WARNING: RTC Board Battery Charging Issue

[MrAl]

Hello there,

Recently i found a problem with the charging circuit on breakout boards that have the DS3231 Real Time Clock chip and also the DS1307 chip, but there may be others.

The problem is that the charging circuit over charges the LIR2032 battery.  This could cause fire, and at best the battery will be ruined and wont function as a backup anymore.

One solution is to remove the 200 ohm surface mount resistor, which stops all charging.  The battery does not need to be recharged in a normal application anyway, as long as it is charged one time only.

Some boards do not have the charging circuit so they are not affected.

Pass the word around.

[SL4P]

This is a really good application for a small Supercap.
Diode to charge and protect from discharge... always in circuit, ready to hold the RTC up for modest outages.

[ralphd]

Is there no diode in series?  The ones I've seen appear to have a 1n4148 in series which will drop the 5v supply down to 4.3V.  The 200 Ohm resistor will limit the current to <1mA at full charge.

[MrAl]

This is a really good application for a small Supercap.
Diode to charge and protect from discharge... always in circuit, ready to hold the RTC up for modest outages.

Hello there,

I am not sure i can agree with that, because the super cap has no voltage limiting.  With a +5v supply the cap could charge up very close to 5v, which is higher than 4.2v, and thus the chip might not accept that.  You could check though to see if it might work anyway.  Maybe check the data sheet, then try it in real life.
Probably not a bad idea though.

[MrAl]

Is there no diode in series?  The ones I've seen appear to have a 1n4148 in series which will drop the 5v supply down to 4.3V.  The 200 Ohm resistor will limit the current to <1mA at full charge.

Hello there,

The first rule with diodes and voltage references is that you can never depend on a diode to drop a certain voltage.  That's because it's voltage can vary for one or more of several different reasons.

The first is current.  The diode drop can be anything from around 0.4v for a very light current up to around 0.8v for a heavier current.  Applying this to the 5v circuit, that means the battery could get up to 4.6v and that is much too high for a standard Li-ion battery and even a high capacity type.  That's a bit extreme, as we would probably see at least 0.5v or even 0.6v in real life where there is more than a tiny tiny current flowing.  So if the leakage of the battery is a little higher we could see 0.6v.  But even with that we see the battery voltage go up to 4.4v, which is still far to high.  The limit is 4.225 max.  And then what happens if the 5v supply is a little higher, like 5.2 volts?  Many voltage regulators have a tolerance of at least 5 percent even at room temperature, so with 5.2v we could see 4.6v which is far, far too high.

Then we have the temperature coefficient.  If the diode gets warm the voltage drop goes down.  If it goes down by only 0.1v we have 0.1v more on the battery.  That could make the difference between charging well and destroying the battery.

So you can begin to see why this type of charging circuit is no good.  At best it could easily ruin the battery when operated from a regular 5v power supply.  Luckily, when run with a 3.3v power source there will be little or no charging so with that power supply it's ok.

The 200 ohm resistor is in series with the diode, and that limits the current yes, but not the voltage.  My measured results from the past few days show the battery voltage rising first to 4.29 volts and then today up to 4.37 volts.  That's much too high for the Li-ion battery and will shorten it's life or maybe worse.

[SL4P]

This is a really good application for a small Supercap.
Diode to charge and protect from discharge... always in circuit, ready to hold the RTC up for modest outages.

Hello there,
I am not sure i can agree with that, because the super cap has no voltage limiting.  With a +5v supply the cap could charge up very close to 5v, which is higher than 4.2v, and thus the chip might not accept that.  You could check though to see if it might work anyway.  Maybe check the data sheet, then try it in real life.
Probably not a bad idea though.

I've been using this exact setup on DS3231 in use for around four years.... no problem so far.

[ralphd]

4.37 volts causing a lithium ion battery to catch fire?  Drama alert!
To avoid reducing the life of the battery, put a ~1K resistor in parallel.  That will keep the diode Vf higher up the I-V curve so the battery doesn't get more than 4.2V.

[amyk]

Yes it may cause a vent and (small) fire, depending on how long it's kept overcharged. In any case it's not good for the battery at all.

