Kindle Paperwhite Died - Can't figure problem

[WaseemAlkurdi]

But i would need to strip the jumpers and solder the cables to the board right?

I'd really prefer using a cable from an old computer case, like the one in my setup. This way, you would have something that you could plug and remove.

Thanks for the details it did help me understand the process better and also added some reassurance now i feel more confident doing the process

From reading that thread, I remembered something: Do avoid wiping /dev/mmcblk0. Why?
U-Boot is supposed to copy itself from RAM to eMMC, and from eMMC to RAM, depending on which one has a missing copy.
But on some Kindle firmware versions, that doesn't work.
So avoid at all wiping /dev/mmcblk0. Keep your eyes open so you don't end up doing it by accident. You have been warned.

[Yorker]

Originally Posted by WaseemAlkurdi

Yeah, it is indeed messy. But look at my setup ... don't solder stuff directly to the Kindle. Solder a cable like I did, then solder stuff to the cable as not disfigure the Kindle itself.

Looking at your setup you still soldered something to the kindle itself, i did the same, only mine didn't have the jumper extension you have.
But yeah i understand that soldering multiple times is bad, of course it's a bit too late it too tiny and i got it burned a bit...

Originally Posted by WaseemAlkurdi

At which point did you check the voltages? At the adapter or at the Kindle?

Both i checked but i was referring to the cable coming from the kindle that i soldered

Originally Posted by WaseemAlkurdi

Why are you expecting that dmesg entry to disappear? dmesg is a log ... plug events don't disappear from there when unplugged. Try with a USB drive and see the entries for yourself.
By the way, the adapter is usually ttyUSB0.

Sorry i forgot it doesn't disappear but it didn't say usb and if i plug the adapter nothing is shown, also after clearing dmesg nothing appears there. does it appear with just the apdater plugged in?

Originally Posted by WaseemAlkurdi

If Tx were to read 1.8 volts, then yeah, but reading 0.8 volts is strange. Unless you weren't reading from the Kindle, or you have something causing an inaccuracy.

I know it's strange but it's from the kindle itself (hopefully i didn't ruin it with my bad soldering)

Originally Posted by WaseemAlkurdi

These are good. I'm not an electrician, but I do remember Ohm's Law being useful for these. Not sure of the exact calculation needed.

So i can check if it relevant or does have to have 1.8 written?

Originally Posted by WaseemAlkurdi

I'd really prefer using a cable from an old computer case, like the one in my setup. This way, you would have something that you could plug and remove.]

Can you elaborate i still fail to see what you mean, i also used a cable but something always has to be soldered at all times no?

Originally Posted by WaseemAlkurdi

From reading that thread, I remembered something: Do avoid wiping /dev/mmcblk0. Why?
U-Boot is supposed to copy itself from RAM to eMMC, and from eMMC to RAM, depending on which one has a missing copy.
But on some Kindle firmware versions, that doesn't work.
So avoid at all wiping /dev/mmcblk0. Keep your eyes open so you don't end up doing it by accident. You have been warned.

I don't actually need to wipe anything i just want it working again so i will only copy the new FW why wipe at all?

[Yorker]

I tried with windows and as mentioned RX works i got GIBBERISH on putty but TX is still dead.
I followed the guide i mentioned earlier and set all the setting correctly, if RX is the receiving side why is it messed up and not in a clear text form?

[WaseemAlkurdi]

Originally Posted by Yorker

Looking at your setup you still soldered something to the kindle itself, i did the same, only mine didn't have the jumper extension you have.
But yeah i understand that soldering multiple times is bad, of course it's a bit too late it too tiny and i got it burned a bit...

Wait, what? Burned the serial pad headers? No wonder you aren't getting a full 1.8 volts. Desolder, clean the pads, measure the voltage to make sure you're getting 1.8 volts. Once you do, you can try again.
If you cooked the pads permanently, don't worry. There are identical pads on the other side of the motherboard, or so people say.

Sorry i forgot it doesn't disappear but it didn't say usb and if i plug the adapter nothing is shown, also after clearing dmesg nothing appears there. does it appear with just the apdater plugged in?

