PDA

View Full Version : Urgent help needed flashing a BIOS with a Genius G540



Shadow Lord
October 16th, 2013, 11:27 AM
So why in Off-Topics? Because this is on a modern system - An Asus P5E64 WS Evolution (http://support.asus.com/knowledge.aspx?SLanguage=en&p=1&s=22&m=P5E64 WS EVOLUTION&os=&hashedid=UWjwJogKctaj0MAN) board.

The back story:

I was trying to flash the BIOS on the MB to the latest version. It was late, and I was being lazy, so I used the Asus WinFlash utility under Win7 x64, The utulity finished, said everything was okay and verified. Rebooted and nothing no POST beep, no display, nada. Fans spin up full blast, lights on MB come on but thats the end of it. Try all the BIOS recovery methods (cleared the CMOS, tried loading the BIOS off of a CD, off of a USB stick) and still nothing. It seems as if the BIOS is wiped and there is no functionality left in it... Check on the internet and many people are complaining about the same issue with the Asus WinFLASH utility under x64.

Okay, what is it that THGTG says? "Don't Panic". Yet...

I look at the BIOS chip and it is a tiny SST 25VF080B (http://ww1.microchip.com/downloads/en/DeviceDoc/S71296_05.pdf) soldered directly on the board. A few expletives later and I started searching on the internet about what are my options and discover the SPI (http://en.wikipedia.org/wiki/Serial_Peripheral_Interface_Bus) bus. A ray of sunshine comes through. I look around a bit more and I find a number of pages describing how to program the BIOS using the SPI bus and a hacked together LPT interface. Not having a LPT port handy (i.e. no easy way to get the 5170 next to my main WS) I check out my trusty G540 and indeed it can program Serial EEPROMS and does support the SST chip.

Here is the SPI Header map out vs. the SST chip:
15596 15595

I've hacked together a cable and attached the SPI header to the G540. Doing continuity testing with my voltmeter shows continuity from the G540 to the PINS ON THE BIOS CHIP. So I am pretty confident the cable is okay. Also the G540 does its own check and it reports the following:

15597

Line 7 (hold) is marked as disconnected but this is to be expected as the SPI header does not map out to this pin on the chip.

I load up the BIOS file as a BIN and start the programming and this is where things fall apart - the erase and program portions report as OK but the verify portion gives a very informative "error" (yes that is it: "error", nothing more). Rebooting the system shows no change in behavior and trying to read the BIOS produces a blank.

Here is what I have tried so far:

1. I've tried both unlocking and not unlocking the chip before starting programming.
2. I've checked and during read/write 3V are being supplied to the chip.
3. I've tried programming with the CMOS backup battery in and out of the MB (the MB is completely powered down during this process and unplugged from the PSU).
4. I've tried blanking the chip before writing (as opposed to just erasing)

None of these have managed to program the chip. I have two radical options left:

1. I could buy and try to use a SOIC8 test clip to program the chip - maybe the G540 needs line 7 to program? It shouldn't though if my understanding is correct.
2. I could buy a preprogrammed chip and see if I can solder it on but this may well be out of my technical skills...

Any other ideas, tips, or whatever from the veterans out there? I really don't want to ditch a perfectly good (and expensive MB), CPU, and RAM over a bad BIOS. Any help is appreciated!

dabone
October 16th, 2013, 11:49 AM
1st, the bios file you have is not a bin file for the chip, you would have to extract the correct bin file from another identical motherboard.
2nd, I've never really had good luck getting the bios chips to program in circuit, I've always had to remove them, and then flash, then reinstall.
3rd, go to ebay and search your motherboard model and bios, alot of people sell replacements, but be sure to dump the eeprom before you install it, that way you have a good backup.

Later,
dabone

Shadow Lord
October 16th, 2013, 11:59 AM
Thanks for the pointers.

1. I did not realize I could not just use the BIOS file form Asus. (bang head against wall).
2. As I said originally I am not confident enough (nor do I have a solder with a fine enough tip) to remove the chip. That would be a very last resort for me.
3. I had already done that :). Here is what I found on eBay (http://www.ebay.com/itm/380539010409?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2649). So since this is a "replacement bios" if I was to read the BIOS using my G540 and create a BIN file then that could be potentially put back on my original chip, correct?

