Advice for Solar Powered Board

Projects General Guidance
1

My senior design team and I are creating a research buoy (DO, pH, RGB, temperature sensors), which will run off of solar power and needs to be able to run during nighttime/no sun periods (roughly 24hrs off battery is the goal). This will only be a demonstration project so for now, 24 hour battery life and short transmitting distances are acceptable. The board is a SparkFun redboard. Can someone advise me on how to build a circuit that will do this? We were gifted a 9W solar panel so we're hoping to be able to use that. Here are the loads we are running:

Solar panel we were gifted: 9 Watt 6 Volt Solar Panel | Small Solar Panels

Loads:

Microcontroller

Temperature Sensor

pH Sensor

Oxygen sensors

RGB sensors

Transmitter

Ambient Light sensor module

LEDs

Thank you in advance!

2

How often do you need to do a measurement? You can achieve you goal only if microcontroller will be in a sleep mode most of the time. Otherwise you need a bigger solar panel (and a batteries).

3

alesam:
How often do you need to do a measurement? You can achieve you goal only if microcontroller will be in a sleep mode most of the time. Otherwise you need a bigger solar panel (and a batteries).

Thank you for taking time to respond!

Two questions:

  1. What do you define as asleep in our case? We plan on having the board monitoring the sensors pretty much 24/7, but not transmitting unless certain thresholds are detected.

  2. Assuming our board is awake 24/7, how would you recommend powering this board using solar?

4

The Redboard is a poor choice, because it is so difficult to implement an effective, extremely low power sleep mode. The sensors and the processor should all be in sleep mode or powered down until measurements need to be taken. If LEDs are there for warning purposes, have them only occasionally blink for very short intervals. Otherwise remove them.

You need to implement a "bare bones" Arduino, or (what I do) take the voltage regulator and LED resistors off of a Pro Mini.

Study this power saving tutorial and this solar powered Arduino tutorial for the best approaches.

5

jremington:
The Redboard is a poor choice, because it is so difficult to implement an effective, extremely low power sleep mode. The sensors and the processor should all be in sleep mode or powered down until measurements need to be taken. If LEDs are there for warning purposes, have them only occasionally blink for very short intervals. Otherwise remove them.

You need to implement a "bare bones" Arduino, or (what I do) take the voltage regulator and LED resistors off of a Pro Mini.

Study this power saving tutorial and this solar powered Arduino tutorial for the best approaches.

Thank you for your response!

So for our project, which uses 6 sensors and transmits, which Arduino board would you recommend?

Also, the single LED is simply there for night time operations so that the buoy can be seen.

Thank you again!

6

The Pro Mini has the same number of I/O pins as the Redboard, so if the Redboard will work, so will the Pro Mini.

However, you need to first get all your sensors hooked up and running before you can settle on any sort of final design, and you may find that more I/O pins are required.

Forget the solar panel, batteries and low power stuff until you have a prototype actually working.

7

Have you checked / calculated if the LED and lens combination will have enough brightness to be seen on the water? The U.S. Coast Guard I think has minimum visibility standards for how far a light on a buoy needs to be seen.

Not good if some boater hits you poorly lighted buoy and damages or sinks their boat.

8

With a 9W solar panel you have about 1-2W of continuous power at your disposal - provided you have a big enough battery. Assuming a 3.7V LiPo battery (that voltage should be fine for all those sensors) you can draw some 300 mA continously. Your LED is rated 350 mA but is on only half the time. That'd give you a power budget of some 125 mA.

The rest is probably <50 mA continuous - less if you use sleep modes. Transmissions use a lot of current but should be short, so if used sparingly (transmitting once an hour or once a day) don't add much to the total current draw.

Battery: you should budget for like 3 days of absolutely no sunlight (yet even on overcast days you will get power from your solar panel, and out on the water you're very exposed which is also good), so 3 x 24 x 250 mA (generous average, you probably don't need as much as 200 mA for the LED) = 18,000 mAh. Pretty big battery but it should be no problem to find one of this size.

