Solar Power Monitor

Hey everyone!

I have a project that I'm interested in taking on that is essentially monitoring the output of a solar panel over time. I'm fairly new to Arduino and electronics in general so I apologize if it's painfully obvious what the solution is.

I want to mount a solar panel on the roof of my shed and monitor the maximum possible power it generates over time. I'm going to be tracking a bunch of data (light intensity, temperature, humidity, time of day, etc) and I want to plot all the data. I have a decent grasp of the components and sensors needed, my only question is how do I actually monitor the possible max power continuously?

I understand that for the power to flow, and for me to monitor that flow of power, I'll need a load, but how do I have a load that adjusts depending on the output of the solar panel?

I'm not trying to power anything specific with this panel, just trying to figure out how much power a square foot could generate, then I could scale that up to my whole roof and have a decent idea of how much power I could potentially generate for my roof.

For a complete picture of what I'm doing, I'll have an x by y sized solar panel on the roof, hooked up to some sort of sensor that can continually monitor voltage and current, which is being stored on an SD card. I will also have the other sensors hooked up, but i'm confident that I know how those all work!

Thank you for your help!

If you want to measure the maximum power thats available , you will need a MPPT charge controller of some sort.
You can either buy one or build one , but if you arnt familiar with the electronics involved then buying one is the easiest.
The power thats being generated can either be dumped into a resistor which will get hot , or used to charge a battery.
Rather then reinvent the wheel, its a lot easier to simply go here https://pvwatts.nrel.gov/ and type in your latitude, and it will give you the answer you are seeking.

The maximum power available is simply proportional to the illumination intensity at the relevant wavelengths.

You can measure that with a suitable photodiode using a transimpedance amplifier.

You have to decide what voltage you need to get from the panel, then you can decide how many amps you can draw and maintain that voltage. Open circuit will give you the maximum voltage with NO current. Short the output wires to get the maximum amps, but voltage is zero. You need to decide something in between.

Back, years ago when we were building PV charge controllers for RVs, there was a panel of incandescent light bulbs in parallel, that could be screwed in-out to change the load and determine the current capacity of a PV panel. The bulbs are still in a box, I think. I could check to morrow, but I think they were 24 volt bulbs because some panels were 24 volt. They worked on 12 volt panels, as well.

You could do something similar, but use relays to turn the lights on/off while measuring the voltage from the PV panel.

Paul

There might be a weather station in your area that already logs solar radiation, like this one.

Measuring short circuit current of a sample panel could give an indication of what can be harvested.
What will be harvested depends on more than just solar power. MPPT charger, batteries, etc.
Leo..

jremington:
The maximum power available is simply proportional to the illumination intensity at the relevant wavelengths.

You can measure that with a suitable photodiode using a transimpedance amplifier.

Thank you for this. I read through that powerpoint and I believe this will be useful. If you are familiar with photodiodes, can you suggest a relatively precise yet cheap one? I’ll have to check through my arduino kit sensors but I believe I have this one.

Will this be suffice?

mauried:
If you want to measure the maximum power thats available , you will need a MPPT charge controller of some sort.
You can either buy one or build one , but if you arnt familiar with the electronics involved then buying one is the easiest.
The power thats being generated can either be dumped into a resistor which will get hot , or used to charge a battery.
Rather then reinvent the wheel, its a lot easier to simply go here https://pvwatts.nrel.gov/ and type in your latitude, and it will give you the answer you are seeking.

Perfect! Thanks for this. I did read through some past posts and someone had a similar project in mind and they were suggested a MPPT, but I guess this doubles down on that. I'll have to look into these. I appreciate the suggestion!

Paul_KD7HB:
You have to decide what voltage you need to get from the panel, then you can decide how many amps you can draw and maintain that voltage. Open circuit will give you the maximum voltage with NO current. Short the output wires to get the maximum amps, but voltage is zero. You need to decide something in between.

Back, years ago when we were building PV charge controllers for RVs, there was a panel of incandescent light bulbs in parallel, that could be screwed in-out to change the load and determine the current capacity of a PV panel. The bulbs are still in a box, I think. I could check to morrow, but I think they were 24 volt bulbs because some panels were 24 volt. They worked on 12 volt panels, as well.

You could do something similar, but use relays to turn the lights on/off while measuring the voltage from the PV panel.

Paul

Thanks for the reply!

I apologize if this is obvious, but if I have a solar panel rated for 12V 3W and I hook it up to just a current sensor, will the output change depending on the output of the solar panel, or will I need to hook up another load, like you mentioned with the bulbs?

You don't need an MPPT controller to measure the energy output but you may need one in order to maximize the energy output.

An MPPT controller can be expensive so care is needed to ensure it makes economic sense. I don't have one because to get the benefit from it I would need to invest heavily in larger batteries and sunshine in the UK is not regular enough to provide a return on that investment.

Measuring the energy output just requires measuring the current and voltage at regular (and frequent) intervals. Amps * volts = watts (power) and watts * time = energy.

You need to take measurements frequently because the amount of solar radiation can change very quickly if a cloud passes in front of the sun or if birds perch on your solar panel.

...R

Robin2:
You don't need an MPPT controller to measure the energy output but you may need one in order to maximize the energy output.

An MPPT controller can be expensive so care is needed to ensure it makes economic sense. I don't have one because to get the benefit from it I would need to invest heavily in larger batteries and sunshine in the UK is not regular enough to provide a return on that investment.

Measuring the energy output just requires measuring the current and voltage at regular (and frequent) intervals. Amps * volts = watts (power) and watts * time = energy.

You need to take measurements frequently because the amount of solar radiation can change very quickly if a cloud passes in front of the sun or if birds perch on your solar panel.

