Help with Intel Mobo and Fan...

[kony]

I just came across something.. could it be the cpu is overheating????

If the heatsink isn't installed properly, sure that's
possible. Otherwise, it's doubtful as a Tualatin at stock
speed and voltage is VERY easy to keep cool relative to any
more modern CPU- a cheap and small $1 heatsink could do it
in most cases. However, if you can determine the vcore
currently used, (or at least double-check that if possible),
that could be useful. So long as the heatsink is making
good contact you could also touch-test the 'sink - if it's
not quite hot, it's not overheating.

I also
found something, go into BIOS and change the settings from automatic, to '3'
(versus '2'), this is uncharted territory for me.. does it make sense??

Hmm.
Depends a lot of what that setting is you're referring to.
If it's memory CAS setting, you should not need that
changed, providing the memory bus is still at same speed as
it was previously, while the old CPU was in use.
Regardless, running memtest86 for a few hours should help
identify memory errors- and only a moderate level of errors
should not prevent the system from turning on, beginning to
POST... though it could certainly bluescreen it later.

 

[Uncle Vinnie]

I can take a look into that.. what program would you suggest?

Also, in terms of speed, what's is the difference between an 800 piii at
133mhz versus a 1.2 piii at 100mhz?

 

[kony]

I can take a look into that.. what program would you suggest?

Take a look at what?
Test memory with memtest86, if that's what you meant. It
would be easier to follow the thread if replies were
embedded under the relevant phrases.

Also, in terms of speed, what's is the difference between an 800 piii at
133mhz versus a 1.2 piii at 100mhz?

Depends on the use of the system. On most tasks constrained
solely by CPU, there would be a nearly linear performance
increase. Naturally each task may have other mitigating
factors that vary per task.

 

[Uncle Vinnie]

I wanted to thank you folks for all your help, 'kony', etc..

The problem I had was that it was 2 problems! Undersized power supply plus
the need to run the 1.2 piii Tualatin at 1.55 volts - at least.

Swapping out the 145 Power supply for a larger one would still result in
BSOD because the Upgradware adapter locks the 1.2piii at 1.50v with a Vcore
of
1.475, which happens to work fine in my Compaq but not my Intel D815EEA.

Switching the Upgradware 370 for the LinLin adapter, even moving the jumpers
to a higher voltage of 1.55+, still was not good without a larger Power
Supply- the Tully, I guess, needs more power.

Seems all I read here, help at the Overclockers N/G, plus what caught my
attention at PowerLeap's site, held the
answer - their same adapter locks voltage at 1.55, not at 1.475 as
Upgradeware's does.

I got a larger Power Supply from New Egg- RaidMax 380 ($8 after rebate, not
bad!). I also set the LinLin adapter jumpers at 1.55, which is what the
m'board shows, with a vcore of 1.52 and change... (I have Intel Monitor).

I can't say what the exact problem was, who knows ( it's been suggested the
regulator in the 815).
All I know is that it works!

I would imagine I could get creative and up the voltage a little further to
1.6 or 1.65, but I am
satisfied, it has been running steady for the past few days.

This has been a frustrating ordeal - 2 years... 2 years...maybe a little
more...

Just a quick note, thank you for your help and patience!

(As you may recall, I have the right BIOS, swapped memory sticks, ran
memtest, swapped motherboards (D815EEA for D815EEA), cards, processors,
Upgradware adapters (which they claim works, whihc I now highly doubt,
unless it's a certain rev), etc...)

 

[kony]

I wanted to thank you folks for all your help, 'kony', etc..

The problem I had was that it was 2 problems! Undersized power supply plus
the need to run the 1.2 piii Tualatin at 1.55 volts - at least.

Swapping out the 145 Power supply for a larger one would still result in
BSOD because the Upgradware adapter locks the 1.2piii at 1.50v with a Vcore
of
1.475, which happens to work fine in my Compaq but not my Intel D815EEA.

Switching the Upgradware 370 for the LinLin adapter, even moving the jumpers
to a higher voltage of 1.55+, still was not good without a larger Power
Supply- the Tully, I guess, needs more power.

Yes it may, though I don't recall what the previous CPU was.
I have a box here still running a Tualatin Celeron 1.1
overclocked to 1.5GHz (vcore increased to 1.65V, IIRC) on a
150W mATX PSU, but the system mainboard has integrated
video, only one HDD and CDRW- not much of a load.

