car usb charger not durable

On Sunday, March 30, 2014 9:32:07 PM UTC-4, Jeff Liebermann wrote:
> Like a complete idiot, I decided take apart one of my cheapo USB car
>=20
> chargers and see what makes it work.  I know they are junk, but I
>=20
> never expected them to be this bad.  Start here:
>=20
> <http://802.11junk.com/jeffl/LTspice/Cig%20Lighter%20USB%20charger/>
>=20
>=20
>=20
> What's inside:
>=20
> <http://802.11junk.com/jeffl/LTspice/Cig%20Lighter%20USB%20charger/Circui=
t%20view.jpg>
>=20
> <http://802.11junk.com/jeffl/LTspice/Cig%20Lighter%20USB%20charger/Top%20=
view.jpg>
>=20
>=20
>=20
> They ground off the numbers from the switcher chip so I couldn't
>=20
> identify it.  I *ASSUME* that it is a common MC34063A, but that's
>=20
> apparently not the case because there's a heat sink on the bottom of
>=20
> the chip that is suppose to be sweat soldered to the PCB.  I therefore
>=20
> have the wrong chip number.  (Sorry about the focus problem but you
>=20
> can see the metal bottom on the chip).
>=20
> <http://802.11junk.com/jeffl/LTspice/Cig%20Lighter%20USB%20charger/SOIC-8=
%20with%20heat%20sink.jpg>
>=20
> Of course, it's not soldered, so heat dissipation will suffer.
>=20
>=20
>=20
> I'm still not sure it's an MC34063A, but if that's the chip, it's only
>=20
> rated to 1.5A, not 2.1A.  Argh.  Duz anyone have a clue as what other
>=20
> chip it might be?
>=20
>=20
>=20
> So, I then traced out the schematic.=20
>=20
> <http://802.11junk.com/jeffl/LTspice/Cig%20Lighter%20USB%20charger/Cig%20=
Lighter%20USB%20schematic.jpg>
>=20
> It's not an MC34063A because it doesn't oscillate and the schematic is
>=20
> somewhat different from the test circuits in the app notes and data
>=20
> sheet.  The LTspice simulation results in zero output.  I'm also not
>=20
> certain of some of the part values.  The resistors are probably
>=20
> correct, but the small cap and inductor are bad guesses since my LRC
>=20
> meter is not available.
>=20
>=20
>=20
> I then noticed something odd in the resistor network around the USB
>=20
> connector.  If you look at R3, R4, R5, and R6, on the schematic, it
>=20
> makes no sense.  The traces between the data pins on the two
>=20
> connectors should have been seperated (or cut).  The way it shows on
>=20
> the PCB and schematic, the pins are shorted together.  No way is that
>=20
> going to work for both Apple and Non-Apple products.
>=20
>=20
>=20
> To be continued, possibly with some measurements, as time permits.
>=20
>=20
>=20
>=20
>=20
> --=20
>=20
> Jeff Liebermann     jeffl@cruzio.com
>=20
> 150 Felker St #D    http://www.LearnByDestroying.com
>=20
> Santa Cruz CA 95060 http://802.11junk.com
>=20
> Skype: JeffLiebermann     AE6KS    831-336-2558

It's not a 34063, the connection at the inductor and the cap to SWC tells m=
e it is an internally boosted N-CH FET switch. The oscillator is internal s=
omething like 500KHz, the current limiting is achieved by sensing VDS acros=
s the FET. Something like this http://www.ti.com/lit/ds/symlink/lm3404.pdf,=
 but not exactly, or http://cds.linear.com/docs/en/datasheet/13756fd.pdf , =
but not exactly. Probably an Asian semi.

On 3/30/2014 6:32 PM, Jeff Liebermann wrote:
> I'm still not sure it's an MC34063A, but if that's the chip, it's only
> rated to 1.5A, not 2.1A.  Argh.  Duz anyone have a clue as what other
> chip it might be?

