Lithium battery voltage tolerance

G

Gary Lynch

Guest
I have been troubleshooting a circuit involving a Maxim
DS1306 real-time clock (RTC) with battery back-up. I
connected Vcc of the RTC to the same 3.3 V supply as
required by the micro that talks to it. The battery is a
Panasonic BR2330 (allegedly a 3 V part):
-
http://www.panasonic.com/industrial/battery/oem/images/pdf/Panasonic_Lithium_B
R2325_BR2330.pdf

The RTCs appeared to be failing in droves, in that I could
program them once, but after a single power cycle they
refused to respond to the micro any more. Replacing the
chip always worked, but only for another power cycle.

After barking up a lot of wrong trees, I found postings on
other forums indicating that the DS1306 switches from normal
mode to battery-backup when Vcc drops below Vbat, but it
doesn't switch back to normal until Vcc rises above
VBat + 0.2 V.

The Maxim data sheet:
- http://datasheets.maxim-ic.com/en/ds/DS1306.pdf

says that a device in battery mode cannot be written (page
9, last sentence of paragraph under "POWER CONTROL"). On-line
user comments indicate that it cannot be read in this mode
either. At this point I realized that if the second
statement is true, a DS1306 in battery mode cannot be
distinguished from a dead chip.

I measured the battery voltage and found the 3 V battery
supplied 3.22 V, and my 3.3 V supply is unlikely to rise
0.2 V above this--ever. I pulled 7 more boards off the shelf
to compare battery voltages:

#1: 2.9995 V #5: 2.9887 V
#2: 3.1975 V #6: 2.9919 V
#3: 3.0579 V #7: 2.9919 V
#4: 2.9894 V

The expected load on the battery of the DS1306 is 550 nA @
3 V (pg 16).

If you look over the BR2330 data sheet you find no
indication it can rise above 3 V, unless you hike up the
ambient temp well above 20 °C (which I don't).

I was wondering what range of voltages other members of this
forum have observed from this battery.
============================================================
Gary Lynch | To send mail, change my
| domain name from "no$pam"
gary.lynch@no$pam.org | to "ieee".
============================================================
 
I measured the battery voltage and found the 3 V battery
supplied 3.22 V, and my 3.3 V supply is unlikely to rise
0.2 V above this--ever. I pulled 7 more boards off the shelf
to compare battery voltages:

#1: 2.9995 V #5: 2.9887 V
#2: 3.1975 V #6: 2.9919 V
#3: 3.0579 V #7: 2.9919 V
#4: 2.9894 V
Click to expand...
Are you charging the battery? The Maxim data sheet said something
about charge but I didn't read it carefully.

Batteries with a (really) full charge often go significantly above
their nominal voltage. (I'm not a wizard on Lithium chemistry.)

What's the voltage on the battery on a working setup? Does
it change if you let it sit there for a while?

What happens if you let the unit with 3.2V sit around
for a while? Is that the voltage when the board is
powered off?

--
These are my opinions, not necessarily my employer's. I hate spam.
 
Gary Lynch <bookworm@execpc.com> wrote in
news:20090925-022542.300.0@Your-Name-Here.News:

I have been troubleshooting a circuit involving a Maxim
DS1306 real-time clock (RTC) with battery back-up. I
connected Vcc of the RTC to the same 3.3 V supply as
required by the micro that talks to it. The battery is a
Panasonic BR2330 (allegedly a 3 V part):
-
http://www.panasonic.com/industrial/battery/oem/images/pdf/Panasonic_Lith
ium_B R2325_BR2330.pdf

The RTCs appeared to be failing in droves, in that I could
program them once, but after a single power cycle they
refused to respond to the micro any more. Replacing the
chip always worked, but only for another power cycle.

After barking up a lot of wrong trees, I found postings on
other forums indicating that the DS1306 switches from normal
mode to battery-backup when Vcc drops below Vbat, but it
doesn't switch back to normal until Vcc rises above
VBat + 0.2 V.

The Maxim data sheet:
- http://datasheets.maxim-ic.com/en/ds/DS1306.pdf

says that a device in battery mode cannot be written (page
9, last sentence of paragraph under "POWER CONTROL"). On-line
user comments indicate that it cannot be read in this mode
either. At this point I realized that if the second
statement is true, a DS1306 in battery mode cannot be
distinguished from a dead chip.

