having some serious cpu issues

[breakingnews]

Originally posted by SteveDallas
Yeah, that's kinda what I was thinking in suggesting disabling the onboard network.

But call Dell.... hold their feet to the fire.

I did call Dell late last week ... I tried to be very firm and pissed off about the problem, but the schmuck from India with whom I spoke didn't seem to understand my pissed-off-ness (nor did he speak very coherent english).

He had me repeat steps literally two or three times, but he was serious about it. As in he made it sound like we were trying something totally new each time.

This ridiculous interchange went on until he told me to disassemble the whole computer and only plug in the critical components. I suddenly got this feeling that if i followed the tech's instructions and fucked something up, Dell would somehow manage to screw me over and say my system is no longer covered.

Next time I am going to call back and pretend to be an idiot.

[wolf]

Oh, the horrors of outsourcing.

What you need to do is throw a party and invite your geekiest friends. At some point, before there have been more than three beers distributed to each geek, scratch your head and mention conversationally, "you know, I've been having this problem ..."

If you have chosen your geeks well, you will have all the remaining beer to yourself and just have to supply them occasionally with Twinkies and Chinese Food until your problem is resolved.

[SteveDallas]

Originally posted by wolf
Oh, the horrors of outsourcing.

Well, this doesn't help breakingnews.... But our solution at work has been to start standardizing on IBM stuff, and become an IBM authorized warranty provider. They'll pay us to perform repairs on IBM equipment that's under warranty. There's a cost associated with it, but we only have to make 10 warranty repairs per year the first two years and 5 after to break even. Between faster access to parts and a more direct line into their second-level tech support, it ought to be a better solution than just coming in off the street.

[breakingnews]

So, I got new RAM chips last night and installed them - problem solved.

Now if I can only get my cable modem to work ...

[breakingnews]

Okay, so I replaced the RAM and it worked fine for ... three days, but now my computer is on the blink again.

Now it won't even start up at all. I left it on all night, but it was still stalled in the morning. What a fucking waste.

Last call to Dell, I was on hold for at least 35 minutes. I'm so fucking tired of dealing with this shit.

Question for you computer folks out there - if I give up and buy a barebone system, would I have any problems swapping out some of the components (CD-RW, HD, video)? Or is that too risky given what's going on with my computer now? Other than the starting-up issue, all the parts seem to work fine.

-bn

[SteveDallas]

You should have no trouble. But if you're going to do that, it would be more economical to just buy a motherboard & CPU as a base and add your working components to it.

[breakingnews]

Good idea. But there's something creepy about the Dell box that makes me think anything that goes inside won't work. I really wonder if it's a mobo issue too.

Plus I was thinking about getting one of those neat glowie CPU cases I see on eBay. :P

[Undertoad]

System building itself is not all that difficult, it's knowing what components go with what and which are the best choices and such, so that if you have to support it yourself you can.

The only advantage of a bare bones system is that the trickiest processes in system building are kinda done for you. Mounting the CPU, mounting the heat sink on the CPU, bolting the whole thing down correctly and attaching the basic IO ports and selecting the right DIP switches -- that's 90% of the battle right there.

But if you can convince someone to help you out with building the first time -- say, by bringing them a case of Red Hook ESB and/or a sack of something similar -- I, I mean they might be convinced to walk you through some of the heavy lifting so that you can get a truly elite system and not settle for a barebones which will probably have a truly crappy case and power supply. Then you'll know what to do next time, or at least with a question to the Cellarites. You could even score a Cellar T-shirt out of the deal, any color as long as you like black.

[Pie]

Originally Posted by SteveDallas

It would be more economical to just buy a motherboard & CPU as a base and add your working components to it.

Add a new psu to that list. Most power supplies are utter crap.
Better fan = less noise
More than enough power = fewer booting problems caused by transients

- Pie

[xoxoxoBruce]

Be sure and check with TW on the power supply.

[BrianR]

I suggest eMachines. Those are cheap and a good start if you strip them down and replace the junk you have a half decent system

[tw]

Originally Posted by breakingnews

But there's something creepy about the Dell box that makes me think anything that goes inside won't work. I really wonder if it's a mobo issue too.

If you have not first measured those DC voltages from power supply using a 3.5 digit multimeter, then you may suffer a classic example of why that measurement are the first numbers obtained.

As for locating the problem - that is what the Dell Diagnostics accomplish. Responsible computer manufacturers provide those diagnostics for free (I don't believe Gateway nor Emachine even offer such diagnostics).

I don't have a clue why the memory chips were replaced. I don't see what the reason was. However intermittent memory, like most other computer parts, are best identified by using heat. Every part should work just fine even when the room is 100 degrees F. IOW any computer that fails in a 100 degree room is defective (and the naive then cure the symptoms by installing more fans).

For example, the memory diagnostics are run repeatedly AND the memory can be heated with a hair dryer. Those SIMMs should be uncomfortable to touch but not burn skin. Intermittent (and defective) memory will most likely fail diagnostics when heated to this temperature - which is normal operating temperature to properly working SIMMs. Same applies to other system components. Heat is the tool to locate currently intermittent and failing components before they completely fail.

But again, until you have taken those power supply voltage readings and confirmed they are in the upper 3/4 of the limits in those tables, then anything else you do to fix the machine is better known as 'spinning wheels'. Only then are you ready to do the next step - manufacturer's diagnostics with heat. BTW, I don't recall reading any information from the system logs or from the Device Manager. This is an NT OS? Then important information may also exist there.

[Carbonated_Brains]

I'm fairly positive tw has been hitting the bong this morning.