[solved] X62: Bios setting change -> No boot incl. splash?

[flyingfishfinger]

Copied over from the other X62 thread to avoid hijacking / thread ballooning.

Anyway, I was playing around with some BIOS settings (SATA in particular, trying to get a Hackintosh installer to recognize any HDD - common problem with that stuff, but irrelevant to the current dilemma). On rebooting, I now have a fully black screen, but fan & HDD seem to run. No boot screen or anything. I had this issue once before while trying this same thing last week but the problem went away after I powercycled including removing the main battery.

This time, however nothing works. Tried removing and reseating the BIOS battery after a few minutes, remove / replace HDDs / RAM / main battery / basically everything that could be unplugged, but still no luck. It's not just the backlight either, nothing (dark) on the screen.

I'll let it sit overnight and see if it's any better in the morning, but does anyone have better ideas in the meantime? I hope this isn't a brick, that'd be odd for a BIOS setting I'd think..
Thanks!

R

[xiphmont]

Copied over from the other X62 thread to avoid hijacking / thread ballooning.

Anyway, I was playing around with some BIOS settings (SATA in particular, trying to get a Hackintosh installer to recognize any HDD - common problem with that stuff, but irrelevant to the current dilemma). On rebooting, I now have a fully black screen, but fan & HDD seem to run. No boot screen or anything. I had this issue once before while trying this same thing last week but the problem went away after I powercycled including removing the main battery.

This time, however nothing works. Tried removing and reseating the BIOS battery after a few minutes, remove / replace HDDs / RAM / main battery / basically everything that could be unplugged, but still no luck. It's not just the backlight either, nothing (dark) on the screen.

I'll let it sit overnight and see if it's any better in the morning, but does anyone have better ideas in the meantime? I hope this isn't a brick, that'd be odd for a BIOS setting I'd think..
Thanks!

R

w.r.t. bricking, I'll have a look to see if the BIOS is being stored in a standard microwire EEPROM... if so, we can always grab a clean image from another board and you can reflash yours (the hard way if necessary)

That also gives us a safe way to try to duplicate.

[edit: just had a quick look, I see two fairly standard 25Q serial EEPROM chips on the board. I'll look a bit harder (and dump them both) later...]

[flyingfishfinger]

No change this morning. Luckily I made a clone of my old drive and have enough spare X61s..

EDIT: I'm not sure if the X62 has the beep feature but I get the exact same reaction (nothing) if I try starting it with or without RAM installed...

EDIT: Could anyone possibly do me a favor and post a pic of the BIOS screen? I want to see if I can get to the "Reset to defaults" in the case it's actually running, but since I can't see anything I need to do it blind and I don't remember the layout. Thanks for any help!

EDIT: Behaviour update. When I plug in a 2.5" HDD, it seems to read briefly from it (corresponding LED flashing), it also seems to react to pressing "Enter" but then stops. Boot menu selection? The internal mSATA slot shows no LED activity whatsoever if I connect one. I may be stuck in UEFI boot only, or perhaps RAID. Also, when there is a 2.5" HDD connected, pressing the power button shuts the machine down instantly, whereas when the mSATA drive is installed I have to long-press it.
Could these be indicative of anything in particular?

R

[mdancer]

Do you have an updated bios version? Are you sure this is not a hardware failure like dead or dying video, memory etc? Have you tried an external monitor?

[flyingfishfinger]

It might be, so far I've tried shuffling RAM & HDDs but I don't have an external monitor or TV I could connect unfortunately.
The only monitor I have is VGA and I don't have a converter around. Or a miniHDMI adapter, for that matter.

Should I be going for the miniDP or HDMI outputs? I'm not sure which one it might try to use...

R

[flyingfishfinger]

Tried an external screen and replacing everything I could think of, no go.

Anyone with a working X62 board, might you send me a screenshot of the BIOS layout so I can try and navigate to "Reset defaults" in case that does the trick?

I'm rapidly running out of ideas and it's kind of a pain in the butt because that's my only machine with an SSD

R

[el-sahef]

Attention, image links!

Press ESC to enter BIOS, press "right" four times, "down" six times, "enter", "up" four times, "enter". This will restore defaults, save and reset.

Additionally, you can also try to change "vga support" to "auto".

http://abload.de/image.php?img=cimg2322egsm0.jpg

http://abload.de/image.php?img=cimg2323s0s7q.jpg

I have dumped the contents of the two BIOS chips before using the mainboard because I anticipated that something like this could happen to me also. PM me if you have the ability to flash the chips with a hardware programmer and need them.

Did you test the motherboard outside of the case (only screen and keyboard attached)?

[flyingfishfinger]

Thanks for the key combination, no luck though. Either I'm doing it wrong or it really isn't accomplishing anything. Maybe Konami code will do something??

