Experiments with solar panels

[alanmalk]

I ran a little solar experiment this morning to answer a couple of questions that have been on my mind. The motivation behind this is to determine how much electricity I can realistically expect to produce under less than ideal circumstances. Those include having a non-movable horizontal panel, low winter sun and the possibility of shade on the fixed panel.

What I wanted to learn was:
1. What are the measurable benefits of tilting a panel.
2. What are the drawbacks of using different size (voltage) panels at the same time.

A quick search of the internet shows: #1 - good, and #2 - bad.

I didn't agree with #2. To elaborate, a couple of solar panel dealer web sites stated explicitly that a lower voltage panel in parallel with a higher voltage panel would "pull down the voltage" of the high voltage panel. My own thought process concluded that if the panels in question had the normal blocking diode then it was impossible for any panel to "pull down" another.

So, out comes the soldering iron, a handful of eBay solar cells, a two ohm power resistor as a load, and a couple of multi-meters to measure voltage and amperage. I assembled two panels from 18 identical cells, one 6 cell (3v) "auxiliary" panel and one 12 cell (6v) "main" panel. The 2:1 ratio was intentional to match the real world where big panels are typically 18v or 36v. ( ETI is using a 18v panel.) The parts in the photograph, starting at the bottom going clockwise: power resistor, meters, 3v and 6v panel.

Note that the main panel is horizontal and the small auxiliary panel is tipped. This is to model a real world example where the ETI panel is permanent on the roof, and a smaller panel is propped up near the trailer.

One big assumption is necessary. My two ohm resistor was a substitute for a discharged trailer battery. Unfortunately not exactly true because a resistor behaves according to Ohm's Law - E=IR, and a mostly discharged battery has a starting voltage of at least 10v and the voltage changes according to the lead-acid chemical reaction. But, that said, I didn't have a discharged battery available today.

Observations #1: The two ohm resistor is enough to pull down the horizontal panel voltage from a nominal 6v to 1.66v here in Denver on a sunny day at noon while pulling .84amps. Adding in the output from the tipped 3v panel raised the system voltage to 2.54v and 1.27amps. So much for the small panel pulling down the big one! Pure nonsense. But only pure nonsense when the load is bigger than the output of the bigger panel.

Observation #2: Under no load conditions, the small panel pulled down the voltage of the big panel by 0.002v. Essentially nothing, and I attribute the microscopic drop to leakage of the blocking diode. (FYI, a 5amp 30v Schottky unit. Schottkys leak more than traditional diodes but have a smaller voltage drop. Drifting off topic now...)

Observation #3: Any tip up from horizontal is a big, big help in fall/winter/early spring up here in Colorado, and more so the further north you go.

Observation #4: Any shade is a no-no. Putting just one cell of my setup into the shade results in a loss of 95%. A shaded cell acts like a blocking diode - not good.

My conclusions (and it sure took me long enough to get here). Mixed size panels in parallel are not an issue. Smaller (or bigger) than the roof mounted ETI panel is not a factor - if the load is "big". A pair of 6v ETI batteries drawn down to the 50% level do in fact represent a huge load on a solar panel. They will pull down the panel voltage to a nominal 11-13 volts until the charge gets up somewhere above 80%.

Second conclusion: The extra panel will benefit greatly from being tipped towards the sun. But it does not have to be exact - anything in the general direction will be a great help. Set it up once, pointed more or less south at 45 degrees tip, and you are in business. If you have any shade on the horizontal ETI panel then the auxiliary panel is doing all the work.

I hope this answers more questions than it raises.

--
Alan

[Parker]

This was really interesting. We hope to be camping on Big Pine Key some this winter, and our campsite this year will be more shaded than in the past. I'm thinking about taking along a portable to add to the roof-mounted ETI panel, and I think you just convinced me that it's worth the effort.

[hotfishtacos]

Alan, great work! Thanks for sharing the details! I may redesign my permanent roof mountings based on your study.

Steve

[Ron in BC]

Good job on quantifying the differences. I've seen lot's of emperical evidence that supports your findings but it's nice to see your results support my observations.

Ron

[alanmalk]

I'll try to quantify 'the benefit of white paper in front of a vertical panel' tomorrow. Maybe worth the effort to put a reflector (snow??) on the ground, vs. grass or dirt.

--
Alan

[tractors1]

Often I wonder whether or not ground-based solar panel owners have dogs or not. That alone led me to install roof mounted units. Two tethered dogs could easily disturb optimum settings, or just plain trash the whole affair.

