***EVGA Precision X and Windows 7/8/8.1 and especially 10 bricking systems***

If the vBIOS was working for some time and now it is suddenly not working, you probably have a bad GPU. If the problems all began at the time of flashing, it may be a bad flash. I just helped someone today with a failed flash on a P870DM-G with 200W 980 GPU and he is back in business. Before I helped him he had a black screen with no backlight. I walked him through the flashing process blindly.

Prema has instructions on how to handle this that are very similar to what I recommended.
@Prema has that posted here: [HOW-TO] Boot & blind-flash a (non-booting) GPU in Windows

Below is what I suggested that worked in his situation.

Take the vBIOS mod package and extract the contents to a USB thumb drive. Make sure the files from the archive are in the root directory. You do not want them in a folder on the USB stick. The USB stick does not need to be DOS bootable. Rename the Modme.bat file to 1.bat.

1. Boot up and log into Windows blindly. You may be able to tell by listening to sounds or the keyboard will change color and the number lock light may come on once you are on the desktop
2. Type Win+R, followed by CMD, then hit enter key (Windows logo key and R at the same time will open a Run dialog box, CMD followed by enter will open a command prompt)
3. You are going to need to guess at the drive letter assigned to the USB stick. If D: is the highest drive letter assigned to internal disks, the USB should be E: (one drive letter higher)
4. Assuming the USB is drive E: type E: and hit enter, then you may briefly see the activity light on the USB stick shows signs of read activity. Type 1 (number one), hit the enter key and wait about 5 seconds
5. Type Y and hit enter, then you should hear a loud beep from NVFLASH… if so, continue to wait and when the flash is finished you may hear an audible chirp
6. If you never hear anything at all, try F: as the drive letter and see if there is any USB activity light action, then repeat the above step
7. If you hear a loud beep after pressing Y you know the flash process has started… wait up to 3 minutes and if you hear no indication the flash has finished go ahead and reboot

If the process was a success, you will see text on your screen as the system goes through the BIOS POST process and the LCD backlight will be working.

Disabling the GPU in Device Manager before flashing (or using Safe Mode for flashing) and manually opening an Admin command prompt in the folder where the NVFLASH contents are located, then typing the Modme.bat command, rather double-clicking the batch file, seems to work better for me. Sometimes, especially with Windows 8.X and 10, double-clicking the batch file does not correctly handle the Admin rights and the flash process can get stuck. Having Windows UAC crap enabled may also contribute to complications.

 

I will give this a shot! All my problems began after ticking the "startup with windows" option on Evga precision. That subsequent reboot yielded 8 beeps. Before installing evga i had a few reboots post-flash that were successful.

Currently I can play games just fine with an external, would you recommend disconnecting the external and going through blind? Or does the external have any affect on the flashing?

My M15x is also suffering from this issue. Not from EVGA/Win 10 though. Granted I can hear the inverter boards on the 15's LCD buzzing. That ol' boy may actually have a fuse or those LCDs may be busted. Question, though, my M15x will display POST even on an external; Is there a setting to allow the 18 to do that?

 

I don't know. You've got a lot going on there. If the 18 is having 8-beep problems it could be a bricked GPU. Since you had the LVDS cable and LCD replaced, unless one of those parts are defective, the next logical culprit would be GPU. If it is a dead GPU, it should POST from Intel HD Graphics with the MXM modules removed.

Not sure about the M15x issues, but it doesn't sound like the same problem.

 

Think it's possible that when the EDID was corrupted, Win10/EVGA stuck it's fingers in the vBios too? Then again, the 18's LCD on the M15x didn't display anything either (prior to M15x faulting out too). My current theory is that the M15x is suffering from power / inverter board issues and the 18 has something silly going on in the gpu or an IC (as odd as it sounds to have an IC prevent a single signal out 30 ish from working).

 

Removed the GPUs. Booted with Intel's graphics... Still no backlight. Shining a flashlight at the monitor reveals that it is in fact displaying. Just no backlight. That effectively rules out the GPUs yeah?

 

Hi guys, my M17X R4 LCD was bricked a few days ago
My OS was Windows 7 (but it had the win 10 upgrade files)
Here is what's happened:
Installed Evga precision, messed with the gpu clock for about an hour and then i've shut down the system.
The next day i power on the system but all i've got was a black screen without any beep codes.

