Capacitor with the lowest possible leakage current

Thread Starter

cmartinez

Joined Jan 17, 2007
8,218
I've been working on a project for a while now that requires it to be powered from a pack of 4 AAA batteries in series. The output from those batteries is connected to a TPS709 3.3V regulator. Per its datasheet recommendation, I installed a 2.2 µF tantalum bypass capacitor at its output. I didn't bother placing a cap at its input other than a standard 0.1 µF because the batteries powering it are at close proximity and are not subject to sudden demand surges.

Anyway, the circuit being powered by said batteries draws very little current (just as I want it to), with my instruments showing a 14.7 µA draw from the battery pack. But sometimes I have to connect what I call a "comm card" to said circuit, which consists of a MCP2221A USB to UART converter and a couple of LEDs. When said card is connected to the circuit, it draws its power from the TPS709 regulator. Well, I discovered that connecting the card caused the MCU installed in it to sometimes reset itself. And I attributed said phenomena to the sudden current demand from the TPS709 when it tried to start up the MCP2221A chip.

My solution was to change the 2.2 µF cap at the regulator's output and replace it with a 10 µF cap, also tantalum. Lo and behold! the spurious MCU resets that sometimes manifested themselves whenever I plugged the "comm card" to my circuit were no more!

Problem solved, or so I thought. After changing the tantalum cap, I re-checked how much current my circuit was drawing, and I now had a reading of 15.1 µA instead of 14.7 µA. I am blaming the 0.4µA difference in the new cap's intrinsic leakage current.

Question, am I using the best capacitor for this application? Is there a better type of cap other than tantalum out there with an even lower leakage current value?
 

crutschow

Joined Mar 14, 2008
34,282
Is there a better type of cap other than tantalum out there with an even lower leakage current value?
Well, a film or ceramic type capacitor would generally have lower leakage than an electrolytic type (which a tantalum is).
But you are worrying about an inconsequential current.
14.7μA will take about 7.76 years to discharge the typical 1Ah AAA alkaline, whereas 15.1μA will take about 7.56 years so, unless a difference of 0.2 years out of 7+ years is important, I think you are good with the capacitor you have. ;)
 

jpanhalt

Joined Jan 18, 2008
11,087
The DS even recommends ceramic capacitors on the output. That is not always the case.

Source: Datasheet,p.15
The effective capacitance is the minimum capacitance value of a capacitor after taking into account variations resulting from tolerances, temperature, and dc bias effects. X5R- and X7R-type ceramic capacitors are recommended because these capacitors have minimal variation in value and ESR over temperature.
 

Thread Starter

cmartinez

Joined Jan 17, 2007
8,218
Well, a film or ceramic type capacitor would generally have lower leakage than an electrolytic type (which a tantalum is).
But you are worrying about an inconsequential current.
14.7μA will take about 7.76 years to discharge the typical 1Ah AAA alkaline, whereas 15.1μA will take about 7.56 years so, unless a difference of 0.2 years out of 7+ years is important, I think you are good with the capacitor you have. ;)
Thanks, crutschow. May I ask what calculator you used to arrive at the numbers you've described? I'm asking because I've tried three different sources online. And none of them gave me the same answer.
 

MisterBill2

Joined Jan 23, 2018
18,170
Every battery chemistry is different, and many different interpretations of battery life are used. And the self-discharge of many battery types will have them stone dead long before 7 years pass.
 

Kjeldgaard

Joined Apr 7, 2016
476
Just be careful with what sort of ceramics you choose.

X5R and X7R, in not too physically small sizes, will in most cases be usable in μF values.
 

Bernard

Joined Aug 7, 2008
5,784
I find a large variation in leakage in a batch of same value caps. The lowest leakage caps are marked for timing use. One remarkable cap, 10,000 uF @ 10 V = 21 meg. ohms.
 

crutschow

Joined Mar 14, 2008
34,282
May I ask what calculator you used to arrive at the numbers you've described?
My (virtual) HP42.
I just divided the nominal 1Ah capacity of a typical AAA alkaline by the load current to get the time (in hours) for the battery to discharge.
Of course that time doesn't include any self discharge of the battery over that long time frame, which likely would be significant.
 
Last edited:

MisterBill2

Joined Jan 23, 2018
18,170
My (virtual) HP42.
I just divided the nominal 1Ah capacity of a typical AAA alkaline by the load current to get the time (in hours) for the battery to discharge.
Of course that time doesn't include any self discharge of the battery over that long time frame, which likely would be significant.
I mentioned the effect of self discharge back in post #5. There are some super long lif batteries available but they are not AAA size or style, and they are not at all cheap.
 

jpanhalt

Joined Jan 18, 2008
11,087
The Energizer Max AAA is rated for 10 years shelf life and 1200 mAH (https://data.energizer.com/pdfs/e92-1119.pdf). I don't know to what percentage of charge that 10-year life is based, but if it is something like 90% of original, then it would lose 120 mAh capacity in 10 years (87600 hours) or 1.4 uA per hour average. Of course, if the state of charge at 10 years is more than 90%, then the self-discharge rate is even lower. If it is based on 20% discharge ( a value on the net), then the self discharge rate is 2.8 uA/hour.

Edit: Changed 1.4 uA per hour to 1.4 uA per hour average.
 
Last edited:

jpanhalt

Joined Jan 18, 2008
11,087
This is a nit, but the drain is 1.4μA.
Whether is per ns, ms, hour, or year it's still just 1.4μA.
Agreed.

I was thinking that the assumption self-discharge was linear may not be true. Writing out "per hour" was meant to emphasize that averaging. Obviously, mAh/h = mA.

This morning, I tried to find a self-discharge curve for alkaline batteries without success. We do know that even with relatively small loads (a few mA), the curve is not linear. I suspect a shelf-life curve is known but not readily available, since the claimed life is 10 years.
 
Last edited:

Audioguru again

Joined Oct 21, 2019
6,673
You can buy cheeeeep leaking batteries at The Dollar Store or high quality alkaline Name-Brand batteries everywhere.
Energizer Battery Company has spec's and graphs on their website for all their batteries (look at Tech Info) at the bottom of the first page. They guarantee their batteries last for 10 years if not used. They have a Lithium AAA 1.5V battery that has a mAh rate that is 3 times the rate for their alkaline battery at a very high current of 1A. Its rate at low currents is the same.
 

Attachments

Thread Starter

cmartinez

Joined Jan 17, 2007
8,218
You can buy cheeeeep leaking batteries at The Dollar Store or high quality alkaline Name-Brand batteries everywhere.
Energizer Battery Company has spec's and graphs on their website for all their batteries (look at Tech Info) at the bottom of the first page. They guarantee their batteries last for 10 years if not used. They have a Lithium AAA 1.5V battery that has a mAh rate that is 3 times the rate for their alkaline battery at a very high current of 1A. Its rate at low currents is the same.
I've even seen a new line of batteries that promise a shelf life of 20 years!:

1605718408258.png
 
Top