2 hardware issues

[Stonegiant]

I have 2 computers both running Windows XP Pro.Machine 1 is a bit faster than Machine 2. Both are AMD Athlons (1 is a 1.8 gig and 2 is a 1.53 gig (I think)).2 started randomly rebooting on me thirty seconds to five minutes. I've switched the power supply with Machine 1 and I had the same symptoms.1 was running when I lost power (some clown ran into a utility pole and dropped the power lines onto himself. He was going fast enough in a parking lot to break a utility pole...). On reboot, there was no indication that video functioned. I replaced the video with a new card (machine 2's video card was hot-glued in by an idiotic computer repair person while I lived in Georgia. He never told me he did it, so I coun't swap video cards for verification). Same symptoms with the new video card.I think 1 is suffering from a bad main board and 2 a bad processor.Any insights?

[GolfProRM]

Swap processors and see what happens

[nlinecomputers]

I'd say both units have bad motherboards.

[Stonegiant]

Swap processors and see what happens

I know I'm going to have to eventually do that, but do they have to make those so freakin hard to remove? I tried for half an hour to remove either one. No success.

[teacher]

Usually it is not the processor itself that is difficult to remove but the cooler/fan combination on top. It helps if you know where to look to release those. Some are screwed in and others clip on. I have virtually trashed some trying to figure it out. At least they are cheap to replace if you don't have an expensive one already. I am still choking over the glued in card. I have never heard of anyone doing that. It really makes me wonder why on earth someone would do that.

[Stonegiant]

Yeah, I meant the heatsink/fan. Good catch.They are both flat pieces of metal that go across the interior of the heatsink. Both sides hook onto one of the three protruding squares. One side is spring returns to hold onto the square (Yes, I know there's no springs. It's a just a spring effect of the bent metal ). The problem arises when I am only able to push that side down. I have no tools that will let me pry the slot A off tab B. Maybe I'll borrow something from work to see what I can do.

[NRD]

Look at the opposite end of that metal retention bar. There should be a small metal tab that you can fit a flat head screw driver into. You want to gently push down and away. When that end is free from the tabs, you can remove the other end by hand. Another trick is to pull the ink tube out of a Bic type pen. You then fit the hollow pen body over the clip and push down and away. This aviods the chance that the screw driver will come out of the slot and damage the motherboard.When the retention clip is free from the tabs, you can GENTLY wiggle the heatsink from the processor ( it may be stuck if a thermal pad was used) or you can run the PC for a minute or so to warm up the pad. This will allow easier removal. Just make sure the PC is unplugged before you attempt removal. Gently applying heat from a hair dryer to the heatsink can also work.Here are a few videos from AMD that show how to remove and replace a heatsink.AMD Video

Edited December 28, 2004 by NRD

[b2cm]

Also check for cpu overheating.

[Stonegiant]

Also check for cpu overheating.

Unfortunately, I don't have any software (that I know of) on machine 2 that I can use to monitor temp. It's quite difficult to install any when I never know how long I have. Oh, one other thing about machine 2. I was going to check if it were some kind of scumware/virus (same thing in my book), so I was booting up in safe mode. It rebooted on me when it was showing the files it was loading.Thanks for the tutorial, NRD. I'd thought of using a flat-head screwdriver, but didn't have one small enough at home (read: I didn't have one at all ).

[b2cm]

Open the case and train a desk fan at the insides. Start the computer, run some programs and see if the reboot occurs.

[Stonegiant]

I was at tbird's house (my older brother for those that still don't know ) on New Year's Eve. My twin brother was also there (he works in IT for the US Army for NASA). We spent about 5 hours working on machine 1. Tbird had known goods for all pieces/parts. Here's what we learned (if I remember correctly): Sometimes the main board worked with all known good parts. Sometimes the video worked with all known goods. Sometimes the processor worked with all known goods. I think the RAM was the same way, but not sure (it was quite a trying time for me).So, I had a power surge (likely caused by inductive kick from a power loss) that is causing intermittents on my CPU, M/B and my video all at the same time? :'( Tbird will correct me (hopefully) if he remembers better than I do.

[Grasshopper]

Nope, you said it right. There were no consistencies found from our testing. We were very scientific. We had plenty of known good parts to swap out for testing, including a motherboard, which "proved" that all his parts were good, but then, when the peripherals were tested on his old mobo and the new one he bought, results were not consistent at all. Sometimes 50/50.I find it very hard to believe that he's had two bad mobo's in a row, even if one was his for year or so and one was new. Good RAM was tested and had mixed results.His old vid card and one of my known goods...mixed.We had enough processors to go around....you guessed it.Power supplies???....yep.This one has gotten the best of three good computer minds.

Edited January 4, 2005 by tbird

[rbdietz]

AC Power cord?Correctly wired AC (proper ground)?

[Ed_P]

Low voltage? Static due to other appliances/equipment on the circuit? Furnace/AC motor kicking on/off, etc.

[NRD]

Maybe try hooking the system to a UPS? It may help rule out line trouble rbdietz & Edp suggested

[Grasshopper]

Good points, but I kind of doubt that's the answer. The same symptoms showed up at his apartment, the OK Dept. of Ag building and also at my house.

[nlinecomputers]

Monitors? Keyboards, Mice, Printers. Any item that is plugged into the system is suspect at this point.

[Stonegiant]

Different monitors (home, work, and tbird's house).Different Mice, K/Bs at home and work, none at tbird's house. Never any printers.

[nlinecomputers]

Could you find any part that is for a certain dead. You may have the bad luck of a "tainted" part. One that appears to work but kills anything it is plugged in to. So that when you do a swap out test you fry other componants on connection. It is also possible that you have more then one part bad.The only other thing to suggest is that you try again making sure to trade only ONE part at a time and put into a fully assembled working system.

