Searching \ for '[EE] Very simple portable iPhone charger and low v' in subject line. ()
Help us get a faster server
FAQ page: www.piclist.com/techref/ios.htm?key=port
Search entire site for: 'Very simple portable iPhone charger and low v'.

Exact match. Not showing close matches.
'[EE] Very simple portable iPhone charger and low v'
2010\12\05@062534 by

So I'll probably try making a portable charger for my iPhone using 4 AA NiMH
batteries (2000 mAh). That should be around 4.8 volts under a load (will be
setting the iPhone to use a 1 Amp load).

I want to set a cutoff voltage of around 4.4 volts for the pack. When the
battery pack voltage drops to this level, I want a simple circuit to cut off
charging in some way. The purpose of this is to prevent damage to the
batteries via over discharging. I'm not even sure if the iPhone will charge
at such a low voltage (but I don't see why not, based on previous replies).

1. Is 4.4 volts/pack = 1.1 volts/cell a good cutoff voltage for NiMH? If
not, what would be a good cutoff voltage? I'm using Eneloops so I want them
to last a long time. I don't have a need to squeeze every last mAh out of
the cell. I need to make sure the voltage doesn't drop below a safe level.

2. How would I go about designing a simple circuit for this?
2a. It should have low power consumption.
2b. It should be reliable. When the set voltage is reached, it should
terminate current draw from the battery somehow. What's a good (efficient,
easy) way to control around 1 Amp of current? Transistor? Solid state relay?
2c. I was thinking of somehow using a comparator with a voltage reference to
signal that the shutoff level has been reached. I'm pretty sure most PICs
have a CCP module, are low power, and will run on the voltage range that the
pack will be supplying. However, where would I get such an adjustable
voltage reference? It needs to be accurate enough for this application.
2d. Maybe this doesn't even NEED a pic. Seems that a voltage reference,
comparator, and a relay of some sort will do the trick. I actually wouldn't
mind using CMOS ICs for this. Haven't touched them in a LOOOONG time
On Sun, Dec 5, 2010 at 6:25 AM, V G <x.solarwind.xgmail.com> wrote:

{Quote hidden}

Note: I just realized that an LM317 can be used as a good voltage reference..
Any comments on this? I want most of the battery's energy to go into
charging the iPhone, not be wasted on maintaining protection circuitry
V G wrote:
> What's a good
> (efficient, easy) way to control around 1 Amp of current? Transistor?

A single low side FET will do nicely, like IRLML2502.

Keep the gate passively pulled low with something like a 100KOhm resistor,
and have the circuit drive it high when the battery is high enough.

The voltage threshold detector sounds like a job for a TL431 and a PNP
transistor.

> Solid state relay?

Definitely not.

> I'm pretty sure most PICs have a CCP module,

There is no need for a programmable part like a PIC here.

********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000
V G wrote:
> I want most of the battery's energy
> to go into
> charging the iPhone, not be wasted on maintaining protection
> circuitry.

At 1A charge current, you'd have to go out of your way to make the current
draw of the remaining circuitry relevant.  You should however put a simple
mechanical on/off switch in series with the battery, and keep it off when
you know you're not charging.

********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000
tinyurl.com/mintyboost

It's a solved problem and you'll learn a lot looking at how it was solved.

Note that charging an iPhone (depending on make) can be non-trivial-
Apple has done some (minor) shenanigans (resistors to the data lines)
to set and verify the charge current.

Mike H.

On Sun, Dec 5, 2010 at 5:25 AM, V G <x.solarwind.xgmail.com> wrote:

{Quote hidden}

>

On Sun, 5 Dec 2010 06:25:19 -0500, "V G" <x.solarwind.xgmail.com> said:
> So I'll probably try making a portable charger for my iPhone using 4 AA
> NiMH
> batteries (2000 mAh). That should be around 4.8 volts under a load (will
> be
> setting the iPhone to use a 1 Amp load).
>
> I want to set a cutoff voltage of around 4.4 volts for the pack. When the
> battery pack voltage drops to this level, I want a simple circuit to cut
> off
> charging in some way. The purpose of this is to prevent damage to the
> batteries via over discharging. I'm not even sure if the iPhone will
> charge
> at such a low voltage (but I don't see why not, based on previous
> replies).