Am I doing the procedure correctly? Should I be unlocking the chip or it does not matter? Thanks again.

dabone
October 16th, 2013, 12:24 PM
First off, can you read the device?
Does it show any data?
Is the write protect pin being held active?

Later,
dabone

Chuck(G)
October 16th, 2013, 12:52 PM
Why not remove the chip and program it?

Alternatively, just order a new one or two (they're cheap) and program it using your G540 and then remove the old one and solder the new one in.

Shadow Lord
October 16th, 2013, 02:20 PM
First off, can you read the device?
Does it show any data?
Is the write protect pin being held active?

Later,
dabone

Can you speak a little slower for us non EE people? ;)

As I said in the original post I have executed the read command and according to the program the read is working. Unfortunately the program indicates a blank EEPROM after the read. Now does that mean it is really empty or it is not reading?

As for WP I have no idea. How would I check this? There is no power to the board so if it needs power to enable WP then the chip should not be WP.

I've tried programming with another programmer and that programmer says there is no chip connected (when the chip is clearly connected). I've tried the second programmer with a known good chip and it has no problem detecting it. This is a cheapy model programmer ($12 from china shipped) and it does not explicitly state support for the SST 25VF080B. I am not sure then if it is not detecting it because it does not "know" the SSt chip or if my BIOS chip on the MB is physically bad. Anyway, I can test it with the tools I have on hand?

Shadow Lord
October 16th, 2013, 02:20 PM
Why not remove the chip and program it?

Alternatively, just order a new one or two (they're cheap) and program it using your G540 and then remove the old one and solder the new one in.

Chuck(G) I can get a preprogrammed ready to go chip for the MB. Problem is I don't trust my soldering skills to be able to solder something this small/delicate.

Chuck(G)
October 16th, 2013, 06:12 PM
To remove the chip, see Chip-Quik (http://www.chipquikinc.com/). Low-temperature alloy that will allow you to remove the chip without making a mess out of the board. After that, a bit of solder-wick to put the replacement back and you're good (Youtube vids on surface-mount).

njroadfan
October 17th, 2013, 03:32 AM
1st, the bios file you have is not a bin file for the chip, you would have to extract the correct bin file from another identical motherboard.
2nd, I've never really had good luck getting the bios chips to program in circuit, I've always had to remove them, and then flash, then reinstall.
3rd, go to ebay and search your motherboard model and bios, alot of people sell replacements, but be sure to dump the eeprom before you install it, that way you have a good backup.

Later,
dabone

Why wouldn't the .ROM file from Asus work? Its a straight ROM image. I'm surprised a board made in 2008 has a SOIC BIOS. My C2D board still has a PLCC style chip.

Shadow Lord
October 17th, 2013, 07:06 AM
To remove the chip, see Chip-Quik (http://www.chipquikinc.com/). Low-temperature alloy that will allow you to remove the chip without making a mess out of the board. After that, a bit of solder-wick to put the replacement back and you're good (Youtube vids on surface-mount).

Chuck(G):

I checked out the videos and the info on the site. While it sounds straightforward enough I am still hesitant. Again this has to do with my skills as the most soldering I've done is wire to wire. I may give it a try as a last resort but will try to program the BIOS in circuit first as that would be the safest route for me. I can always whip out the soldering iron if in circuit programming fails.

Having said that do I need to do anything different to program the chip? I.E. am I approaching it procedurally correctly?

Thanks.

Chuck(G)
October 17th, 2013, 07:45 AM
As far as I can tell, you're okay, but in-circuit programming makes me a bit hesitant to put my stamp of approval on your procedure. As others have said, do check the WP# pin state and see if you can read the current contents of the device.

dabone
October 17th, 2013, 08:01 AM
Also, I noticed that it's reporting a 1meg device but the rom file I downloaded from your link is 2meg.

I would also try disconnecting the power pin from your cable, and using the computer power supply to power the motherboard/chip.