Yeah, ttyUSB0 only shows up when you plug the adapter in, even if the Kindle is not connected to the adapter.

So i can check if it relevant or does have to have 1.8 written?

Yes! It has to have about 1.8 volts (give or take 0.01 volts) to send output over serial. Earlier in this thread, you confirmed that it did read 1.755 volts or something. Why doesn't it read 1.8 volts now? You have to know why, and fix it, before we can get serial output. My money is on dirty or corroded contacts.

Can you elaborate i still fail to see what you mean, i also used a cable but something always has to be soldered at all times no?

You have to have something soldered. A cable or something, because the headers are too tiny, as you have seen.

I don't actually need to wipe anything i just want it working again so i will only copy the new FW why wipe at all?

Good ... but be careful, that's the point

I tried with windows and as mentioned RX works i got GIBBERISH on putty but TX is still dead.
I followed the guide i mentioned earlier and set all the setting correctly, if RX is the receiving side why is it messed up and not in a clear text form?

It's normal. The gibberish is because electricity is going in, but the "translation" by the serial cable results in gibberish because the input isn't "correct" serial protocol. Until you connect it to proper serial input, it's going to do that, and it's nothing out of the ordinary.
As for the Rx side not accepting input, that's also normal until you fix up Tx by feeding it proper serial.

[Yorker]

Originally Posted by WaseemAlkurdi

Wait, what? Burned the serial pad headers? No wonder you aren't getting a full 1.8 volts. Desolder, clean the pads, measure the voltage to make sure you're getting 1.8 volts. Once you do, you can try again.
If you cooked the pads permanently, don't worry. There are identical pads on the other side of the motherboard, or so people say.

I did that, after cleaning as best i could i was still getting 1.765 or so on each pad, i tried soldering again and did a pretty good job but it still didn't work.
First of all each connection only gets 1.7 every now and then so it's annoying to measure unless i plug in the usb cable that make sure it's 1.7 for the next couple of mins for sure.
Logical since i assume it transmits every now and then or only during bootup.
Hooking it up to the adapter yielded the same results as before GND lights up RX seems to work but TX doesn't.
I checked the jumpers this time and got 3.1 and 2.1 on RX and TX.
For some reason i only get TX to light up if i short it with GND or with RX and the gibberish i got was from any place on the board if i touch it with RX including the shields so it's just electrical gibberish at this point.
No matter how i played with it including manually without soldering it didn't work...
The adapter shows up on linux but different from the guide and only 1 connection seems to be showing in tti0, on windows it's not better.

I'm beginning to think the voltage or adapter are the problem as i tried enough things to make sure it's not my fault

[knc1]

That is one confusing post.
Which Rx and which Tx are you speaking of in each instance?
(There are two (2) of each.)

You should see the connections as diagrammed:

Rx  <- Tx
Tx  -> Rx
Gnd <> Gnd

[WaseemAlkurdi]

I did that, after cleaning as best i could i was still getting 1.765 or so on each pad, i tried soldering again and did a pretty good job but it still didn't work.

When I said that you have to get a "full 1.8 volts", I didn't really mean that you should get 1.800 volts. 1.765 is close enough. But 0.8 volts, like what you said a couple of posts ago, isn't. Perhaps you actually meant you were getting 1.8 volts, but mis-typed it as 0.8? Anyhow, you're good with 1.765 volts. I've seen a close reading on my Kindle (it was in the 1.700 range, never quite hitting 1.8) The issue isn't there.

First of all each connection only gets 1.7 every now and then so it's annoying to measure unless i plug in the usb cable that make sure it's 1.7 for the next couple of mins for sure.

Are you saying that it stops giving a signal at some point? That is a sign of trouble. The Kindle is not supposed to totally stop transmitting data via serial while it's turned on. It's either a kernel panic or a power issue ...

Logical since i assume it transmits every now and then or only during bootup.

That is right ... but it should send a "burst" of output every now and then. It shouldn't just die off to the point that you have to plug it in?
Two important questions:
- When you plug it in, does it restart? Not restarting is a good sign.
- Tried to test whether other areas on the Kindle's board are still powered up when the serial output voltage dies? This would tell us whether the whole Kindle has gone off or not.

