Is my K3 have hardware errors?

[felonwan]

My Kindle 3g +wifi is bricked for quiet a long time. Now, it is a white board normally. When I connect it to a PC, Windows shows "SE Blank Ringo", and Linux shows "15a2:0030 Freescale Semiconductor Inc.". When connected to a charger, the LED light is always yellow(the same as the battery is taken out) and then went out after several hours but never turn green.

I've tried to charge the battery out of the device and nothing happened when I put the battery on.

So, is there any hardware errors? The battery or the flash memory?

[hawhill]

The LED probably does not tell useful info, I think it is toggled by a running kernel or even by userland. But nevertheless, a broken battery isn't unlikely. Hard to tell. When you can take it out, a simple way to make a test is to measure its voltage. A voltage way beyond the specs likely means that the battery is gone forever.

As for the flash - you didn't provide any empiric information I could possibly base a diagnosis on...

[geekmaster]

What hawhill suggested can find a "totally dead" battery.

But in many cases, voltage measurement alone is not enough. A battery needs to be tested under load over a time period to determine its remaining useful capacity. Lithium batteries are considered "dead" when their remaining mAh capacity is too small to be useful. The portion of lithium in the battery that no longer holds a charge is called its "rock content". A battery that is almost entirely "rock" still can show full voltage, but it will disharge quickly under load. Lithium battery chargers usually refuse to even charge a battery that is beyond its useful life, to prevent even a minimal risk of fire

For more info, here is a good place to start:
http://batteryuniversity.com/learn/a..._ion_batteries

According to this, most "portable device" batteries are only good for 300 to 500 charge/discharge cycles:
http://batteryuniversity.com/learn/a...e_battery_life

[qlob]

@felonwan: posting in three places is considered bad manners here. In your other thread, I posted this:

Originally Posted by qlob

I'm actually having the same problem at the moment -- We aren't sure what is causing it. My speculation is that Amazon pushed out an update Over The Air, and that broke something...

What hacks did you have installed, if any? What happened before it bricked? I ask because we are trying to see what may have caused this to happen.

[twobob]

Desperation makes us do crazy things ; )

[qlob]

Originally Posted by twobob

Desperation makes us do crazy things ; )

True... I can barely survive without my Kindle here to have a book to read whenever I have time...

[twobob]

Kindle = Read books. Novel idea, I must try that!

[qlob]

Originally Posted by twobob

Kindle = Read books. Novel idea, I must try that!

Hahaha yeah, I will admit most of my kindle time is eaten up by hacks...

[knc1]

Originally Posted by twobob

Kindle = Read books. Novel idea, I must try that!

That is one I haven't tried yet.
Is it a standard feature of the K3 I used to learn ARM coding on?

[qlob]

Originally Posted by knc1

That is one I haven't tried yet.
Is it a standard feature of the K3 I used to learn ARM coding on?

I think it is. I only found out I could read with mine when I dropped a Git manual on the root of UserStore, and it showed up in the menu of kindlets and kite scripts...

[felonwan]

The voltage is 3.07V after a night stay in the device. So, the battery seems OK.

Using the k3flasher, I can get some info now.

$lsusb
...
Bus 002 Device 005: ID 15a2:0030 Freescale Semiconductor, Inc.
$ sudo ./k3flasher mx35to2_mmc.bin info
I: found suitable device
E: wrong transfer length, wanted to receive 8 bytes but received 4 bytes.
I: above error can be ignored, it's due to the device being in ROM kernel mode
I: RAM kernel should be running now. Trying to re-open device: .
I: got it.
I: read info
E: wrong transfer length, wanted to receive 8 bytes but received 4 bytes.
E: aborting. It is suggested you power-cycle the device.

Now I just charge the battery outside the device. I plan to using k3flasher to programing the kernel and rootfs, hope that will help.

[knc1]

3.0volt, no load, on a 3.7volt Li-polymer battery after a night stay is NOT "OK".
Continue to charge it, but if it has been left discharged for "some time" that your first post seems to imply, it may not recover.

The specific order of the messages quoted indicates that the ROM kernel was not successful in starting the eDDR2 RAM, loading the RAM kernel, and transfering control to it.
I.E: it fell back to continuing to run the ROM kernel (perhaps after a low voltage reset).

If charging the battery in the Kindle - -
The battery management has probably defaulted to the 24hour curve - -
Leave the battery to charge to the 3.7 .. 3.8 volt range - which may take days in this case.

O.T:
There is work in progress to help deal with this situation in another thread here.
Yours is at least the third k3 reported in this condition.

[felonwan]

Originally Posted by knc1

The specific order of the messages quoted indicates that the ROM kernel was not successful in starting the eDDR2 RAM, loading the RAM kernel, and transfering control to it.
I.E: it fell back to continuing to run the ROM kernel (perhaps after a low voltage reset).

Ok, I got it, thanks.

Originally Posted by knc1

If charging the battery in the Kindle - -
The battery management has probably defaulted to the 24hour curve - -
Leave the battery to charge to the 3.7 .. 3.8 volt range - which may take days in this case.

Because the led light always went out in the charging process, I do not trust the charging in the Kindle. The charging using my charger takes about one day and night to "full" charge the battery. But I'm not sure whether the battery is really fully charged.

Originally Posted by knc1

O.T:
There is work in progress to help deal with this situation in another thread here.
Yours is at least the third k3 reported in this condition.

I will check the thread, thanks again.

[knc1]

Originally Posted by felonwan

I will check the thread, thanks again.

Nothing interesting there now - I am the one staring at the code.
I will make a really big announcement if I get the problems worked out.

[knc1]

Originally Posted by felonwan

Because the led light always went out in the charging process, I do not trust the charging in the Kindle. The charging using my charger takes about one day and night to "full" charge the battery. But I'm not sure whether the battery is really fully charged.

My K3 is nine months old and has lead a very sheltered life.
I just put it into "DIAGS" mode and pulled up the factory battery management report screen:

Edit: The original numbers here where taken while the K3 was connected to a bad wall charger.

Charge level: 100%
Terminal voltage: 4.185 volts
Current draw: 73.899 ma. (positive == charge, negative == discharge)
Charger draw limit: On
Charger mode: On
Battery temperature: 80 degrees F
Charge cycle count: 15
Charge learn count: 14
Battery LMD: 1823 mAh
Battery Flags: 0xA4

The above copied at the point where the led went from orange to green.
The charge curve had started out at an 800 ma. rate on the (good) wall charger and has been tapering off starting at about 93% of full charge.
EDIT: The above observation indicated that the charge current limiter had failed. Which lead to the death of the wall charger. When a new wall charger was used, it lived long enough for the USB charge circuitry to self destruct.
The main problem here (other than the hardware failures) is that the control firmware for the charge limit checks for: Off, 100ma, 500ma, Other.
And the author assumed that "Other" would be somewhere in range, not over-range. (The wall charger can supply 850ma but the K3 circuitry is only good for 500ma).

The good: It failed while I was watching it - so I know how and why.
The bad: I now need to return my K3 with its dead wall charger and smoked USB port circuitry.
I guess that is progress.

One of the "to do" items for K3Flasher is to pull this information out of the Kindle using the iROM code.

That way K3Flasher can advise the user if the machine needs to be re-charged before preceding with the flash operation.

But while waiting for all of that to happen, see if you can get your battery re-charged to the 4.0 .. 4.2 volt level.

Heck, it might even "just work" after that (others have).

Note: added the file name: ENABLE_DIAGS on user storage, zero length file is ok.
home -> menu -> settings -> menu -> restart
wait
either t or g on the diagnostics menu