<Overclock>E2140 temp problems

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

BonzaiDuck

Lifer
Jun 30, 2004
15,699
1,448
126
I have only cpu speed , cpu voltage , ram voltage , pci-e clock , memory frequency , i dont found multiplier or RATIOS or RAM DIVIDERS..

Here's another possible problem source. You need to set the PCI_E clock to a FIXED speed -- that is "THE DEFAULT" fixed speed -- 100 Mhz I think.

There may be two PCI-E speed settings -- which at default CPU settings would be 100 and 200 respectively if they were set to "auto." When overclocking, you'd need to take these off "auto" and set them to fixed (but default, spec) values.

Also, the safest thing I can imagine (since I can't see your BIOS screens) is to underclock the RAM to 533 while attempting to OC the processor/front-side-bus. It shouldn't matter if you don't have a menu of "ratios," though. With the right settings, you should be able to achieve the 4:5 ratio I mentioned and run the RAM at their spec DDR2-667.

Post a link to the Biostar web-page and PDF for the user-manual of that board, and I'll have a look.
 

wowempire

Member
Jan 12, 2014
140
0
0
Here's another possible problem source. You need to set the PCI_E clock to a FIXED speed -- that is "THE DEFAULT" fixed speed -- 100 Mhz I think.

There may be two PCI-E speed settings -- which at default CPU settings would be 100 and 200 respectively if they were set to "auto." When overclocking, you'd need to take these off "auto" and set them to fixed (but default, spec) values.

Also, the safest thing I can imagine (since I can't see your BIOS screens) is to underclock the RAM to 533 while attempting to OC the processor/front-side-bus. It shouldn't matter if you don't have a menu of "ratios," though. With the right settings, you should be able to achieve the 4:5 ratio I mentioned and run the RAM at their spec DDR2-667.

Post a link to the Biostar web-page and PDF for the user-manual of that board, and I'll have a look.
I did it , i changed my DDR freq to 533 , wont boot with 230+ fsb , i set to 400 STILL won't boot .. PCI-E CLOCK IS FIXED AT 100 , IF I SET IT AUTO THE MULTIPLICATOR WILL VARY WITHIN 9x-11x .. I still dont find the source of my problem ..
 

BUnit1701

Senior member
May 1, 2013
853
1
0
Intel lists his 945GC chipset as having a max FSB of 800Mhz (200Mhz base clock). 230 is probably a realistic max for that board.
 

SPBHM

Diamond Member
Sep 12, 2012
5,056
409
126
Intel lists his 945GC chipset as having a max FSB of 800Mhz (200Mhz base clock). 230 is probably a realistic max for that board.

older revision or outdated description, Intel website have many inaccuracies

his motherboard officially supports FSB266,
and many 945GC boards supported 333 via OC (with PCIE clock overclock)

wowempire,
as I said on my previous post, make sure you lower to 533 and set manually the timings to what would be the default values for 667, if you leave on AUTO some boards will used the values for 533 (lower) which might make your mem unstable once the clock goes up because of the FSB.

try with PCIE clock at 101, or something a little higher,

what multi 11x is that? the CPU multi for the e2140 is 8x and no higher (can go as low as 6x with C1E/EIST enabled)
 

wowempire

Member
Jan 12, 2014
140
0
0
older revision or outdated description, Intel website have many inaccuracies

his motherboard officially supports FSB266,
and many 945GC boards supported 333 via OC (with PCIE clock overclock)

wowempire,
as I said on my previous post, make sure you lower to 533 and set manually the timings to what would be the default values for 667, if you leave on AUTO some boards will used the values for 533 (lower) which might make your mem unstable once the clock goes up because of the FSB.

try with PCIE clock at 101, or something a little higher,

what multi 11x is that? the CPU multi for the e2140 is 8x and no higher (can go as low as 6x with C1E/EIST enabled)
Oh happy to hear somethin' like that ! When i go to "Memory settings " in bios , if i set Frequency manually , i cant let timings auto , they take their default value ( idk which is , i will tell you after 30-40 mins ). So , i will set 533 and pci-e clock to 101 or better , i will come with results :D
 

wowempire

Member
Jan 12, 2014
140
0
0
Thank a lot, i reached 245 fsb raising up the pci-e clock to 110, CPU voltage +0.012, ram freq to 533, and i set timing "BY SPD" ( there are 2 choices "MANUAL" and "BY SPD" ) .I cant go over 245 fsb :( ANY idea ?
I TRIED PCI-E CLOCK TO 120, voltage to +0.05V AND STILL dont boot, even with that values.
 