A voltage detect IC(reset IC) like the Microchip TC54 series(there are
plenty of others including from Microchip) and a mosfet will do all of
that. With no other parts!

Note that protecting a 4 cell nimh pack from undervoltage doesn't really
matter until well under 4 volts, since zero volts doesn't hurt them,
only negative voltage which you can get when you continue discharging
and one cell has gone to zero and starts negatively charging. In a
relatively balanced pack that doesn't happen when there is much juice
left in the remaining cells, so figuring 1.1 or 1.2 volts in those
remaining cells is quite reasonable.

Therefore if you have one cell at zero and three at 1.2 then you only
need to protect at 3.6 volts. In practice that is quite conservative.

Whether your iPhone thinks it's funny or not is another question, but it
shouldn't be damaged by low voltage.

Cheers,

Bob

-- http://www.fastmail.fm - The professional email service
On Sun, Dec 5, 2010 at 1:27 PM, Bob Blick <bobblickftml.net> wrote:
> Note that protecting a 4 cell nimh pack from undervoltage doesn't really
> matter until well under 4 volts, since zero volts doesn't hurt them,
> only negative voltage which you can get when you continue discharging
> and one cell has gone to zero and starts negatively charging. In a
> relatively balanced pack that doesn't happen when there is much juice
> left in the remaining cells, so figuring 1.1 or 1.2 volts in those
> remaining cells is quite reasonable.
>

Hi Bob,

Why do you say that overdischarge (as long as it is not below 0V) does
not hurt NiMH cells?

In my experience, voltage below about 0.7V for an extended time can
damage them. I've seen this with Energizer 2500 mA AA cells. I'm not
sure whether this is different for the Eneloops or not.

Sea
On 05/12/2010 15:05, Mike Hord wrote:
> tinyurl.com/mintyboost
>
> It's a solved problem and you'll learn a lot looking at how it was solved..
>
> Note that charging an iPhone (depending on make) can be non-trivial-
> Apple has done some (minor) shenanigans (resistors to the data lines)
> to set and verify the charge current.
>
> Mike H.
>

Nice project, very well written up.
Slightly different to what is being proposed here in that with 4 AAs there is no need for the boost converter,
so the circuit can be even simpler (just a voltage detector/switch needed as Olin and Bob have mentioned)
Definitely no need for a PIC (unless some more complex stuff not mentioned is being planned)

On Sun, 5 Dec 2010 14:04:38 -0500, "Sean Breheny" said:
> On Sun, Dec 5, 2010 at 1:27 PM, Bob Blick <bobblickftml.net> wrote:
> > Note that protecting a 4 cell nimh pack from undervoltage doesn't really
> > matter until well under 4 volts, since zero volts doesn't hurt them,
> > only negative voltage which you can get when you continue discharging
> > and one cell has gone to zero and starts negatively charging. In a
> > relatively balanced pack that doesn't happen when there is much juice
> > left in the remaining cells, so figuring 1.1 or 1.2 volts in those
> > remaining cells is quite reasonable.
> >
>
> Hi Bob,
>
> Why do you say that overdischarge (as long as it is not below 0V) does
> not hurt NiMH cells?
>
> In my experience, voltage below about 0.7V for an extended time can
> damage them. I've seen this with Energizer 2500 mA AA cells. I'm not
> sure whether this is different for the Eneloops or not.

You're right, it was sort of the "abridged version" of battery care. But
deep discharge does not hurt them "very much". The higher capacity cells
like the 2500 mah ones you have experience with are very fragile.

A battery pack that has gotten so far out of balance that one cell is
zero and the others are 1.2 is near the end of its service life, so the
one cell that is already weaker than the others is protected from
serious harm.

In a well balanced pack, 3.6 volt cutoff is .9 volts per cell and so all
cells are protected from deep discharge.

