1
\$\begingroup\$

Im designing a product which needs around 5 to 7v. I need to power it from a rechargeable battery.

It spends most of the day in an off state, consuming about 0mA. Sometimes is powered by the user, draws 40mA for 4 seconds and 200mA for 1 second, then it shuts down.

The peculiarity of the project is that it needs to run without any charging for 5 months a year. It will be subjected to ambient temperatures up to 50°C.

I've excluded, for one reason or another, pretty much every type of battery i know:

  • lithium based batteries are not happy to stay at full charge and almost full discharge for such long time
  • nichel based ones seems to have a big self discharge
  • lead acid batteries seems like the more suitable type, but it also seems that they are very heat sensitive. I understand that there are also some "deep cycle" ones, that would be more suitable for my application, but are pretty rare.

I don't have any weight or space constrain (well... As long as they are reasonable)

At this point I'm quite lost, i hope for your help. I'm not an expert on batteries, some of my considerations can be wrong.

\$\endgroup\$
11
  • \$\begingroup\$ I do not know where you live, but deep-cycle and gel-cells are common here. Deep cycle batteries tend to be bulky and heavy. Gel cells are sealed, but need a compatible charger. You cannot over-charge a gel-cell without risk. \$\endgroup\$
    – user105652
    Commented Mar 28, 2018 at 23:23
  • \$\begingroup\$ @Sparky256 well, good to know. It was reported to me that are not so common. Thank you. The charger will be integrated into the product, and from the outside will be connected only a power supply, so i'll design the appropriate charger \$\endgroup\$
    – valerio_new
    Commented Mar 28, 2018 at 23:26
  • 1
    \$\begingroup\$ Not so sure that lithium isn't the answer here. If you don't want them to stay fully charged for a long time, just over-dimension them, and charge to x%. Power density of modern lithium cells would still allow them to be competitively compact. \$\endgroup\$
    – Marcus Müller
    Commented Mar 28, 2018 at 23:31
  • 1
    \$\begingroup\$ How many of these 360mAs cycles does the battery have to offer? Can you specify "about 0mA" a bit better? \$\endgroup\$
    – Marcus Müller
    Commented Mar 28, 2018 at 23:33
  • 1
    \$\begingroup\$ So 360 mAs × 30/d × 30d/month × 5month = 3600 × 0.45 As = 0.45 Ah. That sounds very doable with normal lithium cells at your temperatures and low currents. In fact, how long are you planning for that device to live, at all? You might not even need to be rechargeable... \$\endgroup\$
    – Marcus Müller
    Commented Mar 28, 2018 at 23:50

4 Answers 4

Reset to default
2
\$\begingroup\$

Total energy per day (Joules=AmpsVseconds/efficiency*reps/day) (40x4+200x1)*5.5*30/0.85 =70J/day, 25kJ/year,11kJ/5 months.

If you only get to recharge after 5 months, why use rechargeable at all if the energy is low? D-Alkaline=75kJ, AA=10kJ (55C max) (alkaline shelf life 5 years) or lithium-irondisulphide duracell AA (1.6V, 10kJ) 60C max ( shelf life 20 years)

Alkaline performance is excellent at 50C, (poor when cold) Shelf life will be hugely degraded at 50C of course, but still manages 50% at 5 years see fig 15, self discharge graph In your case the choice of cell size is about peak current at lowest temperature, not capacity.

I think your understanding of lithium rechargables is not quite right: To maximise the energy stored in them, they are recharged to a terminal voltage. The higher that is, the more energy you can get, but the more damage you will do. 4.3V = short life, 4.0V = long life. NiMH batteries have also improved with versions called LSD - low self discharge, which gives shelf life to 2 years.

An arrangment with 3 cells and stepup to 5V is good for this sort of job. A small micro can run directly from the batteries (2.7-4.5V) without needing the switchmode, to keep track of sleep time etc. A (cheap) stepup, with poor idle current can be turned on for transmissions.

