Go Down

Topic: ZS-042 DS3231 RTC module (Read 59081 times) previous topic - next topic

sterretje

With my interpretation of the RTC spec: If powered down, the RTC draws 3 uA from the battery. The battery can deliver 210mAh; that would give 210mAh / 3uA = 70000 hours.

Found a post somewhere on the web: http://forum.allaboutcircuits.com/threads/rtc-battery-backup.80592/#post-572321; states 5 to 10 years and matches more or less my interpretation / calculation.
If you understand an example, use it.
If you don't understand an example, don't use it.

Electronics engineer by trade, software engineer by profession. Trying to get back into electronics after 15 years absence.

Koepi

#31
Sep 11, 2016, 11:59 am Last Edit: Sep 11, 2016, 12:06 pm by Koepi
Jeez, some have the info right, others claim dangerous BS.

The "charging circuit" is really evil. The battery is otherwise connected correctly.
See the schematics here:

Remove the 1N4148 and/or the 200R resistor and everything is fine.

I also removed the power LED as I use the RTC for a DIY clock. Works great this way with 3.3V supply.
(click for bigger image)

joshjayk

#32
Sep 12, 2016, 04:25 am Last Edit: Sep 12, 2016, 07:33 pm by joshjayk
Jeez, some have the info right, others claim dangerous BS.

The "charging circuit" is really evil. The battery is otherwise connected correctly.
See the schematics here:

Remove the 1N4148 and/or the 200R resistor and everything is fine.

I also removed the power LED as I use the RTC for a DIY clock. Works great this way with 3.3V supply.
(click for bigger image)
So... Dumb question... How do I remove the battery that I currently have in the module? I've got an LIR2032 in there and it's a pretty tight fit, not really much room for me to stick something underneath to pry it out. Also, the battery should go in with the negative terminal facing down and positive facing up, is that correct?

EDIT: So I figured out how to remove the LIR2032 and replaced it with a CR2032 I bought on amazon... The LED did not light up. I'm assuming that the LED should be lit when it is powered via the battery, and I probably ended up with a bad batch of batteries...?

floresta

Quote
I'm assuming that the LED should be lit when it is powered via the battery, and I probably ended up with a bad batch of batteries...?
Your assumption is incorrect!  As mentioned in a previous post the purpose of the battery is to keep the internal timekeeping functions running when the power is removed from the module. 

Don




sterretje

#34
Sep 14, 2016, 10:43 am Last Edit: Sep 14, 2016, 10:43 am by sterretje
EDIT: So I figured out how to remove the LIR2032 and replaced it with a CR2032 I bought on amazon... The LED did not light up. I'm assuming that the LED should be lit when it is powered via the battery, and I probably ended up with a bad batch of batteries...?
Ah, you want the battery to only last 210mAh / (e.g.) 10mA equals 21 hours :D
If you understand an example, use it.
If you don't understand an example, don't use it.

Electronics engineer by trade, software engineer by profession. Trying to get back into electronics after 15 years absence.

joshjayk

Your assumption is incorrect!  As mentioned in a previous post the purpose of the battery is to keep the internal timekeeping functions running when the power is removed from the module. 

Don




Yeah I realized it was still working without the LED lit!

Ah, you want the battery to only last 210mAh / (e.g.) 10mA equals 21 hours :D
What do you mean? You mean the batteries I got from Amazon are bad?

sterretje

No, if the battery would power the LED as well, it will have a very short life.
If you understand an example, use it.
If you don't understand an example, don't use it.

Electronics engineer by trade, software engineer by profession. Trying to get back into electronics after 15 years absence.

robertew

Many interesting comments here on the ZS-042 RTC module.  For the most part, this is a great module and very reasonably priced.  I was asked to address an issue with this RTC for a high school science project.
These issues are common with most ZS-042 modules.

Battery swells up after a few days of use.  This is true for the LIR2032 and the CR2032 when the module
is supplied with 5V.  If supplied with 3.3V, the LIR2032 will not receive a charge due to low voltage and the CR2032 still receives over voltage to its terminals.  This excess voltage is coming from the DS3231 IC.

As stated many times in previous posts, the charging circuit on this module is very poorly designed and should
be disabled by removing either the diode or resistor in the charging path.  The LIR2032 should be replaced with a CR2032.  Swelled CR2032's must be replaced before case failure occurs.

After this, one problem still remains.  If you look at the voltage on the CR2032, when connected to 5V or 3.3V
supply, it will have about 3.3 volts at the terminals.  This is not good for the CR2032 and should be addressed.
If you supply 3.3V to the RTC module through a simple diode, the voltage will be reduced to about 2.75V.  This
is within spec for the DS3231 and will prevent excess voltage going to the CR2032 battery.  Not all modules have this problem but most do.

