## AA alkaline discharge curve @5mA

27 posts / 0 new
Author
Message

Can anyone point me to a typical discharge curve for a AA alkaline battery with a 5mA discharge rate?  A quick Google search only uncovered curves down to 50mA; all others are at higher rates.

Background:  An upcoming app will drive a small fan periodically.  The drain of the fan is ~4mA so with AVR etc. it is about 5mA.  The aim is to drive the fan at the same speed even when the battery is draining down.  Driving the fan with bench supply or fresh battery indicates an effective PWM duty cycle of ~50% to drive the fan at desired speed.

If the curve is very flat until near "dead" then there is no need to fuss with the duty cycle.  But if it drops 25% or more I might want to bump up the duty cycle a bit to try to achieve the desired speed.

A "typical" curve I've found:

for Duracell "Ultra"; a 50mA test

Now, what does it look like at 5mA?

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

The first curve of the datasheet for a Panasonic LR6XW AA Alkaline battery will give you the number of hours at 0.9, 1.0, 1.1, 1.2V for 10 mA constant current:

From that you can roughly extrapolate to the number of hours at 5 mA. Does that help?

- John

Bob -

I've been though this exercise, myself. It gets even more uncertain when you try to factor in alternating run/sleep cycles. The question is really how the discharge looks at very low rates. Here is what I did: (1) Estimated what the curve looks like for sitting on a shelf. They give you some numbers for self discharge, I think. On that curve, you have two points (new point and "discharged" point) so you draw a curve that is similar in shape to the others, but going though those two points (2) Estimate where your discharge falls between the lowest current curve they provide and the "zero discharge" curve you just fantasized.

IMHO, that is about as good as you can do.

Jim

Until Black Lives Matter, we do not have "All Lives Matter"!

The datasheet for a Duracell MN1500 starts its constant current graph at 5mA.

#1 Hardware Problem? https://www.avrfreaks.net/forum/...

#3 All grounds are not created equal

#4 Have you proved your chip is running at xxMHz?

#5 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand."

jfiresto wrote:
From that you can roughly extrapolate to the number of hours at 5 mA. Does that help?

Maybe.  I'm really after the shape of the curve as in the "Kirkland" chart above.

Here is my thinking:  If the curve shows that the battery V holds at, say, 1.4V to 1.3V for most of the lifetime then I'll just set my duty cycle for that and be done.

But if the curve is more like the Kirkland curve, then I might well want to adjust the PWM duty to try to keep the desired motor RPM as the battery V drops.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

Brian Fairchild wrote:
The datasheet for a Duracell MN1500 starts its constant current graph at 5mA.

This one?

But that doesn't really tell me the shape of the curve.

  But from the above we can do a bit of extrapolation.  As expected, the black line Service Hours is roughly linear at low draws -- 500 hours at 5mA; 250 hours at 10mA; 125 hours at 20mA; ...

I'm looking for something like the above.  The blue curve is definitely flatter.  Now, what does 0.25W represent in current draw?  At 1.25V that would be 200mA, right?

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

Last Edited: Tue. Jul 5, 2016 - 03:50 PM

Try this one:

- John

jfiresto wrote:
Try this one:

We also just found that here, labeled "Procell" instead of "Coppertop". ;)

That's the one needed.  No need to go further right now because we don't have the job yet.  But given that, I think indeed I might want to boost the PWM duty a bit as battery V drops from 1.5 through 1.4 to 1.3 and lower to keep the fan speed more-or-less constant.  Virtually "free" as I need to measure battery V occasionally for low-battery warning.  Thanks all.

Lee

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

You are welcome, and thank you Brian for pointing out the Duracell AA datasheet. Its 9V sibling has a curve for 2 mA constant current, a curve I have been seeking for another project.

- John

jfiresto wrote:

Its 9V sibling has a curve for 2 mA constant current, a curve I have been seeking for another project.

Gotta ask:  We avoid 9V because of the cost and [lack of] power density.  Any particular reason you focus on it?

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

Last Edited: Tue. Jul 5, 2016 - 04:43 PM

small fan periodically.

Then you will get a bit more time, because the battery will recover in between.

You do not have a fan that you can read back?

costwise or lack of free IO pins?

and if you want constant fan speed throughout the battery life you could just put a smps in front of the fan.

we have been playing with AAA batteries and it seems that every vendor has its own discharge behaviour we have seen celles that keep the voltage steady for a long period and then suddenly start dropping rather fast and we have seen cells that hold for a long period and then slowly drop.

