In case you missed it....

[mightyfacundo]

This was posted in the comments section of Jerry's most recent "Android and Chill" article. I don't know enough about enough to know if this is accurate and possibly the cause for the Note7's battery issues. I'm hoping one or more of you knows. Thanks.

Joe J1

Interesting thread on reddit :

–]elephantbutt69 673 points 3 days ago*

I inspected the kernel source code. Samsung charges the note 7 battery at just over 4.35v

Older models have 4.35v batteries but only charge to 4.3v.

If you put a generic 4.2v lithium ion battery in, they survive 4.3v with a reduction in charge cycles but anything above that causes formation of metallic lithium dendrites, a non reversible process that eventually causes a fire. This policy makes sense if you have removable batteries as there are lots of fake genuine samsung batteries. But you want maximum battery life using special 4.35v cells. Perhaps they thought with a sealed battery they could hit 4.35v?

The chemistry is not hugely different, just uses more expensive elements. So a factory trying to minimise cost could easily introduce a defect from changing from one type to another etc.

So this tiny change could cause a defective battery to catch fire where it normally would not.

Doesnt matter how you charge it, more how long it sits above 4.30v at ~60-100% float charge.

So samsungs 60% charge resolves 2 issues, peak voltage above 4.3 and energy release on failure.

Imho they need to go back to 4.3v policy rather than just replace and assume its a once off.Lithium batteries used in phones dont catch fire without physical damage like a knife or overvoltage. That's why this is so unusual.

Ps kernel source is available from samsungs website because android is based on linux.

Tldr: policy change from 4.3v float charge to 4.35vwhich I have confirmed by examining source code available on samsungs website would explain the fires

 

[Milt K]

Can they fix the issue by knocking down the float charge back to 4.3v via a kernel update? Or is the only way to fix this is to gimp the battery max capacity to 60% that they have already done?

 

[Dm1120]

This was posted in the comments section of Jerry's most recent "Android and Chill" article. I don't know enough about enough to know if this is accurate and possibly the cause for the Note7's battery issues. I'm hoping one or more of you knows. Thanks.

Joe J1

Interesting thread on reddit :

–]elephantbutt69 673 points 3 days ago*

I inspected the kernel source code. Samsung charges the note 7 battery at just over 4.35v

Older models have 4.35v batteries but only charge to 4.3v.

If you put a generic 4.2v lithium ion battery in, they survive 4.3v with a reduction in charge cycles but anything above that causes formation of metallic lithium dendrites, a non reversible process that eventually causes a fire. This policy makes sense if you have removable batteries as there are lots of fake genuine samsung batteries. But you want maximum battery life using special 4.35v cells. Perhaps they thought with a sealed battery they could hit 4.35v?

The chemistry is not hugely different, just uses more expensive elements. So a factory trying to minimise cost could easily introduce a defect from changing from one type to another etc.

So this tiny change could cause a defective battery to catch fire where it normally would not.

Doesnt matter how you charge it, more how long it sits above 4.30v at ~60-100% float charge.

So samsungs 60% charge resolves 2 issues, peak voltage above 4.3 and energy release on failure.

Imho they need to go back to 4.3v policy rather than just replace and assume its a once off.Lithium batteries used in phones dont catch fire without physical damage like a knife or overvoltage. That's why this is so unusual.

Ps kernel source is available from samsungs website because android is based on linux.

Tldr: policy change from 4.3v float charge to 4.35vwhich I have confirmed by examining source code available on samsungs website would explain the fires

If the fix were that simple it would have been done by now.

This is a hardware issue.

 

[akin_t]

People on the internet must think they're smarter than Samsung's entire engineering division.

 

[mightyfacundo]

If the fix were that simple it would have been done by now.

This is a hardware issue.

Are you speaking from an understanding of what the guy on Reddit is saying or basing your statement on the presumed likelihood that Samsung would be smart enough to easily fix it? Please understand, I'm not questioning you, just trying to gain more knowledge of the subject.

How easy would it be to make that change? Is it a software fix or hardware fix? Thanks again.

 

[Dm1120]

If the issue were truly with the software allowing too much voltage and software alone, they would lock all their programmers in a room until they could develop a fix and push it out worldwide ASAP.

It seems like there's a design flaw of some kind that's letting these batteries overheat.

 

[dadsterflip]

People on the internet must think they're smarter than Samsung's entire engineering division.

Kinda like how everybody on this forum is smarter then Samsung on what they
should do about the whole situation.

 

[caveman84]

People on the internet must think they're smarter than Samsung's entire engineering division.

Evidently, Samsung engineering division was not "smart" enough to catch an apparent design flaw