Update eink waveform

[Supra James]

Interesting stuff, thank you. After doing a bit of troubleshooting on the command line I eventually worked out that the problem I have is that there was no chip on the ribbon cable at all - the eink driver was trying to load the waveform from eeprom, failing due to no eeprom and loading a "default" waveform.

Ripped open the Kindle again and saw the distinct lack of chip on the ribbon cable. Contacted the seller and they've admitted sending me the wrong screen and have sent a replacement, complete with eeprom.

I am hoping for a plug-and-play transplant next time. It's weird that on some of your screens you have to go through this process of extracting the chip data from a similar screen. It's my understanding that each screen or batch of screens can be subtly different, and the waveform data for that particular screen is written to EEPROM at the factory to be read by the device at boot, so the device knows how to address that particular screen.

A little bit of evidence I found to support the above - when I tried your procedure of booting up with the old screen attached and then hot-swapping to the new screen, I found the colours had returned to normal, but text was a little darker than normal, and menu ghosting was VERY apparent, things just generally looked uglier. I am pretty sure that is due to the waveform of the old screen not being quite right for the new one.

I'll report back when the new screen eventually makes it's way over from China

[sert]

Thank you so much tccsargent I never thought I could solve this problem. I can't believe I have finally solved the problem with images in my kindle 3 after replacing the broken screen (yes I typed it because I want anyone who uses google to find this thread )

Again THANK YOU

[swansonator]

Originally Posted by tccsargent

This problem of the bad waveform almost exclusively lies with the 3G models of the Kindle Keyboard (and also with the 3g Kindle Touches). Here is a practical solution to your problem which, although not perfect, works most of the time: (1) Save the old, broken screen that you have removed from a 3G model of K3. (2) While the Kindle is still open, attach the new screen to it and navigate to "Reset to Factory Defaults" in the system submenu of the Kindle, and press Enter. (3) Press the arrow to navigate to "OK" of the Cancel-OK prompt which appears in the popup window for Reset to Factory Defaults, but do not press Enter yet. (4) Remove the new screen and re-connect the original broken screen; you will now be running blind, but you know that the Kindle is poised to reset with the pressing of the enter key. After the old screen is re-attached, go ahead and press Enter. (5) The Kindle will now reset to factory defaults, (hopefully) accessing the eprom chip on the ribbon cable of the old broken screen and loading the good waveform that resides there. Since you cannot watch the progress of this reset, give the Kindle at least FIVE minutes to complete the process before unplugging the old screen and re-plugging the new screen. Then slide the power bar and take a look: your 'wireframe' pictures should now be restored to solid rich grayscale!

I have repaired hundreds of Kindles and followed this procedure with around a 90% success rate. Unfortunately I have had some units (all 3G) which persistently reset to a bad, or missing, waveform, but the process has worked most of the time. Due to those failures, however, I have built up an inventory of around 50 broken 3G screens so that I can keep trying to find one to boot to which the subject Kindle will successfully read.

By the way, you should follow this procedure only after upgrading to 3.4 OS, since a later system upgrade (with a replacement screen installed) can again lead to a 'bad waveform' problem.


Good luck!

This totally worked for me as well! Thanks so much!

[englishwithcraig]

Thank you very much.

Send me your email address, I'd like to buy you a drink via Paypal.
I am extremely grateful for your helpful post, it fixed my Kindle after I replaced the screen.

The screen was looking like it was missing information, or was of a lower resolution, when I followed your guide it worked perfectly.

I'm so thankful.

Many thanks

Craig

[flameheadrod]

Originally Posted by tccsargent

This problem of the bad waveform almost exclusively lies with the 3G models of the Kindle Keyboard (and also with the 3g Kindle Touches). Here is a practical solution to your problem which, although not perfect, works most of the time: (1) Save the old, broken screen that you have removed from a 3G model of K3. (2) While the Kindle is still open, attach the new screen to it and navigate to "Reset to Factory Defaults" in the system submenu of the Kindle, and press Enter. (3) Press the arrow to navigate to "OK" of the Cancel-OK prompt which appears in the popup window for Reset to Factory Defaults, but do not press Enter yet. (4) Remove the new screen and re-connect the original broken screen; you will now be running blind, but you know that the Kindle is poised to reset with the pressing of the enter key. After the old screen is re-attached, go ahead and press Enter. (5) The Kindle will now reset to factory defaults, (hopefully) accessing the eprom chip on the ribbon cable of the old broken screen and loading the good waveform that resides there. Since you cannot watch the progress of this reset, give the Kindle at least FIVE minutes to complete the process before unplugging the old screen and re-plugging the new screen. Then slide the power bar and take a look: your 'wireframe' pictures should now be restored to solid rich grayscale!

