The K6-II that I fried was a Retail Box with the official AMD cooler.
I kept it clean too. It fried one day when the room I was running the
machine in got a bit warm - about 85F. The fan was working - nothing
else got hot, not even the disk drive. But the K6-II got hot enough to
cause the Epox motherboard to issue an alarm - set at 60C. I looked up
the spec for the K6-II and it said maximum temp is 60C. The machine
would suddenly crash for no reason.
Sure, if your case is poorly cooled and the CPU overheats it
will crash. That has NOTHING to do with it being a K6-2.
Take your pick- most CPUs are subject to instability when
they go past 60C.
I traced it to a defective L2
cache. Then I replaced the K6-II and it didn't even have an L2 cache.
Yes, none of them did, that was the K6-3. Therefore, if you
did accurately "trace it to a defective L2 cache" it wasn't
the CPU at all, it was the board that failed... due to poor
case airflow since a properly working retail 'sink will NOT
result in the CPU getting that hot in an 85F ambient
environement.
I recall one box I built for my aunt that had a K6-2... she
didn't have air conditioning at the time and in summertime
it got up past 95F in her house. Her system remained stable
even though she typically had lots of junk running in the
background like AOL IM, a 3D screensaver, etc... so her box
was never really idle and yet no problems aside from user
discomfort at the ambient temp.
I got fed up with the monkey business and decided to give my business
to Intel.
You blamed the wrong parts and still are susceptible to same
types of problems if you run a box with inadequate airflow.
Heck, FAR MORE SO, since Intel's current-gen parts produce
about 3X the heat of any K6-2.
I did not get into the intricacies. The spec said 60C was the max
temp.
Well the details are fairly important, since the 60C temp
doesn't damage the CPU. Huge difference between a crash and
hardware damage. Regardless, the box wasn't set up right,
it was overheating due to poor design and that had nothing
at all to do with a K6-2.
Even if a K6-2 produced 1000W of heat instead of only 1/3
that of a modern Intel CPU, If/when someone builds a box
with (any particular CPU in it), they are _REQUIRED_ to
engineer the cooling solution as needed per the environment.
Any monkey can fit part A in socket B, but because of other
factors that must be considered, we tend to select someone
other than a monkey to build systems.
If the CPU overheated it was anything except the CPU
manufacturer's fault. Truely ironic that this is your
reason for NOW preferring the manufacturer of the hottest PC
CPUs that mankind has ever seen.
I use the term "fried the CPU" but in reality I fried the L2 cache.
The CPU itself ran fine when I disabled the L2 cache in the BIOS.
I really can't help it if you had damage or hardware config
problems... there were plenty of boxes that didn't.
All I have to go in is that 2.7 Amp rating on the back electrical
sticker.
All I really care about is how long to set the UPS before shutdown. I
have two shutdowns set, one for general battery drain (15 min) and one
for low battery condition (1 min). I am assuming that if the batteries
crap out in 5 nimutes there will be enough reserve to go 1 more
minute.
IMO, unless you NEED to keep the system running for some
reason, it's best to go ahead and shut it down ASAP, rather
than draining the batteries. That is, if the system is
unattended, it might as well shut down in 1 minute, but if
you're using it, you can then shut down as soon as you
finish (whatever), but still have the maximal manditory
shutdown that occurs right before the battery is exhausted.
I am going to run some tests as soon as I can find the time to chase
down the circuit breaker for my office.
There is no need to find the circuit breaker, just pull the
UPS plug out of the wall socket.
I rely on the considered opinions and experience of my primary vendor,
Directron. They sell enough to know if something is going to cause a
problem. The last thing they want is for customers to throw crap back
in their face.
No you are completely wrong.
Directron sells good parts and crap, both.
Particularly they sell crap PSU in cheap cases that come
"with generic PSU", just like other vendors do.
You have an unrealistic bias towards Directron for whatever
reason- perhaps their proximity, perhaps you work there. It
really doesn't matter, they're just another vendor that
largely made their mark due to having a more polished
website earlier than many other vendors.
While they occasionally have decent deals on a few parts,
their service is pathetic. I won't even go into all the
times I received broken cases from them - cases that had
large obvious damage with pieces broken off but the missing
parts weren't even in the (pristine) box.
After conversing with tech support, I'd be informed after a
couple weeks of waiting that they were then "back from
vacation and could take care of (the problem). Directron
wouldn't have survived the past few years if it weren't for
their fancy website.
Certainly there ARE far worse vendors out there, but
Directron is hardly a company to have this much confidence
in.
As mentioned the importer of that case (with PSU included) is just a
couple of buildings away from Directron so the people from both
businesses know one another. The case supplier is reputable.
That's more than can be said about auto dealerships.
Why would anyone care if the case importer is two doors
down or 3 states away?
"Reputable" means close to zero in this context. It can
still be good, or junk. The importer can be a really great
company simply importing whatever sells. Whatever sells,
often sells because of the low price point. There is plenty
of junk out there, one can without question go to
Directron's 'site right now and find (IMO) fraudulently
rated power supplies.
Here's one example
http://www.directron.com/directron400w.html