Which AA battery configuration for longest runtime?

lbenson

Senior Member
I want to run an 08M2 with HC-11 wireless configured for lowest power consumption. The picaxe would report temperature changes and battery level, report in 4 times a day, and once a day ask for a response from the controller telling it what time it is. It would wait for up to a minute for a response, and then continue.

4 possible configurations for the AAs that I have thought of are:

1. run straight off of 3 AAs (easiest)

2. run off of 4 AAs with an LP2950 5V regulator, understanding that LP2950 gracefully passes through input voltages below 5V

3. use diodes to drop the voltage of 4 AAs to 5.5V, with the understanding that at very low currents, you won't get the "standard" .6V diode drop

4. run down 4 AAs to under 5.5V, and connect them straight to the picaxe.

Which, according to theory, would run longest?

Other suggestions welcome. I could use 5 AAs and the LP2950 in a 6-AA box.
 

srnet

Senior Member
Check where the regulator is on the HC11, you may be able to remove it and run everything from 3V, 2AAs.

If it wont run on 3V then 3xAAs and a decent 3.3V low drop out regulator, such as the MCP1700 (LP2950 uses a lot of current in standby) would allow you to run from 3xAAs.

Or run from a single Lithium Ion camera\phone battery and feed it with a small solar panel and one of those 20p lithium chargers off eBay.

I ran up a remote environment sensor doing something similar with a LoRa device, 2 AA Lithiums (3V) would power it for around 5 years.

Its not the size of the battery that counts, its what you do with it.
 
Last edited:

neiltechspec

Senior Member
The HC-12 has a 3.3v LDO regulator (I measured its output) that seems to go into some
kind of pass through at lower voltages to enable running the module down to the specified 3.2v minimum.

I would have thought the HC-11 would be the same, although according to the spec the minimum is 3v.

Neil.
 

Hemi345

Senior Member
If you're using surface mount, I'd go with 4X AA and a MC78LC00 3v3 vreg because of the ultra low quiescent current (my favorite vreg). Since HC11 only consumes 35mA max, that single vreg can power it too. Then put the HC11 on a switch so you can get idle power to just a few uA while nothing is happpening. I use Ti's TPS27082L load switch.
 

manuka

Senior Member
Ibenson:You've not said the circumstances,but for prolonged applications (especially outdoors) your biggest concern with multiple cells may well be corrosion,leakage,dirt,salt spray & moisture at the many contacts. Cell banks of 1.2V NiMH/NiCd or 1.5 V C-Zn/Alk have driven me crazy over the years due to these connection woes - recommend you avoid such setups if after lengthy service! In Nova Scotia I'd say temperature extremes may be an issue too ...

Cell phone style Li-ion/Li-Po (nominally 3.7V) may tempt, but they're prone to tantrums & need smart voltage monitoring. They also show a terminal drop off as they discharge.

Consider instead rechargeable Lithium Iron Phosphate cell types. Although not as energetic as Li-ion, the LiFePO4 ("LFP") variant are near bullet proof, super long standby & charge/discharge life & deliver an extremely flat 3.2V per cell until ~80% discharged. Very neatly a white LED allows simple voltage testing, as such LEDs cease glowing as supplies drop under 3V - just the level that LiFePO4 will need recharging.

My initial use of LiFePO4 was in a 433 MHz Dorji ASK beacon (below), which became both a SiChip article (June 2013) & an informative Instructable. Since then they've shown their merits in all manner of grunty applications - digital cameras, electric shavers, power LED torches etc.

But for your application -I've no HC-11s on hand, but confirm the superior HC-12 (PICAXE-08M2 driven) runs a treat from just a single LiFePO4 cell. As 3.2V may be on the low side for the HC-12's regulator, extended trials should however first be made.

Many superior (~US$15 range) solar security systems are now single cell LiFePO4 fitted. Here in coastal NZ I've a swag of these smart lamps installed over various outdoor sites, & not one has missed a beat in 3 years of service. This is in marked contrast to the pathetic outdoor service life of past multicell NiMH/NiCd setups...

