I was preparing to install Windows XP, and one of the things I had noticed on the various web sites I had visited during my preparations was that a great many of the problems users were having with XP were due to old and outdated motherboard BIOSes. My own ASUS A7V133 motherboard was one of the earlier production units, and, as such, it still had BIOS revision 1001C in place, which was clearly out of date given that the most recent update available was revision 1007. I figured it was time for a BIOS upgrade, so I visited the ASUS web site and downloaded the proper BIOS update file for my motherboard.
As someone who has long been involved with computers and technology, I've flashed dozens of different devices, and though I felt pretty confident I knew what I was doing, I took the time to read carefully the directions provided by ASUS. Experience has demonstrated to me again and again that reading manuals, instructions, etc. is the very best preventative "medicine" to avoid problems with technology. In this case, however, the directions were more trouble than they were worth for a couple of reasons. First, the ASUS web site article on BIOS flashing indicated clearly that I must use the ASUS pflash.exe utility for my Plug-and-Play empowered system, yet this utility was nowhere to be found on their download page. Fortunately, the user's manual for my motherboard suggested the correct utility, aflash.exe, which was available.
Armed now with both the proper flashing utility and BIOS update file, I proceeded as always when flashing a motherboard. That is, I created a fresh, bootable floppy disc for my operating system (Windows 98 at the time), disabled all memory resident programs and so forth by temporarily renaming my config.sys and autoexec.bat files, and even took care to disable GoBack to ensure that it wasn't lurking in memory. I doubted GoBack would be a problem, but when flashing a BIOS, it's better to be safe than sorry. After copying the necessary files to C: and placing the bootable floppy in the drive, I restarted the system.
My efforts were rewarded relatively quickly with a completely clean DOS environment in which to proceed. Thus, I launched the flashing utility, noting with satisfaction that it correctly identified my current motherboard and BIOS revision. I then used the utility to copy my current BIOS to a file on the hard drive just in case something should turn out to be wrong with the new file. Finally, taking a deep breath, I used the flashing utility to update the motherboard with the new BIOS revision. It found the file I specified, confirmed that it was correct for my motherboard, confirmed that I really wanted to proceed and then flashed the chip without incident.
Once the flashing was complete, the utility suggested that I reboot and enter the BIOS setup program to reset the checksum value. Since everything had gone as expected, and the utility reported success, I figured my work was done. I gave my machine the three-fingered salute (i.e., Ctl+Alt+Del) and looked forward to a smooth installation of Windows XP.
Unfortunately, I found myself staring at a black screen. My gaze lingered for some time until I was certain that my computer wasn't just taking a little extra time to restart. When it became obvious something was wrong, I then wondered whether perhaps a simple "warm" boot was insufficient. I had just flashed my BIOS, after all, and it wouldn't be unreasonable for a motherboard vendor to expect a full system reset. Thus, I pressed the reset button, held it for a few seconds (just to be sure) and let it go. Lights began to blink, the hard drive began to spin up, and I figured my work was done.
Unfortunately, I found myself again staring at a black screen. Fear began to creep into the corners of my consciousness at this point, but I suppressed it and retained my composure. I decided that perhaps a complete power-cycling was in order. I had just flashed my BIOS, after all, and it wouldn't be unreasonable for a motherboard vendor to expect a completely "cold" boot... would it? I powered the system down completely, waited a ten-count for capacitors to discharge, and restarted the computer. Again, lights began to blink, the hard drive began to spin up, and I really hoped my work was done.
Yes, you guessed it: I was still staring at a black screen. My monitor's subsequent "decision" to quit trying to acquire a video signal and go completely dark provided an odd sort of reassurance to me that the death knell for my system had indeed just been sounded. The fear began to turn to anger as I continued (barely) to retain my composure and took the very last step of which I could think. That is, I powered the system down completely, opened up the computer's case and disconnected the power connector from the motherboard itself. You never know with these "smart" systems, I tried to reassure myself; the mere fact that I've turned it off might not mean it's really off, you know. After waiting a full minute—that was a pretty tense minute, I must say—I reconnected the motherboard power connector and started the system.
Dead. My motherboard was officially brain-dead simply because I had the gall to use a utility and file provided by the motherboard manufacturer to update their own product. What could I have been thinking!?! Fear was no longer my problem; rage was now my problem. I knew from my past dealings with ASUS that I would get nowhere with their support department. The state of the hardware and software industries is such right now that manufacturers take precisely zero responsibility for anything they produce, using any excuse to avoid providing help. Even if I could reach an ASUS representative, which would be near-miraculous in itself, I knew exactly what to expect. The company line would surely be that BIOS updates are an unsupported tool, the use of which voids the warranty for your motherboard. In short, I was screwed.
Since it was early on the morning of Christmas Eve 2001, I figured that my only hope to fix the system before presents were unwrapped—hey, I might get a neat video game for Christmas (grin)—was to head to my local Fry's Electronics and buy a new motherboard. I was surprised to discover that Fry's had dozens of ASUS A7V133 motherboards on sale for $99 each. I was originally thinking that this situation might provide a good opportunity for me to update my motherboard to one of the new boards that support DDR RAM, but after doing some pricing, I decided that simply replacing my existing motherboard as cheaply and quickly as possible was definitely the thing to do. Thus, I purchased another A7V133 and headed home.