Some of the chargers that I have use 34063 chips. A few others use what 
appears to be a AD84064Q (certainly not Analog Devices). I didn't spend 
much time searching, but the only reference I found describes it as a 
12V to 5V buck converter. I have not yet found a data sheet for it.

On Sun, 30 Mar 2014 19:41:28 -0700 (PDT),
bloggs.fredbloggs.fred@gmail.com wrote:

>It's not a 34063, the connection at the inductor and the cap 
>to SWC tells me it is an internally boosted N-CH FET switch. 
>The oscillator is internal something like 500KHz, the current 
>limiting is achieved by sensing VDS across the FET. Something 
>like this http://www.ti.com/lit/ds/symlink/lm3404.pdf, but 
>not exactly, or http://cds.linear.com/docs/en/datasheet/13756fd.pdf , 
>but not exactly. Probably an Asian semi.

Thanks.  I'll keep digging.  I found some photos of the insides of
other USB car chargers, but they're mostly MC34063 chips.

I couldn't resist tearing apart the "better" USB charger:
<http://www.ebay.com/itm/191118057397>
I like the part where they "upgraded" the specs from 3.1A to 4.2A.  I
wonder if only the spec was change, not the circuitry.

See photos at:
<http://802.11junk.com/jeffl/Powergen%20USB%20car%20charger/>
There are two chips labeled MK2347 and MK2348 which are nowhere to be
found on the internet.  The board construction is creatinly better
than the $3 variety.  It seems like far too much work to reverse
engineer the schematic.  No LTspice model without the chip numbers.
Dead end, I guess.

-- 
Jeff Liebermann     jeffl@cruzio.com
150 Felker St #D    http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann     AE6KS    831-336-2558

Jeff Liebermann wrote:


> I bought some of these.
> <http://www.ebay.com/itm/390759035788>
> on a recommendation by a friend.
> <http://tips.navas.us/2014/01/best-usb-car-charger.html>
> It seems to charge my Samsung tablet at the rated 2A, but I haven't
> bothered to measure it.  No failures so far.
> 

Just a FYI, given the non-standard USB of Apple products, I don't think you 
can make a device that is capable of high current for both Apple and normal 
USB devices. Probably that device works well for normal USB but not the odd 
Apple type. 

Your best bet to getting a good charger is to buy OEM devices. Since a real 
company (say Samsung, HTC) has a name to protect, they are less likely put 
their name on a junky device. Now you can go on ebay and get real branded 
OEM chargers at a good price. I see OEM Samsung car chargers on ebay now.

My last deal, though not for car use, was OEM 1.8A Blackberry branded 
charger for $2 each in a quantity of three. Ebay item 221391879004. 
Blackberry follows the USB standard, so it will charge any phone, well other 
than Apple.  Now I can leave a charger at all times in my suitcase, one in 
the notebook bag, and one in the car. The microusb charger standard is 
command economics at it's finest. People will junk a phone but not the 
chargers because they can always use a spare charger. Anything Blackberry is 
enterprise grade, kind of like Lenovo. It will work.