There is a tradeoff when setting the cutoff voltage, I tend to go for a
fairly low voltage so false disconnects are less frequent. And with a
simple circuit having just voltage hysteresis, less chance of it
becoming a power oscillator :)

Friendly regards,

Bob

-- http://www.fastmail.fm - Faster than the air-speed velocity of an
part 1 4270 bytes content-type:text/plain; charset="iso-8859-1" (decoded quoted-printable)

Note that the subject line describes something which you want but may not
easily be able to achieve.
Note the caveats and don't-don't-be-evil arcane knowledge re iPhone charging
below.

1.  The MintyBoost which Mike suggested is a "reasonably excellent" solution
to look at.

On 6 December 2010 04:05, Mike Hord <mike.hordgmail.com> wrote:

> tinyurl.com/mintyboost
>
> It's a solved problem and you'll learn a lot looking at how it was solved..
>

Note that it would be LESS efficient than what you were proposing but has

- "Having had all the work done for you

-  Will charge at whatever input voltage you wish. Your solution with 4 x
NimH is marginal wrt Vin and will stop charging some devices before you have
used as much battery capacity as you would want to (as you "just won't have
enough voltage under all Vbat and Vp[hone situations)..

He's been though several iterations and has learned a lot along the way.
I've attached his latest V3 circuit.
If your solution uses a switching regulator and is functionally much simpler
than that then it probably won't do what you want. Your original proposal
would probably use a linear regulator (as a buck converter would be only
very marginally more efficient and only when Vbattery was high)

gives you an outline of why the changes and the rest of his nots are worth
looking at.

He uses an LT1302 IC

Not the cheapest IC.
Available ex stock Digikey in 1's.

Efficiency versus Vo/Io/Vin curves give a fairer indication of real world
achievable results than most claims do. The typical claims you see elsewhere
of  "Up to 95%" or 90% or whatever is usually a spot performance downhill
with the wind behind you.

His V3 mentions 1A out somewhere in the notes (AIR) BUT the LT1302 is
notionally good for only 500 mA out at 5V and that only with 3V+ in.

LT claim this IC to have the largest switch capacity available in this p[kg..
An external FET would allow more capacity.

2. The LM317 which you mentioned is wholly unsuitable for the purpose
mentioned in this application and not a good solution in many others. A
careful look at the data sheet will show you why. Look at dropout,
differential voltage, minimum current, % accuracy, ... .

3. While NimH cells will not be grossly damaged by discharge to say 0.8V,
this will substantially reduce their cycle lives. Stopping discharge at 1.0
- 1.05V or so will enhance cycle life. Note that Vterminal drops under load
due to internal voltage drop so what you measure will depend on load current
as a fraction of C. With 4 cells (as you originally proposed) Iin = Iout so
at eg 100 mA a 2500 mAh Eneloop with be operating at 0.4 C which is nicely
conservative for long life.

Similarly, charging to somewhat less than full capacity will also enhance
cycle life.

4. If I was doing this from scratch and the primary specs were "use AA NimH
and be as efficient as possible" I would strongly consider 5 NimH cells and
a buck converter. 6 or more MAY be able to be as efficint for various
reasons but 5 allow operation at any Vbat and as Vbat is above but close to
Vout you can expect good efficiency as of right (unless you are
exceptionally good at wasting energy :-) ).

5. Anyone interested in charging iPhones and other Apple products may find
This may or may not be the sort of thing that Steve Jobs was thinking of
when he scathingly criticised the stupidity of Google's inclusion of "don't
be evil" in their mission statement (dramatic video available on YT).

(Besides this the single charge preventing resistor in Motorola's original
Razr phones and since  then much else, looks tame.

Those who see this as simply a low cost technical alternative to using an
enumeration device may also be happy with the "Charging is not supported
with this accessory" message.

Anyone is free to  buy evil products. But, then, you'll have to eat them.

R

part 2 8793 bytes content-type:image/png; name="mintyboostv3sch.png" (decode)

part 3 181 bytes content-type:text/plain; name="ATT00001.txt"
(decoded base64)

--
http://www.piclist.com PIC/SX FAQ & list archive
mailman.mit.edu/mailman/listinfo/piclist
> ...  what is being proposed here in that with 4 AAs
> there is no need for the boost converter, ...