Seems all I read here, help at the Overclockers N/G, plus what caught my
attention at PowerLeap's site, held the
answer - their same adapter locks voltage at 1.55, not at 1.475 as
Upgradeware's does.

I got a larger Power Supply from New Egg- RaidMax 380 ($8 after rebate, not
bad!). I also set the LinLin adapter jumpers at 1.55, which is what the
m'board shows, with a vcore of 1.52 and change... (I have Intel Monitor).

While you got a good deal on the PSU, it's not a very good
PSU. It's essentially an old-school generic that's worth
around 200W, has poor fan, and poor air intake (unless
they've recently added cutouts on the underside, previously
they've only had small slits on the rear of the unit.

Even so, it may still provide as much useful service as the
original, it could merely be a situation where you'd get
another 3 years out of it which is a reasonable extension of
the system's lifespan for $8. However, if the entire case
cooling is done by _only_ the power supply exhaust fan, I
would not trust their fan in that role.

I can't say what the exact problem was, who knows ( it's been suggested the
regulator in the 815).
All I know is that it works!

Sometimes the socket adapters do require using a little
higher vcore, though it could have even been a combination
of that and the PSU.

I would imagine I could get creative and up the voltage a little further to
1.6 or 1.65, but I am
satisfied, it has been running steady for the past few days.

No need if it's stable. I'd run Prime95's Torture Test for
at least a couple hours to verify proper error-free
operation.

This has been a frustrating ordeal - 2 years... 2 years...maybe a little
more...

Just a quick note, thank you for your help and patience!

(As you may recall, I have the right BIOS, swapped memory sticks, ran
memtest, swapped motherboards (D815EEA for D815EEA), cards, processors,
Upgradware adapters (which they claim works, whihc I now highly doubt,
unless it's a certain rev), etc...)

IIRC, they also have compatibility lists per motherboard,
perhaps that particular make and revision of board wasn't as
suitable as others? Doesn't really matter now so long as
it's working for you.

 

[Uncle Vinnie]

Yes it may, though I don't recall what the previous CPU was.
I have a box here still running a Tualatin Celeron 1.1
overclocked to 1.5GHz (vcore increased to 1.65V, IIRC) on a
150W mATX PSU, but the system mainboard has integrated
video, only one HDD and CDRW- not much of a load.

The original CPU was 800 mhz Coppermine, which ran fine....

While you got a good deal on the PSU, it's not a very good
PSU. It's essentially an old-school generic that's worth
around 200W, has poor fan, and poor air intake (unless
they've recently added cutouts on the underside, previously
they've only had small slits on the rear of the unit.

Even so, it may still provide as much useful service as the
original, it could merely be a situation where you'd get
another 3 years out of it which is a reasonable extension of
the system's lifespan for $8. However, if the entire case
cooling is done by _only_ the power supply exhaust fan, I
would not trust their fan in that role.

I have to agree- I was disappointed when I opened the box.. the PSU is very
lighweight, has only holes in the back- poor fan system. However, I do have
an intake fan in the front of the case, and an oversized Intel heatsink/fan
on the CPU. Intel's Active Monitor Software seems to indicate all is
running fine.. It's supposedly rated at 380 Watts... thought I doubt it....
200 seems quite low, BUT anything over the 145 I had is an improvement... it
definitely runs more accurately voltage-wise, at least according to my
software... it's dead on at 12.0 volts, etc...

Sometimes the socket adapters do require using a little
higher vcore, though it could have even been a combination
of that and the PSU.


No need if it's stable. I'd run Prime95's Torture Test for
at least a couple hours to verify proper error-free
operation.


IIRC, they also have compatibility lists per motherboard,
perhaps that particular make and revision of board wasn't as
suitable as others? Doesn't really matter now so long as
it's working for you.

To the end, Upgradware claims their adapter works with the D815EEA- I
strongly doubt it after going through 2 boards... unless, of course, there
are certain revisions, which Upgradware does not indicate... Powerleap makes
the same claim however their adapter runs at 1.55, which is what I found to
work... also, I was told theirs also includes a voltage regulator...
Powerleap, apparently, has done their homework...