Hooking it up to the adapter yielded the same results as before GND lights up RX seems to work but TX doesn't.

Is GND securely connected to a screw or an RF shield? It is required for serial output.

I checked the jumpers this time and got 3.1 and 2.1 on RX and TX.

My Kindle is sealed-up at the moment, and the adapter is not with me at the moment, so I can't verify, but I guess that's correct for a 3.3-volt adapter.

For some reason i only get TX to light up if i short it with GND or with RX and the gibberish i got was from any place on the board if i touch it with RX including the shields so it's just electrical gibberish at this point.

100% correct ... you have now observed what I told you a post ago:

It's normal. The gibberish is because electricity is going in, but the "translation" by the serial cable results in gibberish because the input isn't "correct" serial protocol.

No matter how i played with it including manually without soldering it didn't work...

Don't give up. It's bound to work eventually.
Look, once you get 1.765 volts at the header and something close to that at the end of the cable you soldered, tape the cable and don't play with it. You might rip off the headers by accidents.
"Manually without soldering" is how I initially did mine until I went to a shop and had it soldered. You have to press the cable so hard on the pad that your thumb would hurt like crazy and the board would flex ... Definitely not recommended, and you might break the board in two. Just stick to soldering it once and for good, as I said above.

The adapter shows up on linux but different from the guide

It's because your adapter uses a different chipset. If it appears when you plug the adapter in and disappears when you unplug it, then simply don't mind it.

and only 1 connection seems to be showing in tti0, on windows it's not better.

What do you expect to happen otherwise?

I'm beginning to think the voltage or adapter are the problem as i tried enough things to make sure it's not my fault

The voltage at the headers is absolutely fine. Just check whether the end of the cable has approximately same voltage as the headers or not. And it probably will, but please check it.
I'm tending to think that the adapter is to blame ... except that I seen a post (either a "success report" or a tutorial) on this very forum that used the same adapter chipset as you do. If I hadn't seen that post, I would've told you to get an adapter with the recommended FT232 chip.

That leaves out only one real possibility. Perhaps your adapter is broken?
To see, go to a local store, ask for an adapter (whatever you can get ,either like the one you have or one with a FTDI chip), and test it on the spot, before buying. That would tell you whether your adapter is working or not, and if the adapter in the store worked, you could simply buy that.

[Yorker]

Originally Posted by knc1

That is one confusing post.
Which Rx and which Tx are you speaking of in each instance?
(There are two (2) of each.)

You should see the connections as diagrammed:

Rx  <- Tx
Tx  -> Rx
Gnd <> Gnd

Thanks, in that case i messed it up it did each to itself

[Yorker]

Originally Posted by WaseemAlkurdi

When I said that you have to get a "full 1.8 volts", I didn't really mean that you should get 1.800 volts. 1.765 is close enough. But 0.8 volts, like what you said a couple of posts ago, isn't. Perhaps you actually meant you were getting 1.8 volts, but mis-typed it as 0.8? Anyhow, you're good with 1.765 volts. I've seen a close reading on my Kindle (it was in the 1.700 range, never quite hitting 1.8) The issue isn't there.

No, there was no typo :\ it was 1.8 all the way until it showed below 1 as with 0.8

Originally Posted by WaseemAlkurdi

Are you saying that it stops giving a signal at some point? That is a sign of trouble. The Kindle is not supposed to totally stop transmitting data via serial while it's turned on. It's either a kernel panic or a power issue ...

It's not like it goes down and thats it, i saw it go up again it's like in cycles every few mins. i thought it's normal as there is no way it will continue to send the signal after booting up. could it be the battery also?

Originally Posted by WaseemAlkurdi

That is right ... but it should send a "burst" of output every now and then. It shouldn't just die off to the point that you have to plug it in?
Two important questions:
- When you plug it in, does it restart? Not restarting is a good sign.
- Tried to test whether other areas on the Kindle's board are still powered up when the serial output voltage dies? This would tell us whether the whole Kindle has gone off or not.