For some values of "no need".
Lithium Ion cells with have ~= 4.2V terminal voltage.
At his target 1.1 V/cell endpoint he has 4.4-4.2 = 0.2V differential.

This must allow for wiring and connector drops, switch drop in charger
and anything that happends in the phone (or target). Doable but
unlikely.  In practice several tenthsof a volt more will be need and
if it happened to need 0.5V or even 1V more than battery voltage for a
full charge it would not be surprising.

If one wanted full control over the result with this type of
operation, adding a 5th cell would be annoying but very effective.
Worst case result would be 80% of prior efficiency, but in practice
closer than that to 4 cell case, and with full capabilities.

Russel
Hi Russell,

I think it's a "she" in this case :)

Sean

On Sun, Dec 5, 2010 at 8:54 PM, RussellMc <apptechnzgmail.com> wrote:
>
>
On Sun, Dec 5, 2010 at 6:50 PM, Bob Blick <bobblickftml.net> wrote:

> In a well balanced pack, 3.6 volt cutoff is .9 volts per cell and so all
> cells are protected from deep discharge.
>
> There is a tradeoff when setting the cutoff voltage, I tend to go for a
> fairly low voltage so false disconnects are less frequent. And with a
> simple circuit having just voltage hysteresis, less chance of it
> becoming a power oscillator :)

Oh yeah, that's something else I forgot to consider.

1. Is it necessarily a bad thing if it becomes a power oscillator?

2. How do I prevent it from becoming one
On Sun, Dec 5, 2010 at 8:54 PM, RussellMc <apptechnzgmail.com> wrote:

> Note that the subject line describes something which you want but may not
> easily be able to achieve.
>

1. I have seen the MintyBoost several times and have considered it, but
don't want to build it. I want to keep it relatively simple.

2. I want to use 4 NiMH cells since the iPhone 4 has a high capacity battery
and and two NiMH cells wouldn't have enough energy to fully charge the
phone.

On Sun, 5 Dec 2010 22:15:26 -0500, "V G"  said:

> 1. Is it necessarily a bad thing if it becomes a power oscillator?
>
> 2. How do I prevent it from becoming one?

Oscillation, while not harmful, can cause transient heat issues in your
mosfet if the switching time is significant(unlikely), or maybe cause
some EMI issues, also probably minor, mostly just annoying from an
esthetic standpoint.

While I have no experience with the iphone, most phone-like things I've
seen have a delay before engaging charging, so if they decide to stop
charging, they won't start it back up again until after a delay of
several seconds. So that's good.

Also, if the current flow when the undervoltage cutoff cuts off is low
enough, the battery voltage will not immediately rise above the
hysteresis voltage, so you won't get obvious oscillation. It may
reconnect at some later time because the batteries recover, but it won't
be because of the internal resistance of the batteries and the current.
Predicting that is probably near impossible, so you might want to get a
spread of voltage detectors in case the first voltage doesn't work out.

Cheers,

Bob

-- http://www.fastmail.fm - Access all of your messages and folders
wherever you are
> 2. How would I go about designing a simple circuit for this?
> 2a. It should have low power consumption.
> 2b. It should be reliable. When the set voltage is reached, it should
> terminate current draw from the battery somehow.

I suspect you will find all these attributes in chips designed for
monitoring such batteries. I would suggest you check out the range of
chips at Linear Technology, looking for 'battery meter' chips. Linear
Technology have also been good in the past at supplying sample chips
free of charge, although it is a while since I got anything from them,
so their policy may have changed since.
-- Scanned by iCritical.
RussellMc wrote:
>> ...  what is being proposed here in that with 4 AAs
>> there is no need for the boost converter, ...
>
> For some values of "no need".
> Lithium Ion cells with have ~= 4.2V terminal voltage.

The OP has mentioned several times that he is using a specific model of NiMH
cells.  4 of those in series does work out just about right over the valid
range of USB voltages.

********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000
V G wrote:
> 1. Is it necessarily a bad thing if it becomes a power oscillator?

Large repetitive unintended voltage and current swings in your circuit are
rarely a good thing.

