Searching \ for '[EE] 2 year battery life with zigbee modules' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/power/batterys.htm?key=battery
Search entire site for: '2 year battery life with zigbee modules'.

Exact match. Not showing close matches.
PICList Thread
'[EE] 2 year battery life with zigbee modules'
2007\07\13@164913 by Rafael Vidal Aroca

flavicon
face

   Hi people,

   i'm playing with these great ZigBee modules from Maxstream (XBEE).
They are really interesting and powerful..

   My problem is related to battery life.

   According to the manufacturer battery life calculator available at:

   http://www.maxstream.net/support/battery-life-calculator.xls

   i could have a life of more than 3 years depending on the conditions
i use my modules. Well, my conditions are quite good:

   - The module is in sleep all the time, waking up each 4 minutes,
sampling ADC ports, transmitting data, and going back to sleep

   Theorically i should get at least a year, but when i tested it, my
batteries did not last for 2 months! I'm using 2 alcaline AA batteries
which gives 3V.

   Well, seems like the batteries still have energy, but their tension
has gone to less than 2,7v, which will not power XBEE.

   The problem, is that there are no commercial batteries available
with 2000mAh and 3 ou 3,3V with low voltage drops in the long time.

   Some engineers suggested me 9v batteries with a voltage regulator.

   Is it a good idea? Won't i loose too much in the voltage conversion
(9-3=6v). At 50mA i'll spend 300mW just to drop the voltage???

   Well, i really don't know how to manage the batteries to give me 2
years of life to my little remote sensor.

   Does anyone have any idea or clue?

   thanks a lot

Rafael

2007\07\13@171549 by alan smith

picon face
1. Have you tried talking to Maxstream live chat about this?
 2. Have you tried reducing power (tx power on the Xbee)?
 3. Have you actually measured the current draw when your up and alive?  Are you sure
 the module is asleep? What else is in your circuit besides the Xbee? What is your
 quiesent current draw when its asleep?
 4. You can do a efficient battery boost circuit, where the input can drop quite low and you boost to 3V.  Lots of design examples from LT and others.

Rafael Vidal Aroca <spam_OUTrafaelTakeThisOuTspamvisionbr.com.br> wrote:
 
Hi people,

i'm playing with these great ZigBee modules from Maxstream (XBEE).
They are really interesting and powerful..

My problem is related to battery life.

According to the manufacturer battery life calculator available at:

http://www.maxstream.net/support/battery-life-calculator.xls

i could have a life of more than 3 years depending on the conditions
i use my modules. Well, my conditions are quite good:

- The module is in sleep all the time, waking up each 4 minutes,
sampling ADC ports, transmitting data, and going back to sleep

Theorically i should get at least a year, but when i tested it, my
batteries did not last for 2 months! I'm using 2 alcaline AA batteries
which gives 3V.

Well, seems like the batteries still have energy, but their tension
has gone to less than 2,7v, which will not power XBEE.

The problem, is that there are no commercial batteries available
with 2000mAh and 3 ou 3,3V with low voltage drops in the long time.

Some engineers suggested me 9v batteries with a voltage regulator.

Is it a good idea? Won't i loose too much in the voltage conversion
(9-3=6v). At 50mA i'll spend 300mW just to drop the voltage???

Well, i really don't know how to manage the batteries to give me 2
years of life to my little remote sensor.

Does anyone have any idea or clue?

thanks a lot

Rafael

2007\07\13@172620 by Sean Breheny

face picon face
These are very good suggestions except #2. Assuming you have an XBee
and not XBee-PRO, it actually draws MORE current in receive than
transmit, so reducing transmit power (whose max is 1mW anyway) will
make no difference.

It sounds to me as though it is probably drawing more than it should
be. At these current levels (about 50uA) even a slight mistake or
oversight could swamp the 50uA with a much higher quiescent current.

I definitely do not think that a 9V battery would be a good idea for
long-term use, as their energy density is poor and you probably won't
find a linear regulator which would have low enough quiescent current
(and of course, as you pointed out, it would waste at least 2/3 of its
power).

Is there a microcontroller or something else in your circuit besides
the XBee? The extra drain could be due to that.

