT TWO PANELS) - SERIES - PARALLEL (RIGHT TWO PANELS)
.................................................................VOLTAGE......CURRENT
NO CELLS COVERED...................................44.1V...........3.9A
8 CELLS COVERED ON ANY PANEL............43.7V...........3.9A
8 + 8 COVERED ON PARALLEL PANELS......40.7V...........3.9A
8 + 8 COVERED ON SERIES PANELS...........43.3V...........2.0A
SERIES (LEFT TWO PANELS) - PARALLEL - SERIES (RIGHT TWO PANELS)
...................................................................VOLTAGE......CURRENT
NO CELLS COVERED....................................44.0V...........3.8A
8 CELLS COVERED ON ANY PANEL.............43.5V...........3.8A
8 + 8 COVERED ON SERIES PANELS...........43.3V...........2.0A
8 + 8 COVERED ON PARALLEL PANELS......40.5V...........3.8A
For the second set of readings I connected the MPPT controller to the solar panel array, both with and without the LED connected across the battery. I ensured that the battery was fully charged before starting but the problem with using a battery is that because of the 7A load the battery quickly starts to discharge and then the controller drops from Float into Bulk and the readings change. What is needed is a load which does not influence the controller but if the controller doesn't see a load which behaves like a battery it just hunts around as happened last time. I don't know the answer to this.
PARALLEL (LEFT TWO PANELS) - SERIES - PARALLEL (RIGHT TWO PANELS)
NO LOAD ON BATTERY......................14.4V...........0.7A ON FLOAT / 2.0A ON BULK
LED LOAD ON BATTERY.....................13.0V...........7.2A ON BULK
SERIES (LEFT TWO PANELS) - PARALLEL - SERIES (RIGHT TWO PANELS)
NO LOAD ON BATTERY......................14.4V...........0.6A ON FLOAT (CONTROLLER DID NOT SWITCH TO BULK)
LED LOAD ON BATTERY.....................13.1V...........7.7A ON BULK
Firstly, I'm inclined to dispense with the controller results. A battery is simply not a stable load so although the controller will hold a stable voltage depending upon which "cycle" it's on, the current varies depending upon what it senses from the battery. For what it's worth, the results from either panel configuration are similar. I've no idea what the four 100W panels would deliver into a more demanding load but at 7.7A I'm near the limit of my meter anyway.
The open panel results are interesting and confirm my conclusion from the last time i.e. using the S/S-P-S/S configuration for my boat would be a mistake.
It doesn't matter how you connect the panels, the same results apply:
Covering 8 cells out of 36 (5.5% total shading) has virtually no effect on the voltage and zero effect on the current.
Covering 8 cells on each of two panels that are in parallel with each other (11% total shading) has, surprisingly, zero effect on current but produces a 10% reduction in voltage.
Covering 8 cells on each of two panels that are in series with each other (11% total shading) has a huge 50% reduction on current and produces a small 2% reduction in voltage.
On my boat the two panels on the same side of the boom will be in the boom's shadow at the same time so it seems obvious that those two panels should be connected in parallel with these pairs then in series to boost the voltage. This assumes that the controller can use volts and amps as rough equivalents to generate a maximum current at 14.4V since, on that basis, choosing between a configuration which produces a 50% reduction in current at constant voltage or one which produces a 10% reduction in voltage at constant current seems like a no-brainer.
Thanks to those who commented last time and, of course, I would be interested in opinions this time as well.
Richard
The setup was the same but this time with an 7A LED array as a load on the car battery:
The other big difference is that this time the tests were done in sunlight starting at 1:00pm. At this time of the year the trees around my garden cast long shadows so I set everything up and waited for the shade of the first tree shadow to move out of the way. I knew I would have about an hour before shadow of the next tree came across. The photo was taken with the righthand two panels in sun and the two lefthand still in shade. It was interesting to watch the meter as the sun moved across .... in full shade the voltage was still around 40V and this only increased to 44V when all the panels were in sun. However, the current steadily increased from around 1A in full shade to 4A when all were in sun.
Perhaps this excellent voltage performance is why MPPT controllers are highly regarded and why Victron recommend using a high a voltage as possible within the controllers capability?
