Firmware upgrade using Linux tool, requires no hardware

[radekpw]

I have a XL2420Z I installed V4 to and I believe it's done for. I can't get anything but 60hz now, except on a much lower 4:3 resolution with two options of 60hz or 120hz. My native 16:9 at 1080p can only do 60hz. I guess there is no real reason to even upgrade the XL2420Z to V4 since there isn't any benefit from doing so. This part should be sticked so more people don't mess up their gaming monitors.

I went from V2 to V4, noticed the issues and so I used the backup firmware to go back to my original version, to no avail. Guess I'm stuck with this expensive brick or I'd have to buy MStar and see if I can fix it this way.

[Falkentyne]

Scroll down to the VERY BOTTOM to try a last ditch effort BEFORE BUYING the Mstar ISP device and BEFORE buying a Displayport cable (Displayport does not follow the normal DVI/HDMI/VGA EDID rules and is never corrupted).

--------------

You can fix this by using a displayport cable.

The reason you are limited to 60hz is because the DVI EDID got corrupted somehow.
I don't know why this has been happening to people recently.
When the Linux flasher first came out, there were no failures like this at all.

May I ask exactly what hardware you used to flash the monitor?
Like, exact motherboard, Bios type (UEFI, Legacy (award / phoenix, AMI, etc), video card/hardware used, and so on?

You can recover the DVI EDID through Linux, by using either edid-rw or https://github.com/tomka/write-edid
But that would require you to have the backup (uncorrupted) EDID from the XL2420Z, and the EDID chip not being write protected, and then using the commands shown similar to this at the bottom of this page (Note theres a typo on edid, it says "edit" instead).

On the Asus VG248QE, the EDID write protect is removed temporarily by entering the Service Menu and enabling "burn in" to On.
Could possibly be the same on the Benq.

http://www.blurbusters.com/zero-motion- ... dware-mod/

(a full hardware firmware flash will write the EDID regardless of write protection since the protection doesn't apply if the firmware is blanked. The Mstar ISP device does a full complete flash. I believe part of this is because the Mstar ISP device BLANKS the firmware (erases it) before flashing, then flashes, then verifies. I don't know if Linux blanks it. Could be another write protect issue if it's not blanking it even though I had no problems with 4 Linux flash runs).

This would 100% recover the EDID IF there was a backup somewhere but I don't know of anyone who has EDID bin files.

The yoycart/cart100 mstar utility would avoid all of these problems because it's an actual hardware flasher that communicates with the flash chip through dedicated hardware and parallel port emulation (through USB), not a low level driver through Linux, so the EDID would always be written correctly.

I *KNOW* that the EDID is stored in the Benq firmware itself, because when I bought my Mstar ISP unit (to toy with, basically), I intentionally BRICKED my Asus VG248QE by flashing a XL2420Z firmware into it! Monitor was as dead as a rock (note I have two VG248's and I DID back up the Vg248 firmware (there are no backup FW files on the internet for this monitor). I then flashed the backup into the dead VG248 and it came RIGHT UP perfectly, logo and all.

Except there was 1 problem.
IT HAD A BENQ DVI EDID !
The refresh rates matched the Benq meaning 85hz refresh rate was gone (this is default on the Asus) as well as the low res VGA refresh rates just said 60 and 120 in ToastyX CRU, instead of the 85hz entries.

To test that further, I this time flashed an XL2720Z firmware into the Asus, bricked it, flashed the backup, recovered perfectly and sure enough...now it had an XL2720Z EDID !

Note: The EDID over a displayport connection is handled by the firmware and displayport spec, so over displayport, it was normal.

With some difficulty, I backed up the 2nd VG248QE's EDID through Linux, then disabled the write protect by enabling burn-in menu, and flashed the EDID with Write-edid.sh (edid-rw was giving errors). Everything back to normal.

The fact that the Asus was flashed with the Benq firmware and NOT limited to 60hz over DVI (displayport does not apply to these problems) shows that the Mstar utility will work perfectly.

So yes, buy that and your monitor will be fixed. Or you can ask and see if someone can dump their XL2420Z EDID for you.