BonzaiDuck

Lifer
Jun 30, 2004
15,699
1,448
126
Thank a lot, i reached 245 fsb raising up the pci-e clock to 110, CPU voltage +0.012, ram freq to 533, and i set timing "BY SPD" ( there are 2 choices "MANUAL" and "BY SPD" ) .I cant go over 245 fsb :( ANY idea ?
I TRIED PCI-E CLOCK TO 120, voltage to +0.05V AND STILL dont boot, even with that values.

Per your remarks, I think we're sensitive to your budgetary limitations, and we also think your quest to OC this E2140 is reasonable -- even enlightened.

I only repeat that increases in PCI-E bus speed may lead to HDD corruption problems down the road. Also, I'm a bit surprised that the Biostar board has some linkage or dependency between the PCI_E bus-speed and your ability to bump up the CPU base speed.

Like I said -- the Gigabyte mATX boards we used were bottom-of-the-line, very cheap units. Never had a problem reaching CPU=266Mhz and FSB=1,066Mhz. Too bad those GA-73VM-S2 boards aren't available anymore. You could search for one, but I doubt you'd find it. If we paid $85 each for them brand-new, I think you could get a used one for chump-change. But then -- "used" is risky. . . .
 

wowempire

Member
Jan 12, 2014
140
0
0
Per your remarks, I think we're sensitive to your budgetary limitations, and we also think your quest to OC this E2140 is reasonable -- even enlightened.

I only repeat that increases in PCI-E bus speed may lead to HDD corruption problems down the road. Also, I'm a bit surprised that the Biostar board has some linkage or dependency between the PCI_E bus-speed and your ability to bump up the CPU base speed.

Like I said -- the Gigabyte mATX boards we used were bottom-of-the-line, very cheap units. Never had a problem reaching CPU=266Mhz and FSB=1,066Mhz. Too bad those GA-73VM-S2 boards aren't available anymore. You could search for one, but I doubt you'd find it. If we paid $85 each for them brand-new, I think you could get a used one for chump-change. But then -- "used" is risky. . . .
Ok, i understand i cant overclock it anymore.. whatever 2Ghz it's better than 1.6, are my setting safe ? pci-e clock to 110, voltage +0.012, cpu 245 and ram 533 ?
Btw i dont buy from ebay or any international comercial website.. i worry about shipping.. And if the e2220 ( example ) is 10$ the shipping in 10$ too ..
 

BonzaiDuck

Lifer
Jun 30, 2004
15,699
1,448
126
Ok, i understand i cant overclock it anymore.. whatever 2Ghz it's better than 1.6, are my setting safe ? pci-e clock to 110, voltage +0.012, cpu 245 and ram 533 ?
Btw i dont buy from ebay or any international comercial website.. i worry about shipping.. And if the e2220 ( example ) is 10$ the shipping in 10$ too ..

Yeah -- shipping -- I hear that. . .

Before we stopped twiddling with bus-speed, I think the common wisdom was that 5 to 10% over the stock speed was "more or less safe." YMMV.

With the newer, "post-LGA775" processors, I don't fiddle with this anymore. And with the motherboard I mentioned, we didn't have to do it either . . . .

When I OC'd my current Sandy Bridge system, I played around with the bCLK and pushed it from stock 100Mhz to about 105. But -- not worth it to me.

Truth be told, for our own E2140, we OC'd it by 33% above the 1.6Ghz. You're only short by about 130 Mhz from that level.
 

BonzaiDuck

Lifer
Jun 30, 2004
15,699
1,448
126
+0.012 voltage??

my E6600 ran at 1.375V.

Well, I think he's talkin' about the increase over stock voltage.

I can vouch for this fact: With the E2140 and E2180, the stock voltage reported in the BIOS monitor was 1.30V. We only had to fix the voltage at that number, and we gave it an extra 0.01V just for good measure -- to get a rock-solid 33% overclock.