Sean

On 7/13/07, alan smith <.....micro_eng2KILLspamspam@spam@yahoo.com> wrote:
{Quote hidden}

> -

2007\07\13@173152 by David VanHorn

picon face
>    Theorically i should get at least a year, but when i tested it, my
> batteries did not last for 2 months! I'm using 2 alcaline AA batteries
> which gives 3V.

First, are you certain that the batteries you used have the capacity
that you think they have?  Counterfeit batteries are more common than
you might think.

>    Well, seems like the batteries still have energy, but their tension
> has gone to less than 2,7v, which will not power XBEE.

Check the discharge curves provided by the manufacturer, and in the
McGraw-Hill "Handbook of Batteries".  A very useful reference.

>    The problem, is that there are no commercial batteries available
> with 2000mAh and 3 ou 3,3V with low voltage drops in the long time.

Lithium 123A cells can probably do this. Available on Ebay for about $1 each.
You might use larger batteries, which will droop less.
Also think about temperature effects if your device will be exposed to
low temperatures.

>    Some engineers suggested me 9v batteries with a voltage regulator.
>    Is it a good idea? Won't i loose too much in the voltage conversion
> (9-3=6v). At 50mA i'll spend 300mW just to drop the voltage???

Depends on the quiescent current of the regulator.
I wouldn't worry about efficiency while you are on, it sounds like
your duty cycle is very low, but the "off-time" drain will be a HUGE
impact on battery life.. I'd say this approach is unlikely to result
in a good outcome.



>    Well, i really don't know how to manage the batteries to give me 2
> years of life to my little remote sensor.

Overkill is GOOD!

Seriously, the consumer probably dosen't care wether the batteries
cost $1 or $2, but they do care if they have to replace them very
often.

2007\07\13@173309 by Timothy Weber

face picon face
Rafael Vidal Aroca wrote:
>     i'm playing with these great ZigBee modules from Maxstream (XBEE).
> They are really interesting and powerful..
>
>     My problem is related to battery life.

1. Are you sure the XBee is in deep sleep, not... whatever they call it,
shallow sleep?  (Don't have the data sheet in front of me.)

2. If your alkaline batteries are putting out anything over 3.0 V, note
that the XBee increases its current usage rapidly as you go over 3.0 V.
 3.0 is great, 3.15 is not good; 3.3 is right out.

>     Some engineers suggested me 9v batteries with a voltage regulator.
>
>     Is it a good idea? Won't i loose too much in the voltage conversion
> (9-3=6v). At 50mA i'll spend 300mW just to drop the voltage???

Yeah, a linear regulator on a 9V will have a theoretical maximum
efficiency of 33%.  Not great.

Your best bet is to use a switcher, but of course that's not a simple
solution.

Maybe 3xAA = 4.5V nominal with a 3.0V linear regulator wouldn't be too bad.
--
Timothy J. Weber
http://timothyweber.org

2007\07\13@173331 by Marcel Duchamp

picon face
Hi Rafael,

Looking at the XBEE webpage, they spec the supply voltage at 2.8V to
3.4V so you cannot discharge a 3 volt battery very far before you are
too low to operate.

Battery manufacturers rate amp-hours with no regard to terminal voltage.
 Especially for consumer primary batteries.  Their curves usually show
running the battery down to half of it's original voltage and measuring
how many amp-hours that produces.

I did not see any information at the XBEE site about what they mean
concerning battery life.  But it looks like they expect you to figure
out how to extract all the amp-hours regardless of terminal voltage.

From your information you posted, it looks like a switcher could help
you out.  A boost configuration should get you where you want to be.

Good luck!

Rafael Vidal Aroca wrote:
{Quote hidden}

2007\07\13@173510 by Marc Nicholas

picon face
Rafael,

Have you looked at these?

http://www.bodhilabs.com/vpack5aa2.html

(You can change a couple of the components to get 3.3vdc).

This will allow you to use the AAs down to 1vdc.

