Weathershield with supercap

[LukaQ]

I will soon give this a go, got my capacitors today. My question would be, I don't need to worry about same level of charge, do I?

I got two HB1625-2R5256-R
http://www.farnell.com/datasheets/1640984.pdf

[Uncle Buzz]

your HB1625-2R5256-R have the following specs :

Working Voltage 2.5V
Surge Voltage 3.0V

you have to use both in series to accept the max voltage and supply your moteino

[LukaQ]

that is the plan, 12.5F @ 5v. But since they will be in series, does one really need to have balance circuit or not

[Felix]

that is the plan, 12.5F @ 5v. But since they will be in series, does one really need to have balance circuit or not

I would say no, not as long as they charge uniformly.

BTW there are now some 400F caps at 2.7V, 1.8mOhm ESR, 68mm length! Two in series to make 5.4V at 200F is not bad huh!
Their size is also very acceptable.
https://www.avx.com/docs/techinfo/General/AVX-SuperCap-Comparison-Matrix.pdf

[LukaQ]

one problem though, their price is about 20$ for one... I think at that point you would be better off with batteries unless you temperature is REALLY low. Also 1mA discharge current, that seems a bit much already? Anyway, there are some that have lower current. Also at what point are you wasting money, 200F seems a bit much, for low power device

[Felix]

LukaQ,
It's $11 at mouser, I just got a few to experiment with: https://www.mouser.com/ProductDetail/581-SCCY68B407SSBLE
Where do you get the 1mA discharge? Isn't that directly proportional to the ESR, which at 1.8mOhm is the lowest I've seen in supercaps?
400F is not a lot in terms of mAh, there's a formula to calculate it, it comes to a few hundred mAh. There are many very important advantages to supercaps over even the most energy dense batteries. Here's what comes to mind:

- virtually unlimited charge cycles - compare to something like 1000x for a LiPo
- very consistent performance in very extreme temps - HOT and COLD - both where LiPos fail miserably
- can be charged and discharged at MUCH higher currents than a LiPo without any risk of catastrophic failure/explosion - LiPos can only take 1C reasonably safely

[perky]

Where do you get the 1mA discharge? Isn't that directly proportional to the ESR, which at 1.8mOhm is the lowest I've seen in supercaps?

It's the DCL spec in the datasheet, and a 400F does seem to have 1mA self discharge current. Self discharge is due to a finite large resistance between the pins, ESR is a small resistance in series with the pins, I'm not aware of any relationship between the two.

However, the DCL spec is scaled with capacitance value, it's like having multiple smaller capacitors in parallel (slightly better actually), so the rate voltage drops due to leakage is roughly the same for smaller caps as it is for larger ones. That's also why  I think the ESR is low too.

Mark.

[LukaQ]

LukaQ,
It's $11 at mouser, I just got a few to experiment with: https://www.mouser.com/ProductDetail/581-SCCY68B407SSBLE
Where do you get the 1mA discharge? Isn't that directly proportional to the ESR, which at 1.8mOhm is the lowest I've seen in supercaps?
200F is not a lot in terms of mAh, there's a formula to calculate it, it comes to a few hundred mAh.

Short answer:
With a 1 volt change in output voltage (for example if you charge it to 5 volts and your circuit works when the voltage drops to 4 volts), a 1 Farad capacitor will give 277 microamp-hours.

The relationship is linear. 10 Farads gets you 2.77 milliamp-hours, etc.

so, if I have 12.5F and I can have 1.7v drop from 5v to 3.3v, that gives me about 5.8mAh. Compared to 1000mAh of batterys...
Even yours "only" come to 94mAh.

Equation from datasheet is 0.5*C*v^2*1000/3600 = mWh

I guess question here is, what is your device consumption over time. With a lot of sleep, that goes down to or below 1mAh

And yes, here are differences in caps: my has 28uA leakage current, if I would scale that up to 400F, that would be below 450uA, less than half of one of yours. Dave Jones would say, you are pissing power away 
While true, you can charge supercaps faster, just about no one would give more than 1-2w solar cell on it, that is still low IMO

Without testing, I'm just not sure, how long would moteino/sensor run before power would need to be put in (charge supercap)

[Felix]

Mark & Luka, I am learning something every day!
You are right, 1mA DCL, but compared and rated at capacity it is much lower than the 7.5F cap discussed in this thread (70uA DCL). The 7.5F is 55mOhm ESR, many times more also, and i'm sure it's actually made of two 15F capacitors in series, rated at 2.5-2.7V to get the 5v+ total rating.

Without testing, I'm just not sure, how long would moteino/sensor run before power would need to be put in (charge supercap)

We shall see won't we!
To me they are an exotic power source.
Everyone seems to be using the overpriced volatile no-name made-on-a-ship on planet zorg chinese crap Lipos which fail if you look at them wrong. I guess it's the New Products video mantras that make them ubiquitous.
Why doesn't everyone use safe chemistries like LiFePo4 instead? Or at least some brand name LiPos? Even end-of-life smartphone LiPos seem to be reliable and run a Moteino for years.
Or like me and a few uPower fans - supercaps : they run 4evaaaa! 

[LukaQ]

they run 4evaaaa! 

well I don't know about that, they do have 500k charge cycle rating 
Even don't think esr is super important as long as it is not too high. Max power draw could be <100mA at any time.

Did that test...
Moteino without flash + adt7420 eval board with continuous sampling, sending every 1min to gateway = 6.4xmA all the time consumption for exactly 60min.
Took the led off, everything else the same, it is sitting now at 0.25mA, this  should be then running at least 25 times longer I think.
Tomorrow I'll even try to send sensor to sleep, that would bring consumption WAY down still.

I have some LiFePo4, nowhere near rated capacity, but if it can run led torch for hours, then yes, could be looking at years with moteino

[LukaQ]

After 550min, voltage on cap is 3.508v, that could run for few hours still. So take away here is, either make low power or have big supercaps or both

Should also mention, that SI1145 and tsl2561 were connected, but not used in code...

[Felix]

they run 4evaaaa! 

well I don't know about that, they do have 500k charge cycle rating 

That's pretty much forever in technology terminology