[Grasshopper]

Nathan, Yeah, we did the one at a time method. We didn't before and it didn't get us anywhere so we had to make time...We didn't notice a tainted part. It seems that every part worked at some point in time, including PS, chip, mem, mobo, vid. If a part worked we tried to go with it and change something else. When that didn't work, we'd go back to the first set up and it wouldn't work. Then the second step, that didn't work...would work...including my known good parts. Just nothing to follow. You are prolly right though and it is just very tedious to find.

[b2cm]

Disassemble and begin only with the board (atop the box of the new board), processor, a memory stick, keyboard and video card. Clear the CMOS.

[patio]

I was just going to suggest the same.I have a rubber mat just for this purpose...Start over with all components out of the case to eliminate a possible short/ground issue...patio.

[Stonegiant]

I had meant to mention that before. When we were doing the "chop and swap" of the parts at tbird's house, we were doing it outside of the case (on a foamy-type mouse pad). Prior to that, everything happened inside the case (at my house and at my work). Same results.

[b2cm]

It looks like you have two failing mobos. Can you post the brand and model/type of these mobos?

[Stonegiant]

No, that would be 4 failing mobos.My original was a Soyo Dragon Platinum KT400.The one I used at work was an Asus A7N266.The one I bought new was a Soyo Dragon Plus KT600 v2.0.I don't know what the one at tbird's house was off the top of my head.All 4 boards were experiencing the same intermittent problems as long as at least one piece of hardware was from the original machine (RAM, CPU, P/S, Mobo or Video Card).

[b2cm]

These are good mobos/vendors. I'd look more closely at the processor or mobo settings for the processor (and maybe for memory).

[Stonegiant]

Please don't take this wrong, but have you read the previous posts on this thread?With a known good P/S, CPU, RAM and Video, sometimes the video wouldn't work.With a known good P/S, CPU, RAM and Mobo, sometimes the video wouldn't work.With a known good P/S, RAM, Mobo and Video, sometimes the video wouldn't work.With a known good P/S, CPU, Mobo and Video, sometimes the video wouldn't work.That's where it stands.If it were just a single (or even 2 components), all the rest would always work. If one or two of these components are cause issues with the others, there's no real way to determine the bad components without costing hundreds (thousands?) of dollars attempting to isolate it. Again, I'm not trying to be a jerk. I am more than a little frustrated by all this. I originally had two computers (when I could afford to buy two) because I was (still am) a hardcore gamer. If one goes down, I'd swap to the other without missing a beat. Now, I probably won't be able to get one computer up for several weeks.

[NRD]

EDITED by MESo let me get this straight, you are saying the every component from the original machine is bad? And no matter what component you try in a good machine, the same error occurs? Failing video.

Edited January 5, 2005 by NRD

[Cluttermagnet]

A pretty vexing problem. FWIW, the time I had suspected inductive pickup of not-so-close-by lightning (rode in on the phone cable, all power cables were disconnected), I ended up buying a new mobo, winmodem, hard drive, and video card. ESD damage can be subtle. It drives you crazy because sometimes your devices start out working OK and then crap out unpredictably. Not in my case. The mobo could still talk to the floppy, but never to any IDE devices. The modem was 'flaky', but worked somewhat, at greatly reduced data rate (sometimes). My video card failed a short time later, and the screen display at the moment of failure was unambiguously spectacular. And so on. You replace one part, another proves to be flaky. In the case of lightning hits, entire computers sometimes end up getting scrapped, after much frustrating waste of time and money.Power line glitches? Now that's a bit fuzzier area. It's entirely possible a big, fat spike rode in and got past the power supply and onto supply busses and just randomly took out various semiconductor devices (transistors, ICs, diodes, etc.) Or maybe scorched little intermittant carbonized paths between closely spaced conductors in ribbon cables. You take a big inductance (refrigerator compressor, etc.) and suddenly interrupt a large current flowing in it, and you can get some spectacular voltage spikes that can punch through insulation and bludgeon their way into sensitive circuits A UPS plus a separate line conditioner might well have prevented that sort of tragedy. About a hundred bucks each. Have you ever directly measured your P/S voltages? How about load testing them? Can you do that somehow? At the time of the incident, just what sort of power line protection were you running (if any)?Your situation almost defies logic. One starts to suspect very basic things such as even plain old cables. That has already been suggested in this thread. I'd mass-replace 'em in a heartbeat, just for grins. Especially the IDE cables. One almost has to suspect you have enough damaged silicon that your 'known good' parts maybe aren't really as good as you think? I'll be very interested to know if you ever isolate any definitive cause(s). You might never do that. You might just have quite a lot of 'tainted' silicon sprinkled around the various devices. I'd be very much suspecting multiple bad parts at this point. Your overall problems might just be similar to those from a lightning hit. It doesn't take much to kill a lot of silicon. And insulation, too (cables, PC boards, connectors, etc.)

[b2cm]

You may have a BIOS issue. If this issue bears on the I/O bus, including the video subsystem, then this would explain why you have failing video as well as with other components. It could be due to incorrect settings, or to a corrupt BIOS code, or, the worst case, a flaky mobo nearing its end, preventing a successful POST, or if the machine manages to boot, only to crash. (Via chipsets also have been known to have PCI latency issues that can cause a system crash.) It could be anything (or everything), but I'd focus first at the processor, memory, video card, and motherboard (BIOS, CMOS). I'd check processor vis mobo settings, video card vis mobo settings, and then memory vis mobo settings. Start with default/factory settings, and disabling integrated peirpherals such as USB, IDE, NIC and audio.