Solar power info

Rsadeika

I went ahead and purchased the 10 pack of Radio Shack 9V 1 watt solar panels that erco had found, now what to do with them. I have a spare Parallax Li-ion Power Pack/charger unit with a couple of batteries that I can spare for this experiment, the question is, can it work with the solar panel(s) that I have? Maybe there is a different kind of charging unit that I should be looking at?

Since my knowledge for this kind of stuff is non-existent, what would the setup look like using the solar panel(s) that I have? In the documentation for the Power pack/charger it shows a requirement of +5-12VDC @ 1amp minimum input, does this mean that I could hook ten of the solar panels in parallel(1.1amp), and this would be sufficient to get the charger to work, or am I way off base on this?

I tested one of my panels with a volt meter connected by putting it up on the inside window sill, getting light through the window, and it was showing ~2.7V, here in Michigan it is cloudy and raining, so I am thinking that maybe with a ten panel exposure it could get above a 5V input, but not sure about that. So, I guess I need some knowledgeable people to way in on this.

For another experiment I am thinking about using an Activity Board setup where I hook up one of my panels and have it do some data logging of what kind of voltage I am getting on a daily basis, that would be a 24/7 period for maybe the whole fall/winter/spring/summer period. Not sure how involved with this I want to get.

Thanks

Ray

Domanik

Ray,
Try this:
http://forums.parallax.com/showthread.php/86333-Solar-power-supply...

Dom..

erco

In playing around, I noticed that if any shade fell on the panel, the location was critical in affecting the voltage output. The panel consists of numerous cells hooked up in series-parallel to give 9V nominal input. If you shade along the short edge, overall output voltage drops drastically. If you block along the long edge, voltage stays up but probably current output falls. Thus, the panel appears to me to be two long 9V arrays in parallel.

So if you have partial sun or shade in your situation orienting the panel properly may boost your useable output over the course of a day.

As I mentioned previously, my plan is to experiment with multiple panels in a switchable series-parallel arrangement, using low-power latching relays to switch the connections. A micro will handle solar tracking for maximum power output and monitoring ADCs to measure panel voltages for relay control.

koehler

Erco, the drastic power drop off is known, and thats why MPPT has been used to mitigate such losses.
Up until just recently, they've been proprietary, and not available under $500 or more for the DIY'er.

http://www.solar-electric.com/mppt-solar-charge-controllers.html


Recently, someone started experimenting with an Arduino, and was getting pretty good results on Youtube.
I haven't looked in a couple of months, but I think there are a lot of DIY efforts going on.

I picked up a couple of 100w+ panels, and will probably get around to researching more in the near future as work slows down.
The best thing for you though, would probably to get them mounted on 1 panel, in direct sunlight, and throw a small tracker mechanism on it.

There are some decent, cheap DIY hacks you can use to follow the sun using LED's.

Just a quick google, and this little girl's simple 1 axis tracker is pretty neat.

https://www.youtube.com/watch?v=ljTJqQYSJ8g

jmg

erco wrote: »

As I mentioned previously, my plan is to experiment with multiple panels in a switchable series-parallel arrangement, using low-power latching relays to switch the connections. A micro will handle solar tracking for maximum power output and monitoring ADCs to measure panel voltages for relay control.

Latching relays cost more than FETS and Inductors, and a MUX system will always waste cell area, and power.

MPPT has many solutions
http://en.wikipedia.org/wiki/Maximum_power_point_tracking
but you can get quite close with a simple tracking load line, that is a simple V + kI - if you look at the MPPT locus line above, for a given panel and temperature, at useful 8:1 power range a straight line locus is quite good.
Almost any switching regulator can do (V+kI) feedback, with an opamp added.
If you know the cells you will be using, this can be set very close to real MPPT.

Of course, a real MPPT can tolerate a wide range of cells/ setups, but it is more complex to implement.

Another problem with MUX switching, is making the decision - because you want to jump settings only when the alternative has more power, you need to move into the voltage clamping region before the change.
The problem is, in that region it is not so easy to sense where you are - see how close the lines are together, and Temperature will move them as well.

Rsadeika

Same cloudy/rainy conditions today, and a test reading I am getting ~2.8V. Now I am really curious as to what kind of reading I would get if I had nine panels wired up. That would be 9V 9Watts 1Amp, that should be a decent amount of power to work with to maybe charge up a 6V battery. From what I understand about charging, for a 6v battery you would need at least 8V incoming, which would leave 1V wasted? Now if you were to charge a 3.7V Lipo battery you would need at least 6V incoming which would leave 3V being wasted. Is this a correct assumption?