It doesn't die completely it cycles eventually.
1. No idea if it restarts i got no screen on it
2. i did test and the power is on for other parts

Originally Posted by WaseemAlkurdi

Is GND securely connected to a screw or an RF shield? It is required for serial output.

it is, i unscrewed the screw and then screwed it on the cable so it's really tight

Originally Posted by WaseemAlkurdi

My Kindle is sealed-up at the moment, and the adapter is not with me at the moment, so I can't verify, but I guess that's correct for a 3.3-volt adapter.

That's ok i ordered all the various parts just in case so if this won't work there are several back up plans.

Originally Posted by WaseemAlkurdi

100% correct ... you have now observed what I told you a post ago:

Don't give up. It's bound to work eventually.
Look, once you get 1.765 volts at the header and something close to that at the end of the cable you soldered, tape the cable and don't play with it. You might rip off the headers by accidents.
"Manually without soldering" is how I initially did mine until I went to a shop and had it soldered. You have to press the cable so hard on the pad that your thumb would hurt like crazy and the board would flex ... Definitely not recommended, and you might break the board in two. Just stick to soldering it once and for good, as I said above.

This is exactly what i did since the connections are tiny! , there is no other options it rips so easily, the manual method is just to check it works before putting the effort into soldering it (which takes some time due to poor skills and tiny pads)

Originally Posted by WaseemAlkurdi

It's because your adapter uses a different chipset. If it appears when you plug the adapter in and disappears when you unplug it, then simply don't mind it.

Ok

Originally Posted by WaseemAlkurdi

What do you expect to happen otherwise?

I thought it either works completely or not at all but in general since i did all the connections perfectly why won't it show the others?

Originally Posted by WaseemAlkurdi

The voltage at the headers is absolutely fine. Just check whether the end of the cable has approximately same voltage as the headers or not. And it probably will, but please check it.
I'm tending to think that the adapter is to blame ... except that I seen a post (either a "success report" or a tutorial) on this very forum that used the same adapter chipset as you do. If I hadn't seen that post, I would've told you to get an adapter with the recommended FT232 chip.

That leaves out only one real possibility. Perhaps your adapter is broken?
To see, go to a local store, ask for an adapter (whatever you can get ,either like the one you have or one with a FTDI chip), and test it on the spot, before buying. That would tell you whether your adapter is working or not, and if the adapter in the store worked, you could simply buy that.

The shops here are pretty bad and if they have the parts it's super expensive but i did order all the other options including your adapter from ebay just 2 weeks to go or so.

based on the comment from KNC1 i connected it incorrectly so i wanna try to mess with that next.

[WaseemAlkurdi]

Originally Posted by Yorker

It's not like it goes down and thats it, i saw it go up again it's like in cycles every few mins. i thought it's normal as there is no way it will continue to send the signal after booting up. could it be the battery also?

No, it should still send signals even after it boots. Each time a line is written to dmesg, a matching spike in voltage happens as that line is sent over serial.
What happens after it finishes booting is that the number of lines that are written to dmesg decreases dramatically, and so do the voltage spikes.

Originally Posted by Yorker

It doesn't die completely it cycles eventually.
1. No idea if it restarts i got no screen on it
2. i did test and the power is on for other parts

That's a good sign.

Originally Posted by Yorker

it is, i unscrewed the screw and then screwed it on the cable so it's really tight

Good.

Originally Posted by Yorker

That's ok i ordered all the various parts just in case so if this won't work there are several back up plans.

I hope that includes the 1.8v adapter as well. It's better to use an adapter that runs at 1.8 volts than to "fix" a 3.3 volt adapter to run at 1.8 volts.

Originally Posted by Yorker

I thought it either works completely or not at all ...

Not really.

but in general since i did all the connections perfectly why won't it show the others?

That's the part I don't understand. What "others"? Do you mean the lights for the Rx and Tx terminals? It's because you apparently wired them incorrectly.
Anyhow, just a small quote to help you remember the terminals (I don't remember the exact wording, but the general idea is there):

In order to hear somebody talking, your ears (Rx) have to be close to their mouth (Tx).

Originally Posted by Yorker

The shops here are pretty bad and if they have the parts it's super expensive but i did order all the other options including your adapter from ebay just 2 weeks to go or so.

