OT: Computer dust protection

I don't have a list - but over the years I've installed (and replaced) numerous power supplies rated at 84-250 volts. They were switchless "universal voltage" power supplies. Can't remember the last one - I suspect they've been"cheapened " out of the market.

There's a few that have come danged close!!!!!

anorton Inscribed thus:

> Remove a heatsink from an Intel CPU. No semiconductor damage

It does in toms world !

The CPU can overheat far far more quickly than the protection circuits can act to protect them. The protection circuits on the cpu chip are designed to reduce the risk of damage if the fan failed for some reason.

Now a 486-50 or 66 would survive without a heatsink for a short while !

>>

I went through 35 supplies and all were rated from 100 volts to 240. Years ago, there were some that were rated to 90 volts, but even then, they weren't common.

> >>

Well I know when I worked for Trillium Computer Resources back before the ATX standard was introduced, MANY of our high end systems shipped with 84-250 volt universal voltage power supplies. You've likely never heard of Trillium, but they were the first computer manufacturer in North America to provide a 3 year warranty as standard. Based in Waterloo Ontario, later taken over by EMJ, which is now Synnex.

>> >>

I've heard of them, but I've never seen one. 90 volts used to be the minimum in the minicomputer days, like the old Data General Nova I had. Of course, that was a huge linear power supply. Switching power supply OEMs can get by with smaller transistors, electrolytics and heat sinks at 100 volts. The on times are shorter, which reduces the waste heat, as well.

> >> >>

We actually MADE the Data General PC line for several years. And I was responsible for all the CD towers sold under both the Trillium and Data General names.(back before CD Rom was supported by Novel, SCO, Microsoft, or Vines.)

>> >> >>

There was a guy near here supporting the early Data general minicomputers, until Y2K fever killed his market.

I've never seen a Data general PC. I did use a DG mini with a 4 bit processor back in the '70s. It was part of a SATE system for the PRC77 Manpack Radio. It had a teletype machine for input, and a Data Products high speed drum printer for output. Talk about heavy metal! That printer would spray the ceiling with paper during a memory dump, and shake the computer room when a test failed, and all 132 hammers fired on the asterisks at once.

My oldest CD-ROM drive was an external unit that required a custom interface card and cost ^ $1000. I kept it to show to the 'Know it alls'. ;-)

> >> >> >>

A Hitachi by chance? Trillium was the first Hitachi CD ROM distributer in Canada, and for a number of years the largest (not only Hitachi, but CD ROM Period) IIRC the 1504? external proprietary interface. I designed and had built an interface adapter to put internal Hitachi interface drives in external cases (4 at a time? cannot remember for sure what the limit was, before the SCSI interface drives were available (7 in a case, then later 14)

>> >> >> >>

I think so, but I'd have to dig it out to be sure.

I don't have the interface card. I found some once but the jerk wanted $350 for one, untested.

We also never put heatsinks on a 486. You should have known that. Those other accusation are also bogus. Baron, if you are going to be credible, then do not invent things so outrageous. You demonstrate knowledge from hearsay.

Most every claim of death due to overheating are from techs who could not even say which part failed. Or say what temperatures are normal for any part. They see a failure. Then somehow know that must be due to heat. Which part failed? Does not matter. They just know it must be heat. That proves it.

How to filter out so many posts based only in feelings? Where are their numbers? He could not even quote the manufacture spec number that says heat is destructive. Wild claims based in speculation prove heat must be destructive.

How did they find design defects on one Intel chipset? Hardware was operated at much greater temperatures. Even at temperatures well above spec, hardware should have been working. But a transistor failed at higher temperatures. That still working transistor that fails at higher temperatures says that semiconductors will fail years later. Heat is an ideal disgnostic tool. Posts based in reality also include the numbers. A transistor that would work fine in the first year would have become worse in 3 years. Then only 5% of those semiconductors would fail. Heat was not destructive. Heat (near zero temperature increases that other hype as disastrous) is an ideal tool to find defective hardware. And does not harm those semiconductors. That heat only causes timing and threshold changes.

Any Intel CPU can operate without damage with a heatsink removed. That Intel CPU will simply protect itself. But those educated in hearsay just know that is not true. Hearsay is fact for so many who did not bother to first learn the numbers.

Manufacturing defects are a most common reason for electronics failure. But those who cannot even identify the failed part will automatically blame heat, surges, and maybe Saddam Hussein. =93I feel it is true. Therefore Saddam must have done it.=94 Please. Replace Saddam with heat to have the same bogus claim based in hearsay or emotion.

westom Inscribed thus:

Wow ! Tom, you're dangerous !

Never put heat sinks on 486 processors? You are full of it. 486 processors require a good heat sink and reasonable air flow

No, if you replace the processor and it fixes it, the processor is dead. If the fan is seized and the heat sink is stuffed with cat hair, the processor overheated. Putting in a new processor and a clean heat sink, with proper heat transfer medium and a good fan makes the computer run properly again untill the heat sink gets plugged with hair again.

You want numbers??? Intel spec for 486 processors is available at

indicates a MAXIMUM operating temperture (chip, not ambient) of

78c on one variant, with many being limitted to as low as 67 and 69C.

There is a complete writeup on "thermal management" and it states it is the MOTHERBOARD design that can limit the temperature, using temp sensors provided by Intel in the chip.

It will ONLY protect itself if the BIOS (and motherboard) supports it and has not had the protection turned off.

And AMD processors are even more succeptible to overheat damage. They run hotter. They will often keep running longer (if not reliably) at higher temperatures and cause motherboard failures. I've had them so hot the motherboard was charred right through under the chip/socket. Eventually they just go "PFFFT!!!!" and die.

Manufacturing defects generally show up as "infant mortality". Overheating generally shows up after about a year or more, or after some fool decides the heat sink is not required (or the fan) or after someone has played with the system and broken or dislodged a heat sink clip, causing the heat sink to loose thermal contact with the chip.