> 2. How do I prevent it from becoming one?

A little hysteresis might help.

********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000
> >> ...  what is being proposed here in that with 4 AAs
> >> there is no need for the boost converter, ...

> > For some values of "no need".
> > Lithium Ion cells with have ~= 4.2V terminal voltage.

> The OP has mentioned several times that he is using a specific model of NiMH
> cells.  4 of those in series does work out just about right over the valid
> range of USB voltages.

His wishful  thinking nor vain repetition doth not a summer make :-) -

He has said that he wishes to use Eneloop AA NimH cells to charge an
iPhone, that he expects them to provide about 1.2V/cell at 1 A load
and that his desired endpoint is 1.1 V/cell.

He may have arrived at the 1.2V/cell at 1 A figure from datasheets et
al - and maybe not. Sanyo's webpages suggest

http://www.eneloop.info/home/performance-details/voltage.html

that at 2A he can expect 70% of capacity at over 1.2V/cell  and that
the cells will drop under 1.1 V/cell at about 90% of rated capacity.
SO the batteries meet his spec reasonably well.

If he's careful and somewhat magic he MAY be able to achieve 0.2V drop
outside the cells (wiring, battery connectors, electronics, output
connector, ...) and as IPhone is made by Apple and so must be magic it
can probably manage 0.2V drop as well.

So at 4.8V battery voltage you get about 4.4V available for  LiIon
charging (enough) but you hit 4.2V for LiIon at about 4.6V battery =
1.15 cell, which the eneloop graph says is well up the discharge curve
- perhsp 80% - 90% of total capacity, which is highly commendable.

If he or Apple blink and you get more than 0.4V total drop at 1A from
his battery pack to the LiIon pack then "there will be problems"
[tm][probably].

I feel (rightly or wrongly) that my original analysis stands re 4 x AA
NimH  cells (even Eneloops) as being marginal but usable for USB
charging apps. I'd personally use 5+ cells, or a boost converter, if I
wanted a fully designable result. I suspect that you would too, on
reflection.

Russell
On Mon, Dec 6, 2010 at 10:07 AM, RussellMc <apptechnzgmail.com> wrote:

{Quote hidden}

I was actually looking at this:

http://www.eneloop.info/home/performance-details/discharge-current.html

Specifically, the 1 Amp discharge curve. Anywhere between 1.1 and 1.18
volts/cell seem to be a good value according to that graph. I want to keep
it *just* before the sharp drop in voltage.

I'll probably stick with 1.15 volts/cell for now and adjust it later if I'm
not satisfied
I don't want to hijack the discussion, just expand it a bit...

While shopping for a car charger for a Blackberry Storm, both the
Verizon store AND several electronics mega-stores warned me against the
cheap chargers (and at Verizon, only their model would do). Their
argument was that the cheap ones would 'force' more current into the
phone than was good for the battery and severely shorten the life of the
battery.

None of the clerks were up to a discussion of engineering
fundamentals...

However, one service tech who seemed trustworthy did claim to personally
know that the non-Verizon charged units came back with batteries that
wouldn't hold a charge any more. He could only speculate on the
engineering reason.

It would seem unwise to design a phone with the battery directly
connected to the usb +V. Some sort of regulation is required, if only to
prevent exploding LiIon batteries.

There is also talk of 'fast charge' adapters, versus regular. Hmmm.

So,

Is there any truth embedded in this 'street' knowledge?

I can't see how one could 'force' more current into the phone, except if
it was a constant current supply (or maybe +12v on the usb cable), but
that would violate the usb spec. Even then, the internal regulator
should prevent problems. Am I missing something?

How is 'fast charging' accomplished?

Now the funny part. I started using a non-Verizon adapter. It charges
other usb rechargeable things fine. The Storm says it likes it fine too.
No excess heat generation during charge. Done a few times over a month
and suddenly the battery went from 24+ hours on a charge to 12, then 7,
then 3, then 1 hour life per charge over the course of a week or two. I

Not sure if iPhone and Blackberry share the same type supply design (the
Storm seems to need more than a simple cable to charge as does the
iPhone). It seems a lot like cause and effect. I'm not willing to buy
into that until I can see an engineering reason though.

