Fried Asus A7V Motherboard

Here's a repeat of the original article first posted on or about Sep 03, 2001:

Let me start out by saying that, aside from a few minor glitches here and there, I've had no trouble with the Asus A7V motherboard, and had every intention of using it for several more years to come. Unfortunately, that doesn't seem to be an option anymore.

In my Tower of Power project (now defunct), I was considering networking several computers together into one big processing machine to augment my main system. I had finally obtained my first "node" of this network, an old Asus P5A motherboard and an AMD K6-2 400 MHz system which I knew would overclock to at least 450 MHz, as seen in Overclocking the AMD K6-2. (The fact that I've gotten to the point to where I reference my own articles more than other people's scares me.) This MB/CPU was left over when I upgraded my brother Tony's computer to a Shuttle AK-11 board running my old 800 MHz Athlon Thunderbird.

To use my new "node" I'd need some extra equipment. I had some RAM laying about, as well as an extra monitor. I was able to purchase a new mouse, keyboard, floppy drive, and power supply quite cheaply. I could do without a case for now, just putting things together on the workbench. For the last components, I begged/borrowed a Trident 4MB video card and a little 1.2 GB harddrive. I figured these would be perfect for a "node" that would mostly just crunch numbers.

I put the system together on the workbench and even got it too boot from floppy, as the harddrive didn't seem to have an operating system installed on it. It did have a whole bunch of software and utilities and other stuff I wanted to keep. So, I took it inside to hook it up to my main system (A7V, 1.2 GHz AMD Athlon 200FSB) and copy everything over to a folder in my 30GB Maxtor.

I set the 1.2 GB drive to slave mode, hooked it to a spare IDE cable end, and rebooted. Hmm, no new harddrive. I checked the BIOS settings, found that I'd set all my slave drives to NONE to speed up boot time, so I changed that. Still no new harddrive. I tried a different connector. Still no new harddrive. Hmm.

Okay, so this wasn't working. The A7V has two ATA-100 IDE headers and two ATA-66 IDE headers, each header supporting two devices, for a total of 8 drives. I'd tried plugging the 1.2 GB drive into each of the two ATA-66 headers in turn as slaves to the DVD and CD-R drives, respectively, with no results. I tried disconnecting the CD-R drive and hooking the 1.2 GB drive as master on that cable. Still didn't work. I was starting to get very confused. I mean, it had worked fine with the other motherboard.

I decided to try connecting it to one of my ATA-100 connectors. As I reached for the drive, I noticed that some of my case fans were turning, which they shouldn't have been doing because I'd switched them off before I started messing around with the innards of my system. But here a fan was turning. I looked at my switch panel, the display of which showed that 12V was going through the circuit (green LED was flashing), even though the switch was plainly in the off position. Not thinking, I switched it to 12V, 7V, and off a few times. No reason. Just fiddling. I then heard something sizzling. I reached to unplug the 1.2 GB harddrive and burned my fingers on it. I powered down the system and reconstructed what happened.

I use a Baybus with 7V/12V Switch to quiet down my case fans by running them at 7V or even turning them off now and then rather than run them at 12V all the time. I was a little sloppy with the wiring of one of the switches, so I'm not sure what the 5V line and its ground coming out of that switch is hooked up to. This wasn't important, because I was only planning to plug in fans, which only used the 12V line and its ground anyway. I do know that the 12V line's partner ground goes to 5V to run the fans at 7V.

What I had done was plug the harddrive into the first empty power connector that I found, which just happened to be one controlled by the 7V/12V switch with questionable 5V and partner ground connections. At best, it was getting 7V where it was expecting 12V. At worst, it was getting 7V where it was expecting 5V, or maybe even 5V reverse current where it was expecting 5V. I haven't pulled the switch out to check yet. However, hooking it back up to the P5A shows that it's dead, dead, dead. The drive will still spin up, but the IDE controller says "What harddrive?"

I gave a prayer of thanks that I didn't flip that switch when I had the 40GB Seagate hooked up the day before. I return the 1.2 GB drive to the guy I borrowed it from, hoping he'll be able to fix it, and tell him I'll pay for it if he can't. Things get busy, so it's the next day before I get around to booting up my A7V system again.