I thought it was the GPU so i've tried to remove it but i've got a pink screen with a grey bar at the bottom and 8 beeps
So i've tried another GPU: same thing, pink screen and 8 beeps

Then i found this thread and put all the pieces togheter, ordered a new screen (because i was unable to boot with an external screen), and followed the procedure hot-swapping the displays and now it's back to life!

The only problem is that i didn't found the EDID for my screen (B173RW01) so i've flashed the B173HW01 one.
Now i'm able to use the older screen but it has the wrong resolution and even the alien logo at POST is cut off

Spoiler: Photos

 

That's not too surprising; it's a 900p panel, so a 1080p edid will not play nice. Added this model's edid (pnp id AUO149E) to the archive. Flash that and things should be peachy once more:

archive correct edids, v12

Spoiler: pnp id - panel nrs

Code:

pnp id  notes   interf  panel nr.      edid eeprom
------- -----   ------  -------------  -------------------
AUO10ED !       LVDS    B156HW01 V0
AUO11ED         LVDS    B156HW01 V1
AUO12ED         eDP     B156HAN01.2
AUO149D !       LVDS    B173HW01 V4
AUO149E !       LVDS    B173RW01 V4
AUO159D         LVDS    B173HW01 V5
AUO219D !       LVDS    B173HW02 V1
CMO1719         LVDS    N173O6-L02
CMO1720 !       LVDS    N173HGE-L11
HWP3122 M       LVDS    LM240WU8-SLA* or LTM240CL01
LGD01CA         LVDS    LP173WD1-TLA1
LGD0226         LVDS    LP173WD1-TLC2
LGD0285         LVDS    LP173WF1-TLC1
LGD0289         LVDS    LP173WD1-TLA3
LGD02C5 X       eDP     LP173WF2-TPA1
LGD02DA !       LVDS    LP173WF1-TLB3
LGD02FC C       LVDS    LP173WF3-SLB2
LGD0343         LVDS    LP173WF1-TLB5
LGD037E         LVDS    LP156WF4-SLB5
LGD0391         LVDS    LP173WD1-TLE1
LGD03FB         LVDS    LP173WF1-TL**
LGD0469 E       eDP     LP173WF4-SPF1
LGD046C E       eDP     LP173WF4-SPD1
MEI96A2         eDP     VVX16T020G00
SDC3654         LVDS    LTN173KT03-W01
SDC4852         eDP     LTN156FL02-L01
SDC4C48 !       LVDS    LTM184HL01-C01
SEC314A         LVDS    LTN184HT03-001
SEC4A4B         LVDS    LTN184KT01-J01
SEC5044 !?AWX   eDP     LTN173HT01-301  Winbond 25X20BLNIG
SEC5044 !?AWX   eDP     LTN173HT02-D**  Winbond 25X20BLNIG
SEC5044 !AX     eDP     LTN173HT02-D02  "" ?
SEC5044 AX      eDP     LTN173HT02-P01  "" ?
SEC5044 AX      eDP     LTN173HT02-T01  "" ?
SEC5443         LVDS    LTN170CT08-D01
SEC5448 !       LVDS    LTN184HT02-S01
SEC544B B       LVDS    LTN173KT01-***
SEC544B BD      LVDS    LTN140KT**-***
-----------------------------------------------------------------------------------------------------
!  = known bricked panels
!? = bricked, but unknown which one
*  = unknown part id
A  = highly suspect: multiple variants exist, perhaps the others are safe ...
B  = multiple variants, flash the correct one!
C  = EliteBook 8**0w DreamColor, 10-bit, for fun ^^
D  = 14.0" version for M14x, just in case
E  = G-Sync approved panel
M  = desktop monitor (HP EliteDisplay E241i)
W  = write-protect possible
X  = 256-byte edid; use write-edid-256.sh instead of write-edid.sh

If multiple edids exist for one PnP id; flash the most recent edid, unless indicated otherwise.
-----------------------------------------------------------------------------------------------------

 

Add my Clevo P750ZM with the LGD037E panel to the list of bricks.

Any suggestions for getting the unbricker working on the 750/770? I can boot the image by adding "nouveau.nomodeset=0" to the kernel line (no Optimus) but the sensors detection fails to detect a supported I2C bus.

 

Install nvidia's linux driver and give it another try.

