# Solar Panel Power Output DAQ

Discussion in 'The Projects Forum' started by jerrold-1, May 13, 2010.

1. ### jerrold-1 Thread Starter New Member

May 12, 2010
2
0
My group is working on a project that involves measuring of integrated power output of a solar panel. We are just wondering if it is okay to use a fixed resistance instead of a variable resistance.
Are there any methods you can suggest on how to measure integrated power output from solar cells?
Is it ok to run daily continuous Data Acquisition using a fixed resistance whose dissipated power is measured?

2. ### retched AAC Fanatic!

Dec 5, 2009
5,201
313
Yes to all of your questions.

For more exact measurements, get a highly tolerant resistor for measurement. At least 1%.

To measute the power output, you can measure the voltage across a resistor of high enough wattage rating to handle the current. Measuring the voltage across the resistor, you can convert that also to amperage using ohms law. This way, you only need to record 1 piece of data per sample. That makes logging much easier.

Build or use a voltmeter datalogger and you only need log the voltage to have the whole equation.

I would not leave the resistor in-line full time. You may want to use a relay to switch the output line to the resistor line for measurements, then switch back when done.

jerrold-1 likes this.
3. ### Bernard AAC Fanatic!

Aug 7, 2008
4,233
414
If using a fixed load resistor, try one that loads output voltage to about 66% to 72% of open circuit V. This will track peak power quite well. Example from 15W panel: R- 20 ohms, V- 16..5V, I- .88mA, P 14.52W[ peak ] @ 22.8V open ckt.

jerrold-1 likes this.
4. ### pochd New Member

May 6, 2010
13
0
Won't a fixed resistance prevent one from obtaining the maximum power output since it forces the panel to operate at a specific point of the IV curve? What then would be the proper resistance value considering the limiting effect of the fixed resistance?

Also, since you may be evaluating a commercial system that usually includes a charge controller, wouldn't it be appropriate to approximate the load that such a device places on the system? If i am not mistaken, charge controllers have a variable resistance component... I'm curious about what type of effect it has on the pane IV operating point... Can anyone answer this?

5. ### retched AAC Fanatic!

Dec 5, 2009
5,201
313
Are you trying to design a MPPT (Maximum Power Point Tracker)?

If so, you also have to track the load. (The MPP changes as the batteries or motor draw more or less from the panels.)

There are many schematics available on the in-tro-net.

Here is some reading on what helped me understand the technique.
(See Attached)

File size:
451.1 KB
Views:
34
File size:
307.7 KB
Views:
29
6. ### pochd New Member

May 6, 2010
13
0
I am trying to verify the optimal tilt angle for PV panels computed with a mathematical model that is based on maximizing extraterrestrial radiation at certain latitudes.

The data gathered will be used to compute for the percentage gain in power output over the industry setup... The systems being tested are intended for rural electrification hence they have a low power rating 20W and utilize charge controllers. I wonder if the the computation of percentage gain using a fixed resistance load would be a credible basis... It would be nice to avoid purchasing charge controllers and batteries for each system since that would be costly...

Btw, DAQ will be continuous for a six month period and power will be measured in kWh.

So would the fixed resistor do?

Last edited: May 14, 2010
7. ### retched AAC Fanatic!

Dec 5, 2009
5,201
313
Yes. A fixed resistor, over a set time per sample will give you a logical representation.

The dataloggers I built were a pretty simple 1-cell from the panels I was using, and a low power PIC controller (And 8 AA batteries)

Wake up for 1 minute every 15 min. use the ADC to gather the data from the cell. Go to sleep. Wake up in 15 more min. LOOP.

using a RTCC and proper sleeping methods, it is quite easy.

pochd likes this.
8. ### pochd New Member

May 6, 2010
13
0
The location of the test systems is at 15deg latitude above the equator. (Philippines) Would the range 66-72% of Voc still apply? Literature puts the value at 80% Voc under standard irradiance conditions. (Markvart, 2000)

May 11, 2010
241
38
I am curious about your application. Is the final product going to dynamically tilt a panel while it's in place? Is this for big panels too? Will the final product have electronic MPPT as well as physical tilt? I personally think that physical tilt system does not have enough payback until you get to 100s of Watts, because of the cost.

If you want to get a relative comparison -- a way to find the peak on a curve -- then the fixed resistor will work fine. If you want to get numbers that you can cite for % improvement in power output, then there is no substitute for actual maximum power point tracking. If you really want real numbers, then you want to determine the power gain by the tilt of the panel, by comparing the maximum power point for each setup.

If you use a fixed resistor, you will get a changing power point... this may or may not be close to the maximum power point, but it will not be the same when you tilt the panel toward or away from the sun... the more current, the higher the power point will appear over the resistor.

