This post is to address a couple of the questions HS posed in his earlier posts after doing some testing. I finally found the time to re-read everything and understand it better and do a few calculations. I don’t think anything below changes the conclusions we seem to have reached but hopefully it will tie up some loose ends…HS’s statements are in bold type:
Although I was in the bulk charge phase...is it possible that I would have seen more amperage out of the controller IF the batteries had been discharged? Maybe it was giving me higher voltage and less amperage, because that's what the battery needed more in the existing charge state?
As you pointed out HS, it is really wattage that matters so the totals would have been the same either way. Also, since we know we never want to discharge our batts beyond 50% I think your conditions were good ones since they reflect reality and your batts were 30-40% discharged and ready for a good bulk charge rate. You were not seeing a resistance to the charge rate and what was being put into the batts in WATTAGE was close to what was available from the panels.
What IS puzzling to me is that we never saw 14.4-14.5 V level charging out of the controller. I think the closest you got was one 14.2 reading. Is this a function of the controller that can be changed or was mis-adjusted. If not…I can see a sulfation issue cropping up over time. Your thoughts?
**************
While looking for solar output /vs/ temp, I found http://userwww.sfsu.edu/~ozer/engr300-solar1N.pdf
Where some engineering students were measuring output versus angle to the sun. They made the SAME MISTAKE I DID, and others have, of measuring the panle output under no-load conditions. So bear in mind--when they show output voltage from the panels, they are probably WAY WRONG and the voltage probabyl drops WAY SOONER. Dunno, I didn't see any link to check with them or their professor.
But check it out, they do show how both voltage and current plummet as the panel faces away from the sun, if as I've noticed it drops much faster under load--they are way over-optimistic and panel orientation makes a much bigger difference.
The student experiment with regard to angle of incidence vs. power is a bit flawed for our purposes since they eliminated all reflection and did not place a load on the cells …but it shows a 10% loss in power for a 25 degree off axis shift and a 23% loss when off axis by 45 degrees. These are less than the losses you showed for off axis positioning, but…it does show BOTH that the ability to tilt and position matters a great deal during the day AND that normal swinging at anchor can also have a great effect. No surprises there.
My guess is that with a real load…the voltage would behave in a similar manner but we have no solid data to support that…though your data would suggest that is true. You ought to write them and suggest THAT experiment for next year!!
Caution….MATH FOLLOWS!! (G)
According to the experiment…both POWER and CURRENT are a COSINE function of the angle of incidence.
Example…Power= Maximum power times the COSINE of the angle of incidence.
Lets take a 10 degree angle of incidence on a 100 watt panel (i.e. 10degrees off vertical) . Cosine of 10 degrees is .985 so multiplying by 100 watts gives you a maximum of 98.5 watts…a 1.5% loss.
20 degrees=6%, 30= 13%, 40=24% etc. etc.
Note that this jibes well with what USERS of the solar stik have said about how MANY adjustments are necessary per day AND what Tomaz projected way early on in our earlier Stik debates. Way to go Tomaz!
But if as HS suggests…both amps
and voltage drop in a cosine like manner…then power drops quite a bit quicker since power then would be:
Wattage = Cosine angle * max voltage TIMES Cosine Angle * Max AMPS…
Example at 20 degrees off vertical:
100Watt panel… (Cosine 20degrees* 17.5V) TIMES (Cosine 20degrees* 5.72amps)=88.5 Watts… a 12% loss which is about what HS measured and DOUBLE the experiments “no load” prediction.
Since we don’t have actual data measurements of both approaches in a controlled environment, it seems to me that we only can safely say that getting off axis by 20 degrees…(or about 100 minutes without adjustment) causes between a 6 and 12% drop from maximum output.
******************
Also found [IMG]http://www.locoengineering.com/images/Current_vs_Time.gif
[/IMG] a graph which lives at
http://www.locoengineering.com/Solar%20Info.htm
and gives a display of output-vs-time-of-day, for fixed panels. Do some calculus for me, how does their output from 10-2 compare to the entire day's output? (Nothing radical or new there, just the first nice graphs I've seen for that.) They're showing surprisingly good output at 6AM/PM, and again that's with "fixed" panels...if I were more enterprising I could combine that data with the angle-of-incidence bit from the first site to get some numbers for us, on how much gain can be gotten from angling the panels between 6AM-10AM and 2PM-6PM. Without doing the math...I suspect it's more than 20% power gain during those hours?? Take a look.
OK…When faced to the noon sun at the ANGLE of Latitude…
from 10AM-2PM panels produce 80-100% of rated power. Average 90% of power times 1/3 of a 12 hour day.