 

So simple I didn't think of it. Thanks. I'll try it this evening.

Edit: You used lubuntu so the binary blob driver is available from the additional drivers installer. Much easier than trying to assemble the tool chain needed to compile the kernel modules. Also, I used UUI to write the image to a 2GB stick and TopoResize to resize the casper file in order to fit. Haven't actually tried it on my P750. I'm just getting the stick prepared during my work break.

 

Flashed the AUO149E this morning and it worked, thank you
I've also had the same problem where the i2cdetect didn't show all the busses but i solved it by enabling Optimus

 

No Optimus on the Clevo P750/770 series notebooks.
On the other hand, the HDMI and DVI outs work even when the panel is bricked so hopefully no panel swapping will be necessary.

 

You ROCK (lots of exclamation points).

 

Glad to see you were able to get that fixed. Always impressed by how the members of the forum here go about helping each other out.

 

Hi everyone.... Thx for all the work put into this and other various threads. I am having similar issues after a win 10 upgrade fail. I think it might be a backlight issue..... As was stated a few posts back if i shine a light i can see very faintly that windows is running. I can hdmi out to a monitor. I was able to install Ubuntu and Windows 7 but it didnt make any difference to my backlight issue. I even changed the lcd screen. I havent changed the cable yet. I have taken the entire thing apart twice and have not found any burnt looking areas on components around the mobo lvsd connector. The connection is tight etc. Is it worth trying the edid rewrite ?? Or am i grasping at straws and shoiuld be trying to find a burnt component? Ive been up for too many hours trying to diagnose thx for any help dudes.


I am running an Asus x750l with intel 4000 integrated. I do not think nvidia graphics are present on this particular model. I havent tried to install any and they never seem to show up in the device manager. I bought the laptop used in this condition for dirt cheap really hoping i can get it functional.

 

I just did the eeprom edid write and it completed successfully. But I still have win 10 installed so I will have to install something else and then write it again blah blah. It didnt restore my monitor even for 1 second.

 

Yes, as you've pointed out already; a backlight issue. The edid does not in any way control a single parameter of the backlight, it only deals with the lcd itself. These things are really two separate devices stacked on top, same as a touchscreen is placed over the lcd (assembly).

Already having tried a new display that leaves the cable or motherboard. With a multimeter it's possible to measure between the backlight voltage+ground pins on the lcd connector and at the motherboard connector. If the lcd shows 0V and the mb 12-14V then it's the cable, otherwise it's the mb.

The lcd end has several voltages and grounds side-by-side, so there's no risk of causing a short with the multimeter's pins; connect to the center-most of the pin groups and you can't go wrong. The same thing applies to the mb, but then you need to have the schematics in order to know which pins are, in fact, used for backlight power. See, it's quite easy to bridge two pins with a single tip of the multimeter and, depending on their respective purpose, that results in nothing, a spark and/or permanent damage to the lcd and/or mb. Anyway, the specific lcd pins can be looked up using the specification sheet; google ' panel nr. + pdf' usually results in a hit. For the mb the same thing would apply; try the stamped marks on the mobo, since ' brand+model laptop' is probably a dead end.

In a pinch you could diy your own baclight power source; a 12V@1A adapter is very common for a lot of external hdds and this should do nicely. If not as a practical, permanent solution, then at least for testing purposes.

 

Hey can someone comment on this, underside of my board directly underneath the lvsd connector does the bottom one look like the culprit? A little solder connector to fix ?


crap how do i add an image or upload a file

 

Hey can someone comment on this, underside of my board directly underneath the lvsd connector does the bottom one look like the culprit? A little solder connector to fix ?
http://postimg.org/image/kylsfkhfl/8c5ca7ac/

 

Couple things..... could not find a schematic for the board. pn 60nb04b0-mb1100 or hannstar j mv-4 94v-0 e89382 1352

i also found a bad copper line on the board picture here
http://postimg.org/image/gshh8cj5t/50da1582/

likely thats the major problem but again cant determine without a sheet

 

Found a few comparable schematics, but these had a separate connector for the backlight. Does this system have a split-off type cable? If so, check the voltage on the pins of the smaller, secondary connector. No need to verify which pins are voltage and ground; it will have much larger solder blobs, so don't have to be as careful as with the main lcd connector. Which specific panel is this, btw? Not a CCFL, per chance?