[MrAl]

This is a really good application for a small Supercap.
Diode to charge and protect from discharge... always in circuit, ready to hold the RTC up for modest outages.

Hello there,
I am not sure i can agree with that, because the super cap has no voltage limiting.  With a +5v supply the cap could charge up very close to 5v, which is higher than 4.2v, and thus the chip might not accept that.  You could check though to see if it might work anyway.  Maybe check the data sheet, then try it in real life.
Probably not a bad idea though.

I've been using this exact setup on DS3231 in use for around four years.... no problem so far.

Hello there,

That's good to hear but you might have an exceptional case there.  Perhaps you can measure your battery voltage with a digital meter and report that value back here.  We can then work from there.
There are many possible variations here, but we want to make sure we cover all the bases so we never see the worst one.

[MrAl]

4.37 volts causing a lithium ion battery to catch fire?  Drama alert!
To avoid reducing the life of the battery, put a ~1K resistor in parallel.  That will keep the diode Vf higher up the I-V curve so the battery doesn't get more than 4.2V.

Hi,

Sorry to say, you can not do that.  Placing a 1k resistor in parallel to the battery means it discharges in 10 hours when the power goes out.  Normally it will last for years like that, but with a 1k resistor it shortens the backup time from maybe 5 years down to 10 to 20 hours.  Not good.
Drama?  Perhaps by you stating that in this place that is the real drama here.  Think good engineering not good sarcasm.

[ralphd]

Yes it may cause a vent and (small) fire, depending on how long it's kept overcharged. In any case it's not good for the battery at all.

And you may get hit by a meteor if you go outside, so stay in an underground bunker to keep safe...

[ralphd]

4.37 volts causing a lithium ion battery to catch fire?  Drama alert!
To avoid reducing the life of the battery, put a ~1K resistor in parallel.  That will keep the diode Vf higher up the I-V curve so the battery doesn't get more than 4.2V.

Hi,

Sorry to say, you can not do that.  Placing a 1k resistor in parallel to the battery means it discharges in 10 hours when the power goes out.  Normally it will last for years like that, but with a 1k resistor it shortens the backup time from maybe 5 years down to 10 to 20 hours.  Not good.
Drama?  Perhaps by you stating that in this place that is the real drama here.  Think good engineering not good sarcasm.

No sarcasm.  Warning about LIR2032s catching fire by charging to 3.37V IS drama.  Good engineers scientist don't go chicken little.

[SL4P]

Hello there,

That's good to hear but you might have an exceptional case there.  Perhaps you can measure your battery voltage with a digital meter and report that value back here.  We can then work from there.
There are many possible variations here, but we want to make sure we cover all the bases so we never see the worst one.

5V supply, single blocking diode... 5V Supercap - not measured, I'd guess around 4.3v on the cap... feeding the battery backup pin of the DS323x

[ralphd]

The parallel resistor was a bad idea.  Having given it a bit of thought now, a tl431 with 680K and 1M resistor divider would give a 4.2V shunt.  The 1n4148 could be left out.
Total leakage would be <5uA, at a cost of 5c in parts.

[MrAl]

4.37 volts causing a lithium ion battery to catch fire?  Drama alert!
To avoid reducing the life of the battery, put a ~1K resistor in parallel.  That will keep the diode Vf higher up the I-V curve so the battery doesn't get more than 4.2V.

Hi,

Sorry to say, you can not do that.  Placing a 1k resistor in parallel to the battery means it discharges in 10 hours when the power goes out.  Normally it will last for years like that, but with a 1k resistor it shortens the backup time from maybe 5 years down to 10 to 20 hours.  Not good.
Drama?  Perhaps by you stating that in this place that is the real drama here.  Think good engineering not good sarcasm.

No sarcasm.  Warning about LIR2032s catching fire by charging to 3.37V IS drama.  Good engineers scientist don't go chicken little.

Hello again,

And where is that good engineer who says that?
In other words, show me the proof.  If you dont have proof then its' still just an Li-ion battery and there is plenty of data on these kinds of batteries.

Show me the proof that shows factually that it can never get hot and start a fire, or even get hot and melt something.  There's less energy storage luckily, but would you trust it with your kids for example?

