Solar Cell Audio and their sounds

Audio Amp

This was a little audio amp from Radio Shack. This little amp has an input capacitor, so you can hook it directly to DC of a solar panel. I turned the gain up all the way and connected the amp to the output of any selected solar panels. I had previously established the capability of isolating any panel with switches. Now, not only can any panel be isolated for measurements and performance, but also for noise and audio. I was hoping to hear solar noise from the sun, or shooting stars, or blinking airplane lights.

I could only hear two things:

One was the noise of the controller. Quite irritating! And this is the reason for the 4700uF capacitor. The solution worked well. That basically ended the noise, and probable increased the efficiency of the solar panels too. I had no idea my panels had so much electrical noise.

The second thing that I emediatly heard was artificial lighting. 60 Hz hum, found in parking lots. And it is so loud that I have to turn off the amp. And this applies to the yellow sodium lights as well as florescent lights from store windows. And it does not have to be night time either. Those lights are loud, day or night.

It was not going well. I have not heard any shooting stars or lightning flashes! But I do not know when to listen either. So it may work after all, and I have not witnessed it yet. Also, I am interested to know if I can hear sun spots as they pass in front of the sun. Actually, what I want to hear is the light from solar mass ejections, if the light exists. I could have a day of lead time of a EMP. If the sound exists, I do not even know what it would sound like. No, probably will not work. I would need to hear below 0.1 Hz. To slow to hear.

Update:
After several months of listening, I now have heard flies and bees. Quite unexpected! Every time I hear bees, I climb to the top of the roof and see absolutely no bees. Then I realized they are in tree limbs above, 15 feet away. And it is definitely bees in flight as you can also hear the doppler shift as they fly toward or away from my Solar Cell sensor. Also, I "hear" flies. Flies are much loader, give only a single pass, and have a distinctive doppler and buzz that is a "fly". The acoustic signature in light from a solar pannel sounds identical to what you would hear from the acoustics in air.

Also can hear at times what I describe as nats or mosquitoes in a swarm. I have not identified this noise yet. It has musical qualities. It is high frequency: about 5khz. But some nat swarms are almost impossible to see: They can be high in the sky, and you have to know they are there before you know where to focus. The human eye can not "see" 5 khz at all. Vision is limited to only about 60hz. This strange sound can be rendered to sound and consequently heard but the original identity is masked. At least for me. I have no idea! The one and only time that I heard this sound was at a lake. And there were rain drops on the solar cells. By rain drop lensing, I know that the cells were actually seeing horizontally through these rain drops.

To further investigate these strange sounds, I purchased a small 6 volt solar cell from RadioShack, and an additional amp. The large solar panels on the roof can not be aimed, or pointed at any one target. A small portable version is the only way to see where these sounds are coming from.


I have verified the bees, flys, and other flying insects with this little setup. Ofcourse, I have to be about 3 feet away to hear these things. The big solar panels on the roof have such huge gain, that they can hear a bug over 20 feet away.

I have found that solar cells produce strange sounds, perhaps from internal resistance as the cell heats up or cools down. I can identify them, as they are usually below one hertz. And produce a "waving sound" or "pulsation" in the white noise. The white noise will wax and wane at an audible rates. The white noise is modulated, and produces rapidly changing frequencies, even up into the audible range.

And/or, these sounds were waves or ripples in the upper atmosphere. With two different densities of air layers, waves could exist in theory. But highly unlikely that I would have discovered something that no one else has seen - or should I say heard.

Here is another phenomenon:
If I am behind or in front of a automobile, I can usually tell if they have LED taillights or blue headlights. I can hear alternator whine. All incandescent have slow reacting filaments and are silent, and this corresponds to about 90% of the cars on the road.

If I am to the side of a car, I can hear the hubcaps if they are shiny and if they are moving.

Solar cells can hear the wind! This is amazing. If the wind is blowing in tree limbs that have no leaves, you can hear the wind. The small twigs are vibrating and reflecting sun light. The sound is identical to the acoustic sound of wind, and is immediately identifiable to the human ear. And it is louder inside the RV through the speaker, than outside. Some trees are louder than others, and some make it sound as if a hurricane is outside.


I did not know the frequency response of a solar cell. After all, my investigation could be hampered if the cell has restrictions on its response. Here is a way to tell approximately: By using feedback!

Reflect sun light off of a surface that will vibrate. Here I am using a tea bottle. (It is not positioned correctly in the picture.) Aim the solar cell at a reflection from the tea bottle, and have the amp positioned near the tea bottle too. Feedback reflections work great, and are very sensitive. The cell will pick up a change in the light, and cause the amp to produce a sound, and the sound will be felt by the side of the bottle. Oscillations instantly build up to produce a high pitched frequency. The loud squeals are ear piercing, and sent my dogs running for cover. So be prepared: It will hurt your ears. By changing the positions and such, the highest frequency that can be manuevered is about 10 kHz. But this tea bottle is only a crude setup. Cells may go to several hundred kilohertz, as far as I know. And I don't know. What I do know from this tea bottle is that they can go to at least 10khz.