Anyway, these general schematics won't help with the scraped copper thread as they're not detailed enough for that. Doesn't matter, either; it's obviously a broken connection, so run the system and bridge the exposed traces with a wire and see if that fixes the problem. And, if so, solder on the wire for a permanent solution.

Can't see anything wrong in that first image you've posted, though. Could you specify which component you're suspecting to be faulty?

Hmm ... and just to rule out the obvious; tried Fn+F7?

 

I was hoping the 1st image was a picture of the fuse with a blown filament not connecting the 2 different solder points. On close inspection it was not. My bad. I wish the fn f7 worked but alas it is not the issue. The lvsd does have a split in it. However I think one side goes up to the Web cam and the other is the lcd ribbon. Lcd is LP173Wd1(Tl)(G1
Which I think is led backlit. I could be totally wrong about all this. This is my first laptop and my first set of repair attempts. Thx again for any help. I'll try and do a temp bridge shortly once I get off of dad duty.

 

Well ... it might very well have been a blown fuse, but these should look a little different on a modern system. Shame the motherboard isn't labelled; a good "F" would be all you want. It's probably two empty solders pads, but try bridging it anyway to make certain. For regular fuses, look for things like these single-letter thingies:

Spoiler: smd fuse, fast acting

Most are white, but other non-black colours are also common. Measure these for ohms or just bypass them with a wire or screw driver to find out if it has indeed been triggered.

That panel is indeed a led and, unlike replacing a ccfl tube, you can't easily repair the wled strips. Do measure if first; pins #32 and #39 are perfectly safe to test with the multimeter. Do the same for #32+#36, the latter functioning as the backlight on/off pin.

 

Any chance to the immediate right of the pad in the first pic is the fuse? it looks like what you posted except off colored.... i have an lvsd cable coming should arrive tomorrow. Gonna swap it out before soldering. The guy i bought the notebook from said this all started after a failed win10 upgrade says screen went black and he thought it was dead. Guess he was close enough. I just tried to bridge those fuses and nothing. I am pretty scared to try the multimeter out on different pins but i will..... gotta dig the bugger out first

 

I confirm the issue with a Macbook 13 2010 too, Windows 10 + evga precision + last nVidia drivers.
The lucky thing is that I can boot the system. In windows with image but not in OSX (it runs but doesnt recognize the LCD)
The system boots with no backlight and image until nVidia drivers loads.
Now I must to look for a somebody with the same machine for dump the correct EDID and try to flash it.

 

Apple doesn't make displays, so you just need to know the panel id and we can find that edid somewhere (it'll have been used in non-Apples, too). Alternatively, since it's still running, read the current edid with MonInfo and copy/paste the ' Raw data' section (or, better, use the edid tools). Can either use that for the .bin, or fix the values if it happens to have been corrupted.

 

Hi, I tried to boot the CD image but was impossible. Moninfo gives this info:

Raw data
00,FF,FF,FF,FF,FF,FF,00,06,10,C3,9C,00,00,00,00,1C,13,01,03,80,1D,12,78,0A,00,00,00,00,00,00,00,
00,50,54,00,00,00,01,01,01,01,01,01,01,01,01,01,01,01,01,01,01,01,52,1C,00,8F,50,20,2E,30,30,20,
36,00,1E,B3,10,00,00,18,00,00,00,01,00,06,10,20,00,00,00,00,00,00,00,00,0A,20,00,00,00,FE,00,4C,
50,31,33,33,57,58,33,2D,54,4C,41,33,00,00,00,FE,00,43,6F,6C,6F,72,20,4C,43,44,0A,20,20,20,00,20


And the rest:

Monitor
Windows description...... Generic PnP Monitor
Manufacturer............. Apple
Plug and Play ID......... APP9CC3
Data string.............. Color LCD
Serial number............ n/a
Manufacture date......... 2009, ISO week 28
Filter driver............ None
-------------------------
EDID revision............ 1.3
Input signal type........ Digital
Color bit depth.......... Undefined
Display type............. RGB color
Screen size.............. 290 x 180 mm (13,4 in)
Power management......... Not supported
Extension blocs.......... None
-------------------------
DDC/CI................... n/a