I have repaired hundreds of Kindles and followed this procedure with around a 90% success rate. Unfortunately I have had some units (all 3G) which persistently reset to a bad, or missing, waveform, but the process has worked most of the time. Due to those failures, however, I have built up an inventory of around 50 broken 3G screens so that I can keep trying to find one to boot to which the subject Kindle will successfully read.

By the way, you should follow this procedure only after upgrading to 3.4 OS, since a later system upgrade (with a replacement screen installed) can again lead to a 'bad waveform' problem.


Good luck!

Thank you!!!
I broke my Kindle 3 Keyboard (no 3G) screen and I order one from ebay and I got it and it worked except the grayscale wouldn't really render. It was super messy and sometimes the words would mess up.
I followed these instructions and WALA! it worked!!
Thanks!!!!!!

[Nodrog]

I am so glad to have found and read this thread - I replaced my screen and had exactly this problem (of greyed out pdfs/images/screen savers) but using this reset method worked for me!

Nodrog

[zombic]

It work, but after restart Kindle the problem returns.

PS. Kindle firmware v 3.4

[HarryT]

Originally Posted by zombic

It work, but after restart Kindle the problem returns.

PS. Kindle firmware v 3.4

That is entirely to be expected, but why would you need to do a restart? The only time I ever restart my Kindle is after a firmware update.

[zombic]

Originally Posted by HarryT

That is entirely to be expected, but why would you need to do a restart?

I just checked

[Ian.M]

Hopefully the setting might be kept somewhere in the system partition. Immediately after a restart with the cable with the good waveform EEPROM connected, dumping the partition to an image file on a PC, then repeating the process after a restart with the new screen then a binary compare of the two image files might give sufficient clues to locate what is changing or which waveform file is being read so a software fix can be developed.

If anyone has tried this previously, what was the conclusion?

Otherwise only Uber-Geeks with serial EEPROM programmers or SMD rework facilities are able to make a permanent repair. If you are doing more than just reading pre-packaged ebooks, you *DO* need to be able to restart when the system becomes unstable, and I certainly wouldn't want to open any of my Kindles every month.

[HarryT]

Unfortunately I don't think anyone ever suggested that this was a permanent solution. Unless you have a way to reprogram the EEPROM on the display, there's nothing you can do.

[Supra James]

Originally Posted by zombic

It work, but after restart Kindle the problem returns.

PS. Kindle firmware v 3.4

It took me a while to realise that this was an "acceptable" solution to the rest of the guys

It's a bodge, but it's ludricous to expect that your hardware will NEVER EVER get restarted again.

I've just fixed my mum's kindle this way, and not really looking forward to telling her she must NEVER let the battery run out completely...

I thought for a long time that I was doing something wrong, or the new screen I had wasn't suitable, but nah - this won't ever persist through a restart, sadly.

[dsmid]

BTW Ian's idea sounds reasonable.
A little experiment on affected Kindle could solve this problem for good.
Creating a hack that would write the waveform data on every boot is the easier task (I can take care of that).

[dsmid]

I've done a little research and found this:

• Various waveforms are stored in /opt/eink/images/ in *.wbf files.
• Helper links to these files (named after waveform versions) are in /var/local/eink/.
• If file /var/local/eink/dont_update is present, the waveform won't be flashed. Ever.
• If file /var/local/eink/update_flash is present, the waveform will be flashed and then the file will be removed.

So all it takes to stop flashing the waveform is to create file /var/local/eink/dont_update . Easy peasy.
I can create a launchpad .ini file if someone's interested.

[Nephiel]

Hi there. I just managed to copy the waveform from a broken panel into a new one!
This thread has provided me with very valuable info so I thought I'd share the process here.
Full picture album here.