Later,
dabone

Shadow Lord
October 17th, 2013, 11:03 AM
As far as I can tell, you're okay, but in-circuit programming makes me a bit hesitant to put my stamp of approval on your procedure. As others have said, do check the WP# pin state and see if you can read the current contents of the device.

Chuck(G),

How can I check the state of the pin?

Edit: Nevermind - I think I found it:


A digital signal is low (a digital ‘0’) if its voltage is less than about 0.4V. This is measured using a multimeter set to the correct voltage range. The black (negative) lead is connected to the negative terminal of the battery or power supply. The red (positive) lead is connected to the digital signal.

A digital signal is high (a digital ‘1’) if its voltage is nearly the same as the supply voltage. This is measured using a multimeter in the same way as for a low level signal.

Shadow Lord
October 17th, 2013, 11:07 AM
Also, I noticed that it's reporting a 1meg device but the rom file I downloaded from your link is 2meg.

I would also try disconnecting the power pin from your cable, and using the computer power supply to power the motherboard/chip.

Later,
dabone

Dabone:

Good catch there. I looked up the specs on the MB at Asus's site and indeed it uses a 16Mbit part (matching up with the 2MB BIOS). I wonder if these older eyes of mine misread a 6 for an 8 (the 16Mbit part number is 25VF016B). I'll have to check with a magnifying glass when I get back home tonight!

Shadow Lord
October 19th, 2013, 04:31 PM
Why wouldn't the .ROM file from Asus work? Its a straight ROM image. I'm surprised a board made in 2008 has a SOIC BIOS. My C2D board still has a PLCC style chip.

I wish mine had a socketed PLCC BOS as well, things may have been infinitely easier. The ROM image was my impression too but what Dabone says can make sense too: for example the BIOS "supposedly" has recovery options built in. Well if those get erased during the programming process (i.e. the first step in the flash procedure) then you could not make use of them. I know BIOSes are modular so perhaps the complete BIOS is bigger and contains the recovery modules and the BIOS modules that are regularly updated. Of course I could be completely wrong in this and you should wait for Dabone to weigh in.

A slight update: I looked at the BIOS again and the model number is wiped like this 25VF***B. I can make out a little curved pattern before the B which is why I assumed it was the 8. However, setting the programmer to 25VF016B (or even 25VF032B) made no difference. I've ordered a replacement BIOS chip but it is coming from Europe and will take a while to get here. I've also ordered a SOIC8 clip, but I am not sure it will be any help as there are resistors around the BIOS and I am afraid the clip won't be able to get a good grip.

Shadow Lord
October 19th, 2013, 04:32 PM
I would also try disconnecting the power pin from your cable, and using the computer power supply to power the motherboard/chip.

Later,
dabone

What do you mean exactly? Would you leave the computer with standby power and try programming or are you saying to get 3V straigth from the PSU?

dabone
October 19th, 2013, 04:45 PM
Turn on the computer. Check the voltages on the bios, if it's powering the bios, try reading it or programming it without connecting the power pin from your programmer.


Later,
dabone

Shadow Lord
October 20th, 2013, 02:33 PM
Turn on the computer. Check the voltages on the bios, if it's powering the bios, try reading it or programming it without connecting the power pin from your programmer.


Later,
dabone


Dabone,

So I did as you suggested. Turned on the computer (light on, fans on full blast) and checked power to the BIOS. BIOS was getting 3V. So I tried programming it again and ended up with the same results.

I also noticed something. If I erase the BIOS and try to boot I get into a loop of boot-shutdown-reboot with the computer and I have to unplug it to shut it down. However, if I try to program it and then turn on the system the system comes on and stays on. So it seems as if something is being written but not correctly.

Chuck(G)
October 20th, 2013, 02:54 PM
Can your programmer provide details on the verify failure?

Shadow Lord
October 20th, 2013, 03:03 PM
Can your programmer provide details on the verify failure?

Unfortunately no. This is a Genius G540 with the latest SW version. The SW is barely in English and the manual is in poorly written Engrish. There is no other info aside from "verify error".

dabone
October 20th, 2013, 05:40 PM
Do a erase on the chip. Then read the chip. Look at the data to make sure the chip is blank.
Then try programming it. After the verify fails, read the chip again and see what if any programmed.

