#165325  by mattn
 Thu Sep 10, 2020 3:40 pm
As the Tracer is an MPPT controller, in theory you get better efficiency having them in parallel as the panel voltage is closer to the battery voltage allowing the controller to run more efficiently, but this is offset by having a higher current and therefore power loss in the solar panel wiring, which is then offset by parallel paneling providing higher voltages in marginal light conditions, which is then..... The difference is therefore mostly theatrical, in that lots of people will spend many hours debating the very small differences, often in a most entertaining way.

The 8Ax40V gives 320W, so what you would expect from 300W panels, and will be converted to 20x14V for charging by the controller.
 #165360  by Nut17
 Sat Sep 12, 2020 12:44 pm
His test results are pretty much in line with mine with one exception. Our existing seriesed connected panels are not any more adversely effected by shading than the parallel connected set up. I assume there must be by pass diodes in the panel connections to minimise this problem.

Our new rig will be set up with two banks of panels - each comprising 2 x 240 watt and one 190 watt panel = 670 watts wired in series to a 40 amp MPPT controller. The out put of this pair of controllers will then be parallel connected then wired to the 400 AH LiFePO4 lithium battery.
 #165365  by scubadoo
 Sat Sep 12, 2020 1:46 pm
For those already confused the first 30 seconds of this video by the same blogger clearly explains it all. :? :mrgreen:

https://m.youtube.com/watch?v=MxziHKvTRh8q




In the meantime I just let our 3x 270W 24V paralleled panels fed into the Victron 50A MPPT solar controller keep filling our 4 cell battery. 1.88kWh (c135Ah) generated so far here in Whanganui at 1.45pm today and counting.
Votronic = 89% and Juntek = 91% SOC.
 #165412  by Neddy
 Sat Sep 12, 2020 7:58 pm
Ian, the author of that article seems to be unaware of the fact that MPPT controllers lose efficiency as panel supply voltage rises.
Take a look at the performance data of a typical MPPT controller :-
TriStar 60.JPG
See how the higher voltages of Serial panel connection are clearly less efficient than lower voltage Parallel panel connection.
The author's 96 volt(!) array will be particularly inefficient, especially at low light levels.
Notice also how, the higher the panel voltage, the earlier MPPT performance collapses as the ambient light level falls.
Parallel connection is more efficient than Serial connection right across the board, but especially at low light levels - when you need all the power you can get.
 #165414  by scubadoo
 Sat Sep 12, 2020 10:22 pm
Neddy
If I extrapolate your posted graph with my estimates and compare let's say a typical 12V (c18 VMP) panel and my 24V (c36VMP) panels with 50W output for an entire "low light" day.

Solar input to our battery with 12V panels would be perhaps 50W x 6h x 95% efficiency = 285Wh or about 21.5Ah.
Solar input to our battery with 24V panels would be perhaps 50W x 6h x 93% efficiency = 279Wh or about 21.0Ah.
That would run our TV setup for an extra 15 minutes.
The losses at 25W look similar. i.e. 2 or 3% of not much for a 30-60A MPPT controller.

Even with 2 or 3 24V panels in series at 50W (a very bad day for us!) according to that graph it looks like worst case perhaps a 5-6% difference or about 200mA when compared to 12V paralled panels. Our "everything off" current is about 700mA.

I suspect that any tiny efficiency loss would be more than overcome by the resulting reduced cable losses.

Our Victron controller may be more or less efficient than that Morningstar. Victron engineers have stated "The efficiency level is pretty stable througout the whole range (>90-94% )" in the past.
I certainly have no concerns with such small and to me all but unmeasurable differences.
 #165416  by ianganderton
 Sun Sep 13, 2020 1:43 am
Neddy wrote:Ian, the author of that article seems to be unaware of the fact that MPPT controllers lose efficiency as panel supply voltage rises.
Take a look at the performance data of a typical MPPT controller :-
TriStar 60.JPG
See how the higher voltages of Serial panel connection are clearly less efficient than lower voltage Parallel panel connection.
The author's 96 volt(!) array will be particularly inefficient, especially at low light levels.
Notice also how, the higher the panel voltage, the earlier MPPT performance collapses as the ambient light level falls.
Parallel connection is more efficient than Serial connection right across the board, but especially at low light levels - when you need all the power you can get.
I haven’t seen that graph before and the logic flies in the face of the best practice I see on various solar forums I lurk on. Nate works closely with Victron doing online Q&A for them through their channels so will have access to their technical staff.

I suspect efficiencies will vary between different MPPT controllers. It’s interesting and I’m definitely going to look into it.

The cables connecting solar panels to the charge controller are often some of the longest runs in our vehicles and carry the most power during a day. Higher voltage means significantly lower cable costs and significant improvements in efficiency’s.

On top of that the voltage headroom gives the mppt controller more to work with as voltages drop on cloudy days.


Sent from my iPhone using Tapatalk
 #165420  by scubadoo
 Sun Sep 13, 2020 9:04 am
No 12V panels here for a long time so I can't compare.
I have installed 3x 24V solar panels on our motorhome roof normally connected in parallel to a Victron 50A MPPT controller. All 3 are almost identical in surface area.
AASolar 270W rated mono. 6 years old.
270W rated poly. About 3 years old.
305W rated brand name mono "Perc". Added February this year replacing a 270W rated mono. I was curious.

I can connect 1, 2 or 3 panels in series or parallel combinations and measure individual currents when required with little effort via multiple 3way MC4 connectors.
Limited testing reveals that ALL output almost identical energy into the battery! The maximum variation is less than 300mA so far even for the "305W" panel. That extra 35W dissapears somewhere. ;)
Surprisingly and against all theories the poly is consistently the best performer but only just.
Checked occasionally on sunny and very overcast days at random times throughout daylight hours.
Shading. I have not really experimented sufficiently to write anything in stone but when the satellite dish casts a shadow over the nearby Perc panel output drops by very roughly 30% if parallel connected and 90% if serial connected.
Chris has much better luck.
Needless to say we avoid shading whenever possible.
Pay with Paymate Express