Now, you might be wondering: why did I just buy another one of the very boards that failed?! That's a good question. Essentially, I was tired, frustrated, angry and simply wanted to be up and running as quickly and as cheaply as possible. I didn't want to spend time reading a motherboard manual, setting jumpers, fiddling about with a whole new set of problems and so forth. I figured by sticking with the A7V133, I was ensuring that my experience would be of a simple, drop-in replacement. I should only have to rip out the old board, transfer all the connectors directly and restart the system. I would be stuck with the devil I knew, in effect, rather than with a new devil.
That assumption turned out to be correct. Upon arriving home, I opened up the computer, completed the transfer in only a couple of minutes and restarted the system. Lights began to blink, the hard drive began to spin up and... I had a working system with motherboard BIOS revision 1005C in place. I was $99 poorer, and I had lost a couple of hours of my time, but I had learned two important lessons: (1) never flash an ASUS BIOS, and (2) never buy another ASUS product. I hated to accept such things, and I dreaded what problems might come with Windows XP as a result of my out-of-date BIOS, but I didn't particularly feel like potentially blowing another $99 by flashing the BIOS on this new board as well.
As Christmas gave way and the calendar flipped onward toward New Year's Day, I got to thinking about the ethics of my situation. I have certain expectations for the products I purchase, among which is that if they fail during a warranty period under normal use, then the vendor will replace or repair them. I know that applying BIOS updates voids warranties, but is this really reasonable? I didn't think so. After all, BIOS updates are only necessary to fix defects in the original product, and they're provided directly by the manufacturer toward that end. In short, I felt screwed by ASUS, and the more I thought about it, the more I felt like doing what I could to pass on the cost of the new board to them. I talked to a friend whose advice I value, and he suggested something of which I hadn't thought, namely, talking to a Fry's representative about this situation.
So, I placed a quick call to a technician with whom I've dealt before and chatted with him about the situation. Perhaps it's because he's a fellow geek, but he sympathized with my cause, and he suggested that I package up the old board and bring it in to Fry's. He said that I should take it to the guys in customer service, tell them it doesn't work, and have them ship it back to ASUS, crediting me in the process for the returned item. This was quite a relief, really, as money has been a bit tight of late, and the $99 was an expense I really didn't need. Thus, I did what he said. I took the board back on 01/04/2002, explained the situation to the fellow at the desk, and essentially threw myself on the mercy of Fry's.
The fellow in customer service had to call his tech department to make sure this was kosher, but when they signed off on it, he began to process the board as a return. When he hooked the board up to their diagnostic teest-bench to ensure that it wasn't working, however, I was greeted with a rude shock. He connected everything, plugged a known-good Athlon CPU and video card into it, and applied power. The board booted up immediately, showing BIOS revision 1007. You could have knocked me over with a feather as I stood there slack-jawed and dumb-struck staring at the screen. He must have thought I was some kind of idiot who just didn't know how to reboot a computer or something. I was in a complete daze as he repackaged it and put it back on the shelf to be sold as a working product.
When I got home that afternoon I just couldn't leave well enough alone. I couldn't understand how it worked on his diagnostic bench but wouldn't work in my home system. I called the same friend who had served as an ethical reality-check just a day or so ago to ask if he could think of anything I had failed to try. The only thing he could think of was something that I had neglected to consider, namely, manually clearing the CMOS data before attempting the reboot. This didn't occur to me because every other flashing I've ever done automatically clears anything that needs to be cleared for a successful restart.
After hanging up the phone and thinking about it for a bit, I ran into the second major failing of the ASUS documentation, namely, that it said absolutely nothing one way or the other about such clearing. Despite the dearth of evidence, I felt certain that something must have remained undone, and I'm just not the kind of fellow to let such a thing go. The reason for my stubbornness is pretty simple: if I agree not to update my system for fear of hosing something, then I implicitly agree to accept substandard performance and unnecessary limitations on the use of my computer.
Thus, I got out my boot disk, took all the necessary steps (this time with Windows XP already installed and running—more on that hassle elsewhere), rebooted the system to a clean DOS environment and updated the new motherboard's BIOS from 1005C to 1007. Sure enough, when I rebooted, the system was again "dead". I took all of the steps I had taken previously just to be sure I hadn't simply missed something last time. I hadn't. I then powered the system down, opened up the case, pulled the CMOS backup battery and shorted the contacts on the motherboard to force a complete erasure of the CMOS data. I then waited ten, long, harrowing minutes just to be sure that the data would indeed be cleared. Finally, I reinserted the battery, reconnected everything, and restarted the system. Lights began to blink, the hard drive began to spin up and...
Success! The system booted with BIOS revision 1007 clearly in place. I went into the BIOS setup, twiddled some settings to optimize performance, generally inspected everything to make sure I was in good shape, and rebooted the system again. This time I let it go all the way. Within a minute or two, I was staring again at my Windows XP desktop with quite a sense of relief, as well as a feeling that I had accomplished something. I wish I had thought to do that with the other motherboard. I say that because the older board was of higher quality than the new one insofar as it could be overclocked quite a bit and remain completely stable, while the new board fails if I increase the CPU multiplier even a single notch. Oh well. I guess that's the price I pay for rushing to judgement.
The moral of the story is simple: if a motherboard "dies" after having its BIOS flashed successfully, try clearing the CMOS data manually. This shouldn't be necessary insofar as the flash utility should really do this kind of thing for you. As long as I know it has to be done, however, I can't get too upset. What really bothers me in this case is that ASUS' documentation is more trouble than its worth. I can only imagine how many other poor souls get stuck in exactly the same situation, and that's precisely why I'm writing up this incident. Here's hoping that my pains will help someone else avoid the same problem.
01/04/2002