Basic capacitor questions

Hi all,

I am considering powering pro mini sensor nodes from a coin cell, and as a result I started looking into using capacitors to handle peak loads:

http://www.ti.com/lit/wp/swra349/swra349.pdf https://electronics.stackexchange.com/questions/43306/pulse-powering-heavy-loads-with-a-coin-cell

The problem is, I am struggling with the basics :) My most significant peak load is 20mA for 35ms, happening every 3mn (radio sends ping).

The first question is whether to use a charging resistance, and if so what size. I am thinking that it wouldn't make sense to replace the sensor peak load with another peak load when starting to charge the capacitor, so I am coming up with:

-> charging resistance = 2.7V / 0.2mA = 13k (a bit more at 3V but then I assume a bit more than nominal current should be ok).

That will still charge even a 1000uF in a little over a mn so it looks like this should work.

Does this make sense so far?

The next step is to pick the capacitor; the first link above gives me a value of 1000uF through a calculation I don't fully understand...

If instead I assume that the discharge load is 2.2V / 20mA = 110ohms, for a 500uF I get a time constant of 55ms which should be plenty enough? In fact if I discharge down to 2RC I could use 160uF.

Finally I read that electrolytic capacitors have significant leakage, so I am considering this kind of ceramic capacitor:

https://www.digikey.fr/product-detail/fr/samsung-electro-mechanics/CL32A157MQVNNNE/1276-3367-2-ND/3889025

So, what are your thoughts on the plan above?

Thanks! Franck

My most significant peak load is 20mA for 35ms, happening every 3mn (radio sends ping).

One of the more important points in the TI paper is that there is only a small (about 5%) decrease in the coin cell lifetime when the peak current is 30 mA, compared to 15 mA.

Your peak current draw is substantially less than 30 mA. If it is really worth your time to research the perfect value of an added capacitor, go for it. I wouldn't.

Keep in mind that adding a resistor wastes energy.

That's a good point I guess... though the exercise was interesting in itself :)

It seems I forgot to take the voltage drop into account, with

C = I × Δt / ΔV

if I want to limit ΔV to 0.4V (2.2V towards end of life - 0.4V = my BOD setting) I end up with a 1750uF capacitor... not that attractive.

Franck

The most effective strategy to increase battery lifetime is to reduce overall system power consumption to the absolute minimum, which takes some effort.

Here is a great tutorial. Additional strategies include using high value resistors for voltage dividers, have necessary LEDs blink at the lowest practical current level and on time, etc.