Because of the power needs that I think that I will need are in the range of 5V, for instance an Activity Board, Raspberry Pi, cell phone, tablet, ..., etc, it is starting to look like with the 9V panels the most efficient way to go would be to charge up a 6V battery and then have a regulator to provide the 5V power needs. But then these are based on the best conditions of capturing at least 8V.

So if I go with an assumption that the best I can do on average is about 6V, hopefully on cloudy days, then maybe I should go for trying to charge up some 3.7V batteries, assuming that 6V would be the optimum incoming, then the wasted voltage would be only 1V, which I guess is not to bad. And then using two 3.7V batteries to get to 5V, would not be such a big loss when you have to use the regulator. Is any of this making any rational sense, or am I missing some pertinent piece of information?

Ray

kwinn

What you are saying makes sense. A 9V panel connected through a diode to charge a 6V battery is a reasonable compromise between best efficiency and maximum simplicity as long as the battery is not over charged or discharged.

The best one can do is to use an mppt battery charger on the input/charging side of the battery and a switching regulator on the output side to provide the constant voltage output. This decoupling of the input and output will allow the mppt charger to extract the maximum power from the solar panel while the battery and output regulator provide the current and voltage the load needs. The system also needs to be monitored to prevent excessive charging and discharging of the battery.

AlvinWingate

[QUOTE=Rsadeika;1296560]I went ahead and purchased the 10 pack of Radio Shack 9V 1 watt solar panels that erco had found, now what to do with them. I have a spare Parallax Li-ion Power Pack/charger unit with a couple of batteries that I can spare for this experiment, the question is, can it work with the solar panel(s) that I have? Maybe there is a different kind of charging unit that I should be looking at?

Since my knowledge for this kind of stuff is non-existent, what would the setup look like using the solar panel(s) that I have? In the documentation for the Power pack/charger it shows a requirement of +5-12VDC @ 1amp minimum input, does this mean that I could hook ten of the solar panels in parallel(1.1amp), and this would be sufficient to get the charger to work, or am I way off base on this?

I tested one of my panels with a volt meter connected by putting it up on the inside window sill, getting light through the window, and it was showing ~2.7V, here in Michigan it is cloudy and raining, so I am thinking that maybe with a ten panel exposure it could get above a 5V input, but not sure about that. So, I guess I need some knowledgeable people to way in on this.

For another experiment I am thinking about using an Activity Board setup where I hook up one of my panels and have it do some data logging of what kind of voltage I am getting on a daily basis, that would be a 24/7 period for maybe the whole fall/winter/spring/summer period. Not sure how involved with this I want to get.

Thanks

Ray[/QUOTE]
Well I am not sure about this but were you able to find some solution? I seems bit tricky to me..

Rsadeika

I got somewhat derailed with this experiment, but I will probably get back to it very soon. I think what I am going to do next is, on a large piece of stiff cardboard I am going to fasten nine of the panels in parallel and see what kind of power I will be getting in all kinds of light conditions. I know that with one panel I was getting about 4V on a cloudy day, I want to find out if the V increases with the amount of panel exposure, and if it does by how much, real time data as opposed to theory. Just want to get some of the simple ideas straight.

Ray

jmg

A good quick app is here
ww1.microchip.com/downloads/en/AppNotes/00001521A.pdf

Which shows the MPPT points, with differing light levels.
This locus is quite close to a constant-voltage, for a given panel, and panel temperature.
At very low light levels, the MPPT peak drops a little, so you could use a slight angle on the Locus

Building a MPPT dummy load could be an interesting to speed measurements.

evanh

The voltage should be relatively constant. If the voltage drops without being loaded then it's not getting enough light to work. This part I'm confident of.

From my limited experience with crystalline silicon, If the panel is rated for 9 volts then it should consist of 18 cells in series. Optimal loaded power output is around 0.5 volt per cell. Unloaded in bright sunlight will net you maybe 11 to 12 volts - varies a bit with tech I think.

Not sure if amorphous silicon is the same though.

jmg

evanh wrote: »

The voltage should be relatively constant. If the voltage drops without being loaded then it's not getting enough light to work.

Yes, the curves in the link I gave in #10, show the Open (no load) voltage varies with light level, as per the clamping diodes effect, but the optimal MPPT voltage does not vary nearly as much.