The odd thing is that it does TRY to read from the HDD, and the fan kicks in after a minute or so, Thinklight works too so someting must still be working.

One other thought occurred to me. If the CMOS battery is completely shot (i.e disconnected) doesn't that usually generate some kind of error message? Would it stop booting?
Before I screwed something up I noticed that the clock was wrong at every boot which usually implies the BIOS battery, so I'm wondering if I was delivered a faulty battery. I get no difference in behaviour whether or not it's connected.

In the meantime, let's talk about the BIOS flashing. Which chips are they? I've previously rigged a flasher using an Arduino and flashrom, will that method work here?

R

[xiphmont]

In the meantime, let's talk about the BIOS flashing. Which chips are they? I've previously rigged a flasher using an Arduino and flashrom, will that method work here?
R

There are two EEPROMs I see, not sure yet which is the BIOS. I need to get my EEPROM readers set back up after the move (discovered to my chagrin that my W701 has no serial port, need to grab an X61 base...)

[flyingfishfinger]

I got 2 dumps from el-sahef.
I'm going to try and set up the Arduino-based flashing tool either tonight or tomorrow, I'm ashamed to say that I have to go and find some through-hole resistors as a good bit of my kit disappeared under circumstance that shall remain...unspecified.

R

[xiphmont]

I got 2 dumps from el-sahef.

Oh good.

BTW, previous flashing I did via dirt-simple PonyProg with the simple diode-and-resistor interface. I know that works great, so I can't imagine it wouldn't work with a much more refined Arduino process.

[flyingfishfinger]

Problem solved! I only flashed the larger of the two chips, which luckily is on the top side of the board. That means if it happens again, all we have to do is take the keyboard off.

It took me the better part of the day to get the flashing working, ISP with a voltage divider is pretty finicky. It also turned out that the branch of serprog I used for the Arduino Mega was buggy, the latest master version with a tweaked Makefile did the trick. Still, I had to turn down the SPI speed to 1M and the serial baud rate to 115200 for 1) the right chip to be detected and 2) write operation to run without errors.

I'm happy to say that everything is now working as it used to! Thanks el-sahef for the dumps and everyone else for the advice! Let me know if you have specific questions, now we just have to figure out why changing BIOS settings resulted in corruption...

R

[wileE]

Good to hear you got your machine unbricked.
You don't remember which BIOS settings change was responsible?

[xiphmont]

Good to hear you got your machine unbricked.
You don't remember which BIOS settings change was responsible?

Yes, I'm also dying to know if it's a deterministic problem.

[flyingfishfinger]

It may have had something to do with the SATA settings, also with BIOS Legacy modes.

I'm not going to be at home for the next few days (and I'm not bringing the flashing rig with me to my gf's place ), so I don't want to do any experiments until I'm back in case I brick it again. I'll see what I can do next week though.

R

[flyingfishfinger]

One observation and one comment on how I might have broken the BIOS initially.

- First, it now takes around 17 seconds for the boot splash to even appear. I don't recall this being the case before, is there a common explanation for this? The machine works again overall so that's great, but 15 or so extra seconds is kind of a lot these days...

- Thoughts about initial bricking: As some of you may have gathered, I'm a bit of a Hackintosh fan. On Broadwell, there is a UEFI setup variable hack that must be made to avoid kernel panics in OS X relating to DVMT (See http://www.insanelymac.com/forum/topic/ ... ite-10103/). Note that this did work without breaking anything else for some time, but the installer wasn't able to detect any SATA drives. Then I started messing with SATA settings, which is when the brick occurred.

R

[RealBlackStuff]

Under certain circumstances, boot-delay is caused by the search for a bootable drive.
Maybe change your Startup/Boot/Priority in BIOS and exclude unused drives?

[flyingfishfinger]

Disabled everything I can think of / see and removed all storage devices except the mSATA drive; no change. I wouldn't even know how to tell what the machine is doing, but it takes almost exactly 17 seconds before the X62 boot splash shows up.

R

[jaspen-meyer]

It took me the better part of the day to get the flashing working, ISP with a voltage divider is pretty finicky. It also turned out that the branch of serprog I used for the Arduino Mega was buggy, the latest master version with a tweaked Makefile did the trick. Still, I had to turn down the SPI speed to 1M and the serial baud rate to 115200 for 1) the right chip to be detected and 2) write operation to run without errors.

If you could post the commands you used to:
a) compile serprog
b) run flashrom

I'd be much obliged. I spent a few hours with it recently but it never detected the chip.

PM them to me if it's too off-topic.

[flyingfishfinger]

PM sent. Happy to post results here, if they're generally relevant.

R