Do you know what ceel is used always? or is the user able to change the batteries? in the last case there can be alcaline/ zinc / rechargeables inserted and you will not know the curve.

sparrow2 wrote:
Then you will get a bit more time, because the battery will recover in between.

Indeed.  The total time will be what it will be.  The reason for the question is to try to keep the fan about the desired speed.  When the battery drops from e.g. 1.5 to 1.2 that is 20% less, right?  And the simple fan is now ~20% slower.

meslomp wrote:
costwise

Costwise.  There are no parts left on the schematic. ;)  Small AVR; diode across the motor; indicator LED w/ resistor; mode switch.

meslomp wrote:
we have been playing with AAA batteries and it seems that every vendor has its own discharge behaviour we have seen celles that keep the voltage steady for a long period and then suddenly start dropping rather fast and we have seen cells that hold for a long period and then slowly drop.

For the same chemistry?  Interesting.

meslomp wrote:
the user able to change the batteries? in the last case there can be alcaline/ zinc / rechargeables inserted and you will not know the curve.

True enough.  In this application the typical is alkaline.  But one doesn't know whether it is going to be name-brand Ultra or no-name from hardware depot.

But if I feather the PWM based on battery V then I should be able to get approximate desired fan speed even for different curves.  As mentioned,

theusch wrote:
Virtually "free" as I need to measure battery V occasionally for low-battery warning.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

I would consider blinking the led with short on cycle as at 2 to 10 mA steady state will be a significant power drain.

Jim

FF = PI > S.E.T

ki0bk wrote:
I would consider blinking the led with short on cycle as at 2 to 10 mA steady state will be a significant power drain.

Indeed.  The specs are for red/green.  Green blips when inactive; "I'm alive".  IME even a millisecond or so blips the LED nicely.  Red blip for low battery.  (Using current for low-battery indication always seems like an oxymoron, doesn't it?)  Solid green when active but indeed I would pulse that so it "looks" like solid-on.

Anyway, we don't have the job yet so no use doing more experiments.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

theusch wrote:

For the same chemistry?  Interesting.

Same chemistry, maybe. Same composition, probably not.

I've just replaced a couple of AA cells in a torch; they were the ones that came with it. Their weight was significantly less than the Duracell parts I replaced them with. Now, given the volume was the same, the only difference can be with the density.

#1 Hardware Problem? https://www.avrfreaks.net/forum/...

#3 All grounds are not created equal

#4 Have you proved your chip is running at xxMHz?

#5 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand."

Brian Fairchild wrote:
the only difference can be with the density.
May be related to:

http://lifehacker.com/test-if-yo...

David

When the battery drops from e.g. 1.5 to 1.2 that is 20% less, right?  And the simple fan is now ~20% slower.

It's not that simple :)

if the fan was an 300 ohm resistor (5mA at 1.5) it would give 7.5 mW but at 1.2 4.8mW

Then air resistance is in 3th power of speed so I guess it's even less at 1.2V

Again,

theusch wrote:
Anyway, we don't have the job yet so no use doing more experiments.

;)

sparrow2 wrote:
Then air resistance is in 3th power of speed so I guess it's even less at 1.2V

So all the more reason to adjust the PWM duty cycle to keep the fan speed more-or-less constant, right?  (if the customer desires that...)

And indeed, then more power would be drawn and shorten the battery life.  Compromises, compromises...

A bit of background:  Customer specifies motor/fan assembly, with nominal specs to run at 1.5V which draws ~4mA.

Our design has 2x AA alkaline; 3.2V when fresh; ~2V when dying/dead.  Put in a Tiny run directly from battery.  Use PWM on a port pin and when the adjusted PWM duty indicates the same fan speed as when driven directly from 1.5V the AVR+fan is a bit less than 4mA.  [I've seen this before on fan apps -- one can cut the duty cycle from 100% to say down to about 80% and save power with little or no effect on fan speed, which I attribute to inertia and "coating".  Cut it down further, though, and the motor/fan slows enough that when re-energized it takes more current to get it up to speed again.  Anecdotal.]

I hope we get the job so we can spend some time playing with the variables. ;)  Indeed, the 4mA draw off the Tiny port pin introduces a bit of voltage drop.  And it will be interesting to do the bench tests at e.g. 1.2V to see how much fan speed slows.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

Last Edited: Thu. Jul 7, 2016 - 02:19 PM

Indeed, the 4mA draw off the Tiny port pin introduces a bit of voltage drop.