Color characteristics
Default color space...... Non-sRGB
Display gamma............ 2,20
Red chromaticity......... Rx 0,000 - Ry 0,000
Green chromaticity....... Gx 0,000 - Gy 0,000
Blue chromaticity........ Bx 0,000 - By 0,000
White point (default).... Wx 0,313 - Wy 0,328
Additional descriptors... None

Timing characteristics
Range limits............. Not available
GTF standard............. Not supported
Additional descriptors... None
Preferred timing......... Yes
Native/preferred timing.. 1280x800p at 60Hz (16:10)
Modeline............... "1280x800" 72,500 1280 1328 1360 1423 800 803 809 846 -hsync -vsync

Standard timings supported

Report information
Date generated........... 20/5/2016
Software revision........ 2.90.0.1002
Data source.............. Registry
Operating system......... 10.0.10586.2


Thanks for the advices

 

Thanks for the edid (= 'Raw data'). It's an LG; model LP133WX3-TLA3, re-labelled as 'Apple'. You can use this file to flash it.

However ... doesn't this system use a fairly ancient gpu? If so, then the latest-latest drivers from Nvidia will have ditched support for it, so not sure if this really is edid corruption. But if it is then you do need to manage to boot the live usb stick, which might require changing some settings in the bios.

 

I think is a EDID corruption because the panel isn't recognizable at boot, in OSX, and in Win (only after the nVIdia drivers loads image&light appears). I tried all the possible resets, "genius" bar, and reinstall everything. Nothing.
I use RW everything, "Get EDID error". softMCCS says Number of monitors enumerated = 0.

And this situation was after "playing" with eVga precision looking for undervolt/underclock the GPU, then my hunch may be true.

Macbook has not BIOS setting like normal PC, only shortcuts, and boot orders programmable via aplication inside the OS (Win or OSX). I can boot with no problem Win or OSX CD installation disks blindly, then this image must be bootable at some way. I will try again. (The disk was slowly burned and verified, will try other burn soft and disk).

Thank you very much identifiying my panel.

 

Hello,

MSI afterburner still looks safe? It doesn't have EDID OC option.

Regards

 

Hello, I managed to boot an USB with your CD image, but occurs this:


At 513.0077 I disconnected the external monitor.

 

Oddly enough, that ' Raw EDID' is from your MultiSync P221W and is, as the output indicates, corrupt. So ... with both displays out of order, I'd first reflash the P221W when hooked up to a working system. When you've got the monitor working again, it'll be possible to flash the Mac's internal display as well.

Haven't found a good edid for the P221W, but the old, uncorrupted version should still be visible in MonInfo's list. This info is pulled from the registry and might still have the original values from the moment the monitor was first connected to the system.

 

Probably safe... But Nvidia Inspector is a better choice for OC.

 

Hey,

Okay. Thanks for the info Papusan

Regards

 

The .bin file created with Moninfo is the same type that you use with "EDID Tools"?

I saved the correct one, the one with [Registry].

But it seems that registry have less info than actual. I attach it:

Realtime (Presumibly corrupted):
DDC/CI................... Supported
MCCS revison............. 2.0
Display technology....... TFT
Controller............... Not specified
Firmware revision........ Not supported
Active power on time..... Not supported
Power consumption........ Not supported
Current frequency........ 16711,68kHz, 0,00Hz

Registry (Presumibly correct):
DDC/CI................... n/a




Realtime (Presumibly corrupted):
EIA/CEA-861 Information
Revision number.......... 1
IT underscan............. Not supported
Basic audio.............. Not supported
YCbCr 4:4:4.............. Not supported
YCbCr 4:2:2.............. Not supported
Native formats........... 0
Detailed timing #1....... 1280x720p at 60Hz (16:10)
Modeline............... "1280x720" 74,250 1280 1390 1430 1650 720 725 730 750 +hsync +vsync
Detailed timing #2....... 720x480p at 60Hz (16:10)
Modeline............... "720x480" 27,000 720 736 798 858 480 489 495 525 -hsync -vsync
Detailed timing #3....... 720x576p at 50Hz (16:10)
Modeline............... "720x576" 27,000 720 732 796 864 576 581 586 625 -hsync -vsync
Detailed timing #4....... 1280x720p at 50Hz (16:10)
Modeline............... "1280x720" 74,250 1280 1720 1760 1980 720 725 730 750 +hsync +vsync

Registry (Presumibly correct):
Nothing

 

Yes, though I've used a hex editor to make the LG edid. Also to make sure the checksum was ok, as this is an almost 100% guarantee that the edid itself is original and uncorrupted.