From 8-10 AND 2-4 they produce between about 50-80% …of rated or average 65% for 1/3 of 12
From 6-8 and 4-6 they produce between 0 and 50% of output or average 25%
So for the full day we can approximate that a ONCE aimed panel puts out the average of 25% +65% and 90% of its full rated 12 hour value or
60%.
A 100 watt panel gives 1200 watt/hours in a 12 hour day…and 60% would be 720 watt hours.
At 13.6 volts this would be 53 amp hours per day (on a perfect day and assuming no conversion losses.)
It is not clear how much of the hour to hour loss of wattage is due to not aiming the panels and how much is due to atmospheric weakening of the sun as it goes lower on the horizon. So…we can’t go much further than to say that any increase beyond 53 ah’s achieved is likely due to stik adjustment beyond the initial adjustment.
But in fairness...lets also note that many panel installations can't even make that initial adjustment and will get out of whack with full output faster and cut out completely at low angles of the sun and so
would not achieve even 53 ah's.
The other thing I would take away from this link is that HS’s latitude in mid summer is just as capable of producing full output as say Florida is at this time of year. The 63watt peak measured output may be the result of atmospheric absorbtion or high clouds but it is not a function of latitude.
***************************************
At 11:20 I measured 15.81V into the SBoost and 13.36V out of it, against 13.5-13.6 showing on it's panel. In theory the SBoost was providing about a 6-7% net gain, putting out 4.7A @ 13.6V instead of the 4.4A at 15.81V that was coming out of the panels. (Check my math, but that's what I wrote at the time.)
OK…so you have panels putting out 69.6 watts. The Sboost…converts the excess voltage into amps and puts out 63.92 watts or an 8.2% loss from perfect conversion. What we really need to compare to is what a normal regulator would do in the same circumstance…it would preserve the amps and chop any voltage over 13.6 converting it to heat. So we would have 13.6 Volts times 4.4 amps or 59.84 Watts. Thus in this example the Sboost provides a 6.8% gain in power delivered to the batteries over a conventional 3 stage regulator.
Even though I think we each got there differently….out numbers agree! (G)
*********
11AM
13.5V @ 3.1A angled to sun
moved 45d away dropped to
12.5V @ 2.5A
10% loss.
….umm …that is a 25% loss of power AND 12.5V means nothing is going into the battery. Is that voltage a typo? Something is screwy there…the loss of power makes sense but I would expect the amps to drop rather than the voltage unless you were measuring panel output and not after the controller.
(then)
13.4V @ 2.1A aimed direct
13.3V @ 1.9A aimed 15d off
10% loss of output in an hour of sun movement!
1:10PM
14.2V @ 3A /vs/ 14.1 C @ 2.7A (42.6 Watts)
14.2V @ 3A direct /vs/ 14.1V @ 2.7A aimed 10deg. Off
Another 10% loss but finally we see 14+ Volts to the battery. Any time the panel is being fully driven in bulk charge mode we should be seeing over 14V if we are to expect our batteries to last. We need to understand better why you are not seeing 14+ more of the time. 13.6V is absorbtion level voltage.
*******************
Finally, you responded to Amy saying:
batt voltage: 12.2 to 12.3
solar panel current: 6.1
output charge current: 7.8 "
The numbers tell me that your batteries are nowhere near full, so the controller is easily able to take excess voltage from the panels (probably over 16V) and squeeze out some extra amperage instead. No way to tell, without using an extra meter, what VOLTAGE it is charging at though. For all we know, it could be charging at 12.6 volts times 7.8A, which would be 98 watts--within the spec for your panels in blazing sunlight. No miracle--just good efficient charging.
I would suggest that the real answer lies in the fact that there was stuff running on the boat and that the numbers reflect that...remember ZERO charging is going on if the VOLTAGE is less than about 13V at the battery terminal during charging. You can't rely on that 12.3 rating with other stuff running and thus you can't rely on the amps either. I may be making the wrong assumption here but my guess is the fridge or other stuff (chargers, pc etc.) was on. Amy can correct me if that assumption is wrong and everything n the boat was shut down.
**************
So for HS and others with an interest in this death spiral of a post… I hope the above was somewhat useful or provocative to your thinking. Obviously I speculated a lot on some things and am quite open to critiques if you think I am off base in my assumptions. This continues to be most interesting. Thanks not only for your work but those earlier links HS…they were most interesting and thought provoking.
Sorry...no pictures Giu!
Connect with Facebook