[hillary]

Originally Posted by tractors1

Often I wonder whether or not ground-based solar panel owners have dogs or not. That alone led me to install roof mounted units. Two tethered dogs could easily disturb optimum settings, or just plain trash the whole affair.

We have dogs and a portable panel, but we use an x-pen as a temporary holding cell for the dogs if they need to be contained, so we don't have an issue with tethers. They're off leash a lot when we're camped places that's possible, which is our ideal.
Our 80w briefcase solar panel has been amazing over the last year and a half and never left us without a full charge even on dreary, cloudy northwest days.

[Parker]

So, Alan, since it appears that you're measuring the performance of the panels directly, without a charge controller in between the panel and the load, what does this say about connecting two dissimilar panels in parallel before a single controller? I think it says that it's no problem. I've been advised by the Go Power! people that I can go either way.... parallel two dissimilar panels ahead of a single controller, or run each panel through its own controller, then parallel the two outputs to the battery. I questioned the latter idea, and he said it's no problem, probably because of blocking diodes in each controller. Sound reasonable?

[dave macrae]

I am not personally interested in dealing with a portable panel , but a friend of mine engineered a adjustable tilting bracket for his roof panel that works really well. I will
See how we do with the factory 160 w setup next year and go from there

[alanmalk]

Originally Posted by Parker

... what does this say about connecting two dissimilar panels in parallel before a single controller? I think it says that it's no problem. I've been advised by the Go Power! people that I can go either way.... parallel two dissimilar panels ahead of a single controller, or run each panel through its own controller, then parallel the two outputs to the battery. I questioned the latter idea, and he said it's no problem, probably because of blocking diodes in each controller. Sound reasonable?

Good question. To paraphrase - two panels, one controller or two?

Let me start by saying that my primary interest is to draw conclusions from actual measurements, and I don't have one controller, much less two, to play with. But I'm always willing to guess...

The "which-is-better" question depends on many variables. For example, are the controllers absolutely identical output - voltage, amperage, charging algorithms, etc? Plus, how do they respond to different input voltages from the panels? Just too many variables.

I'm not sure how blocking diodes come into the picture when talking about
controllers. To my way of thinking, a controller should not allow reverse current to flow - without depending on a diode. But I could be completely wrong... Maybe a diode is necessary to prevent one controller from fooling the other?

Personally, I would go with a single controller with the two (or more) panels combined in parallel before the controller. If nothing else, it would be cheaper.

I can say with some certainty that a controller in the bulk charge mode (maximum charge) can be modeled reasonably well by a low value power resistor. Meaning simply, a heavy load on the panels, pulling down one panel in favor of the other due to sun angle, panel size, shade, etc. If the controller is not in maximum charge mode, say float or equalization, then the load is light and it doesn't matter that one panel is not playing its part. Or to say it differently; adding more solar won't charge an almost full battery any faster.

Hope that helps.
--
Alan

[tractors1]

Originally Posted by dave macrae

I am not personally interested in dealing with a portable panel , but a friend of mine engineered a adjustable tilting bracket for his roof panel that works really well. I will
See how we do with the factory 160 w setup next year and go from there

We picked up our 21 with the 160w a couple of months ago and I've very pleased with the charging that continues as Oregon moves into fall/winter. Output is less, but even on gray cloudy days it still feeds my big single 12V.

[Vermilye]

From my experience using 2 controllers, I'd stick with one. I added a portable panel to my built in GoPower 95 watt system, feeding the 160 watt portable through the trailer connector. In full sun both worked fine, however when the trailer was parked in shade, the GoPower & built in panel still managed to produce 14.5 volts (at a trickle of current). While I don't have a easy way of proving it, I believe the high voltage fooled the portable panel controller (which was in full sun). While the portable panel was capable of producing 8.5 amps, it produced less than an amp.

My solution was to add a connector to the portable panel & the trailer so that I could go directly from the panel (bypassing the controller) to the input of the GoPower controller. Since the trailer is in storage, I haven't had a chance to pull the batteries down enough to see if this solved the problem, but I suspect it will.

[Ron in BC]

Originally Posted by tractors1

We picked up our 21 with the 160w a couple of months ago and I've very pleased with the charging that continues as Oregon moves into fall/winter. Output is less, but even on gray cloudy days it still feeds my big single 12V.