Please upload the dspinfo.bin and/or copy paste its hex values on the EDID Reader site. Can also use ' File -> Open' with MonInfo and point to the dspinfo.bin; if the checksum doesn't match, then it'll yelp about this and makes an auto-correct. This, mind, does not fix the edid, it merely appends the re-calculated checksum (the last byte).

 

What a mess! Both are checksum valid.

The "corrupt":
Valid Checksum: TRUE
EDID Header: OK
EISA ID: NEC
Product Code: 674A
Serial Number: 16843009
Manufacture Date: 44/2009
EDID Version: 1.3


Number of Extensions: 1
Checksum: 0x3E

and has 4 detailed timing descriptor

The "good":
Valid Checksum: TRUE
EDID Header: OK
EISA ID: NEC
Product Code: 674A
Serial Number: 16843009
Manufacture Date: 44/2009
EDID Version: 1.3


Number of Extensions: 1
Checksum: 0x3E

and this one hasnt detailed timing descriptors but has CEA Data block.




I tryed again boot the EDID tools and I thing the corrupt message was due the disconnection, but the stored EDID and actual differs.


EDIT: I boot the PC with EDID tools and after the i2cdump and online EDID check, the checksum is valid. What happens here?

EDIT2: I boot the MAcbook with other monitor. Totally "new". The same "DCC Responded but no EDID for LDVS-1"

 

Windows 10 Tinekring the BIOS

I think it's important to add these 0.02c to the discussion. This is first somewhat documented report that Windows 10 does something it shouldn't do to the BIOS. Which might be one more clue as to what exact piece of code renders the screens unusable unless the user rewrites their EDID.

 

Ah, you're right, of course; the last few 'FF's must've been the disconnect (doh ). As for the difference in edids; the one in Windows might've been pulled from the (v)bios or from a secondary eeprom (for various 'reasons' ...), whereas i2cdump can ignore that copy. Doesn't matter which it was; as long as the edid is still ok and compatible with the display then it'll work.

Ok, it's safe to say the monitors are fine, yet the corrupt edid of the internal lcd is preventing boot. It's a little better than some of the Alienwares, though, which refused to boot altogether when there was no valid, internal edid available.

What you need to do is disconnect the lcd's cable, either from the motherboard or the lcd side, whichever is easier to reach. With the monitor hooked up it can probably still boot, at which time you can reconnect the lcd cable (insert straight, not slanted) and it'll be flashable. If the system does not like that option, then you're good to disconnect the lcd while that 'Nouveau' output is complaining about the missing edid. It should then default output to the monitor and proceed to boot happily.

 

if someone would help me, I have a alienware 17 r1 but no idea wich .bien to use

 

ok so i found out wich display is the one i have but its not overwriting the edid its leaving it as it was, dont know why...

 

tried to do it directly and this is what i get
root@lubuntu:~/EDID/write-edid# i2cset -r 19 0x50 0x0c 00
WARNING! This program can confuse your I2C bus, cause data loss and worse!
DANGEROUS! Writing to a serial EEPROM on a memory DIMM
may render your memory USELESS and make your system UNBOOTABLE!
I will write to device file /dev/i2c-19, chip address 0x50, data address
0x0c, data 0x00, mode byte.
Continue? [y/N] y
Warning - data mismatch - wrote 0x00, read back 0x05

 

Please copy/paste the output starting from ' sudo bash ./write-edid.sh'. Also make sure not to skip a step, some may be redundant, but not on every system. Also list the panel id, just to exclude another possibility.

Hm ... from the looks of that address you have 5555-corruption, right?

 

my monitor is a LGD02DA2, here is what i get on the console
https://onedrive.live.com/redir?resid=A03CD54FBE815D1A!7617&authkey=!AGZCAhALh7gc5pE&ithint=file,txt

i did every step multiple times but no luck :/, other things i shoud mention is that after the 8 beep error i was able to reset the bios and stop geting the beeps, after that i restored windows to windows 8 (was on 10 when this happened) . Also im able to use the screen on windows when i use the generic microsoft drivers, aparently they sends a analog 64Hz signal that my monitor is able to recibe (better than nothing i guess), but im unable to use hdmi or adjust anything on the display (to log on lubuntu i have to use the intel drivers and conect the hdmi to another monitor)

regarding that, if you're refering to the corrupted parts of the edid being change to 05 then yeah
thanks for the help!