Anyone care to comment?

-Skip
> While shopping for a car charger for a Blackberry Storm, both the
> Verizon store AND several electronics mega-stores warned me against the
> cheap chargers (and at Verizon, only their model would do). Their
> argument was that the cheap ones would 'force' more current into the
> phone than was good for the battery and severely shorten the life of the
> battery.

This will only happen if either

- The design of the device is bizarrely incompetent or
- It's a built in design feature intended to cook your batteries if an
"unapproved" charger is used so they can sell more batteries, sell
their own chargers, scare users off doing the same AND make it all

Both of the above are possible.

> However, one service tech who seemed trustworthy did claim to personally
> know that the non-Verizon charged units came back with batteries that
> wouldn't hold a charge any more. He could only speculate on the
> engineering reason.

As above.

> It would seem unwise to design a phone with the battery directly
> connected to the usb +V. Some sort of regulation is required, if only to
> prevent exploding LiIon batteries.

Absolutely.
Any device that did that would be so incompetently "designed" as to be
arguably criminal.
About the best "excuse" for doing that, if it was done, would be to
say "it's not actually a USB port - it just uses the same plug - it
needs a proper charger to work".
ie what part of "don't don't be evil don't you understand?"

> There is also talk of 'fast charge' adapters, versus regular. Hmmm.

If the battery is LiIon of some sort then this should only be possible
under the device's own control
A given LiIon battery has a manufacturer fixed maximum charge rate. It
MAY be possible to exceed this rate with known and acceptable
consequences (for some values of "acceptable") BUT the device should
be the controller of this behaviour.

Following suffixes are mine:

Normal LiIon practice is to charge at Ichg_max until Vbat_max is
reached, then hold the voltage at Vchg_max until Iterminate is
reached. The battery "decides" the voltage profile during the 1st
phase and the current during the second phase.
Ichg_max is usually 1C but some makers or chemistries allow 2C and
some specialist designs coming to market allow MUCH faster. At the end
of this phase charge capacity is about 60%-80% of max. Some "=fast
chargers" terminate at this point - you get faster time to "charge
complete" but less capacity. Rate is better than full charge as you
get say 2/3C in about 1 hour whereas last 1/3C takes typically another
2 hours so is on average 4 times slower rate (ie 2/3C : 1 hour  versus
1/3 C : 2 hour).

A "simple" test is possible IF you can access battery terminals during
charging. Possibility depends on device and your ingenuity.
With a "standard": LiIon/ LiPoly cell, if Vbattery is > ~4.2V at
"normal" temperatures then "you have problems". If Iin is > 1C usually
or maybe 2C (manufacturers spec needed) when cell hits the 4.2V
pedestal then "you have problems". If Vbattery EVER exceeds 4.3V you
have BIG problems. Even at 4.3V you are running on the edge. As
temperature rises Vbattery max should be slightly reduced. Maybe 4.1V
at 40C if 40C is in spec for cell used.

> So..  Is there any truth embedded in this 'street' knowledge?

Potentially yes as above.
But only through extreme incompetence or evilness on makers part.

> I can't see how one could 'force' more current into the phone, except if
> it was a constant current supply (or maybe +12v on the usb cable), but
> that would violate the usb spec. Even then, the internal regulator
> should prevent problems. Am I missing something?

Not apparently.
BUT the "internal regulator should" may not apply if the cell is
directly connected to a "not really a USB USB connector" as above and
the "real" charge control is external.

> How is 'fast charging' accomplished?

As above.
It isn't, OR the device does it with special care OR it is just early
termination.

> Now the funny part. I started using a non-Verizon adapter. It charges
> other usb rechargeable things fine. The Storm says it likes it fine too.
> No excess heat generation during charge. Done a few times over a month
> and suddenly the battery went from 24+ hours on a charge to 12, then 7,
> then 3, then 1 hour life per charge over the course of a week or two. I

Class action time?