-marc

On 7/13/07, Rafael Vidal Aroca <.....rafaelKILLspamspam.....visionbr.com.br> wrote:
{Quote hidden}

> -

2007\07\13@183735 by Rafael Vidal Aroca

flavicon
face
part 0 44 bytes
his is a multi-part message in MIME format.
part 1 114 bytes content-type:multipart/related; (decoded base64)

NƦ[bgs0Gf 2007\07\13@184002 by Rafael Vidal Aroca
flavicon
face
Sean Breheny wrote:

>It sounds to me as though it is probably drawing more than it should
>be. At these current levels (about 50uA) even a slight mistake or
>oversight could swamp the 50uA with a much higher quiescent current.
>  
>
   As i told, i'm doing a test with only and only the XBEE and 2
batteries directly connected.

{Quote hidden}

   no, nothing extra :(

Rafael

2007\07\13@184620 by Xiaofan Chen

face picon face
On 7/13/07, Marc Nicholas <EraseMEgeekythingspam_OUTspamTakeThisOuTgmail.com> wrote:
> Rafael,
>
> Have you looked at these?
>
> http://www.bodhilabs.com/vpack5aa2.html
>
> (You can change a couple of the components to get 3.3vdc).
>
> This will allow you to use the AAs down to 1vdc.
>
> -marc
>
>

I am not so sure if this will save battery power or not depending on
the efficiency. It might consume extra current from the battery
even though it can extract extra power from the battery.

2007\07\13@184813 by Bob Blick

face picon face

--- David VanHorn <microbrixspamspam_OUTgmail.com> wrote:

> First, are you certain that the batteries you used
> have the capacity
> that you think they have?  Counterfeit batteries are
> more common than
> you might think.

I especially love the batteries made by a Chinese
manufacturer that are marked "Made in Japan" - they
have no factory in Japan.

Cheers,

Bob

2007\07\13@190003 by David VanHorn

picon face
> I especially love the batteries made by a Chinese
> manufacturer that are marked "Made in Japan" - they
> have no factory in Japan.

They'll mark them "Made in USA" and "Duracell" if you like.
:(

2007\07\16@124412 by alan smith

picon face
If the Xbee is the only thing in the circuit....no serial connection? or is there a serial connection to a host thats putting it asleep...and waking it up...putting it back....etc.
 
 Are you *sure* its going asleep?

Rafael Vidal Aroca <@spam@rafaelKILLspamspamvisionbr.com.br> wrote:
 Sean Breheny wrote:

>It sounds to me as though it is probably drawing more than it should
>be. At these current levels (about 50uA) even a slight mistake or
>oversight could swamp the 50uA with a much higher quiescent current.
>
>
As i told, i'm doing a test with only and only the XBEE and 2
batteries directly connected.

{Quote hidden}

no, nothing extra :(

Rafael

2007\07\16@124516 by alan smith

picon face
If the Xbee is the only thing in the circuit....no serial connection? or is there a serial connection to a host thats putting it asleep...and waking it up...putting it back....etc.
 
 Are you *sure* its going asleep?  And if no host is connected....whats putting it asleep.  Could be the they now allow you to write code to the micro onboard, last time I checked they didnt, or didnt have much room to put anything else in there except for the stack

Rafael Vidal Aroca <KILLspamrafaelKILLspamspamvisionbr.com.br> wrote:
 Sean Breheny wrote:

>It sounds to me as though it is probably drawing more than it should
>be. At these current levels (about 50uA) even a slight mistake or
>oversight could swamp the 50uA with a much higher quiescent current.
>
>
As i told, i'm doing a test with only and only the XBEE and 2
batteries directly connected.

{Quote hidden}

no, nothing extra :(

Rafael

2007\07\16@185320 by Rafael Vidal Aroca

flavicon
face
part 0 44 bytes
his is a multi-part message in MIME format.
part 1 114 bytes content-type:multipart/related; (decoded base64)

NƦ[bgs0Gf 2007\07\16@191900 by Rafael Vidal Aroca
flavicon
face
alan smith wrote:

>If the Xbee is the only thing in the circuit....no serial connection? or is there a serial connection to a host thats putting it asleep...and waking it up...putting it back....etc.
>  
>
   There are no serial or IO connections.