A friend gave me a Kindle 3 (US 3G model), fully functional except for a broken screen. I ordered a new screen panel online. After replacing it, I had the same problem as the OP, grayscale images not displayed properly:



Broken panel model and waveform chip:

ED060SC7(LF)H2 2.02
Macronix International MX25L2005ZNI-12G

New panel model and waveform chip:

ED060SC7(LF)C1 1.54
Winbond W25X20

So, in my case both panels had a 256KB CMOS serial flash chip on the ribbon to store the waveform.

Firmware version was 3.1, I updated to 3.3 and then to 3.4 as per Amazon's instructions, just to see if this problem was fixed in later versions. (Nope).

Booting the Kindle with the broken panel and hot-swapping it for the new one works, but lasts only until the next restart. So far, nothing new.

The Kindle 3 reads the waveform from the flash chip on every boot, and if it can't (when there is no chip, or when the chip does not contain the proper waveform data), it uses a default waveform, which was not quite right for my new display, so grayscales didn't show as they should. There was evidence of this in the logs right after booting with the new panel:

140206:191308 eink_fb: I ipu_controller_type:def:Detected broadsheet controller
140206:191308 eink_fb: E bs_panel_data_valid:def:Unrecognized values in panel data
140206:191308 eink_fb: E broadsheet_waveform_valid:def:Unrecognized values in waveform header
140206:191308 eink_fb: W bs_get_isis_waveform:def:using built-in waveform as panel-specific waveform couldn't be found
140206:191308 eink_fb: E bs_panel_data_valid:def:Unrecognized values in panel data

Fortunately, the waveform from my old panel seemed to be good enough for the new one - grayscales were fine and text edges were smooth. So I looked for a way to copy the old waveform into the new panel. First, where to find the old waveform?

I poked around the filesystem, dumped it after booting with both panels and compared both dumps, as suggested here. Nothing changed below the first partition (sectors 0-7688). Nothing changed on the first partition (root) either. The second (/var/local) did change, but that was to be expected since it stores log files.

Found the .wbf files dsmid mentioned in /var/local/eink, but those are links to files on the first partition, which did not change:

[root@kindle root]# cd /var/local/
[root@kindle local]# ls -lah
drwxr-xr-x    9 root     root         1.0k Feb  6 02:57 .
drwxr-xr-x    9 root     root          180 Feb  6 12:02 ..
drwxr-xr-x    2 root     root         1.0k Feb  6 02:57 audio
drwx------    2 root     root         1.0k Feb  6 02:57 browser
drwxr-xr-x    2 root     root         1.0k Feb  6 12:03 eink
drwxr-xr-x    4 root     root         1.0k Feb  6 02:58 java
drwxr-xr-x    2 root     root         1.0k Feb  6 12:30 log
drwxr-xr-x    2 root     root         1.0k Feb  6 12:04 system
drwxr-xr-x    2 root     root         1.0k Feb  6 02:58 wan
[root@kindle local]# cd eink
[root@kindle eink]# ls -lah
drwxr-xr-x    2 root     root         1.0k Feb  6 12:03 .
drwxr-xr-x    9 root     root         1.0k Feb  6 02:57 ..
lrwxrwxrwx    1 root     root           44 Feb  6 12:03 V110_044_60_M06 -> /opt/eink/images/V110_B044_60_TE2703_BTC.wbf
lrwxrwxrwx    1 root     root           53 Feb  6 12:03 V110_059_60_M06 -> /opt/eink/images/V110_B059_60_WJ0105_ED060SC5_BTC.wbf
lrwxrwxrwx    1 root     root           53 Feb  6 12:03 V220_004_60_M12 -> /opt/eink/images/V220_C004_60_WJ4405_ED060SC7_BTC.wbf
lrwxrwxrwx    1 root     root           53 Feb  6 12:03 V220_008_60_M12 -> /opt/eink/images/V220_C008_60_WJ0607_ED060SC7_BTC.wbf
lrwxrwxrwx    1 root     root           55 Feb  6 12:03 V220_008_60_M14 -> /opt/eink/images/V220_C008_60_WJ0705_ED060SC7H2_BTC.wbf
lrwxrwxrwx    1 root     root           55 Feb  6 12:03 V220_008_60_M24 -> /opt/eink/images/V220_C008_60_WJ0804_ED060SC7C1_BTC.wbf
-rw-r--r--    1 root     root       236.0k Feb  6 02:57 usid
[root@kindle eink]# ls -lah /opt/eink/images/
drwxr-xr-x    2 root     root         1.0k Jul 31  2010 .
drwxr-xr-x    4 root     root         1.0k Mar 10  2011 ..
-rw-r--r--    1 root     root        46.3k Jul 31  2010 V110_B044_60_TE2703_BTC.wbf
-rw-r--r--    1 root     root        56.8k Jul 31  2010 V110_B059_60_WJ0105_ED060SC5_BTC.wbf
-rw-r--r--    1 root     root        79.4k Jul 31  2010 V220_C004_60_WJ4405_ED060SC7_BTC.wbf
-rw-r--r--    1 root     root        87.9k Jul 31  2010 V220_C008_60_WJ0607_ED060SC7_BTC.wbf
-rw-r--r--    1 root     root        94.4k Jul 31  2010 V220_C008_60_WJ0705_ED060SC7H2_BTC.wbf
-rw-r--r--    1 root     root        86.9k Jul 31  2010 V220_C008_60_WJ0804_ED060SC7C1_BTC.wbf
-rw-r--r--    1 root     root         2.1k Jul 31  2010 cmd0047_c0_fi260_fo5_d0100_s00_cd0b00.bin
-rw-r--r--    1 root     root        24.0k Jul 31  2010 factory_reset_n.png
-rw-r--r--    1 root     root        22.3k Jul 31  2010 factory_reset_s.png
-rw-r--r--    1 root     root        16.9k Jul 31  2010 factory_reset_t.png
-rw-r--r--    1 root     root          138 Jul 31  2010 rectangle.gif
[root@kindle eink]#