See also this answer to a similar question

Use a tiny solar panel (1" square, 40mW) to stretch out the alkalines life if you want. 70J/day / (3hoursSunPerDay *3600secs) = 6.5mW. An alkaline may last years with solar float charge

\$\endgroup\$
8
  • \$\begingroup\$ They asked me specifically to use a rechargeable battery because the product will be almost sealed and replacing the battery, even after 3-4 years would be a pain. The solar panel idea is something we are working on for the next release, because we will have to use a flexible panel. Maybe for this release we will go for the alkalines. Thank you for your answer! \$\endgroup\$
    – valerio_new
    Commented Mar 29, 2018 at 21:39
  • 1
    \$\begingroup\$ For very long product lifetimes, rechargeables don't have a very good track record. With Lithium evolving rapidly (and getting thinner and less reliable separators) there is probably little credible data available, on what cell capacity and leakage will be like in 5 years. Except maybe Tadiran, who do make their livelihood out of long and lifetime batteries. \$\endgroup\$
    – Henry Crun
    Commented Mar 29, 2018 at 23:45
  • 1
    \$\begingroup\$ with recharging batteries inside gas-tight enclosures, you have to consider the outgassing. Gel-cells (lead-acid) are especially bad for this they always outgas. \$\endgroup\$
    – Henry Crun
    Commented Mar 29, 2018 at 23:54
  • \$\begingroup\$ Our enclosure is definitely not gas-thight but is quite a pain to open. At least some tens of screws, for up to a thousand units per owner, we definitely can't ask the owner to open each and every unit, even once in a while. I hope that for this release we will settle with alkalines and in the next release with the solar panel maybe we'll be using lead acid (sealed?) \$\endgroup\$
    – valerio_new
    Commented Mar 30, 2018 at 0:02
  • \$\begingroup\$ Lead acid do not last, and do out gas. Probably your last choice. \$\endgroup\$
    – Henry Crun
    Commented Mar 30, 2018 at 0:06
1
\$\begingroup\$

Don't use a rechargeable battery. Most of them tend to have short self-discharge times.

40mA for 4s followed by 200mA for 1s sounds like a ZigBee or other similar 2.4GHz wireless sensor application...

I built a wireless sensor product with similar energy/duty cycle requirements (compared to yours). We used a lithium thionyl chloride primary battery. Specifically we used the Tadiran TL-5930.

http://www.tadiranbat.com/assets/tl-5930.pdf

It is a D-size battery with a capacity of 19Ah at 3.6V. It has an operating temperature range of -55 to 85C.

To get the 5V to 7V from the 3.6V battery just use one of the many available low power boost converter chips. Texas Instruments has a good selection.

Using this battery and a low power boost converter, our testing has shown that the product we designed can last anywhere from a few months to a few years depending on how often the user accesses it.

\$\endgroup\$
2
  • \$\begingroup\$ you have to watch lithium batteries. Many cannot be shipped by air, even some small (AA) ones, and they have become much tighter in the last year. USA designers tend to forget this, as they can get them by surface. The other major, is you can't get them right-now, or right-here \$\endgroup\$
    – Henry Crun
    Commented Mar 29, 2018 at 7:38
  • \$\begingroup\$ They asked me specifically to use a rechargeable battery because the product will be almost sealed and replacing the battery, even after 3-4 years would be a pain. I'll see if we can switch to a primary battery. Thank you! \$\endgroup\$
    – valerio_new
    Commented Mar 29, 2018 at 21:35
0
\$\begingroup\$

@MarcusMüller i'd say that we can assume 30 as the daily maximum number of cycles. I'm planning to use a mosfet to interrupt the power supply to the board to save more batteri, there will be some leakage but hopefully it will be in the order of tens of microamps at most

So that amounts to 0.45 Ah in 5 months.

You can buy much higher capacity lithium batteries from many distributors (e.g. Mouser), and as long as you keep them within the specified operating conditions, you'd really don't have to worry about damage to the cells. With cell capacities significantly above twice what you need, and mature battery management ICs that prevent deep discharging, I'd also argue that cell aging is not much of a problem over the next decade.

\$\endgroup\$
0
\$\begingroup\$

High density batteries of consumer grade tend to leak more at high density and high temp, just as capacitors but depends on chemistry and quality.

Lower density type chemistry tend to also have lower leakage current.

LiPo's tend to have lowest self discharge between 50 to 66% SoC and that's why they are shipped with products, this way for longest shelf life.

@Ali's suggestion of Tadiran TLI LiPo cells fit this category from a very reputable company. Low mAh rating but low self discharge rating.

Alkalines tend to have 8 yr ~10 , more or less shelf life and I concur with @Henry's suggestion and have done this many times for my own verification using a regulated float voltage on Alkalines but didn't use a PV.

So you have many choices actually, including these from Panasonic

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.