Anyway, for those who are interested, the tests were exactly the same as last time but this time starting with a Parallel/Parallel - Series - Parallel/Parallel configuration and then the S/S-P-S/S configuration used last time.
Once again the pieces of cardboard were used to cover 8 cells at a time and for each shading condition I used every possible permutation of the outermost 8 cells of every panel and averaged the results. once again, the results were highly consistent with a maximum variation of only 0.2V between each set of readings, although the currents were much higher this time, of course. I've not used Vic's method for the tables as I have copied and pasted all this from a text document, so apologies for all the dots.
Once again, the first two set of results were taken with the panels on open circuit connected directly into the multimeter i.e. no regulator or battery in circuit.
PARALLEL (LEFT TWO PANELS) - SERIES - PARALLEL (RIGHT TWO PANELS)
.................................................................VOLTAGE......CURRENT
NO CELLS COVERED...................................44.1V...........3.9A
8 CELLS COVERED ON ANY PANEL............43.7V...........3.9A
8 + 8 COVERED ON PARALLEL PANELS......40.7V...........3.9A
8 + 8 COVERED ON SERIES PANELS...........43.3V...........2.0A
SERIES (LEFT TWO PANELS) - PARALLEL - SERIES (RIGHT TWO PANELS)
...................................................................VOLTAGE......CURRENT
NO CELLS COVERED....................................44.0V...........3.8A
8 CELLS COVERED ON ANY PANEL.............43.5V...........3.8A
8 + 8 COVERED ON SERIES PANELS...........43.3V...........2.0A
8 + 8 COVERED ON PARALLEL PANELS......40.5V...........3.8A
For the second set of readings I connected the MPPT controller to the solar panel array, both with and without the LED connected across the battery. I ensured that the battery was fully charged before starting but the problem with using a battery is that because of the 7A load the battery quickly starts to discharge and then the controller drops from Float into Bulk and the readings change. What is needed is a load which does not influence the controller but if the controller doesn't see a load which behaves like a battery it just hunts around as happened last time. I don't know the answer to this.
PARALLEL (LEFT TWO PANELS) - SERIES - PARALLEL (RIGHT TWO PANELS)
NO LOAD ON BATTERY......................14.4V...........0.7A ON FLOAT / 2.0A ON BULK
LED LOAD ON BATTERY.....................13.0V...........7.2A ON BULK
SERIES (LEFT TWO PANELS) - PARALLEL - SERIES (RIGHT TWO PANELS)
NO LOAD ON BATTERY......................14.4V...........0.6A ON FLOAT (CONTROLLER DID NOT SWITCH TO BULK)
LED LOAD ON BATTERY.....................13.1V...........7.7A ON BULK
Firstly, I'm inclined to dispense with the controller results. A battery is simply not a stable load so although the controller will hold a stable voltage depending upon which "cycle" it's on, the current varies depending upon what it senses from the battery. For what it's worth, the results from either panel configuration are similar. I've no idea what the four 100W panels would deliver into a more demanding load but at 7.7A I'm near the limit of my meter anyway.
The open panel results are interesting and confirm my conclusion from the last time i.e. using the S/S-P-S/S configuration for my boat would be a mistake.
It doesn't matter how you connect the panels, the same results apply:
Covering 8 cells out of 36 (5.5% total shading) has virtually no effect on the voltage and zero effect on the current.
Covering 8 cells on each of two panels that are in parallel with each other (11% total shading) has, surprisingly, zero effect on current but produces a 10% reduction in voltage.
Covering 8 cells on each of two panels that are in series with each other (11% total shading) has a huge 50% reduction on current and produces a small 2% reduction in voltage.
On my boat the two panels on the same side of the boom will be in the boom's shadow at the same time so it seems obvious that those two panels should be connected in parallel with these pairs then in series to boost the voltage. This assumes that the controller can use volts and amps as rough equivalents to generate a maximum current at 14.4V since, on that basis, choosing between a configuration which produces a 50% reduction in current at constant voltage or one which produces a 10% reduction in voltage at constant current seems like a no-brainer.
Thanks to those who commented last time and, of course, I would be interested in opinions this time as well.
Richard