Transmitter: that one is rated 500 ft (150 meters) under ideal conditions, and transmits at a paltry 4,800 bps. Consider HC12 modules instead. They're also pretty cheap, much better data rate and much greater range. Likely using less power overall.

9

Check into the ESP32
It has rather poor ADC specs but has a deep sleep mode that can monitor analog inputs in sleep mode.

I would recomend more time considdering battery power. How long is a storm where there is no recharging? Can you system run for 3 or 4 days on one charge.

10

saildude:
Have you checked / calculated if the LED and lens combination will have enough brightness to be seen on the water? The U.S. Coast Guard I think has minimum visibility standards for how far a light on a buoy needs to be seen.

Not good if some boater hits you poorly lighted buoy and damages or sinks their boat.

Thank you for your response! We reviewed the Coast Guard/NOAH requirements, which calls for a visible white light onboard, and the buoy itself to be yellow to indicate it is a research buoy. Luckily, this is for a senior design project, so the actual project is more of a demonstration on the functionality of the sensor system, transmitting, and solar charging concept. This specific project will most likely never actually be left in the water, and is understood by the professor to just be a prototype with plenty of room for improvement.

11

dave-in-nj:
Check into the ESP32
It has rather poor ADC specs but has a deep sleep mode that can monitor analog inputs in sleep mode.

I would recomend more time considdering battery power. How long is a storm where there is no recharging? Can you system run for 3 or 4 days on one charge.

Thank you for your response Dave!

Unfortunately our team hasn't had any experience with the ESP32, and we are on a very short time timeline here. I will be looking into the ESP32 in my free time. I'm assuming it is an Arduino based platform, and is programmed similarly?

Long term we would love for our project to be able to run full force for 3-4 days off a single battery charge, but as a prototype for demonstration purposes and our limited budget, we're thinking 24 hours will have to be acceptable. Should the system go any longer than 24 hours, it will just have to pick back up once periods of sun return.

12

we are on a very short time timeline here.

That is a real problem, because you apparently don't even have a prototype working, solar powered or not.

And here you are worrying about battery life.

I would plan on several months of hard work to get a USB powered prototype working, let alone complete an ambitious project like this.

13

As a boater and formerly CG Licensed Master - thanks for checking the requirements and clarifying your usage

There are all around white lights with a Fresnel lens you can buy off the shelf, also LED bulbs can be put in them, bad news is they are usually 12VDC so that would be yet another design issue.

Good luck, I have thought about building something similar to yours just for the heck of it

14

saildude:
As a boater and formerly CG Licensed Master - thanks for checking the requirements and clarifying your usage

There are all around white lights with a Fresnel lens you can buy off the shelf, also LED bulbs can be put in them, bad news is they are usually 12VDC so that would be yet another design issue.

Good luck, I have thought about building something similar to yours just for the heck of it

No problem sail! Buoy restrictions and guidelines was actually the first thing we checked on, and prepared a specific slide to discuss the requirements (lights, mooring, color, etc.) in our presentation that our buoy would require SHOULD it be further developed and deployed.

Our thoughts were that the improved design (not our prototype) could use its own battery and light sensor to power multiple bright LED's.

We're excited to see how this project turns out. We only have a few months with it for Senior Design, but we're hoping to have a solid concept prototype functioning at the end.

15

jremington:
That is a real problem, because you apparently don't even have a prototype working, solar powered or not.

And here you are worrying about battery life.

I would plan on several months of hard work to get a USB powered prototype working, let alone complete an ambitious project like this.

We knew from the beginning that we'd be fighting the clock, but unfortunately that seems to be the purpose of Senior Design here at our school. They give us a short deadline for complicated projects to see how we function in groups under stress to complete a deadline. We've got till the end of April to demonstrate a working prototype, so with hard work and dedication I think we will be good to go. It doesn't have to be perfect, just proof of concept that we can program the sensors and board to the task at hand. We have in hand all of the sensors and have almost completed the coding (testing the sensors as we speak) so I was just trying to get a jump start for the solar power aspect planned (we've been powering the board from the barrel jack. Things such as battery life, transmission distance, and buoy designs can always be improved upon in the final design.

