Eureka!

Having been sick this past week with a serious case of a head-cold, you can imagine how antsy I am getting, being cooped up inside all week long, and all. I figured I’ll blog about my latest computer hardware disaster, that has had my Ubuntu 7.04 machine out of commission for the last 2 months.

It all started when I started experiencing random shutdowns, as if someone had just pulled th plug. At times I would literally have to wait 10 minutes for the power button to even turn the computer on again. Being the logical type I naturally deduced that it was the Power Supply. I figured that it had finally given up the ghost, after 2 years of 24/7/365 operation at 80% load. It probably had lost its edge (power supplies do experience a drop in wattage output over time).

So off I went to newegg.com to get something that would last a while: HIPER HPU-4S730-MS 730W. I know that usually modular PSUs are to be avoided, but this one turned out to b e excellent with its industrial-grade connectors.

Silly me. As if that would get rid of the gremlin that had invaded my computer… of course it was still randomly shutting off - no, not shutting down, just OFF … poof - after about 20 seconds. This would happen during boot, while I was in the BIOS, or just being idly waiting for me to insert a bootable floppy (I removed all the hard drives to eliminate any OS or HD hardware). Well, what else could it be, but the motherboard now? Right?

Having no obvious answer in mind (any other piece of defective hardware connected to the motherboard would prompt beep errors) I was at newegg.com in two clicks, buying this baby: ASUS A8N32-SLI Deluxe. Maybe this wasn’t going to be so bad after all, looks like I’ll get a whole new PC out of this when I’m done. (I didn’t want to spend too much money, so I wanted to keep my 939 AMD Athlon64 3500+, which meant I had to go with a 939 motherboard.)

Figured I would throw the gremlin out with my old motherboard. Boy, was I wrong.
Man, I was at my wits end. The only other thing that it could possibly be was the CPU. But the CPU doesn’t just shut a computer off if its bad. Well, I’ve heard of stranger things, so off I went to newegg.com, again: AMD Athlon 64 X2 4400+ Toledo.

Nothing could possibly go wrong now. And on top of it all, I now got myself a dual-core processor, which would help me multi-task the hell out of programming! muahaha… haha.

Having delusions of grandeur, I eagerly plopped in the new CPU … booting … loading OS … nice, I can get back to work! … PFFFT. Blackness. The gremlin was back! I couldn’t believe it. I requested an RMA, having tested all the hardware (memory, video, cdrom, floppy, harddrives, etc.) I made a call to ASUS tech support. After explaining the situation, and convincing them that my CPU was NOT overheating, they admitted it might be the motherboard and authorized an RMA for me.

10 days later I got the package with my now refurbished motherboard in the mail. Unfortunately I was heading to the otherside of the Globe that day for vacation in the Philippines. My computer would have to wait.

Coming back 3 weeks later, I looked forward to assembling my PC and being able to get back to work. The gremlin was still there. Someone must have soaked it in a lake of water, because this was one evil gremlin I could seem to shake: my computer STILL kept shutting off anywhere between 20 and 60 seconds after starting it up. I have tried various PSUs, video cards, memory sticks, removed all peripheral drives and PCI cards. All to no avail. The gremlin still had to be in the motherboard. So I called into ASUS again, and they admitted that it must be the motherboard. Sometimes this happens, they acknowledged, and gave me another RMA number.

I now have to send in my motherboard again (this coming Monday, since I was sick this last week) and hope for the best. My last resort is to go out and buy another motherboard, but I don’t want to waste the $170 I already put into the new ASUS.

Hoping for the best, your aspiring gremlin-slayer.

PS: if anyone has slain a gremlin of this nature in the past, please let me know what I’m overlooking here!

Update: as per Mista_Random’s comment, one of the best suggestions I’ve heard regarding this issue so far, I tried using a surrogate case (plugging in the case connectors from another case, while being connected to the PSU of the original case). Unfortunately it still is pulling the plug on itself after about a minute. Great suggestion, Mista, thanks! Too bad itdidn’t pan out.

While trying to google a fix for this issue I was experiencing, I came accross a post on a forum detailing the resolution to this problem. It appears that this is an issue that most likely only NVidia users will see if they have their Color Correction settings not set correctly. By default (at least for some driver versions) color correction is set to “Desktop”, while the fix for the issue is to set it to “All”, as shown in the screen print:

Last week my climate control unit in my bimmer gave out. Initially I thought it was the computer shutting the thing down because the compressor failed or something, but the engine check showed a-ok. After googling my issue for a bit, I discovered that others were experiencing this problem around the 100k mile marker. It turns out that supposedly a capacitor failure in the climate control computer (located directly beneath the radio for those of you with an E36 series BMW) is at fault for this. His very detailed and accurate instructions let me fix what would have been a $600 replacement at the dealer for $.50 at the local electronics store (in my case Fry’s Electronics) and 2 hours of work.

I picked up the capacitor, removed the one detailed in the howto guide, and replaced it with the new one. Two notes I’d like to add: polarity does matter! So when replacing the capacitor, be sure to sodder the new one in with polarity directed the same way as the old one. You will need a volt-meter to test for polarity using the resistance (ohm) settings by switching the contacts of the voltmeter around when testing the capacitor. One way will result in a connection, the other way will not.

Also, another note: I succesfully used the capacitor that was mentioned in the how-to guide (.47uF50), however, the capacitor that I removed was a (u47F63). The guide mentions that there may be many capacitors that work, and I stuck to the one in the guide, since everyone has reported success with it. So far I have not had any problems.

Thanks, Richard Swope, for you most excellent guide!

[Edit: 8/5/2005 17:23 | Original Post: 7/5/2005 10:06]
I have been experiencing intermitent climate control failures after installing the new capacitor. However, my initial thoughts were that there was a cold sodder contact point, meaning that when I soddered the capacitor, it did not make sufficient contact with the contact on the board, and it cooled without gaining proper contact, leading to an intermittent or even failing electrical connection.

I had initially installed the capacitor by soddering it from the bottom, where one would expect the contacts to be. However, after taking it out and resoddering it, it appears the contacts are on the “upper” side of the board, so when you apply sodder, do it from the top, not the bottom.
After completing this “fix” it seems to be working fine so far. I’ll keep an eye on it and report on any further issues I may find.