I booted the Kindle into diagnostic mode (placing an empty file named ENABLE_DIAGS on the user memory of the Kindle and rebooting) with the new panel attached.



Pressed N for Misc Diagnostics, then V for Video. The new panel couldn't be identified.



I rebooted with the old panel, hot-swapped it with the new one and ran Video diagnostics again. This time it showed the ID of the old panel and passed all tests... until next reboot.



I figured I'd try those .wbf files from /opt/eink/images. After booting with the new panel, pressed S for Settings:



Then pressed T to update the waveform. However, none of the .wbf files I tried matched the panel ID (which was blank), so none worked. It turns out the flash chip on the panel also contains the panel ID, and the Kindle won't accept any waveform unless it somehow matches the panel ID. So back to square one.

My last resort was to read the flash chip directly. I knew it could be done (kudos to Mike Harrison). I had an Arduino Uno microcontroller board on hand, and some Logic Level Converters from Sparkfun. I built a flash reader, fashioned a crude jig to connect it to the chip, and used flashrom on my linux laptop to read the chip from the old panel.

nephiel@laptop$ flashrom -p serprog:dev=/dev/ttyACM0:2000000 -Vr oldpanel.rom
flashrom v0.9.7-r1711 on Linux 3.2.0-58-generic (i686)
flashrom is free software, get the source code at http://www.flashrom.org

flashrom was built with libpci 3.1.8, GCC 4.6.3, little endian
Command line (4 args): flashrom -p serprog:dev=/dev/ttyACM0:2000000 -Vr oldpanel.rom
Calibrating delay loop... OS timer resolution is 3 usecs, 550M loops per second, 10 myus = 12 us, 100 myus = 120 us, 1000 myus = 1058 us, 10000 myus = 9982 us, 12 myus = 18 us, OK.
Initializing serprog programmer
Baud rate is 2000000.
serprog: connected - attempting to synchronize
...
serprog: Synchronized
serprog: Interface version ok.
serprog: Bus support: parallel=off, LPC=off, FWH=off, SPI=on
Warning: Automatic command availability check failed for cmd 0x08 - won't execute cmd
Warning: Automatic command availability check failed for cmd 0x11 - won't execute cmd
serprog: Programmer name is "serprog-duino"
serprog: Serial buffer size is 65535
serprog: Warning: Programmer does not support toggling its output drivers
The following protocols are supported: SPI.
[...]
Probing for Macronix MX25L2005(C), 256 kB: probe_spi_rdid_generic: id1 0xc2, id2 0x2012
Found Macronix flash chip "MX25L2005(C)" (256 kB, SPI) on serprog.
Chip status register is 0x00.
Chip status register: Status Register Write Disable (SRWD, SRP, ...) is not set
Chip status register: Bit 6 is not set
Chip status register: Bit 5 is not set
Chip status register: Bit 4 is not set
Chip status register: Block Protect 1 (BP1) is not set
Chip status register: Block Protect 0 (BP0) is not set
Chip status register: Write Enable Latch (WEL) is not set
Chip status register: Write In Progress (WIP/BUSY) is not set
[...]
Found Macronix flash chip "MX25L2005(C)" (256 kB, SPI).
===
This flash part has status UNTESTED for operations: PROBE READ ERASE WRITE
The test status of this chip may have been updated in the latest development
version of flashrom. If you are running the latest development version,
please email a report to flashrom@flashrom.org if any of the above operations
work correctly for you with this flash part. Please include the flashrom
output with the additional -V option for all operations you tested (-V, -Vr,
-VE, -Vw), and mention which mainboard or programmer you tested.
Please mention your board in the subject line. Thanks for your help!
Reading flash... done.