That sort of sums the situation up. A flat mounted roof panel is less efficient because of the mounting and, also likely, due to the difficulty of cleaning it. But, and it's a big but, at some amount of wattage efficiency becomes a non-issue if even a technically inefficient output still does the job.

I think the diode thing is still around because for those of us who originally were using smaller panels year round did add a diode to prevent overnight discharge. Because the panels were small no charge controller was used.

Ron

[santacruzer]

I was going to wire a second, portable panel into the GoPower Controller of my 17, but never got around to it before selling the 17 because threading the wire and connecting to the controller always seemed a daunting endeavor. I'd think about it over a beer, and postpone!

With the 19 (due to arrive any moment!) I had Reace place the controller down by the inverter switch and converter, and install a thru-hull port. Wiring the second panel into the controller should be a snap, now I just need to research portable panels.

[alanmalk]

Originally Posted by Ron in BC

... I think the diode thing is still around because for those of us who originally were using smaller panels year round did add a diode to prevent overnight discharge.
Ron

Backflow through the panel is something I can measure easily - and did so just now.

Using my 3v and 6v in series to simulate a 9v panel with no diodes and applying 8v to the positive terminal, I measured 13 milliamps of backflow. It is reasonable to scale this up to a 18v panel (ETI) and a well charged battery of 13.8v and speculate that very little would lost to backflow if the ETI panel had no blocking diode.

But applying a voltage to the terminals that is higher than the panel rating results in a current measured in amps - not milliamps. (I didn't want to stress my cells so I stopped at 11v and .5 amps.) Totally unacceptable...

Conclusion: The single panel from ETI would pull out a trivial amount of electricity in the summer from a well charged battery - if it didn't have a blocking diode. In winter or under a cover, the same panel might pull an unacceptable amount of charge out of a weakly charged battery. But in a multiple panel situation it could pull most, if not all, the current out of the second panel - again - if installed without a blocking diode. So that one dollar blocking diode is cheap insurance.

[Parker]

Originally Posted by Vermilye

From my experience using 2 controllers, I'd stick with one. I added a portable panel to my built in GoPower 95 watt system, feeding the 160 watt portable through the trailer connector. In full sun both worked fine, however when the trailer was parked in shade, the GoPower & built in panel still managed to produce 14.5 volts (at a trickle of current). While I don't have a easy way of proving it, I believe the high voltage fooled the portable panel controller (which was in full sun). While the portable panel was capable of producing 8.5 amps, it produced less than an amp.

My solution was to add a connector to the portable panel & the trailer so that I could go directly from the panel (bypassing the controller) to the input of the GoPower controller. Since the trailer is in storage, I haven't had a chance to pull the batteries down enough to see if this solved the problem, but I suspect it will.

Hi Jon,

That's been my thinking all along, but when I talked with the GoPower guy he was thinking mostly about their portable units with the included controller. He said in effect that whichever controller is putting out the highest voltage, it will be the one feeding the battery. I can't really fault that logic, but I do prefer the idea of running everything through the existing built-in controller so I have one place to monitor what's going on. Sometime maybe you could point me to the bulkhead connector you used to connect through the wall. I like their portable units, since they hinge already, and I would be inclined to just bypass the built-in controller.

Parker

[dave macrae]

I get that there are limitations such as trailer parked in a shady area but wouldn't it just be easier and more reliabe to have as big a panel on the roof as is reasonable and perhaps with a mount that can be tilted rather than having multiple planels and controllers.

[Parker]

Originally Posted by alanmalk

Conclusion: The single panel from ETI would pull out a trivial amount of electricity in the summer from a well charged battery - if it didn't have a blocking diode. In winter or under a cover, the same panel might pull an unacceptable amount of charge out of a weakly charged battery. But in a multiple panel situation it could pull most, if not all, the current out of the second panel - again - if installed without a blocking diode. So that one dollar blocking diode is cheap insurance.

So it sounds like a Schottky diode or similar between each panel and the controller would be a good idea.

[yardsale]

I am concerned that our flat mounted 160 watt panel will be able to keep our two 6v batteries charged this winter in the PNW where the maximum sun angle will be about 20 degrees. What about using a reflector adjusted to about 65 degrees to reflect solar energy directly onto the panel?

[gbaglo]

I have a half dozen large mirrors ( up to 2' x 4' ) mounted on a trellis in my garden. They reflect sunlight back onto the raised bed.
Been there for about six years.
I've noticed no improvement in my tomato crop.