I am running an outdoor project, and I came across the following article about using a super cap to power an Arduino.
So I did some research to find a 4F super cap, and I saw the following entry in the datasheet about how it's endurance is 2000 hours. A LiPo battery generally discharges after a few years depending on usage, so I'm wondering what does this "endurance" statistic really mean?
I would love to power my 3.3V solar powered project via a supercap, but not if the capacitor is going to die after 4 months of continuous usage. I get the temperature advantages, but would I be losing endurance/usage lifespan if I switch out to a supercap?
I want to add that my circuit uses around 8mAh. I strobe the power using a hardware timer every 20m. I guess my real question is would it be more feasible to use a supercap instead of a rechargeable battery since I’m only using 8mAh but in outdoor conditions 24/7 (in a case).
Interesting... thank you for that! 2000 hours would be the absolute max lifespan if I exceeded the min/max values of the datasheet. My next question is... are there any significant disadvantages to powering a 3.3V microcontroller with a max power consumption of 40mA (overall it's 8mAh since I strobe it with a TPL5110) with a 4F capacitor vs. an external battery (LiFePO4 probably)?
I will solar be charging the batteries/capacitor. I'm trying to figure out if the lifespan is double/tripled/halved if I use a capacitor vs. a battery given my circuit is running outdoors in the sun 24/7.
What is the actual current your project uses when powered on?
You mention "strobing" it every 20 minutes - I suppose you mean you switch it on shorty? How is that?
Solar & supercap is a viable solution, depending on your actual power draw and the actual hardware you're powering. Note the discharge of a supercap is that of a capacitor, not of a battery, so voltage falls much faster.
czu001:
... my circuit is running outdoors in the sun 24/7.
On average it would be 12/7 max cause, you know, night time.
Add to that clouds, and inefficiency in early morning and late afternoon due to atmospheric dimming. And unless you move your collector, the incidence angle will also diminish available solar power.
I use lipo batteries in all my solar projects, but I do use a supercap in a timer circuit where the purpose of the cap is to keep the timer alive when replacing the primary batteries, connecting and disconnecting, etc. The voltage of a supercap starts diminishing immediately upon drawing current. Don't calculate the time it will last based on it's total capacity, but rather figure the time until it falls below the minimum voltage to work your circuit.
Envelope-back calculations tell me this will be about an hour for your device powered with a 4F capacitor if it's charged up to 5 volts, and your circuit can use it down to 2.5 volts.
Bottom line is that you will need to be real stingy if you want a solar powered project. Or get bigger solar cells.
What is the actual current your project uses when powered on?
My current draw while the circuit is powered on is ~40mA (on the VERY high side and being extremely conservative).
You mention "strobing" it every 20 minutes - I suppose you mean you switch it on shorty? How is that?
I'm using the TPL5111 to ground the regulator for 20m at a time. My circuit will only be powered on at MOST for 2 minutes until it grounds the regulator. On average, my circuit is powered on for about 30s every 20m. I think the TPL is measured at 20uA.
So I calculate the average as 40mA * 90s = 3600mAs/3600s = 1mAh. On the high side it's 4mAh (40mA * 360s, etc.). I add an additional 4mAh for error (doubled).
Envelope-back calculations tell me this will be about an hour for your device powered with a 4F capacitor if it's charged up to 5 volts, and your circuit can use it down to 2.5 volts.
Thank you!!! Assuming the cap can be charged up to 5V sounds like the critical part. I'm wondering if there's some combination of a coin cell battery + solar charged supercap... (I'm not a hardware guy, so I'm not sure if that even makes sense).
What do you think? Would a supercap make sense for this application?
Solar charged batteries are commonplace - usually LiPo type nowadays, but lead/acid is also used a lot in solar installations. Works fine, too, just a bit harder to move around.
If you use a supercap, and charge it to 5V (you need a 5.5V rated one), you have about 330 seconds of power for your project. That's 5.5 minutes. If each on cycle lasts 2 minutes, that's less than three cycles.
This is in reality even less as the current does not drop with voltage in a microcontroller the way it does in a simple resistor. So you'd get two cycles out of your fully charged supercap.