> Not sure if iPhone and Blackberry share the same type supply design (the
> Storm seems to need more than a simple cable to charge as does the
> iPhone). It seems a lot like cause and effect. I'm not willing to buy
> into that until I can see an engineering reason though.

iPhone DOES use "proper" USB connection on power leads but uses non
standard resistor voltage level setting on "data" leads to set charge
rate. This is "safe enough" ([tm] if you understand what is needed and
do it. Odds are SOME people will start making iPhone chargers with
data lines hard clamped in flat out charge mode. Battery life will
suffer. Phone life may.

Russell McMahon
On Tue, Dec 7, 2010 at 7:02 PM, RussellMc <apptechnzgmail.com> wrote:

> Class action time?
>

I would most certainly say so
RussellMc wrote:
> BUT the "internal regulator should" may not apply if the cell is
> directly connected to a "not really a USB USB connector" as above and
> the "real" charge control is external.

But then the phone would have problems charging from a real USB port.

********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000
On Tue, Dec 7, 2010 at 7:02 PM, RussellMc <apptechnzgmail.com> wrote:
> iPhone DOES use "proper" USB connection on power leads but uses non
> standard resistor voltage level setting on "data" leads to set charge
> rate. This is "safe enough" ([tm] if you understand what is needed and
> do it. Odds are SOME people will start making iPhone chargers with
> data lines hard clamped in flat out charge mode. Battery life will
> suffer. Phone life may.

If I understand what you're saying, it doesn't make sense. The
'charger' is just a ~5V power supply. It doesn't tell the battery
charger IC to charge the battery. The resistors just indicate that it
is safe for the phone to draw a certain amount of current.

-- Martin K
> RussellMc wrote:
>> BUT the "internal regulator should" may not apply if the cell is
>> directly connected to a "not really a USB USB connector" as above and
>> the "real" charge control is external.

> But then the phone would have problems charging from a real USB port.

Indeed - that was the exact point being made.

If you look at the prior thread (often not a terrible idea)  you'll
see that the xxx's battery seems to die rathjer quickly if not charged
with brand xxx charger. Indications are that the unit MAY not be
suitable for "real" USB connection. If so, probably purposefully
rather than incompetently.

> On Tue, Dec 7, 2010 at 7:02 PM, RussellMc <apptechnzgmail.com> wrote:
>> iPhone DOES use "proper" USB connection on power leads but uses non
>> standard resistor voltage level setting on "data" leads to set charge
>> rate. This is "safe enough" ([tm] if you understand what is needed and
>> do it. Odds are SOME people will start making iPhone chargers with
>> data lines hard clamped in flat out charge mode. Battery life will
>> suffer. Phone life may.

> If I understand what you're saying, it doesn't make sense.

That's probably true :-).

> The 'charger' is just a ~5V power supply. It doesn't tell the battery
> charger IC to charge the battery.

The power supply per se doesn't. The resistors do.

> The resistors just indicate that it
> is safe for the phone to draw a certain amount of current.

A look at the LTC4066 data sheet (previously referred to) will explain.;

The resistors provide information to the phone / which has the charger IC in it.
While the usual use may be to tell the phone what current it can
reasonably draw from the charger, one available mode is "draw what you
can". As long as "what you can" is less than the battery's maximum
charge rating then all may be well. If the charger is capable of very
large currents and the LTC4066 decides to play along, then such a mode
MAY allow the phone to be used in lieu of a toaster oven for reflow
soldering - especially so once the Lithium metal vaporises,as happens.
As a modern PC 5V supply may be capable of upwards of 50 A, the
attention paid to USB max possible current draw may be all that saves
you - or fails to.

When people will sell you small caps inside large cases, and
"legitimately" sell "D" cells with AA batteries therein, YMMV widely.

Russel
Excellent discussion so far. I have a few observations that may be of
interest, and a clarification.

First, I may not have been clear. I've only used this 'other' charger a
couple of times. All other times are with the approved manufacturer's
cradle. Poor performance is seen independent of the charger du jour.

When plugged in with its own original cable, there is a slight delay,
then it recognizes a power source and commences charging. It uses a
micro usb connector. I then tried a mini-to-micro adapter placed on the
end of a generic usb cable (tied to the pc). It gave a message that it
was not the proper cable and charging did not commence. I figured at
that point that it was doing something similar to the iPhone, from a
long ago discussion here.