16

wvmarle:
With a 9W solar panel you have about 1-2W of continuous power at your disposal - provided you have a big enough battery. Assuming a 3.7V LiPo battery (that voltage should be fine for all those sensors) you can draw some 300 mA continously. Your LED is rated 350 mA but is on only half the time. That'd give you a power budget of some 125 mA.

The rest is probably <50 mA continuous - less if you use sleep modes. Transmissions use a lot of current but should be short, so if used sparingly (transmitting once an hour or once a day) don't add much to the total current draw.

Battery: you should budget for like 3 days of absolutely no sunlight (yet even on overcast days you will get power from your solar panel, and out on the water you're very exposed which is also good), so 3 x 24 x 250 mA (generous average, you probably don't need as much as 200 mA for the LED) = 18,000 mAh. Pretty big battery but it should be no problem to find one of this size.

Transmitter: that one is rated 500 ft (150 meters) under ideal conditions, and transmits at a paltry 4,800 bps. Consider HC12 modules instead. They're also pretty cheap, much better data rate and much greater range. Likely using less power overall.

This was very helpful and gave me some things to think on, thank you!

I'm currently creating an excel sheet to get an idea of the power budget, and will reach back out once I've completed that.

Our plan is to have the buoy sample readings once every hour, and only transmit the readings if certain thresholds are detected by the sensors. I'm hoping this will significantly cut down on power usage. Voltaic (the company we received our 9W solar panel from) has a 24,000mAh battery (link below). Is this more ideal for our project? I've also included the link for their 12,000mAh batteries.

24,000mAh: https://www.voltaicsystems.com/V88

12,000mAh: https://www.voltaicsystems.com/V44

As far as transmitting goes, our transmitter is not completely ideal for actual deployment, but I believe the 500ft range will be good enough for demonstration purposes within a small auditorium room. I did look at the HC12 you mentioned and was very impressed, however it is slightly out of price range with our limited college kid budget, lol.

Thank you again for your input, I hope you'll stay tuned into this thread! :slight_smile:

17

I would make the buoy send data at least once a day regardless of whether you're over a certain threshold, so you know it's still alive.

The HC-12 is going for around USD 4-5 a piece, your budget must be really tight.

18

wvmarle:
I would make the buoy send data at least once a day regardless of whether you're over a certain threshold, so you know it's still alive.

The HC-12 is going for around USD 4-5 a piece, your budget must be really tight.

That's a smart plan that we didn't consider. We will implement a daily "check in" into the code.

I have no idea what module I was looking at price wise then, as I was seeing $300. facepalm

19

NOAA land based weather stations record data pints every 10 minutes and send updated data once an hour (not always on the hour to spread the data load out a bit) - I would expect their buoys to have the same protocol.

20

To keep you guys updated, after working with Voltaic, we calculated our power usage to be roughly 10.69 Wh/d. Taking readings every 10 minutes with transmissions every hour. Our 9W panel has a peak voltage of 6.5V and 1.42A. We sized a 44Wh battery, which will last the system roughly 3.7 days (our goal was 2 so plenty of wiggle room) without solar power (system will consume roughly 27% battery per day). Using suncalc we determined our area sees roughly 11 hours of sunlight during worst condition (winter).

The code is mostly finished (still need to code in data transmissions and data storage).We have function tested 3 out of the 5 sensors already (RGB, temperature, and ambient light). We hope to have the electronics fully functioning and tested by next Friday so we can mount the system to the actual buoy over Spring Break.

Then it's demonstration time. Not bad for less than a months work of work!

I'll create a new thread to go over the details as we get closer to completion. Thank you guys a ton for all the input!