These changes have been made to dozens of projects without issue.  The school continues to purchase this module for their labs and makes the simple changes before distribution to the students.

borland

#38
Oct 22, 2016, 02:02 am Last Edit: Oct 23, 2016, 02:39 am by borland
These design flawed ZS-042 modules are now selling on eBay from sellers in China for as little as $0.75USD, including shipping charges to USA.

All that's needed to fix these is just to remove the 200 ohm surface mounted resistor marked '201'.

I did confirm what robertew was saying about excessive voltage on the coin cell, but couldn't confirm what he said about higher voltages on the coin cell after disabling the charging circuit.  I measured cell voltages both before and after removing the charging resistor.

Before:
Vcc         CR2032
0.0v        3.18v                     -                              -
3.3v        3.22v
5.0v        3.34v

After:
Vcc         CR2032
0.0v        3.18v
3.3v        3.18v
5.0v        3.18v

I also tested the modules I2C serial on both 5.0v and 3.3v Arduinos.  Both Arduinos tested good with modules Vcc pin powered by either 5.0v or 3.3v.  >>> Warning: 3.3v Arduino board processor input pins may be damaged by connecting them to 5v powered sensors.  So, supply the DS3231's module's Vcc with a supply voltage consistent with the Arduino processor's supply voltage. <<<

grjonjon

#39
Oct 30, 2016, 08:02 am Last Edit: Oct 30, 2016, 08:17 am by grjonjon
Hi. I read the previous posts on the ZS-042 RTC board. New here but experienced with electronics.

First off, if you have a CR2032 (non-rechargeable!), by all means, don't put it to work with the board as is. REMOVE the diode or the "201" resistor first, or cut the copper trace between those two. This will ensure the battery won't try to charge and possibly explode.

If you have the LIR2032 (rechargeable) or want to use one, read on...

My experience:
My board came with a LIR2032 rechargeable battery and I've been using it for several months now without any problems. No swelling or such. Also, the battery works great in supporting the DS3231 IC's timekeeping whenever external supply is removed.

My suggestion:
If you have the LIR2032 rechargeable battery on, just leave the ZS-042 board alone. The already installed 1N4148 diode drops the 5V voltage to 4.3V and that is around the maximum voltage your rechargeable battery will ever get (and can take) while charging, plus the 200 ohm series resistor makes sure the charging current is very low.

[ I=(4.3-Vb)/200  (Amps), where Vb is the voltage between the battery poles; will go up to 4.2 V as batt becomes fully charged.]

P.S. If you are the kind of people that want to be extra safe, and feel comfortable and experienced with electronics, you can try connecting a low forward voltage diode (schottky would do) in series with the already installed diode. That will drop the charging voltage around 0.2 V further down. The battery will now end-charge to a less-than-full condition, but will still be fine for keeping the RTC ticking during mains power failures.

floresta

Quote
My experience:
My board came with a LIR2032 rechargeable battery and I've been using it for several months now without any problems. No swelling or such. Also, the battery works great in supporting the DS3231 IC's timekeeping whenever external supply is removed.
My experience is exactly the same.

Don

ceptimus

#41
Dec 06, 2016, 12:42 pm Last Edit: Dec 06, 2016, 12:48 pm by ceptimus
The on-board diode should drop about 0.7V so the battery is charged up to 4.3V when the board is powered by 5V.  This is a little high and could be dangerous when your 5V supply is actually slightly above the nominal value. 

One solution is to power the module's VCC pin from your 5V through a diode.  With that external diode plus the one already on the module it drops the charging voltage down to 3.6V.  This is not enough to fully charge the battery, but it will charge it enough for normal use.

All this only applies to the LIR2032 rechargeable battery.  If your module has CR2032 that is a non-rechargeable battery and recharging it at any voltage is dangerous.  If you have a CR2032 then it's better to remove the '201' resistor on the board, or cut the track that leads to it.

klennex71

>> Manufacture fault!!....
Just noticed my ZS-042 RTC wouldnt respond to I2C .. after usual debug stuff, I started closely inspecting PCB..
Found lifted SDA PIN!!!


pinguy85

I cut the trace and removed the components in Koepi's post, but I'm still seeing about 1v across the battery terminal. Could this just be my meter? Is it safe to use the CR2032?

jasomers

After reading the entire thread and constantly flip-flopping on the correct path of action to take, I have realised that my ZS-042 module was shipped without the 201 resistor or diode.

This would make me think that the manufacturer has acknowledged the fault and rectified it. So anyone else with an older module might consider this being the correct fix.

Go Up