It is very good that you rethought your original modification to the charging circuit.  Maybe after your fire you can rethink this problem too, that's if you are still around to think about it.

[ralphd]

4.37 volts causing a lithium ion battery to catch fire?  Drama alert!
To avoid reducing the life of the battery, put a ~1K resistor in parallel.  That will keep the diode Vf higher up the I-V curve so the battery doesn't get more than 4.2V.

Hi,

Sorry to say, you can not do that.  Placing a 1k resistor in parallel to the battery means it discharges in 10 hours when the power goes out.  Normally it will last for years like that, but with a 1k resistor it shortens the backup time from maybe 5 years down to 10 to 20 hours.  Not good.
Drama?  Perhaps by you stating that in this place that is the real drama here.  Think good engineering not good sarcasm.

No sarcasm.  Warning about LIR2032s catching fire by charging to 3.37V IS drama.  Good engineers scientist don't go chicken little.

Hello again,

And where is that good engineer who says that?
In other words, show me the proof.  If you dont have proof then its' still just an Li-ion battery and there is plenty of data on these kinds of batteries.

Show me the proof that shows factually that it can never get hot and start a fire, or even get hot and melt something.  There's less energy storage luckily, but would you trust it with your kids for example?

It is very good that you rethought your original modification to the charging circuit.  Maybe after your fire you can rethink this problem too, that's if you are still around to think about it.

You say it can catch fire so the onus is on you to prove it.
You are trying to play the old reverse onus fallacy.
You should watch some of Dave's videos to realize he doesn't tolerate dramatic bullshit, and neither do I.
If you can't control your limbic system then this is the wrong forum for you.  Try the Adruino forums where they coddle the emotional fools.

[MrAl]

4.37 volts causing a lithium ion battery to catch fire?  Drama alert!
To avoid reducing the life of the battery, put a ~1K resistor in parallel.  That will keep the diode Vf higher up the I-V curve so the battery doesn't get more than 4.2V.

Hi,

Sorry to say, you can not do that.  Placing a 1k resistor in parallel to the battery means it discharges in 10 hours when the power goes out.  Normally it will last for years like that, but with a 1k resistor it shortens the backup time from maybe 5 years down to 10 to 20 hours.  Not good.
Drama?  Perhaps by you stating that in this place that is the real drama here.  Think good engineering not good sarcasm.

No sarcasm.  Warning about LIR2032s catching fire by charging to 3.37V IS drama.  Good engineers scientist don't go chicken little.

Hello again,

And where is that good engineer who says that?
In other words, show me the proof.  If you dont have proof then its' still just an Li-ion battery and there is plenty of data on these kinds of batteries.

Show me the proof that shows factually that it can never get hot and start a fire, or even get hot and melt something.  There's less energy storage luckily, but would you trust it with your kids for example?

It is very good that you rethought your original modification to the charging circuit.  Maybe after your fire you can rethink this problem too, that's if you are still around to think about it.

You say it can catch fire so the onus is on you to prove it.
You are trying to play the old reverse onus fallacy.
You should watch some of Dave's videos to realize he doesn't tolerate dramatic bullshit, and neither do I.
If you can't control your limbic system then this is the wrong forum for you.  Try the Adruino forums where they coddle the emotional fools.

You seem to be very judgmental here. The question is not just about if it will catch fire, it is about if the battery will be ruined and therefore void the function of the battery in the first place.  People have already reported early failures and battery bloating, so i assume that is not something anyone wants.
But it's up to you how you want to run your charging system on this or any other board.  Do what you wish, but dont ruin it for other people who want to have their battery last a long time which is the very purpose of it.
Check out other sites too, and find ONE site that claims that the battery charge system is GOOD.

BTW, that brings us to the question of what YOU would do.  Are you saying that you would leave the circuit alone as is after reading about the problems?

[Ian.M]

Film at 11: Cheap and nasty breakout boards have cheap and nasty Li-ion charging circuits.

The simplest solution that retains the LIR2032 is to drop the 5V supply to the board to 4.5V.  Either use a LDO, or if power consumption isn't an issue, a silicon diode and a resistor to ground.

Alternatively, if the battery's got a holder, cut the charging track or desolder the resistor and fit a CR2032, saving the LIR2032 for something else.