...R

This was my original approach. If I just have the solar panel connected to a power sensor, that would be a known resistance, the solar panel would output a varying Voltage, and thus it would monitor the current/voltage over time. Does that sound correct to you?

Does that sound correct to you?

No. You still don't understand the issues.

The power available is proportional to the short circuit panel current.

The power delivered depends on the load.

jremington:
No. You still don’t understand the issues.

The power available is proportional to the short circuit panel current.

The power delivered depends on the load.

I understood that and I apologize if my wording is off, i’m new to this still. Maybe I’m confused at what I’m trying to measure. My yard and the neighboring yards have tall trees and I want to know if installing solar panels would be practical. The plan as mentioned before is to put a test solar panel in average exposure area and see what kind of power I generate. Maybe I’m interested in the ratio between measured power vs actual power? If that’s the case, if the rating of the panel is 5V with an efficiency of 80%, and I have it in series with a 1000 ohm resistor, I should have 5mA of current. Meaning I should be delivering 25mW of power. If I monitor the actual power delivered, I should be able to plot the actual power vs time and calculate how much power I’m losing due to coverage issues/light intensity. Then taking efficiency into account, I could also play with those values as well. If that makes sense to you guys, would this sensor work for that?

Here is a link to the sensor

Solar panels work the same, no matter how big they are. So a little tiny cell ripped out of a solar-powered calculator will give you the same result as x by y metres.

Little ones are cheap, so if you want to measure both short-circuit current and open-circuit voltage then just buy two. If you search your local discount stores I expect you could find two calculators or two garden lights for under $1 USD each.

The other advantage of small test cells is they won’t start a fire if you get the wiring wrong. A few kW from a big panel is actually dangerous.

If shadows are important then move the small sensor to the different corners of your ideal panel each day. I expect the difference will be small.

The simplest way to determine whether installing solar panels will be practical is to invite a solar panel installer to visit the location.

He/she will make a map of the skyline, including obscuring trees, consult the annual weather statistics, inspect the possible panel mounting locations and angles and will then tell you what to expect and what it is likely to cost. Usually that service is free, as the installer may bid on the job.

The second method requires to measure the power available (proportional to the short circuit panel current of any panel) at the proposed panel installation site, for at least an entire year. Several years of data would be better.

You do that to take into account the sun's path in the sky and details (like trees) obscuring the panel. You also need to take into account the average annual weather pattern to determine the total expected solar input. That information can be found on line for many locations around the world, like here and here.

jremington:
The simplest way to determine whether installing solar panels will be practical is to invite a solar panel installer to visit the location.

Unfortunately, bylaw states within our district that all renewable energy sources (Wind/Solar) are forbidden. This is 100% a political issue, as our government decided on a multi-billion dollar hydro plant to supply power, and since the population is going to be paying for it for generations to come, they outlawed renewables to decrease the time before it's paid off!

This is more as a personal project to do as I go through my degree, just to familiarize myself with electronics, programming, and renewables, as I plan to work with renewables in the future.

I will look into monitoring short-circuit current. I plan on doing this for at least 2-3 years.

Unfortunately, bylaw states within our district that all renewable energy sources (Wind/Solar) are forbidden

That sounds just like the direction the current administration is trying to take the United States: the quickest possible retreat to the failed past.

TempleofTyler:
If that's the case, if the rating of the panel is 5V with an efficiency of 80%, and I have it in series with a 1000 ohm resistor, I should have 5mA of current.

For a given brightness there is a maximum amount of power that a solar panel can produce. And it is essentially a constant current device - reducing the load beyond the optimum will not cause more current to flow.

At the extremes if it is open-circuit it will generate a high voltage and zero current - hence no power. And with a short circuit it will generate its maximum current at zero volts - also no power. To get the maximum power you need a load that gives the optimal balance between voltage and current. An MPPT controller continually varies the load to stay at the optimum point.

To get meaningful data without an MPPT controller you need to choose a load resistance that is optimal in the most likely level of sunshine. If the resistance is too high the solar panel will show a high voltage even in relatively low levels of sunshine. If it is too low then you will see a low voltage for too much of the time. I have a small panel that I bought for tests. Nominally it is 12v and 1.5 watts. In fact it produces its max power at 17.5v and a current of 86mA. For that it would require a 200 ohm resistor. A 1000 ohm resistor would be far too high. Just guessing (for the purpose of illustration) that when my panel is producing 0.75 watts it produces 15v and 50mA. If so it would require a 300 ohm resistor. With a 200 ohm resistor it would still only produce 50mA but the voltage would fall to 10v so the power would be significantly lower.

The better solar panel manufacturers have graphs that show how the current and voltage vary with the amount of sunshine.

...R

Again, to measure solar energy available, you don't use a load.
You just short the panel with a lowest possible value resistor (current shunt), and measure current.
Leo..

Wawa:
Again, to measure solar energy available, you don't use a load.
You just short the panel with a lowest possible value resistor (current shunt), and measure current.
Leo..

Sorry, but that does not make sense to me. IMHO you need to know the highest voltage that is associated with the current and you can't measure that with a too-low or zero resistance.

If what you say is true what would be the point of an MPPT controller?

...R

OP only wants to predict what could be harvested.
Cell current, without the effect of a varying load, is a single good indicator for that.

If you want to make things complicated, then don’t just use a resistive load (varying voltage).
Use a zener circuit (zener/opamp/power stage) with voltage/power to match the panel.

An MPPT charger constantly adjusts it’s values to get max power into a load or battery.
That can be measured, but not predicted. Battery voltage might be 12.0volt or 14.5volt.
I think MPP current is about 90% of the short circuit current.
Leo…