On Sunday, March 30, 2014 11:56:27 PM UTC-4, Jeff Liebermann wrote:
> On Sun, 30 Mar 2014 19:41:28 -0700 (PDT),
> 
> bloggs.fredbloggs.fred@gmail.com wrote:
> 
> 
> 
> >It's not a 34063, the connection at the inductor and the cap 
> 
> >to SWC tells me it is an internally boosted N-CH FET switch. 
> 
> >The oscillator is internal something like 500KHz, the current 
> 
> >limiting is achieved by sensing VDS across the FET. Something 
> 
> >like this http://www.ti.com/lit/ds/symlink/lm3404.pdf, but 
> 
> >not exactly, or http://cds.linear.com/docs/en/datasheet/13756fd.pdf , 
> 
> >but not exactly. Probably an Asian semi.
> 
> 
> 
> Thanks.  I'll keep digging.  I found some photos of the insides of
> 
> other USB car chargers, but they're mostly MC34063 chips.
> 
> 
> 
> I couldn't resist tearing apart the "better" USB charger:
> 
> <http://www.ebay.com/itm/191118057397>
> 
> I like the part where they "upgraded" the specs from 3.1A to 4.2A.  I
> 
> wonder if only the spec was change, not the circuitry.
> 
> 
> 
> See photos at:
> 
> <http://802.11junk.com/jeffl/Powergen%20USB%20car%20charger/>
> 
> There are two chips labeled MK2347 and MK2348 which are nowhere to be
> 
> found on the internet.  The board construction is creatinly better
> 
> than the $3 variety.  It seems like far too much work to reverse
> 
> engineer the schematic.  No LTspice model without the chip numbers.
> 
> Dead end, I guess.
> 
> 
> 
> -- 
> 
> Jeff Liebermann     jeffl@cruzio.com
> 
> 150 Felker St #D    http://www.LearnByDestroying.com
> 
> Santa Cruz CA 95060 http://802.11junk.com
> 
> Skype: JeffLiebermann     AE6KS    831-336-2558

They should be using this:

http://www.ti.com/lit/ds/symlink/tps54560-q1.pdf
-or-
http://www.ti.com/lit/ds/symlink/tps54360-q1.pdf

but it's waaaaay too expensive. 

On 31.03.2014 03:32, Jeff Liebermann wrote:

> I'm still not sure it's an MC34063A, but if that's the chip, it's only
> rated to 1.5A, not 2.1A.  Argh.  Duz anyone have a clue as what other
> chip it might be?

MP1410 - 2A Step Down DC to DC Converter

Please google for the data sheet.
Best regards, Piotr

On Monday, March 31, 2014 3:07:01 PM UTC-4, Piotr Piatek wrote:
> On 31.03.2014 03:32, Jeff Liebermann wrote:
> 
> 
> 
> > I'm still not sure it's an MC34063A, but if that's the chip, it's only
> 
> > rated to 1.5A, not 2.1A.  Argh.  Duz anyone have a clue as what other
> 
> > chip it might be?
> 
> 
> 
> MP1410 - 2A Step Down DC to DC Converter
> 
> 
> 
> Please google for the data sheet.
> 
> Best regards, Piotr

Really bad, 16V max input in an automotive app? No wonder they pop.

bloggs.fredbloggs.fred@gmail.com wrote:


> 
> Really bad, 16V max input in an automotive app? No wonder they pop.

Actually 16v is a common spec for such chips. The problem with going much 
higher is either the chip gets very complicated, process wise, or has to be 
made on a very old high voltage process. It is cheaper to make a 16V chip 
and protect it externally. 

Of course it is really cheap not to protect it at all, especially if you are 
some Chinese company that can just do a name change if the product is 
unreliable. 

On Monday, March 31, 2014 6:10:48 PM UTC-4, miso wrote:
> bloggs.fredbloggs.fred@gmail.com wrote:
>=20
>=20
>=20
>=20
>=20
> >=20
>=20
> > Really bad, 16V max input in an automotive app? No wonder they pop.
>=20
>=20
>=20
> Actually 16v is a common spec for such chips. The problem with going much=
=20
>=20
> higher is either the chip gets very complicated, process wise, or has to =
be=20
>=20
> made on a very old high voltage process. It is cheaper to make a 16V chip=
=20
>=20
> and protect it externally.=20
>=20
>=20
>=20
> Of course it is really cheap not to protect it at all, especially if you =
are=20
>=20
> some Chinese company that can just do a name change if the product is=20
>=20
> unreliable.

I'm not sure about that, all the major semis have automotive grade chips, a=
ble to withstand that 60V 500ms load dump, which keeps getting worse every =
time you turn around. Someone has to be buying their stuff maybe out of nec=
essity to meet safety regulations.