TANSTAAFL.

[Ice-Tea]

You say it can catch fire so the onus is on you to prove it.
You are trying to play the old reverse onus fallacy.
You should watch some of Dave's videos to realize he doesn't tolerate dramatic bullshit, and neither do I.
If you can't control your limbic system then this is the wrong forum for you.  Try the Adruino forums where they coddle the emotional fools.

Going to disagree with you there. He takes the specification of the datasheet (one I found states 4.1V open circuit after charging, another stops charging well below 4.3V) and assumes that 4.37V is bad. Untill proven otherwise, I'd agree with him. Simply caling him chicken shit is not exactly a worthy argument.

[ralphd]

You say it can catch fire so the onus is on you to prove it.
You are trying to play the old reverse onus fallacy.
You should watch some of Dave's videos to realize he doesn't tolerate dramatic bullshit, and neither do I.
If you can't control your limbic system then this is the wrong forum for you.  Try the Adruino forums where they coddle the emotional fools.

Going to disagree with you there. He takes the specification of the datasheet (one I found states 4.1V open circuit after charging, another stops charging well below 4.3V) and assumes that 4.37V is bad. Untill proven otherwise, I'd agree with him. Simply caling him chicken shit is not exactly a worthy argument.

Another logical fallacy; this time straw man.
I never said the circuit was good.
I said until the warning of catching fire is proven it is emotional drama that doesn't belong in these forums.

[ralphd]

You seem to be very judgmental here. The question is not just about if it will catch fire, it is about if the battery will be ruined and therefore void the function of the battery in the first place.

Yes, I judge you to be an emotional fool who, when confronted engages in logical fallacies instead of scientific discussion.  The fact that you have shown you are capable of logical scientific thought (by way of your understanding of V=IR) makes it even worse that you let your limbic system overtake your behavioral control.

And the question is just about whether it will catch fire.  I never challenged the assertion that sustained charging at 3.37V is likely to reduce the life of the battery.
That's another common high-school level debating trick; advance something that is true along with something that is false, and less intelligent humans will believe both due to correlation bias even without causal connection.

[free_electron]

why use rechargeable batteries at all ? a simple coin cell will run tens of years .
or use a supercap. : note : use very low leakage diode to charge. even better : use collector base junction of a npn. these have an order of magnitude ( sometimes two) less leakage than a regular diode. emittor left floating.

[Simon]

4.37 volts causing a lithium ion battery to catch fire?  Drama alert!
To avoid reducing the life of the battery, put a ~1K resistor in parallel.  That will keep the diode Vf higher up the I-V curve so the battery doesn't get more than 4.2V.

I thought 4.2V was the limit. My boss showed me a charger for a light he has and it looked like a regular wall adapter that you would expect to output 5V but instead it was 4.2V, I'm guessing his battery had no charge control either and the system relied on the power "brick" (pebble) not being able to source enough current to do any harm and the battery voltage not exceeding 4.2V, he wanted to cut the cable off and solder a USB plug to it but I wasn't convinced as I'm no expert in these things.

[Simon]

1) You say it can catch fire so the onus is on you to prove it.
You are trying to play the old reverse onus fallacy.
2) You should watch some of Dave's videos to realize he doesn't tolerate dramatic bullshit, and neither do I.
3) If you can't control your limbic system then this is the wrong forum for you.  Try the Adruino forums where they coddle the emotional fools.

1) Have you even asked what chemistry this is and looked into it ? NO!

2) Oh how fearful everyone should be of a mere member, you speak for Dave now ? You represent him do you ?

3) If you don't mind Dave and I decided who stay's, who goes and what can't be talked about. Fancy seeing how fast you can go ? You have probably not spent much time on the arduino forums where electronic theory and many of the practicalities of it go over many peoples heads as they think "once a programmer that can flash an LED an electronics engineer too".

[Ice-Tea]

I said until the warning of catching fire is proven it is emotional drama that doesn't belong in these forums.

Having regard for maximum ratings is not drama.