Good!

Originally Posted by Yorker

based on the comment from KNC1 i connected it incorrectly so i wanna try to mess with that next.

Try using your current adapter and let us know. I have a feeling that it is going to work. You now have the correct voltage and the correct wiring, and barring an issue with the adapter, it's supposed to work.

[Yorker]

I re-soldered again
Same issue same situation
I did make a mistake when i said i connected TX and RX to themselves i meant i followed the picture in the guide so it was correct, plus honestly with all the attempts it was reveres as well so if that was the problem it would have worked.
the voltage start at 1.7 and raises to 2.1 and 3.1 on RX and TX so no idea if that's the problem.
I tried in both Windows and Linux the only other update is that i had to add the driver in linux to make the adapter working so now i have this:

[ 0.625977] 00:05: ttyS0 at I/O 0x3f8 (irq = 4, base_baud = 115200) is a 16550A
[ 4.722992] usb 5-1.1: cp210x converter now attached to ttyUSB0


Not sure if ttyS0 means that only RX or TX is working while the other isn't, and maybe it is related to the 3.1 voltage since 2.1 is kinda close to 1.8 but 3 is too much.

I did order the other adapter the 1.8 tty cable and the resistor, so i guess we'll wait and see..

[Yorker]

Originally Posted by WaseemAlkurdi

No, it should still send signals even after it boots. Each time a line is written to dmesg, a matching spike in voltage happens as that line is sent over serial.
What happens after it finishes booting is that the number of lines that are written to dmesg decreases dramatically, and so do the voltage spikes.


That's a good sign.


Good.


I hope that includes the 1.8v adapter as well. It's better to use an adapter that runs at 1.8 volts than to "fix" a 3.3 volt adapter to run at 1.8 volts.


Not really.

That's the part I don't understand. What "others"? Do you mean the lights for the Rx and Tx terminals? It's because you apparently wired them incorrectly.
Anyhow, just a small quote to help you remember the terminals (I don't remember the exact wording, but the general idea is there):



Good!



Try using your current adapter and let us know. I have a feeling that it is going to work. You now have the correct voltage and the correct wiring, and barring an issue with the adapter, it's supposed to work.

I got the parts but i couldn't get it to work :\
Not with the USB-TTL and not with the FT232RL Adapter and not with the one i had to begin with.
its all the same i get the tty0 connection just from ground but RX or TX add nothing but they had voltage, i tried swapping them i tried the FT232RL adapter with the 3.3v jumper no luck at all.
i'm going crazy questioning either my soldering skills but i'm so sure its correct or the device itself

Any ideas?

[WaseemAlkurdi]

Originally Posted by Yorker

I got the parts but i couldn't get it to work :\

Not with the USB-TTL and not with the FT232RL Adapter and not with the one i had to begin with.

its all the same i get the tty0 connection just from ground but RX or TX add nothing but they had voltage, i tried swapping them i tried the FT232RL adapter with the 3.3v jumper no luck at all.

i'm going crazy questioning either my soldering skills but i'm so sure its correct or the device itself



Any ideas?

There's definitely something wrong with either your soldering or your adapter.
If you have 1.8 volts on the terminals then you should definitely have output on the console, because you ONLY get that voltage when you have output on the console (no output = no voltage).

[Yorker]

Originally Posted by WaseemAlkurdi

There's definitely something wrong with either your soldering or your adapter.
If you have 1.8 volts on the terminals then you should definitely have output on the console, because you ONLY get that voltage when you have output on the console (no output = no voltage).

Ok i"ll double check again but just to be sure all i need is RX TX and GND right?
all the other v3 v5 etc voltage connectors and cables are unnecessary
Also could it be a dead kindle somehow?

[knc1]

Originally Posted by Yorker

Ok i"ll double check again but just to be sure all i need is RX TX and GND right?
all the other v3 v5 etc voltage connectors and cables are unnecessary
Also could it be a dead kindle somehow?

Yes, No, Maybe.
Some adapters also require a reference voltage.
Photograph what you have so we can see what you have set up.
Written descriptions are just not going to cut it, in this case.