That means if you know the cell you will be using, and are prepared to find that MPPT point on a test bench, that you can design a simpler 'preset MPPT' controller (can be just a SMPS chip ), which can then be compared with a more complex software solution, as a reality check.

macrobeak

There are many threads covering the solar panel MPPT topic.
This is one thread giving some background . . .
http://forums.parallax.com/showthread.php/154879-Solar-Panel-MPPT-methods?highlight=solar+mppt

erco

OT, but I thought this fellow's simple solar tracker was clever:
http://www.josepino.com/?simple_sun_tracker


No electronics. He uses a small secondary solar cell to power the rotate motor. If I understand it correctly, the motor only rotates in one direction. The secondary solar cell is shielded with aluminum foil which allows it to track (power the rotate motor) as the cell is alternately exposed to sunlight and shade from the foil. In the morning, the tracker resets (continues rotating ~180 in the same direction) when the sunlight (coming from an oblique angle) is reflected off the foil onto the cell.

evanh

Ah, I was purely responding to the opening poster's varying voltages. JMG just happened to post his comment moments before me.

Publison

erco wrote: »

OT, but I thought this fellow's simple solar tracker was clever: http://www.ebay.com/itm/256GB-Micro-SD-Card-class-10-with-Adapter-/221592150261?pt=LH_DefaultDomain_210&hash=item3397ebe0f5

No electronics. He uses a small secondary solar cell to power the rotate motor. If I understand it correctly, the motor only rotates in one direction. The secondary solar cell is shielded with aluminum foil which allows it to track (power the rotate motor) as the cell is alternately exposed to sunlight and shade from the foil. In the morning, the tracker resets (continues rotating ~180 in the same direction) when the sunlight (coming from an oblique angle) is reflected off the foil onto the cell.

erco, I think you posted the wrong ebay link.

erco

Thanks Jim. My bad. Fixed link to http://www.josepino.com/?simple_sun_tracker

Not even Ebay for once!

Publison

Pretty neat!

Digikey and Mouser do not care those fotocells he specs. It will be impossible to build this.

A sun tracker allows to get 9 to 12 hours of direct sunlight, enough to charge some batteries. I saw some sun tracker circuits long time ago but those are kinda complicated; sensors, fotoresistors, digital circuits, transistors, etc. As I just need something simple, I decided to build my own.

Rsadeika

So I went ahead today and taped 9 of my panels in a 3x3 configuration on a piece of stiff cardboard and wired them up in parallel. Today is a cloudy day here, a quick reading shows ~5.95V - 6.20V spread, at the cloudiest time . When the sun peaks out. but it does not shine on the panels, I get ~7.20V - 7.85V spread. I also did a one panel test in the same light condition and in the same orientation, I was getting ~5.75V. It looks like by having a 9 panel setup, I am getting at least a 1V pickup. These are just some very quick readings, without really getting into any precise orientation of the panels. Since one panel is rated at 9V 1W, with the 9 panel setup I should be getting, at the lower readings of ~5.95V - 6.20V and 9W or ~0.9Amp, correct?

I guess the next thing I might do is hook up one of my old GG PPUSB boards and see if I get enough real power to drive it. I will have to setup my Activity board and do some data logging to see what kind of readings I would get from maybe dawn to dusk.

Ray

evanh

Next time you test, put an amp meter across them. Yes you should expect up to about 0.8-1.0 amp output in that config.

Shorting out solar panels is perfectly okay to do. They can operate shorted in full sunlight indefinitely. In fact it's pointless putting a fuse in circuit for overload protection.

Rsadeika

I am ready for my next experiment, but I do not know what I should expect to see. I have a Li-ion BoeBot Power Pack-Charger, now, the next time I am reading, lets say, 7.5V on my solar array and I plug it into the barrel plug that I hooked up to the solar array, what would I expect to see? I actually already tried it today when I was getting a reading of about 7.0V, all that I got was the green power LED to come on, but I never got the blue LEDs to come on. I guess there must be something missing in my setup?

Ray

evanh

Open circuit voltage is near meaningless. Measure the short circuit current, this tells you very closely how much can be powered from the panels.

Rsadeika

OK, so, what the heck is a short circuit current, and what does a setup look like to get that measurement?