---------------------
Last ditch effort :
Since the EDID is in the firmware, SOMETHING is failing to write the EDID and it's getting corrupted.
Try the following.

Boot to Linux and get the flasher / USB drive ready and follow all the commands BEFORE you get to the write firmware part.
After you finish the part where it backs up the firmware (you can probably skip that command), power off the monitor and enable "burn-in mode".

Hold button #3 and #4 for 5 seconds, Press Power while holding down #3 and #4, release power and keep #3 and #4 held down until you get a display. The "Eye care" and input prompts should not appear.

Then press button #5 to bring the factory menu up. Enable "burn in" option then exit. the power light should change from white to orange.

Now flash the firmware.
you need to do the burn in setting at the end, right before writing the flash, because if you are reading the flash/backing up, the flashrom will power cycle the monitor after the read, possibly disabling burn-in.

See if this works and if this recovers the DVI EDID.

[radekpw]

It would make sense that the DVI EDID got corrupted. My desktop has a ASRock Z77 Extreme4/AMI UEFI. I used a second monitor plugged into a Geforce 770 and the BenQ monitor plugged into first the DVI on-board then the VGA on board connectors. I guess I should have read through the post more and maybe I might have avoided the issue if I just went with the on board VGA connection the first time I attempted the firmware flash. My laptop's bios is AMI, it does not have a VGA connector, it does have a DVI, HDMI, and Display Port.

I was able to successfully flash the dumped firmware backup into the monitor. One of the other things I noticed is that the resolution within the bios and boot process is very low as well. I was going to attempt to flash the backup again using the laptop, using a DVI to VGA adapter. I will try some of the thing you mentioned above including those with the Display Port. If I purchased the MStar utility do you think it will fully recover the monitor's state to what it was previously, do you think it'll fix the DVI EDID issue?

[Falkentyne]

The MStar utility has a 100% success rate.

remember that I told you that I flashed a Benq XL2720Z (27" monitor) firmware into the VG248QE (24" monitor, completely different firmware chip (winbond) and it bricked it and after I flashed the backup, the VG248QE had the XL2720Z's DVI EDID.

Remember this was on a intentionally BRICKED monitor I restored.
I'd say your changes are quite good.

Can you do what I suggested though?

Please enable BURN IN option in the factory menu right before you get to the "write firmware" command line in Linux and try flashing with Burn-in enabled. The monitor power light will turn amber/orange when burn in is enabled.

[radekpw]

I am attempting to do what you suggested right now. I am willing to mess around with this some more, but I am strongly considering just buying the Mstar utility as I've been a long time BenQ monitor user, and I can see it becoming useful as I upgrade in the future.

[Falkentyne]

Okay.
Did you enable burn-in and the power light turned orange, right?

[radekpw]

I did, no luck with the laptop, trying a couple more times on the desktop. I have a DP cable also coming Tuesday and if worse comes to worse I'll just order that Mstar. I'm supposed to enable burn in and turn off the monitor or leave it on and just close the menu?

[Falkentyne]

I'm not sure.
It was just a random guess.
On the Asus VG248QE, when you enable burn in, you need to keep the menu open or the write protect gets disabled.

I'm still a bit confused as to your setup, though.
The displayport cable is guaranteed to work. I'm just pretty sure that something blocked the DVI EDID from being flashed and it got corrupted.

Can you even read the DVI EDID through Linux ?(you can run the edid rw program and dump it from there but you need the proper bus ID)

[radekpw]

I'm able to successfully get the v4 firemware installed on the 2nd try, however the DVI corruption still persists even after using a VGA cable, homefully the DP cable that's coming might fix this issue. Also the monitor is set to 800x600 resolution when I boot or go into the bios, this is a 16:9 monitor could this also be caused by the malfunctioning DVI connector? I have to test boot with VGA etc another time.

[radekpw]

So after getting that display port cable and identifying the flash chip on the display port bus, flashrom isn't able to identify the firmware correctly. However the VGA and DVI are able to. Not only that, somehow my laptop monitor got damaged in the process, its display is now split horizontally. Anyways this is the panel in my laptop, RIP.

http://www.panelook.com/B173HW01%20V4_A ... r_738.html