On 1/7/2017 1:25 PM, T i m wrote:
On Sat, 07 Jan 2017 11:48:01 -0800, mike wrote:

snip

You have an unknown source suggesting some vague fix that
defeats charging protocols using an unknown
"diode" not clearly defined on the schematic to a DIFFERENT tablet
made by Samsung,

Potentially yes. He *could* know exactly what he's doing (he looks
like he might) but he also may have no idea (so point noted).

MIGHT? Sure, he might...but he also may catch fire. Are you feeling
lucky with your daughter's life? house?

the poster child for lithium battery safety.

Didn't know that, thanks.


Unless you've been living off the grid you couldn't escape the hoopla
over the Samsung Galaxy Note 7 battery fires.


How many bad signs do you need before you determine that this is too
risky??

Well, to be fair I wasn't going to even attempt anything (if I ever
did) until several other avenues had been exhausted but again. I do
get your point / caution.

Not at all clear exactly what he's doing.
If he's adding a parallel charging path

If you have followed his idea and see that's exactly what he is doing
then fair enough.

I couldn't tell where on the schematic he put the diode or the type of
diode. We could certainly guess the type of diode.
If he had the EXACT tablet as yours, with the SAME hardware version
and the SAME firmware version and he disclosed the EXACT connections
on the schematic and we had specs for all the relevant chips, I'd be much
more inclined to give him the benefit of the doubt.

If not (and playing devils advocate for a second)
then couldn't it (possibly?) be a nifty / safe solution?
Possibly. Are you feeling lucky? I suggest that if it's
a widespread problem with a solution that simple, you'd read
a lot more about it from verifiable sources.

If your car heater is busted and someone suggested that you
should put an open container of gasoline on the front seat
and set it on fire, would you try it? I'm quite sure
you'd be warmer if you did that.

I couldn't tell where on the schematic he placed the diode.
The schematic shows the protection chip, so it looks like he
may be defeating some or ALL of the circuitry that controls
the charging process, possibly affecting the full charge detection. That is
an extremely bad idea. Do you think he is likely to more about this
than the people who make/use the chip in production?

Another alternative is to bypass the chip and apply some charge to directly
to the battery until it reaches the threshold where the protection
chip can take over.

There are two serious problems with this.
If you just use a diode, you're gonna supply all the current the
charger can supply. Maybe that's two amps? Let's say it's only one amp.
If the diode voltage is at the max spec, it might be 1.2V. So, at 1 amp
that's 1.2 watts and the diode will melt the solder and fall off, landing
who knows where.

Second, it's risky to try to recover a lithium battery that's been
discharged below the minimum voltage. If you do, you need to use LOW
current. The diode provides HIGH current...very bad.

You fix the current with a resistor in series with the diode.
Now, you have to worry about maximum voltage.
At nominal 5.000V charger voltage and nominal 0.7V diode drop,
that leaves you with 4.3V on the battery. That's not good.
I've seen chargers with nominal 5.3V at low current. That puts your
battery at 4.6V. That's asking for a fire.
If the diode overheated, it may be leaky, making things worse.

If you use two silicon junction diodes and a resistor, you start
to get into the range of possibility. Three diodes is even better,
but then you run into the possibility that the battery will never reach
the voltage needed for the protection chip to take over.

THE DEVIL IS IN THE DETAILS. We don't know ANY of the details of the
charging circuit or what his mod does. But, we can surely guess.




thru the unspecified diode,

Unspecified other than a "a std rectification' diode you mean, or were
you still talking about it's roll?

What is a std rectification diode? How do you know that's what he used?
He did suggest removing a random diode from another tablet/phone.
If it's a schottkey diode, the voltage is lower and might have significant
effect on how it works...maybe...we don't know.

(The other day I fixed an electronic motorcycle speedo for a friend by
replacing the remains (through water ingress / corrosion) of some SM
diode with a 4001 I happened to have a load of. It seems to work ok
but isn't charging a battery of course.

A 1N4001 diode always works, until it doesn't. You had the ability to test
the result and determine that the circuit works.
I suggest that you haven't the knowledge or equipment to validate a lithium
battery charger mod...or you wouldn't have asked the question. Since
it's not your tablet, you'll not be around to notice any telltale signs
of danger.

I'd suggest that it's a time-bomb.

I do understand your caution so thanks for that.

You'd need at least three silicon junction diodes and a series resistor.
I would never, ever, suggest that anyone try such a stupid thing.

Understood ... and that's why I asked if anyone could 'vet' what was
proposed here. ;-)

Another thing that I believe is overlooked.
When you modify a device in an unsafe manner, you risk harming yourself
and those around you.

Understood ... but I would prefer to see it as 'if' rather that when
as not all cases *will* be inherently dangerous (even if they weren't
fully considered or technically correct). Like, loads of people have
simply unplugged and replugged the battery in such 'non-charging'
situations but how do we know if that also isn't potentially bypassing
some safety function?

Wouldn't be the first time that something seemingly simple and impossible
to screw up caused a major screw up. Won't be the last either.
Samsung can't even get it right on their second try of the Note 7.

If you let ANYONE else use it, or steal it, or sell it to someone,
you put THEM at a risk that they're not even aware of. If it's sold
at your estate sale, all your good intentions to remove the mod
are worthless.

Understood. There are many jobs I could easily undertake but where the
thought of something going wrong, even if nothing to do with my input,
doesn't bear thinking about ... so I don't. What I call the 'what if'
factor.

This is one of those times.

Just OOI if I may Mike, do you ever buy any unbranded electronics ...
phone chargers, cables, battery packs etc? I ask because I understand
some of these sort of things are also not built with the same level of
safety in mind that we might assume and pay for in some of the
(hopefully) better branded kit? (genuine question).

I don't think you can tell any more. Paying more doesn't guarantee
better stuff. The name is no longer sufficient. Most everything these
days is bought from the cheapest bidder. The seller slaps his label
on it and charges big bux. The next one you buy from the same vendor
may be completely different.

If you treat the tablet and charger as a system, you can save money
by leaving out some parts. The manual probably says, "don't ever use
any charger other than the one supplied." How many of us heed that
warning?

Even if the designer had it all figgered out, that's still no guarantee.
I've had situations where purchasing decided unilaterally to change a part
to save a few cents. I didn't learn about it until the production line
shut down.
Just like your example. "It's just a std. diode." Unless it isn't.

I buy almost all my stuff busted at garage sales.
I have an electrical engineering degree and 40 years design experience.
I almost always get it close enough to right.
ALMOST!
I once set a laptop on fire while charging the battery. I assumed that
the designers had some common sense. I was wrong. The
charge current limit
was in the external charger, not anywhere near the battery. Label says
19VDC, but if you apply 19VDC, you let the smoke out.

The internet is a very dangerous place.
It's a great place to get ideas, but implementation is often lacking.
Most people who post stuff haven't a clue. They take a narrow
view of the problem/solution. Often it doesn't matter.
Sometimes, it sets you on fire.

Get your ideas from the web. VERIFY the solution and do your own math.
Implement what you can verify and/or test. Even if you think you
know what you're doing, it's extremely difficult to test for stuff you
never anticipated or couldn't simulate. We learned that from the design
of the Hubble Space Telescope.

Are we having fun yet?


Thanks for your feedback. ;-)

Cheers, T i m