[ralphd]

1) You say it can catch fire so the onus is on you to prove it.
You are trying to play the old reverse onus fallacy.
2) You should watch some of Dave's videos to realize he doesn't tolerate dramatic bullshit, and neither do I.
3) If you can't control your limbic system then this is the wrong forum for you.  Try the Adruino forums where they coddle the emotional fools.

1) Have you even asked what chemistry this is and looked into it ? NO!

2) Oh how fearful everyone should be of a mere member, you speak for Dave now ? You represent him do you ?

3) If you don't mind Dave and I decided who stay's, who goes and what can't be talked about. Fancy seeing how fast you can go ? You have probably not spent much time on the arduino forums where electronic theory and many of the practicalities of it go over many peoples heads as they think "once a programmer that can flash an LED an electronics engineer too".

Now someone on an emotional power trip.  Is there too much estrogen around here?
Go ahead and ban me; no loss to me and it will prove you are acting emotionally.

[JimRemington]

Here is some drama from the Arduino forum on that particular topic.

I wish I saw this thread earlier! I've been running this module with the included CR2032 battery at 5V ever since I got it with no issue until today. Today I heard a loud pop noise and looked up to see that the battery exploded and shot across the room. Pretty surprising considering how long it's been operating just fine.

Looks like I'll be getting an LIR2032 or cutting that trace!

[ralphd]

I said until the warning of catching fire is proven it is emotional drama that doesn't belong in these forums.

Having regard for maximum ratings is not drama.

Another straw man... I agreed 4.37V is too high.
Its the warning about catching fire at that voltage that is drama.

[Ian.M]

For anyone who doesn't believe its excess drama, please post a link to a picture or video of a LIR2032 failure that involved an explosion, fire or significant smoke release, that wasn't provoked by a blowtorch, or car battery charger or similar.

[free_electron]

I said until the warning of catching fire is proven it is emotional drama that doesn't belong in these forums.

Having regard for maximum ratings is not drama.

Another straw man... I agreed 4.37V is too high.
Its the warning about catching fire at that voltage that is drama.

why would that be drama ? you are charging the battery beyond it's operating limits. Unless you are a battery chemist you have no way of knowing what is happening electrochemically in that cell. for all you know a dendrite may grow , short the electrodes , conduct a large current , glow and set the thing on fire. Then again , nothing may happen at all. You don't now, i don't know , op doesn't know. So you stay within safe boundaries. 4.2 volt and not a microvolt more.

Op gave a warning about these circuits floating aroudn that do not have a proper cut-off voltage. fair warning.

[ralphd]

For anyone who doesn't believe its excess drama, please post a link to a picture or video of a LIR2032 failure that involved an explosion, fire or significant smoke release, that wasn't provoked by a blowtorch, or car battery charger or similar.

OK, so at least one other person doesn't need their estrogen levels checked... A bit of sanity prevails.

[Ian.M]

I wouldn't be at all surprised if the LIR2032 cell bulged and vented in a controlled fashion by the seal meting and displacing.   Annoying, board damaging due to contamination, but not Youtube worthy.  However a CR2032's chemistry has no allowance whatsoever for charging so explosive venting of uncontrolled gassing is a real possibility.

[Simon]

1) You say it can catch fire so the onus is on you to prove it.
You are trying to play the old reverse onus fallacy.
2) You should watch some of Dave's videos to realize he doesn't tolerate dramatic bullshit, and neither do I.
3) If you can't control your limbic system then this is the wrong forum for you.  Try the Adruino forums where they coddle the emotional fools.

1) Have you even asked what chemistry this is and looked into it ? NO!

2) Oh how fearful everyone should be of a mere member, you speak for Dave now ? You represent him do you ?

3) If you don't mind Dave and I decided who stay's, who goes and what can't be talked about. Fancy seeing how fast you can go ? You have probably not spent much time on the arduino forums where electronic theory and many of the practicalities of it go over many peoples heads as they think "once a programmer that can flash an LED an electronics engineer too".

Now someone on an emotional power trip.  Is there too much estrogen around here?
Go ahead and ban me; no loss to me and it will prove you are acting emotionally.

Your wish has been granted, if you think blackmailing me into not banning you will work think again. You stired half of the drama up in this thread instead of having a resonable response that could even go as far as not responding to it at all. Happy cooling off for 30 days.