Beep! Beepbeep!

WTF? Let's try that again.

Beep! Beepbeep!

I check all my cables, cards, etc. Try to boot again. Same thing. Check to make sure the little green LED on the board is coming on, meaning that the A7V is getting power. I check the A7V motherboard manual. Nothing that matches a long beep followed by two short beeps. There is a long beep followed by three short beeps listed as video card error. Maybe I'm just not hearing the pause between the long beep and the first short beep. I reseat the video card, which doesn't help. I take the Asus V7100 video card and try it in the P5A, in which it works fine. Insert the Trident 4 MB card in the A7V. Beep! Beepbeep!

I decide to drop back and punt. I dig into my system, find the solder pads to short to clear the RTCLK RAM, and incidentally all my CMOS and BIOS settings, and do so. I try to reboot. Beeps.

I call the local computer shop. They suggest reseating all the cards and cables. I tell them that I've already done so, but do again, and then try to boot up while I have them on the phone. System boots to BIOS menu. Thank them and hang up. Check the BIOS settings. Save settings and Exit. Beep! Beepbeep!

For the next hour or so, the pattern repeats. I can get the system to boot, it goes to the BIOS menu, I save and exit, it fails to boot. Turn the system off for a few minutes. Repeat. Eventually, I decide to exit without saving, and low and behold the system boots!

I get online and find a site that says that one long beep followed by two short beeps means damaged DRAM. I have two sticks of generic 256 MB PC-133 RAM. I try it with just one and then just the other. I swap their positions. No good. I try putting one in the slot farthest from the CPU, which I've been avoiding because it's hard to get to under all those IDE cables. I get a series of long beeps. Manual says that this means no DRAM is detected.

Naturally, both sticks of 256 MB RAM test fine in the P5A. And the P5A's stick of 128 MB PC-133 RAM causes the same error beeps when used in the A7V. I put all three sticks into the P5A and marvel at a 400 MHz computer with 640 MB of RAM, remembering back when I had a 4 MHz computer with 640 KB of RAM. Then it occurs to me that I tried to run 640 MB in the A7V back when I first got the RAM in, but I got that series of long beeps every time I tried to boot, which didn't go away until I took the RAM out of the farthest slot.

Hmm. Obviously, the A7V had a problem with the farthest slot some time ago, which I overlooked because I was so happy to finally be up to 512 MB of RAM. Whatever I did to fry the harddrive, it appears to have also put the DRAM controller over the edge.

I read somewhere that in addition to shorting the solder pads, you have to remove the RTC battery to get the CMOS to reset. I do that. I try to boot. Beep! Beepbeep!

I find that by making the attempt 20 or 30 times, I can get the A7V to boot, and that it will run for a little while. Running SETI@home will cause it to lock up quickly, but I can do pretty much anything else, including editing video. Now I reboot as seldom as possible, and everytime that I have to shut the system down, I pray that it starts up again.

I ordered another Shuttle AK-11 board, same as what I used in my brother's system. It was very easy to set up and offered no surprises. I've not given up on trying to fix the A7V, and I've certainly not given up on Asus boards in general, but the AK-11 is cheap and I was kicking myself for not trying my 1.2 GHz AMD Athlon 200FSB in it while I had the chance.

I'll let you know how it all turns out.

Update on Feb 24, 2005:

A recent email reminded me that I never did let you know how it all turned out. The Asus A7V never did get any better, but never got any worse, either. I loaned it to a couple of friends who were convinced that they could fix it, but couldn't. I think we ended up throwing it away.

I've gotten three other emails about the Asus A7V (one from as far away as Finland!) describing the same problems, always after working on hard drives. The problem is very uncommon, apparently, but not unheard of.

The Shuttle AK-11 board (which I purchased simply because it was the cheapest I could find at the time) worked with the 1.2 GHz AMD Athlon processor I had at the time, and even allowed some minor overclocking. I eventually upgraded to 1.729 GHz AMD CPU (I forget the exact model number) which I still use at the time of this writing. The only complaint I have about it is that it doesn't recognize hard drives larger than 136 GB or so.