LiFePO4 cells/batteries can be imported cheaply,but Lithium shipping delays often now arise. If you want to wet your LFP feet perhaps check "Lighting "sections in local hardware stores - here in NZ/Australia Bunnings now sell 14500 sized AA LiFePO4 at ~US$4 each.

THOUGHT: Given their inbuilt solar panel,weather proofing & supplied mounting brackets it's feasible to hack smart LiFePO4 security lamps for PICAXE/433 MHz use ! Stan.
 

Attachments

Last edited:

inglewoodpete

Senior Member
Thanks for the well-researched post Stan. I have been hesitant until now to get too involved with "lithium" batteries because there seemed to be too many variables. You pack a lot of useful information into one small article.
 

lbenson

Senior Member
Thanks for all the replies. I do have some of the rechargeable Lithium Iron Phosphate batteries, and may try those.

Most responders took my "Other suggestions welcome" comment to heart, and didn't address the AA questions.

ZOR ("Yoik! Another one bites the dust.")--I think you misunderstood what I was suggesting. I was wondering about draining 4 AAs down to 5.5 volts (a safe, if maximal range for the picaxe), and then connecting it to the picaxe--wondering how that would compare to 3 AAs (starting at about 4V65) for runtime.

Hemi345--thank you for the suggestion of 4X AA and a MC78LC00 3v3 vreg, with Ti's TPS27082L load switch controlling the HC-11. If that's a good configuration, I'll try it with 4 AAs & 3V3 LP2950s (which I have) without the load switch, using the HC-11 ultra low power mode (if it tests out to have the punch to reach my receiver). Running at 3V3 may be the best suggestion for runtime if the comms work out.

manuka--thanks as ever for sharing your extensive experience.

Is the idea of 4 AAs with diode drops to 5V5 a bad one?
 

ZOR

Senior Member
Sorry lbenson if I misunderstood your idea, I thought you were considering using the run down batteries as a permanent power supply. I recently gave up my burglar alarm project because of the reliable long term power problems. I bit the dust like hippy advised way back with others and bought a commercial alarm. Will always be on lookout for reliable easy powering devices for remote Picaxe projects. Regards
 

srnet

Senior Member
I have been hesitant until now to get too involved with "lithium" batteries because there seemed to be too many variables
One thing is for sure, avoid battery holders where possible, use tagged cells or those where its possible to solder leads onto the terminals without heating the cells of course.

And then arrange for charging that does not require what your building to be taken apart just to charge the battery.
 

AllyCat

Senior Member
Hi,

It's not clear (to me) whether this is an "indoors" or external application (where corrosion may be more of an issue and solar charging an obvious solution). But if the transmitter works at 3.3 volts (the PICaxe certainly can) then I'd suggest a single LiFePO4 cell (topped up by solar power) for external use, or a pair of Lithium primary (non-rechargeable) AA cells (which also have a very "flat" voltage discharge curve) for an internal application. However, note that primary lithium AAs can put out around 1.85 v each at very low drain currents and a LiFePO4 over 3.6 volts if simply "floated" across a solar panel.

Cheers, Alan.
 

srnet

Senior Member
A pair of Lithium primary (non-rechargeable) AA cells (which also have a very "flat" voltage discharge curve) for an internal application.
As a side note, the primary AA (3200maHr) or AAA (1800maHr) Lithiums are one of the very few cells that will work down to -40C.
 

lbenson

Senior Member
-40C you say ? Nova Scotia gets chilly but not quite THAT cool!
The lowest winter temperature I have recorded at this location is -17C. This past winter the lowest I recorded was -12 on Feb 14 & 15. It was -10 on January 5 & 26, and February 13.

Regarding the application, I was looking for something I could put in or on a 4- or 6-AA box and place in a location where mains power is not ready to hand inside my barn. There is nothing critical to be monitored or controlled here--I've just been curious as to what would work. As an exercise, it is possible that a solar panel could be mounted where it could catch afternoon sun--I'm not sure how much useable power you would get in the winter at 44 degrees north (with sunny days a rarity).
 
Last edited:

srnet

Senior Member
I'm not sure how much useable power you would get in the winter at 44 degrees north (with sunny days a rarity).
You dont need much though.

You ought to be able to get the sleep current of the setup under 50uA, so you would be using 1mAhr per day in sleep.