You should be able to hear music playing inside automobiles if you can catch a reflection from window glass. I have not heard this yet...


Originally the SolarAudio Sensor was placed on top of the Air Conditioner. After only a year it was found that connectors could not be "water proofed". All wires were soldered, then silicon past, and heatshrunk, then tape with caulking to prevent tape unwinding. Then the all wiring was wrapped in aluminum foil and metalized tape to prevent ozone oxidation, solar heat degradation, and EMP shielding.


Here is the permanent SolarAudio Sensor, a half watt panel. For about two years it was on the front of the Air Conditioner. The solar cell was totally incased in Silicon Glue. The incasing experiment turned out to be good: no problems with water. However, a horizontal mounting showed only terrestrial events, and nothing of remarkable interest.
Then I mounted it on top of the Air conditioner, back where it had been originally. This works better for "atmospheric" noises, for example "fluttering" with atmospheric heat layers. The SolarAudio Sensor is on the right, and the other solar sensor is the SolarFlux Sensor on the left. It is 15 watts and VOC is 22 volts. It gives the percent of a Sol and is dedicated to the measurement of Solar Radiation (Flux). The rest of the air conditioner still needs more "metalizing" and aluminum coating. A cell boaster antenna can be seen up front, and yes, can throw off the calibration of the SolarFlux with a shadow. Also, the antenna can be heard as wind noise in the SolarAudio. But I will live with it. The AirConditioner is the highest point of the RV, and works good for all three devices.


You can not just simply attach an audio amp to the output of a solar cell. It will not work. You have to load the cell to produce fruitful voltage swings. This regime is the same as the MPP (Maximum Power Point). And as you can see, MPP changes with Solar Intensity.


If you just connect an audio amp to a solar cell. All of the different Solar Intensities converge on the voltage axis, about 21 volts. They converge at the Open Circuit Voltage, OCV. Therefore the voltage does not change much with varying Irradiance. The voltage just stays at OCV, particular at high Sols.


Here I have drawn in three static load lines. They are identified in ohmage as internal load lines, but they are still representative of external loads of my choosing. Also, internal load lines are not straight and linear.
One 0.15 ohms
One 0.38 ohms
One 1.1 ohms
The three external load lines can be represented by three static resistors, and are not like my dynamic automatic attenuator circuit. But the arbitrary load lines do point out that the maximum voltage swing (gain) is at the MPP, evidenced, and read, by the reflection down on the voltage axis.

I have invented a Load Controller. And the loading has to change depending on Irradiance. For example, at low light levels there has to be no loading.

This is only a 5 volt sensor. I think, theoretically, that a 100 volt sensor would be better with a higher signal to noise ratio. But that is a different topic...

I have invented an automatic attenuator that is presented as a variable load to the cell. The section with the transistor and the three LEDs in parallel is an automatic attenuator. It produces a variable load line to the solar cell. The external load mimics the internal load line at all the MPPs of varing solar intensities. The 220k pot adjusts for maximum audio. There is little audio on either side of this adjustment, wither toward Voc (Voltage Open Circuit) or Isc (Short Circuit Current).

I lay claim to two inventions here:
One is a new way to determine the MPP. And that is by my Noise-Tracking. My Noise-Tracking uses a feed back loop to maximise thermal and solar noise from a solar cell.

And two, a resistive network or resistor, that produces less heat. A resistor that radiates thermal heat, or produces less thermal heat in the first place as I have shown, is more efficient and quieter. In my example,I use LEDs.

I discovered these. I may discover for fun, but always still lay claim to the facts.


added a choke made from a relay coil.
Have better side vision. Here, I have added half spheres. This increases side vision, for earthly stuff, and still have sky vision. But it decreases overall sensitivity.
I had a lot of things wrong with the above schematic. For one, by accident I did not have the 33k resister that is included in the bottom schematic. The result was indiscriminate maximum gain when the impedance of the solar cell was low. During the day when the sun is high, the impedance of the solar cell can be a few ohml. This supper high gain is just going to amplify solar noise. In contrast, at night, the solar cell impedance is in the megohm range, and there is not enough gain. For example in the bottom schematic; The gain is the ratio of the 2M ohm feedback resistor devided by (infinity+10k+33k). Not enough gain. But too much gain in the day.
I increased the gain by adding nine solar cells in series, instead of the one old single solar cell. You do not need current in a detector, it just adds noise. Current is used for power production. You need voltage for a better signal to noise ratio!

The spare parts in the upper right corner are there because of an experiment. I built a physical "AGC" with these parts, but never finished the schematic. The problem was that the gain is determined by the 33k ohm resister. And any attempts to effect its effective value with the AGC circuit are cancelled out. In other words, if you short out the signal after the 33k ohm resister with an AGC circuit, the gain of the op amp goes up. Almost exactly! The addition of the 10k resistor was critical for the proper operation of the AGC circuit. I was not happy, and abandoned the AGC project. Besides. I could not find a bipolar or FET that could withstand the 50+ volts. Perhaps I need to order some FETs and approach this from a different direction. I left the parts just scattered on the schematic.