The official panel rating is 1W - power output, 9V - output voltage, and 112mA - output current, that is what it shows on the box. What I am trying to do, using volts = watts/amp, is to determine what the actual amp output would be at, lets say 5.0V. I am assuming that the given 9V 1W 112mA is a hard number, so how would you figure out a new value for a 5.0V number. If I do 1W/5.0V, that is not going to produce the correct figure, because actually, doesn't the watt and amp numbers change when the rating of 9V drops to, maybe 8V, or any other number? I am not sure if I am stating this clearly. It seems like I should be able to calculate how many 9V 1W 112mA panels that are receiving 5.0V of light should produce a 1A minimal amount of power. For starters how do I get the new amp rating given a 5.0V sun input?

Ray

kwinn

Short circuit current is just what it says, the current the panel outputs when shorted. Of course to measure that you need to put an ammeter across the panel. Just make sure it can handle the maximum current the panel produces, which is not a problem for such a small panel. Start on the 10 amp range and go to a lower range if the current is lower than the next range down.

Rsadeika

I am using an old Sears 82417 meter which can measure V, ohms, and it has two settings, one for 200mA and the other for 10A. I am going to make the assumption that these are acting as an ammeter.

Today is not a very good solar panel day, it is overcast and my panel array is showing 3.64V. I decided to see what the 200mA setting would show, and the number was .1mA. Now that I have two values I can make some predictions as to what I should be getting at 5.0V. I guess I have some numbers I can work now, I hope this is real.

Ray

Hal Albach

"I am using an old Sears 82417 meter which can measure V, ohms, and it has two settings, one for 200mA and the other for 10A."
Good heavens, man! Get a Harbor Freight flyer that offers a free DMM from Centech and go get one, usually no purchase required.. It may not be the worlds most accurate but at least it has more ammeter ranges, 10A, 20uA, 200uA, 20mA, and 200mA. And it includes a diode and transistor tester.
You can also get the flyer online.

RDL2004

I have some of those RadioShack solar cells also. As soon as they arrived the first thing I did was check the amps and volts. In direct sunlight they put out very close to their advertised spec, about 9.1 volts at 107 mA. In cloudy but still very bright light the volts remained fairly high, but the current dropped dramatically, to around 8 to 10 mA of current, if I remember right. I checked the voltages open circuit (nothing connected) and I checked the current through a 22 ohm (3 watt) resistor.

Rsadeika

I am trying another experiment, I found a YUASA 6V 4.0Ah SLA battery laying around. It has not been charged for about four years, so I figure I will see what I can do with that. I checked the battery and it was 1.10V, at least it was not reading 0. I also found a blocking diode which I connected to the + side of the solar array. I hooked up the battery to the solar array at midday today just to see what would happen; most of the afternoon the solar array was showing 4.75V. I just checked the battery and it was showing 1.66V, I was surprised, I thought it would still be 1.10V. Now I will leave the battery connected all night to check how the blocking diode is working. At the worst I am thinking I will see 1.10V in the morning, I would interpret that to mean that the battery no longer holds a charge. Tomorrow should be a partly cloudy day around here, I will keep the battery hooked up to the solar array all day and see what happens and the end of the day.

Ray

erco

That Yuasa SLA sounds dead, so don't give up if it won't take a "solar charge". OTOH nothing bad can come of your experiment. You might want to test it first with a real SLA charger just to see if it will hold a charge. But SLAs are great long-term storage batteries. Much better than Lipos, which deteriorate when left fully charged indefinitely.

RDL2004

I picked up a 6 volt 7 AH SLA battery off eBay with the vague idea of using these RadioShack solar cells (I bought 10) to make some kind of LED lighting setup for my balcony, but I haven't made much progress.

The batteries aren't too expensive to buy new. I got mine from seller "expertbattery" for $12.45 including shipping. They sell many other sizes.

Rsadeika

This morning I unplugged the solar array from the battery and the reading was 1.10V, back at the same place that it was yesterday before the solar charging event. Yesterday, the solar array was putting out about 4.75V, it got the battery up to 1.66V, not sure how long that took because I was not manually data logging the process every fifteen minutes. Today it is supposed to be a partly cloudy day, I will guess that maybe some of the readings from the panel should be in the 9V range. Leaving the battery plugged in, all day, I am hopping to see the temporary battery charge to be maybe 3V or 4V.

Now I am starting to wonder what the wear and tear would be on a good SLA battery, considering that a solar charger would be supplying a very unstable power supply, in my case, anywhere from 3.0V - 9.4V, with a lot of fluctuations. I guess now I will have to buy a new 6V SLA battery and start thinking about adding an Activity Board to the mix so I can start data logging to get a better idea as to what the real numbers would be for my indoor window sill location. I think I need a bigger window sill.:-)

Ray