   The only thing connected to xbee is the power linex (+ and GND).
Only 2 wires to the batteries.


>  
>  Are you *sure* its going asleep?  And if no host is connected....whats putting it asleep.  Could be the they now allow you to write code to the micro onboard, last time I checked they didnt, or didnt have much room to put anything else in there except for the stack
>  
>

   Yes, i'm really sure. I have tested with an amperimeter. When XBEE
wakes up the current goes to aprox 50mA, and after 1 second it goes as
low as  less than 1mA.

   I'm using the maxstream firmware with cyclic sleep, with nothing
more than their own firmware.

thanks again for the patience

[]s Rafael.

2007\07\16@210154 by Timothy J. Weber

face picon face
Rafael Vidal Aroca wrote:
>>  Are you *sure* its going asleep?  And if no host is connected....whats putting it asleep.  Could be the they now allow you to write code to the micro onboard, last time I checked they didnt, or didnt have much room to put anything else in there except for the stack
>
>     Yes, i'm really sure. I have tested with an amperimeter. When XBEE
> wakes up the current goes to aprox 50mA, and after 1 second it goes as
> low as  less than 1mA.

1 mA is, of course, way higher than you should expect.  You may need to
get an ammeter with a reliable microamp scale for this, or use a
sensitive voltmeter and a suitably-sized resistor in series with the
power supply to measure current with more resolution.

>     I'm using the maxstream firmware with cyclic sleep, with nothing
> more than their own firmware.

So you should expect at most 50 uA with 2.8-3.0V supply.

You never answered (or perhaps never saw for some reason) my question
about your power supply voltage.  Is it 2.8-3.0, or more?  If it's 3.3V,
for instance, you should expect 100 uA; 3.4V gives 240 uA.  (This is
from the data sheet.)

240 uA from a 2000-mAh battery would give around a year of life, not
counting what it's using while awake.  I can't remember what you said
your duty cycle was or which XBee model you're using.
--
Timothy J. Weber
http://timothyweber.org

2007\07\16@222157 by Rafael Vidal Aroca

flavicon
face

>1 mA is, of course, way higher than you should expect.  You may need to
>get an ammeter with a reliable microamp scale for this, or use a
>sensitive voltmeter and a suitably-sized resistor in series with the
>power supply to measure current with more resolution.
>  
>
   Hmmm

{Quote hidden}

   I'm sorry. I think i did. Anyway, i'm using 3.0 volts as supplu
voltage. 2 AA batteries of 1,5V each in series.

>240 uA from a 2000-mAh battery would give around a year of life, not
>counting what it's using while awake.  I can't remember what you said
>your duty cycle was or which XBee model you're using.
>  
>
   Nice. This little math would help :)
   thanks

   maybe, that shows that my problem is certainly not the batts, but
the fact that they can't maintain more than 2,7 volts for several months.

   I'll try buying VPack. As it comes with a battery voltage regulator,
it would keep the tension for more time with its boost feature....i think

   thanks for all the help up to now

Rafael



2007\07\16@222303 by Rafael Vidal Aroca

flavicon
face

>1 mA is, of course, way higher than you should expect.  You may need to
>get an ammeter with a reliable microamp scale for this, or use a
>sensitive voltmeter and a suitably-sized resistor in series with the
>power supply to measure current with more resolution.
>  
>
   Hmmm

{Quote hidden}

   I'm sorry. I think i did. Anyway, i'm using 3.0 volts as supplu
voltage. 2 AA batteries of 1,5V each in series.

>240 uA from a 2000-mAh battery would give around a year of life, not
>counting what it's using while awake.  I can't remember what you said
>your duty cycle was or which XBee model you're using.
>  
>
   Nice. This little math would help :)
   thanks

   maybe, that shows that my problem is certainly not the batts, but
the fact that they can't maintain more than 2,7 volts for several months.

   I'll try buying VPack. As it comes with a battery voltage regulator,
it would keep the tension for more time with its boost feature....i think

   thanks for all the help up to now

Rafael




2007\07\17@041023 by Alan B. Pearce

face picon face
>Yes, i'm really sure. I have tested with an amperimeter. When XBEE
>wakes up the current goes to aprox 50mA, and after 1 second it goes
>as low as  less than 1mA.