The last time I did this I ended up removing the chip, soldering some kynar wire to it and a socket, and then programming it while it was hanging mid air from the wires. Leaving it in the board didn't work at all for me.



Later,
dabone

Shadow Lord
October 20th, 2013, 10:39 PM
Do a erase on the chip. Then read the chip. Look at the data to make sure the chip is blank.
Then try programming it. After the verify fails, read the chip again and see what if any programmed.

The last time I did this I ended up removing the chip, soldering some kynar wire to it and a socket, and then programming it while it was hanging mid air from the wires. Leaving it in the board didn't work at all for me.



Later,
dabone

It is looking like I may have to do that as well. I had a question: can I piggyback a good BIOS on to the bad one? I.E. Using two tests clip can I connect a good BIOS chip to the legs of the bad one? I would of course erase the bad one first so there would be nothing to load. Would that work? Or would I need to cut the power to the BAD BIOS as well?

EDIT: Looks like theoretically it is possible...

dabone
October 21st, 2013, 09:49 AM
No, you need to remove the old chip. Do not try piggybacking roms.
These are very easy to remove, and easy to put back as long as you have a fine iron and some solder wick.

Here's a quick example.

http://www.youtube.com/watch?v=LHMmG3NuM9U

Later,
dabone

Chuck(G)
October 21st, 2013, 10:30 AM
Again, unless you've had some experience doing it, investigate Chip-Quik. You can also powder up any common fusible low-temp alloy, such as Cerrobend. I used a PAR-38 spotlight to heat the area.

Shadow Lord
October 21st, 2013, 10:53 AM
Again, unless you've had some experience doing it, investigate Chip-Quik. You can also powder up any common fusible low-temp alloy, such as Cerrobend. I used a PAR-38 spotlight to heat the area.

Yes. If I got the removal route that's what I would do.

Shadow Lord
November 9th, 2013, 08:13 AM
Small update: I am still awaiting he replacement BIOS from Europe to arrive. In the mean time I am continuing to try different things. Last night I tried to blank the chip. I am sure this had worked for me previously but now give the following error:

9: BlankOperation: Error ****
Addr:0000000H Device:00H

Could this mean the chip is bad?

I've also made further progress on the verify issue. If I read the chip in its current state and then do a verify immediately after, the verify operation works fine. However it fails after the write operation. This leads me to believe that in fact it is the write operation which is failing (even though it is reported as OK) and the verify operation is working fine (the contents of the chip don't match the BIOS file hence the error!).

I still have not removed the chip from the MB - I am waiting for the replacement to arrive first.

Shadow Lord
November 29th, 2013, 08:40 AM
Well after waiting a few weeks to get my replacement BIOS it arrives and while the correct BIOS code is on it they sent me a DIP8 instead of a SOIC8 chip! Grrr...

I've gone ahead and bought a replacement board and processor. So now that there is no rush/urgency to fix this board I am going to try Chuck's suggestion and see if I can get the chip off and solder on a new chip! Will post more progress as time goes on. Thaks for the help everyone.

dabone
November 29th, 2013, 09:25 AM
Dump the bin file from the dip chip, remove the soic and then reprogram it with the code from the dip, and then reinstall.

Later,
dabone

Shadow Lord
November 29th, 2013, 10:26 PM
Dump the bin file from the dip chip, remove the soic and then reprogram it with the code from the dip, and then reinstall.

Later,
dabone

Step 1. Done
Step 2. Done
Step 3. Error! :D I get the same result with the BIOS removed and the chip connected to the programmer. Namely, It supposedly erases, supposedly writes, but then verify fails. Reading the chip after a write operation produces a blank chip.

So either the chip is bad or I am doing something wrong. Of course at this point it may be moot as I think I took off one of the pads while removing the chip. Yes, it looks real easy on the video with the Chip Quik and to be honest I got 7 pins off but I must have not done so well on the last pin as I was trying to avoid a resistor right next to it... The trace is still there so I know it is possible to save it. I may try and buy a new blank SOIC chip just to see if it can be programmed by my programmer. If I can program it I will try resoldering the new chip in. Other than that I am calling it a bona fide bust!