This produced a 256KB rom file. Comparing it to the one from Mike Harrison, it was not identical, but the data structure was very similar. The chip from the new panel, however, produced a 256KB file with all bits set to 1, meaning the chip was empty. Blank. No wonder the Kindle couldn't read the waveform nor the panel ID.

All that was left was to cross my fingers and write the rom file from the old panel into the new...

nephiel@laptop$ flashrom -p serprog:dev=/dev/ttyACM0:2000000 -Vw oldpanel.rom
flashrom v0.9.7-r1711 on Linux 3.2.0-58-generic (i686)
flashrom is free software, get the source code at http://www.flashrom.org

flashrom was built with libpci 3.1.8, GCC 4.6.3, little endian
Command line (4 args): flashrom -p serprog:dev=/dev/ttyACM0:2000000 -Vw oldpanel.rom
Calibrating delay loop... OS timer resolution is 2 usecs, 514M loops per second, 10 myus = 16 us, 100 myus = 132 us, 1000 myus = 1316 us, 10000 myus = 13060 us, 8 myus = 12 us, OK.
Initializing serprog programmer
Baud rate is 2000000.
serprog: connected - attempting to synchronize
...
serprog: Synchronized
serprog: Interface version ok.
serprog: Bus support: parallel=off, LPC=off, FWH=off, SPI=on
Warning: Automatic command availability check failed for cmd 0x08 - won't execute cmd
Warning: Automatic command availability check failed for cmd 0x11 - won't execute cmd
serprog: Programmer name is "serprog-duino"
serprog: Serial buffer size is 65535
serprog: Warning: Programmer does not support toggling its output drivers
The following protocols are supported: SPI.
[...]
Probing for Winbond W25X20, 256 kB: probe_spi_rdid_generic: id1 0xef, id2 0x3012
Found Winbond flash chip "W25X20" (256 kB, SPI) on serprog.
Chip status register is 0x00.
[...]
Found Winbond flash chip "W25X20" (256 kB, SPI).
Reading old flash chip contents... done.
Erasing and writing flash chip... Trying erase function 0... 0x000000-0x000fff:W, 0x001000-0x001fff:W, 0x002000-0x002fff:W, 0x003000-0x003fff:W, 0x004000-0x004fff:W, 0x005000-0x005fff:W, 0x006000-0x006fff:W, 0x007000-0x007fff:W, 0x008000-0x008fff:W, 0x009000-0x009fff:W, 0x00a000-0x00afff:W, 0x00b000-0x00bfff:W, 0x00c000-0x00cfff:W, 0x00d000-0x00dfff:W, 0x00e000-0x00efff:W, 0x00f000-0x00ffff:W, 0x010000-0x010fff:W, 0x011000-0x011fff:W, 0x012000-0x012fff:W, 0x013000-0x013fff:W, 0x014000-0x014fff:W, 0x015000-0x015fff:W, 0x016000-0x016fff:W, 0x017000-0x017fff:W, 0x018000-0x018fff:W, 0x019000-0x019fff:W, 0x01a000-0x01afff:W, 0x01b000-0x01bfff:W, 0x01c000-0x01cfff:W, 0x01d000-0x01dfff:W, 0x01e000-0x01efff:W, 0x01f000-0x01ffff:W, 0x020000-0x020fff:W, 0x021000-0x021fff:W, 0x022000-0x022fff:W, 0x023000-0x023fff:W, 0x024000-0x024fff:W, 0x025000-0x025fff:W, 0x026000-0x026fff:W, 0x027000-0x027fff:W, 0x028000-0x028fff:W, 0x029000-0x029fff:W, 0x02a000-0x02afff:W, 0x02b000-0x02bfff:W, 0x02c000-0x02cfff:W, 0x02d000-0x02dfff:W, 0x02e000-0x02efff:W, 0x02f000-0x02ffff:W, 0x030000-0x030fff:W, 0x031000-0x031fff:S, 0x032000-0x032fff:S, 0x033000-0x033fff:S, 0x034000-0x034fff:S, 0x035000-0x035fff:S, 0x036000-0x036fff:S, 0x037000-0x037fff:S, 0x038000-0x038fff:S, 0x039000-0x039fff:S, 0x03a000-0x03afff:S, 0x03b000-0x03bfff:S, 0x03c000-0x03cfff:S, 0x03d000-0x03dfff:S, 0x03e000-0x03efff:S, 0x03f000-0x03ffff:S
Erase/write done.
Verifying flash... FAILED at 0x000144e4! Expected=0xaa, Found=0x56, failed byte count from 0x00000000-0x0003ffff: 0x3
Your flash chip is in an unknown state.
Please report this on IRC at chat.freenode.net (channel #flashrom) or
mail flashrom@flashrom.org, thanks!