A few seconds of RF and processor activity 4 times a day ought to use no more than 1mAhr a day also.

AA Lithium lifetime would then be 4 years.

And 2mAhr a day average from a solar panel is not a lot.
 

lbenson

Senior Member
Thank you, srnet. 4 years as a very rough ballpark for lithium primaries is plentiful. It would represent no major inconvenience (or expense) if I had to change the batteries every year.

RF activity monitoring temperature changes of 2 degrees C would be more often than every 4 hours, so I'll just have to see how it goes. I'll be sending the results of calibadc--that may not reveal very much for lithiums.
 

srnet

Senior Member
If you want to read the supply volts, then a resistor divider and the internal VREF should work as well.

To turn the resistor divider off, and prevent it consuming power in sleep mode, you use a spare I\O pin to ground the bottom end of the divider.
 

Jeremy Harris

Senior Member
Like Stan, I've found the LiFePO4 cells to be excellent. Low self-discharge, very robust, long life and very easy to charge with a small solar panel. I've been using one cell running an outdoor transmitter for around 3 years now, and it still seems to have plenty of reserve capacity.

One trick I've used is to build a shunt regulator for the battery charger, using a TL431. This is set to never allow the LiFePO4 cell to charge to it's maximum charge voltage (typically around 3.65V) and clamps the charge from the small solar panel at 3.45V. Looking at the data for LiFePO4 cells, under-charging them massively increases their storage life. The same is true for other lithium secondary cell chemistries too. I've found that limiting the maximum charge voltage of LiCo cells gives a very marked improvement in life, for only a modest reduction in capacity.

In general, if you can run cells like this between 40% state of charge (SoC) minimum, 90% SoC maximum, even LiCo cells will have a cycle life of many thousands of cycles.

For small solar powered circuits I've also had some success using a single, reasonably good quality, protected 18650 size cell, with a shunt regulator set to 4V. The TL431 has a pretty low quiescent current, yet will shunt up to 100mA, so is a good match for small solar panels that can only deliver a few 10's of mA, with no need for anything other than a small Schottky diode in series with the panel to prevent discharge at night.
 

Jeremy Harris

Senior Member
There are very inexpensive Li over / under voltage protection boards [they are tiny and probably meant to integrate on a cell...] for example: http://www.ebay.com/itm/10PC-3-7V-3A-Li-ion-Lithium-Battery-18650-Charger-Over-Charge-Protection-Board-/162025038275?hash=item25b971e1c3:g:eEoAAOSwZ8ZW~htv . I like the TL431s too...
There are, BUT, they allow the max possible charge voltage, and that dramatically shortens the life of the cell when it's being constantly charged, hence the reason for using a DIY shunt regulator that kicks in well below the normal protection board voltage.

This is exactly what the likes of Tesla, Nissan etc do to increase the life of the battery packs in their cars, they never allow them to reach the maximum SoC, nor do they allow them to fall below a fairly modest lower SoC limit. Doing this increases the cycle life of ordinary LiCo cells from maybe 1000 cycles maximum to over 20,000 cycles, at the cost of a bit less usable capacity.

Many model-type lithium chargers have a "battery storage" charge mode that does the same thing, it limits the maximum charge voltage to well below full charge, as this increases the life of the cells.
 

premelec

Senior Member
All true - perhaps put a shottky [sp?] diode in series with the cell in charging direction - probably these boards aren't very smart... ;-0 In solar apps the TL431 can shunt your high end...
 

AllyCat

Senior Member
Hi,

CALIBADC10 allows the PICaxe to measure its own supply rail (Vdd) to better than 20 mV (also I've posted several code snippetts to resolve to around 1 mV and 1 degree C). So the PICaxe can be arranged to "manage" its battery supply (any type) to any desired voltage range and temperature (requiring only a one-shot calibration) without the need for any external components.

At the moment I'm preparing a "fun" application which implements not just shunt regulation, but options for PV voltage boosting (low-cost solar garden lamps usually charge only a single 1.2v NiMH cell) and even MPPT conversion, using just one PICaxe pin (of e.g. an 08M2). ;)

Cheers, Alan.
 
Top