I measured the battery voltage and found the 3 V battery
supplied 3.22 V, and my 3.3 V supply is unlikely to rise
0.2 V above this--ever. I pulled 7 more boards off the shelf
to compare battery voltages:

#1: 2.9995 V #5: 2.9887 V
#2: 3.1975 V #6: 2.9919 V
#3: 3.0579 V #7: 2.9919 V
#4: 2.9894 V

The expected load on the battery of the DS1306 is 550 nA @
3 V (pg 16).

If you look over the BR2330 data sheet you find no
indication it can rise above 3 V, unless you hike up the
ambient temp well above 20 °C (which I don't).

I was wondering what range of voltages other members of this
forum have observed from this battery.
============================================================
Gary Lynch | To send mail, change my
| domain name from "no$pam"
gary.lynch@no$pam.org | to "ieee".
============================================================
Click to expand...
Did you manage to confirm later that the RTCs were not dead, only in a state
that couldn't be written to or read from?

Either way, something I've seen used with backup batteries might work for
you, a series pair of 1n4148 diode, and a 1K resistance. The resistor would
prevent any chance the battery can fry the chip in the absence of main power
(I know, pretty stupid if that occured given the purpose of the IC but in
some devices that can pull a pin low the chance exists of damage that way,
and the 1K will allow enough current for the backup purpose). The diode will
drop the voltage for you, likely enough to get the headroom you need for 3.3V
to rise beyond the 0.2V you need. So long as the backup battery is only held
out of circuit, and does not need a charge current, this should work ok.
 
Lostgallifreyan <no-one@nowhere.net> wrote in
news:Xns9C9138F22E408zoodlewurdle@216.196.109.145:

Did you manage to confirm later that the RTCs were not dead, only in a
state that couldn't be written to or read from?

Either way, something I've seen used with backup batteries might work
for you, a series pair of 1n4148 diode, and a 1K resistance. The
resistor would prevent any chance the battery can fry the chip in the
absence of main power (I know, pretty stupid if that occured given the
purpose of the IC but in some devices that can pull a pin low the chance
exists of damage that way, and the 1K will allow enough current for the
backup purpose). The diode will drop the voltage for you, likely enough
to get the headroom you need for 3.3V to rise beyond the 0.2V you need.
So long as the backup battery is only held out of circuit, and does not
need a charge current, this should work ok.
Click to expand...
A thought on that diode... maybe use Schottky. I was thinking earlier of
something I'd seen on a 3.6V circuit. In this case it's a battery (3.2V coin
cell?) that is often considered dead once it falls below 2.6V so a smaller
voltage drop is better.
 
In article <Xns9C913A9FD8FADzoodlewurdle@216.196.109.145>, no-
one@nowhere.net says...
Lostgallifreyan <no-one@nowhere.net> wrote in
news:Xns9C9138F22E408zoodlewurdle@216.196.109.145:

Did you manage to confirm later that the RTCs were not dead, only in a
state that couldn't be written to or read from?

Either way, something I've seen used with backup batteries might work
for you, a series pair of 1n4148 diode, and a 1K resistance. The
resistor would prevent any chance the battery can fry the chip in the
absence of main power (I know, pretty stupid if that occured given the
purpose of the IC but in some devices that can pull a pin low the chance
exists of damage that way, and the 1K will allow enough current for the
backup purpose). The diode will drop the voltage for you, likely enough
to get the headroom you need for 3.3V to rise beyond the 0.2V you need.
So long as the backup battery is only held out of circuit, and does not
need a charge current, this should work ok.


A thought on that diode... maybe use Schottky. I was thinking earlier of
something I'd seen on a 3.6V circuit. In this case it's a battery (3.2V coin
cell?) that is often considered dead once it falls below 2.6V so a smaller
voltage drop is better.

I was thinking the same thing.
Click to expand...
Test, the "dead" units and see if they come back to life when
reinserted, and put a diode in line with the battery to drop the V a
little bit. I use the DS1307 and the lithium coin cell we use drops to
~3V after about a day of use and then hangs there for the rest of the
life of the cell. A Schottky diode in line with the cell might just do
the trick.
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top