...the first write attempt failed verification, but a second try...

[...]
Found Winbond flash chip "W25X20" (256 kB, SPI).
Reading old flash chip contents... done.
Erasing and writing flash chip... Trying erase function 0... 0x000000-0x000fff:S, 0x001000-0x001fff:S, 0x002000-0x002fff:S, 0x003000-0x003fff:S, 0x004000-0x004fff:S, 0x005000-0x005fff:S, 0x006000-0x006fff:S, 0x007000-0x007fff:S, 0x008000-0x008fff:S, 0x009000-0x009fff:S, 0x00a000-0x00afff:S, 0x00b000-0x00bfff:S, 0x00c000-0x00cfff:S, 0x00d000-0x00dfff:S, 0x00e000-0x00efff:S, 0x00f000-0x00ffff:S, 0x010000-0x010fff:S, 0x011000-0x011fff:S, 0x012000-0x012fff:S, 0x013000-0x013fff:S, 0x014000-0x014fff:S, 0x015000-0x015fff:S, 0x016000-0x016fff:S, 0x017000-0x017fff:S, 0x018000-0x018fff:S, 0x019000-0x019fff:S, 0x01a000-0x01afff:S, 0x01b000-0x01bfff:S, 0x01c000-0x01cfff:S, 0x01d000-0x01dfff:S, 0x01e000-0x01efff:S, 0x01f000-0x01ffff:S, 0x020000-0x020fff:S, 0x021000-0x021fff:S, 0x022000-0x022fff:S, 0x023000-0x023fff:S, 0x024000-0x024fff:S, 0x025000-0x025fff:S, 0x026000-0x026fff:S, 0x027000-0x027fff:S, 0x028000-0x028fff:S, 0x029000-0x029fff:S, 0x02a000-0x02afff:S, 0x02b000-0x02bfff:S, 0x02c000-0x02cfff:S, 0x02d000-0x02dfff:S, 0x02e000-0x02efff:S, 0x02f000-0x02ffff:S, 0x030000-0x030fff:S, 0x031000-0x031fff:S, 0x032000-0x032fff:S, 0x033000-0x033fff:S, 0x034000-0x034fff:S, 0x035000-0x035fff:S, 0x036000-0x036fff:S, 0x037000-0x037fff:S, 0x038000-0x038fff:S, 0x039000-0x039fff:S, 0x03a000-0x03afff:S, 0x03b000-0x03bfff:S, 0x03c000-0x03cfff:S, 0x03d000-0x03dfff:S, 0x03e000-0x03efff:S, 0x03f000-0x03ffff:S

Warning: Chip content is identical to the requested image.
Erase/write done.

Success! Diagnostics show ID and model from the old panel, which isn't quite right, but it works, and grayscales now survive reboots