Interestingly, I tried a Kindle power cord which also uses a micro usb
end. The Blackberry liked it. Perhaps pullups or such are an informal
'standard' among those using micro usb ends? I've not tried data, but
all 3 cables mentioned did get the device recognized in Windows. The
mini-to-micro just didn't prompt a charge scenario.

I also have an app called 'battery watch', which provides some data from
the system on the battery. It showed some interesting data yesterday, so
I'll share and perhaps it will fuel discussion.
On the charger, I happened to invoke the app and see it was at 72% level
(however that is calculated) and 4200mV. Based on your note, I was
interested... It stayed at 4200mV up to 100%. It could either be tightly
controlled, flawed, or that may be the limit of the A/D or the software.
I found such a nice round number suspicious, especially for such a long
way to go. I guess it could also be something the lawyers said to the
designers - "don't let the OS report more than 4200mV or engineer-users
may threaten us with legal action" ;)

This morning, starting at 100%, it showed 85% at 82F temp and 4095 mV
after only a few minutes of usage (email, web), and now is at 87%, 77F,
and 4049 mV. This apparently reveals 2 things: the capacity is much less
than just recently, when it would drop to 95% with even greater usage
and stay there for a while, and if you look at the numbers closely, it
seems to apply temperature to the state of charge equation (note the
temp and voltage relationship vs %.
Note that these 3 numbers are not just from a pre-programmed graph,
since it showed 72% and 4200mV while charging, and 4200mV all the way to
100%. Just to check, I plug it in now and it shows it's charging, and
reports 86%, 75F and 4045mV. In a minute, it is now at 88% and 4200mV.
It seems to pop there quickly.
There is also a 'battery status' parameter that shows 'discharging' when
off charger, was at '101' while charging, and changed to '111' when it
hit 4200mV. I suspect it reports what it sees. It seems to me to be
showing symptoms of higher internal resistance in the battery. Is this
something an external charger could cause, given the observations I've
offered? I've not seen a LiIon degrade so rapidly.

BTW, the 82F was from holding it while using, the 77F is sitting on the
desk. I believe this app (a free one) only reports what the system
reveals and doesn't do much on its own. This would imply that there is
more to the charging circuit than just usb to +V internally.

Still, in an hour or two, it will die and for a short time, try to start
up again, then die repeatedly mid-boot from lack of power.
-Skip

On Wed, 2010-12-08 at 13:02 +1300, RussellMc wrote:
>
> iPhone DOES use "proper" USB connection on power leads but uses non
> standard resistor voltage level setting on "data" leads to set charge
> rate. This is "safe enough" ([tm] if you understand what is needed and
> do it. Odds are SOME people will start making iPhone chargers with
> data lines hard clamped in flat out charge mode. Battery life will
> suffer. Phone life may.
> On the charger, I happened to invoke the app and see it was at 72% level
> (however that is calculated) and 4200mV. Based on your note, I was
> interested... It stayed at 4200mV up to 100%. It could either be tightly
> controlled, flawed, or that may be the limit of the A/D or the software.
> I found such a nice round number suspicious, especially for such a long
> way to go. I guess it could also be something the lawyers said to the
> designers - "don't let the OS report more than 4200mV or engineer-users
> may threaten us with legal action" ;)
>

The 4200 mV is real and comforting - see my prior description of LiIon
charging within the last day.

Briefly, LiIon charges at constant current until a max allowed voltage
is reached and then at that constant voltage with declining current
(as set by characteristics of accepting device, until some fraction of
Imax is reached and charge terminates.

V max is usually 4.2V - as you observe.
4.1V leads to longer cycle life and noticeably lower capacity , 4.3V
leads to a potentially exciting life, and 4.4V leads to vent with
flame.

At the transition from constant current to constant voltage capacity
is in 65% - 80% range - consistent with your meter.
Discharge V is below charge V at same state of charge.

Russell McMaho

More... (looser matching)
- Last day of these posts
- In 2010 , 2011 only
- Today
- New search...