That doesn't sound like it is going to sleep to me.

>I'm using the maxstream firmware with cyclic sleep, with nothing
>more than their own firmware.

I think this is your trouble. I suspect the on-board firmware only goes into
a standby mode, if it goes to sleep it probably requires something external
to wake it up. It may not even be able to go to sleep on its own, it may
need to be commanded to go to sleep.

No experience with these units, but this is where I would start looking.

2007\07\17@084749 by Gerhard Fiedler

picon face
Alan B. Pearce wrote:

>>I'm using the maxstream firmware with cyclic sleep, with nothing
>>more than their own firmware.
>
> I think this is your trouble. I suspect the on-board firmware only goes
> into a standby mode, if it goes to sleep it probably requires something
> external to wake it up. It may not even be able to go to sleep on its
> own, it may need to be commanded to go to sleep.

It has a cyclic sleep mode. The mode it enters is similar to the one that's
called "pin doze" -- it has a higher consumption than "pin hibernate", but
it's still not more than approx. 40uA (@3V).

Anyway, it seems that at least one problem you're having is with the
discharge curve of the AA batteries. You can try to use a Lithium battery
and see where you're getting (check out the discharge curve and compare it
with alkaline; it's flatter). Also, check whether your sleep cycle is what
you'd expect; both in terms of duration and in terms of consumption.

Consumption is a bit tricky to measure, because a resistor that gives you
enough voltage drop during the sleep phase probably has too much drop when
the module is awake. A simple solution would be to have a make-before-break
switchover between two resistors and do it by hand. (With the 4 minutes
between wakeups you said you have, you can easily manually switch to the
higher resistor once it is in sleep mode, measure the current, and then
switch back. Important is that the switchover happens without interruption
of the supply.)

Gerhard

2007\07\17@091210 by Alan B. Pearce

face picon face
>It has a cyclic sleep mode. The mode it enters is similar to the
>one that's called "pin doze" -- it has a higher consumption than
>"pin hibernate", but it's still not more than approx. 40uA (@3V).

OK, that sounds like it should be doing what is required then.

>Consumption is a bit tricky to measure, because a resistor that
>gives you enough voltage drop during the sleep phase probably
>has too much drop when the module is awake.

Hmm, for doing check measurements like this I like to use a device like the
Keithley 2400 Sourcemeter as a power supply, it can measure down to tiny
currents, while having a 1A capability if needed. If the current change is
large enough and the wires fine enough you can even use it with remote
sensing.

2007\07\17@132857 by alan smith

picon face
Also, there is a reset pin...better make sure that you don't need an external pullup on this..bad things happen when resets float ya know.  I don't have the data sheet handy, but im sure the spec is there.
 
 If indeed....it really is in a cyclic sleep mode, then I do agree...look at battery technologies that might be better suited.  Use a LiON battery with a boost to keep the voltage up at 3V even when the input is lower.

     
---------------------------------
Need a vacation? Get great deals to amazing places on Yahoo! Travel.

2007\07\17@171843 by Sean Breheny

face picon face
The reset does have a pull-up built in on these modules.

I agree that it is very likely that the supply undervoltage is the
problem. If battery size is not a real issue, I'd suggest several AAA
cells and a low-current step-down switcher (such as one which runs to
charge a cap and then shuts down until the cap droops below a certain
value - often called burst mode). I say this because if you use two
cells and a step up, you won't be able to run when the input is more
than 3V. (which, by the way, seems to be the optimum voltage for the
XBee modules, even though they are rated for 2.8 to 3.4V).