10. ### Bernard AAC Fanatic!

Aug 7, 2008
4,233
414
.

If you use a fixed resistor, you will get a changing power point... this may or may not be close to the maximum power point, but it will not be the same when you tilt the panel toward or away from the sun... the more current, the higher the power point will appear over the resistor.]

I plotted 3 curves with 1 W panel directly facing the sun, Feb 2009, about 1 PM 33 deg. N, 100%, one blocking screen, & 2 blocking screens to see if peak power would be the same for all conditions.As Sage pointed out there is a drift of peak power upward at higher power. 'decided that a % of OCV was a viable method of using peak power.

File size:
182.4 KB
Views:
28
11. ### pochd New Member

May 6, 2010
13
0
How did you determine that payback will not be enough? What would increase the cost if physical tilt were added to the system? We thought of doing a monthly or seasonal adjustment of tilt angle facilitated by a retrofit to existing panel mounts...

A solar collector installed at a yearly optimum tilt angle
3586 kW h/m2, i.e. approximately 10% lower than the
amount of solar radiation received over a full year by a collector
which is tilted at an optimum seasonal angle and
approximately 12% lower than the amount of solar radiation
received over a full year using monthly optimum tilt angle. (Skeiker,2009[ScienceDirect])

This would be one of our bases for attempting experimental verification for PV panels since the researcher got the gains using the mathematical model only. A similar but older study was done on solar panels but testing was done only for a short period and results were not published. Besides, no one has really verified the optimal tilt in the Philippines... That is why we would also have a setup amongst the industry, monthly and seasonal setups that will be getting daily optimal tilt albeit in a very crude manner.

I was thinking about using a high power FET to do the IV curve trace if we are forced to go for mppt. But i find this strange because the systems marketed just come with charge controllers since they are just 20W systems. Should i try to approx mppt load or charge controller load?
Thoughts anyone?

Btw, i really appreciate all the responses I've gotten so far. You've got a very active community here.

Last edited: May 14, 2010
12. ### retched AAC Fanatic!

Dec 5, 2009
5,201
313
pochd are you jerrold-1?

It seems you are continuing the thread started by jerrold-1, as he has visited today and has not posted again.

I just want to make sure jerrold-1 doesn't feel his thread was hijacked.

13. ### pochd New Member

May 6, 2010
13
0
I actually work with jerrold on the project i have been talking about. He's not available at the moment so i was assigned to take over forum inquiry. I guess ill be the dominant responder since he's working on the system mounts... Sorry for the confusion.

Any thoughts on my latest post?

Last edited: May 15, 2010
14. ### retched AAC Fanatic!

Dec 5, 2009
5,201
313
jerrold-1 likes this.

May 11, 2010
241
38
When I said that I don't think that cost is justified on smaller systems, I mean doing dynamic physical tilting of the panels... with motors and a controller. Because of the cost of the motors and controller, the gain is not enough to justify it. The added components and the complexity and maintenance...

But I see that you are talking about manually adjusting the panels to a tilt angle on a seasonal or monthly basis. I think that is a great idea, if the user will do it.

Your other question is... what is the target?

Let me see if I understand your question. Should you be optimizing the tilt for the maximum power from the solar panel, or should your optimize it for the best performance using a non-MPPT charge controller? Is that the question?

If your clients in the Philippines will be using non-MPPT charge controllers, then that is the final application. I guess that you're maximizing the current after the linear voltage drop from the panel voltage to the battery voltage. I am not sure how the tilt will affect that... I am not sure if there will be different tilts for: (A) maximum power at maximum power point and (B) maximum current into 14.4V linear charger.

But I think a very interesting topic is, can you get low cost MPPT controllers to people in the Philippines?

It has been a recent dream of mine to design a low-cost MPPT charge controller. I think it can be done. I aim for 200W input up to 48V, and output for a lead-acid battery and diversion load (like water heating). I also want to have circuits for battery charging (cell phone 5V and 1.5V cell recharging) and control of a Peltier-effect cooler for medicine and food refrigeration. Also, night-lighting with LED controller. I think that it is possible to make all this for \$25 per unit, and then set up local cottage industry production. I would base it on a microcontroller. I would make all the design open source. I am interested in your thoughts on that.

But, back in the real world as it is now...

If you're working on this for existing linear charge controllers, then I think you can do the sweep just by using a current-sense resistor input into a 12V lead-acid battery as you tilt the panel. Just make sure the battery is not totally topped off, so you don't overcharge it. I think that you can tilt the panel until the current as shown by the voltage on the resistor is at the maximum. I think that would work very simply if the target is to maximize the current going into the battery in a linear charging system.