I suspect this is a pretty cost sensitive product?

(In which case I should stay out of the discussion...)

But an option instead of using the Tiny as a PWM'd high side driver is to put a small NFet, (2N7000 ? or similar), on the low side.

(V+ to Motor to NFet to Ground).

Although the voltage drop from the Tiny's pin driver is "small", you might find the external NFet's saturated D-S voltage drop is even less, giving you more useful battery life.

That's the concept, untested.

The 2N7000 costs roughly \$0.06 qty 2000.

No uC to Gate resistor needed, but probably want a pull down so that the NFet can't go linear as the battery and micro die from low Vcc.

Some calculations would show whether this was even a consideration or not, and hence not needed.

Might have to look at the NFet to make sure you can turn it on at the very low battery voltages...

I also missed whether or not you are actually measuring the fan's rpm, or are just characterizing the fan model RPM vs voltage, and having a little look up table.

Obviously easy to RC filter the supply and measure the average supplied voltage, and hence your estimated fan RPM.

If you have a sample fan on the bench you could also see how much ripple is in the supply voltage and perhaps directly measure the fan RPM, without the need for a true current monitor, (energy wasting sense resistor, etc).

Minimalistic designs aren't my thing.  I'm pretty good at adding parts!

JC

Edit: Typos

Last Edited: Thu. Jul 7, 2016 - 03:49 PM

DocJC wrote:
I suspect this is a pretty cost sensitive product?
DocJC wrote:
Minimalistic designs aren't my thing. I'm pretty good at adding parts!
theusch wrote:
meslomp wrote: costwise Costwise. There are no parts left on the schematic. ;) Small AVR; diode across the motor; indicator LED w/ resistor; mode switch.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

what kind of motor do we talk about, BEMF brush .... ?

theusch wrote:

jfiresto wrote:

Its 9V sibling has a curve for 2 mA constant current, a curve I have been seeking for another project.

Gotta ask:  We avoid 9V because of the cost and [lack of] power density.  Any particular reason you focus on it?

Oh, it is no big deal. I am trying to estimate the battery life of a B&K Precision 390A multimeter I have been evaluating. Someone, presumably at the Chung Instrument Industrial Electronics Co., Ltd., decided to power it with a single 9V battery rather than a couple AA's. The meter draws a constant 1.8 mA down to roughly 6V, before flaking out and drawing much more current at lower voltages. I reckon the meter will run for 300-350 hours, stand alone, and 200-250 hours when logging to a computer over USB. That will be good enough for me.

All in all, it is a good multimeter for the money, with great support from B&K, and reasonable if sometimes uneven specs. I tend to the notion that all the Far Eastern meter makers in its price range pursue what you might call specifications of opportunity. If the product can meet a tighter spec at little added cost, then it will; otherwise, it will meet one that is good enough.

So as not leave you with a post of leaden prose, here is a picture of the multimeter along with a thermocouple amplifier I am assembling to test the meter's temperature ranges:

I will use the meter and amplifier to refine the heating/cooling schedules of a hot plate reflow scheme I have been musing with. It is a fun break from writing software.

- John

sparrow2 wrote:
what kind of motor do we talk about, BEMF brush .... ?

I didn't look closely.  We were given a sample of the motor/fan assembly by the customer.  IIRC it is rated up to 5V, and is 400rpm at 1.5V.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

jfiresto wrote:
Someone, presumably at the Chung Instrument Industrial Electronics Co., Ltd., decided to power it with a single 9V battery...

Indeed, quite common for many bench instruments.

jfiresto wrote:
The meter draws a constant 1.8 mA down to roughly 6V, before flaking out and drawing much more current at lower voltages.

Now, y'all are the sparkies, but I'm guessing a buck-boost regulator not LDO...

For long-term logging use, consider patching in a better power source.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

I am puzzled as to what is going on. Another, similar grade, nothing fancy DMM did not show any change in current as I reduced the B&K's battery voltage from 9 to 6V. I will try a scope on it when I get a chance.

- John

jfiresto wrote:
I am puzzled as to what is going on.

Again, I'm not a sparkie.  If really interested, time to dismantle that bad boy and see if the innards are potted.  If not, decipher some of the power circuit.

It could be different regulator; more LDO.  Or the "good" model has 3V innards and the "bad" model has 5V innards.

============

Back to my topic:  Good news; I get to do the experiments mentioned above as the customer indicated today that they intend to order prototypes.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.