 

That last output shows there's data beyond the edid proper. Furthermore, this display seems to have two eeproms; one with only the eeprom and another with the edid plus ... other stuff:

Spoiler: LP173WF1-TLB3 - U1

Spoiler: LP173WF1-TLB3 - U2

You could use 'sudo i2cdetect x' for the other 19 buses; one of these may show the same, corrupted edid, try overwriting that version. This is the trickiest bit and could do some serious harm, but have tried that on several systems and no fubar as of yet ...

Hmm ... and this is an Optimus-capable system, right? If so, make sure that that is enabled in the bios; this worked for _zero_, as it revealed some more buses (it would also re-enumerate them upon boot). Not sure how this system is hardware-wired, but the 'i915' is Intel graphics (and 'Nouveau' for Nvidia), yet 'DPDDC-A' and 'DPDDC-B' would be Intel's naming convention for DisplayPort. This means that i2c-19/DPDDC-A can't be the right bus for your system, considering the TLB3 is an LVDS panel, not eDP ... so ... either the system bios or the vbios has messed up mapping the various display outputs.

 

Yessssssssssssssssssssssssssssssssssssss

I had to boot it with disconnected LVDS, and before Lubuntu starts, reconnect it, because if I connect LVDS with loaded desktop the panel isn't in any i2c (only detects the extenal one). Then I explore all i2c due no one shows panel or something similar. I choose the correct one, with partial and corrupted panel EDID, and then flash. Fast and perfect.

Thank you very much for your patience. If you ever visit Barcelona, you have a bunch of beers or anything else for you

With permission of the users, I put some tags for future or actual people that have that issue.

bootcamp, macbook, mac, screen, black, lcd, backlight, evga, nvdia, boot, osx, windows

 

Thanks , might take you up on that; there's an astronomy institute in Barcelona that has an extensive library and they have a CD in their collection I'd like a copy of. Perfect excuse for a city trip .

 

I found the issue..... if i bridge pins 37 and 38 the backlight shows up..... what do you recommend i do to permanently bridge them..???

 

A small blob of solder would do. However, this is not a decent solution; try turning off the laptop completely and bridge the same pins. The result could be that it's still turning on the backlight, which would mean there'll be 24h battery drain.

The cause looks more to have to do with the backlight-enable pin and shorting voltage forces bl current to flow again. Whether that is so can be checked by using the lcd-off Fn key while shorting #38; the result would be a 100% white display, the lcd having been turned off, but not the backlight. Also try bridging pins #4 and #36 (taking utmost care not to touch pin #3 !); that might turn on the backlight as well, proving there's something amiss with the bl-enable relay or the bios/ec that is controlling that switch.

Hm ... or something even more prosaic; apart from the bl_enable signal, the lid switch is an override with vetoing power. Lid closed = backlight off (but not the lcd itself). This switch comes in the form of a magnetic switch and is most often integrated as an smd type in the power button board (though mb is possible, too). If that switch has been magnetized in one setting permanently then not even replacing the lcd, cable or motherboard will do any good. You could find out if this is so by disconnecting the power button pcb and taking a strong, permanent magnet (such as a neodymium type) and circling this in close proximity around the small pcb. At one point it ought to force open or close the switch again and turn on the backlight. A new power button board would solve the issue then, or soldering on a new switch, of course, if the board is hard to find, excessively priced or the switch has been soldered to the motherboard itself.

First find out exactly where the switch is located by spotting the magnet inside the display lid:

Spoiler: Clevo W230SS - lid switch and magnet

Just for fun; take a small, strong magnet and hover it around the probably location of the switch (anywhere above the keyboard). That way you can experience the lid-close function in action without having to dismantle anything or, indeed, closing the lid.

 

Unfortunately it appears i have already ruined this motherboard. As i was bridging components it turned off and will not show any life any longer. The usual leds do not light up on motherboard. I have traced the power from the power supply it is getting power but not functioning. Looks like this mobo has had its day. If it comes back to life i will try the magnet trick. Right now I am going to try the battery disconnect and hold the power button fix see where it gets me. Not feeling too hopeful tho.