Sean


On 7/17/07, alan smith <RemoveMEmicro_eng2TakeThisOuTspamyahoo.com> wrote:
> Also, there is a reset pin...better make sure that you don't need an external pullup on this..bad things happen when resets float ya know.  I don't have the data sheet handy, but im sure the spec is there.
>
>   If indeed....it really is in a cyclic sleep mode, then I do agree...look at battery technologies that might be better suited.  Use a LiON battery with a boost to keep the voltage up at 3V even when the input is lower.
>
>
> ---------------------------------
> Need a vacation? Get great deals to amazing places on Yahoo! Travel.
> -

2007\07\17@180121 by Matt Pobursky

flavicon
face
On Tue, 17 Jul 2007 17:18:41 -0400, Sean Breheny wrote:
> I agree that it is very likely that the supply undervoltage is the
> problem. If battery size is not a real issue, I'd suggest several AAA
> cells and a low-current step-down switcher (such as one which runs to
> charge a cap and then shuts down until the cap droops below a certain
> value - often called burst mode). I say this because if you use two cells
> and a step up, you won't be able to run when the input is more than 3V.
> (which, by the way, seems to be the optimum voltage for the XBee modules,
> even though they are rated for 2.8 to 3.4V).

Two AA cells will still give you better energy density. Running a step-up
converter at greater than cell voltage isn't strictly a problem either.
Texas Instruments, On-Semi, Maxim and Linear Technology all make step up
converters that will work with input voltages greater than the output
voltage set point. They just function as a pass device until the output
voltage starts to drop. Some of the ones I've used from T.I. and On-Semi
also have an output disconnect switch that disconnects the input voltage
from the output on shutdown, always a potential problem with a step-up
converter.

This is one of those problems that there is really not a perfect answer
for. If it were me, I might consider trying to convince the powers that be
to use a single lithium ion cell and run a step-down converter. ;-)

Matt Pobursky
Maximum Performance Systems

2007\07\17@181156 by David VanHorn

picon face
Seems obvious that the module is designed to run from single cell lithium power.
4.2V -> 3.0V before the voltage begins freefall.  Also, you get better
life out of the rechargables if you don't let them go as low.

In the McGraw-Hill handbook, they show a 18650 cell curve on page
35.37 that seems just fine.. You'd need an efficient LDO, but you CAN
shut that off when not in use, so it's not always blowing away
microamps.

Use the pic to enable the power to the zigbee, then talk, then shut it
down again.
If you need to receive while "asleep", you can probably do that by
recieving only once in a while, and sleeping most of the time.


Non-rechargable Li cells are an even closer match, with Li/MnO2cells
being essentially flat at about 3.1V till they die.  14.59 shows pulse
discharges of 5.5mA into a 190mAH coin cell.

Sounds like they want to be fed from a 123A cell..

2007\07\18@140912 by Matt Pobursky

flavicon
face
On Tue, 17 Jul 2007 18:11:54 -0400, David VanHorn wrote:
> Seems obvious that the module is designed to run from single cell lithium
> power. 4.2V -> 3.0V before the voltage begins freefall.  Also, you get
> better life out of the rechargables if you don't let them go as low.
>
> In the McGraw-Hill handbook, they show a 18650 cell curve on page 35.37
> that seems just fine.. You'd need an efficient LDO, but you CAN shut that
> off when not in use, so it's not always blowing away microamps.
>
> Use the pic to enable the power to the zigbee, then talk, then shut it
> down again.
> If you need to receive while "asleep", you can probably do that by
> recieving only once in a while, and sleeping most of the time.
>
>
> Non-rechargable Li cells are an even closer match, with Li/MnO2cells
> being essentially flat at about 3.1V till they die.  14.59 shows pulse
> discharges of 5.5mA into a 190mAH coin cell.
>
> Sounds like they want to be fed from a 123A cell..

You and I think a lot alike! The same thoughts crossed my mind as I read
the OP's post.

Both of those batteries get a lot of play in my 3.0V/3.3V designs these
days. One other good cell to consider is the 14500 Li-Ion cell. It's a AA size
cell which is nice for packaging, 800 mAH. I used this cell in a portable
medical device recently and it's worked out quite well. Since it is AA
size, you need to make sure someone can't put a standard AA alkaline cell
in the unit or better yet make sure your design gracefully ignores it. It
is nice being able to use standard battery holders for it though.

Matt Pobursky
Maximum Performance Systems

More... (looser matching)
- Last day of these posts
- In 2007 , 2008 only
- Today
- New search...