Solar noise and solar cell noise are both huge. Shame; I really needed that AGC circuit.

However, on second thought, there is the possibility that an AGC may not benefit that much during the day. The day time high flux will decrease background impedance and increase background capacitance. A small light weight signal does not have a chance during the day.


What about atmospheric air showers? Air showers, originating in the atmosphere, from one or more particles, grow from the subsequent multiplication of particles. One of several manifestations is light. I would be looking for a fast light. I am looking for something that my eyes have never seen. Even on a black night, I have never seen this light. The solar panels have to see light that my eyes can not see.

Not that this is impossible. For example, I have heard atmospheric layers oscillating, from a slow changing, less than a Hz, to a fast 100 Hz. I have heard silver glistening waves on a lake ringing and singing. I have heard many insects, like nat swarms with a hundred tiny beating wings, and hidden insects, with larger wings, busily flying in trees. And some cars produce alternator whine in their headlights. Some of these things can not be seen with human eyes. And some only rarely seen.

To detect subatomic particle showers, I may have to cut out the 22pF capacitor in the feedback. I had to include the 22pF cap for the AGC circuit. Under high gain, and with AGC, there were stability problems with oscillations. I need this cap out anyway for fast events like lightning, which I have already detected. Even if the frequency is above 20kHz, and out of the audible range, I still want the response. I do not know why. Just because.

I might make another comment; It is about my stupidity. I ironiously thought that I might increase the light sensitivity by forward biasing the panels a little, perhaps a half volt. Not seeing the error of my ways, I installed the circuit. The power supply noise was intolerable and the parallel impedance was self defeating. But the biggest problem - in fact a deal breaker - was what Einstein said about such a notion. Einstein addressed this very notion at the turn of the century when he was talking about the photoelectric effect, and how only color determined the ev of the ejected electron. I was NOT going to increase the quantity of current. I was NOT going to increase sensitivity. Instead, I was only going to shift the response wavelength toward the inferred! The damned embarrassment is not worth a little more real estate.

At the output of the OpAmp the DC level is set to operate between the rails, or about 2.5 volts. The AC signal is about 0.4 volts peak to peak.
A closer look at the AC signal is about 0.4 volts pp.
The Solar Cells have a DC component of 50 volts.
The gain of the OpAmp is about 15.
Therefore the 0.4 volt on the display represents about 0.4/15=0.026 volts.
In a terrestrial setting, the background noise from the sun is 0.026/50 =0.00052, or about 0.052% percent. That is here at my RV.

Here the span is slowed down to 40mS/cm. And I can just make out a dominant frequency. It is about 3.5 cycles per 2 cm, 3.5 cycles per 80mS. About 44 Hz. A spectrum analyser would bring out the frequencies better than my imagination. There is no chance to find the 44Hz.
There are large temperature gradients today. 60 degrees ambient in the shade, 80 in the sun. It is overcast, and at the moment, 71% Solar Flux, 0.71 Sols. Lots of tree limbs near by. Some wind. The limbs could be vibrating in the wind. I have heard vibrating limbs before in the speaker. Amongst all the solar noise, This is inaudible. I can not clearly say if I can hear it, in this case.

Set the gain of the preamp to 44k/1.5M = 35.
Changed the feedback cap on the preamp from 22pF to 2pF.
Added shielding.

The Solar Fluctuations are about .8 volts
The preamp gain is 35.
The DC voltage is about 50 volts.
This is hard to estimate because of the terrestrial fluctuations. Nether the suns fluctuations nor the terrestrial fluctuations are actually noise. Both are very real. What would be noise, is Solar Cell Noise.
0.8volts/35/50volts=0.00046, or about 0.046%
the background "power noise" from the sun is 0.046% Percent.
This is almost impossible to estimate: The 50volts DC solar volts does not correspond to the Solar Output.


A possible way would be to measure the noise at low solar flux levels, for example 15%. There the charge carriers have not been used up yet in a solar cell, and there is no 50 volt open circuit limit.

Also, any solar Sol values above 30% causes my Maximum Power Point Loader to engage. This loads, and artificially reduces, the 50 volts to a lesser value.

A quarter of the Solar flux, or Solar Irradiance, measured in space is degraded by the atmosphere of a clear sky. Is this attenuation softening the noise, as in an average, or is it contributing to noise? I do not know. I have not heard a twinkling star yet, but I have heard slowly changing pitches, which I suppose, comes from changing thermal layers in the atmosphere. In these two cases, the atmosphere contributes to fluctuations.

Research Papers report a variance of 0.001, or 0.1%, from hour to hour or day to day. So far, I am seeing about the same. But my next project will be to design an Audio Range AC Automatic Gain, and then there will be no way to know the percent. At least without a switch or two to cancel out the AGC.
Removed the 15 volt zener, to achieve AGC action, in light as low as possible. In bright sunlight, the transistor should operate nowhere near 50 volts. Both the DC Loader, as well as the AC AGC, are reducing the 50 Open Circuit volts. I have just designed and built the circuit, and have not tested it yet. A plus, is that I do not need the zener for protecting the transister anymore.