On the other hand a 2,200 mAh LiPO battery (a common 3.7V 16450 type would offer this) will last for weeks at this use without recharges.
wvmarle:
If you use a supercap, and charge it to 5V (you need a 5.5V rated one), you have about 330 seconds of power for your project. That's 5.5 minutes. If each on cycle lasts 2 minutes, that's less than three cycles.
My calculation of 'less than an hour', actually 41 minutes, was based on the previously stated discharge rate of 8 mAh(sic), and fully charging and discharging the capacitor, as a theoretical maximum life. 5 minutes is a LOT more reasonable based on my experience with smallish (1 and 4F, 5.5 volts max) supercaps. They make much bigger capacitors and capacitor banks these days, so storage is monetarily limited.
This is in reality even less as the current does not drop with voltage in a microcontroller the way it does in a simple resistor. So you'd get two cycles out of your fully charged supercap.
On the other hand a 2,200 mAh LiPO battery (a common 3.7V 16450 type would offer this) will last for weeks at this use without recharges.
I totally agree with the choice of an appropriate capacity lipo. In some low power applications, I use a 500 mAh battery, and in others I use 1200. These handle anything from 8 to a few dozen LEDs, so long as only some of them are on at once. Disclaimer: I live in Arizona.
For very low power applications, I find it necessary to completely turn the device off for most of the time.
And the way solar works, it's cyclic: All night long, the system draws on the battery. Once the solar cell is illuminated, it begins to generate voltage. Your circuit determines the current drawn from that voltage. At a certain point the voltage will be sufficient to turn on the circuit and normal discharge will begin. Let's say this happens at 10AM. The charge rate will be the current in-from-solar-cells minus the current drawn-by-the-circuit. The voltage may drop to less than useful at 6PM, 8 hours (28800 seconds) later. You can determine the current used with ohm's law and algebra. But it all changes based on charge and illumination, so it's really a multi-dimensional calculus problem, so an exact solution is not appropriate here, but ...
You can determine, by trial and error how much solar cell capacity will keep your battery charged.
The actual average draw of your circuit determines the needed current and voltage. This is why I tested my circuits for more than a year. The difference between summer and witnter is like ... night and day? Overdoing it will always work though. I guessed the right amount of cells (4 x 7.2v x 200mA, in parallel) after a couple discharge cycles. After than I had enough data to do the math.
I found as a rule you need roughly 10x the solar panel rated output compared to your current draw. A 20W solar panel with 7 Ah lead/acid battery was able to run a 2W air pump continuously - even through a week of gloomy days, where the panels don't produce much (but still some). After that it took a day or two of good weather for a full recharge.
For your project:
40 mA * 90 s / (20 * 60) s = 3 mA average. Ignoring the quiescent current of your control electronics. At a LiPo's nominal3.7V that'd be 11 mW, requiring a 100 mW solar panel.
To run your project for 5 days without any sunlight you need 360 mAh capacity; a 500 mAh rated battery sounds like a good size.
Back to the solar panels. Such a battery can be charged at 50 mA (iirc 10% of rated capacity in mAh is the maximum safe charge current). So your solar panels should be able to produce 50 mA, any more than that would be waste. A 5V, 50 mA panel would come to 250 mW rated output.
As at 5V you have losses, that 100 mW is too low as minimum. Look for 150-250 mW solar panel, appropriate LiPo charger, and proper LiPo battery. Make sure you have battery protection circuits in place, preventing overcharging and overdischarging, both which can cause serious problems.
40 mA * 90 s / (20 * 60) s = 3 mA average. Ignoring the quiescent current of your control electronics. At a LiPo's nominal3.7V that'd be 11 mW, requiring a 100 mW solar panel.
These calculations are exactly what I've been looking for; however, it's great to see my "trial and error" landed on very close power requirements. I am currently using the following for my circuit, and I've been using for the last 3 months with batteries at a steady 100% (even with the pouring rain).
And for thermal protection charging.. I'm trying out this from eBay.
The batteries already include protection circuitry (no thermal though). It sounds like I can shave off more solar panel size, but for $1.50 it's great.
Having said that, it sounds like supercaps are out. I may need to spin another thread, but I'm actually looking very hard at either a 500mAh LiFePO4 battery or 500mAh LTO. LTO seems like a perfect fit (I'm using the nRF52840), but I get mixed reviews on it. Any suggestions for a battery since the supercap is out? I don't like LiPo simply because it is more voltage than I need and the instability.