Eventually, we replaced the E2140 with an E6700 Wolfdale. This was not the original Conroe E6700 (and I assume your E6600 was also the famous Conroe we all started with.)

The Wolfdale was a whole different enchilada. I think -- despite the potential for much more -- we twisted it up to maybe 3.8 Ghz -- mostly to avoid any problems with sleep-state, if there might be any.
 

DigDog

Lifer
Jun 3, 2011
13,444
2,084
126
with semi decent cooling he should be able to "test" 1.38V, it won't insta-kill the cpu.
it is my opinion that the OC is failing because the processor isn't getting enough voltage.

long story so get comfy.

i had a E6600 conroe, the first stepping, and a gigabyte 965p-ds3; this board was fabled for the "endless" FSB, 450 and more. and it was cheap, too.

at first, my 6600 took 3.2Ghz without blinking. 8x multi and 400 FSB. small but noticeable bump in voltages.


after about 4 years, i started getting BSODs, this was still on XP, so i put it down to bad software, lugging the pc around, and dust. i did a rebuild/clean, and when i went to OC, i found that it needed a smaller bump in voltage to keep 3.2Ghz

time went by, and more BSOD appeared, this time i lowered the OC.

eventually, i found out that the CPU was still stable, but the board was dying - it needed 1.38 to hold 3Ghz, but would even boot in windows at 1.4v for 3.2Ghz

The board was dying, really.
back in 2006, mobos were not as good as they are now, nowhere near.

also, that biostar 945 ... bleeeurgh

mind you, biostar makes nice stuff .. *now*

so, here is what. the CPU isn't getting enough voltage to hold the OC - the electricity gets lost somewhere in the mobo.
the FSB also probably isn't getting enough voltage - my DS3 had a +0.1~3V selector for the FSB alone
the E21 series CPU were dogs
the mobo stinks

he's got a frikkin DeLux PSU, for chriss' sake. that thing is just waiting to explode and kill the dog!

all in all, i see two possible ways to solve this; one involves the use of MOAR* voltage across everything, while it still lives.
the other involves a plastic bag and a trashcan.
 

BonzaiDuck

Lifer
Jun 30, 2004
15,699
1,448
126
with semi decent cooling he should be able to "test" 1.38V, it won't insta-kill the cpu.
it is my opinion that the OC is failing because the processor isn't getting enough voltage.

long story so get comfy.

i had a E6600 conroe, the first stepping, and a gigabyte 965p-ds3; this board was fabled for the "endless" FSB, 450 and more. and it was cheap, too.

at first, my 6600 took 3.2Ghz without blinking. 8x multi and 400 FSB. small but noticeable bump in voltages.


after about 4 years, i started getting BSODs, this was still on XP, so i put it down to bad software, lugging the pc around, and dust. i did a rebuild/clean, and when i went to OC, i found that it needed a smaller bump in voltage to keep 3.2Ghz

time went by, and more BSOD appeared, this time i lowered the OC.

eventually, i found out that the CPU was still stable, but the board was dying - it needed 1.38 to hold 3Ghz, but would even boot in windows at 1.4v for 3.2Ghz

The board was dying, really.
back in 2006, mobos were not as good as they are now, nowhere near.

also, that biostar 945 ... bleeeurgh

mind you, biostar makes nice stuff .. *now*

so, here is what. the CPU isn't getting enough voltage to hold the OC - the electricity gets lost somewhere in the mobo.
the FSB also probably isn't getting enough voltage - my DS3 had a +0.1~3V selector for the FSB alone
the E21 series CPU were dogs
the mobo stinks

he's got a frikkin DeLux PSU, for chriss' sake. that thing is just waiting to explode and kill the dog!

all in all, i see two possible ways to solve this; one involves the use of MOAR* voltage across everything, while it still lives.
the other involves a plastic bag and a trashcan.

Too bad -- for those options. Or he could live with 2.0Ghz until the system just up and croaks.

Sure -- I've seen "voltage-creep" occurring in good motherboards over a few years. You wouldn't think it, but my Striker 680i -- a $350 purchase -- gave me 1.32 with a "1.32" setting. Eventually, I had to notch it at 1.38V to hold 1.32. Maybe "capacitor aging?" Can't say.

I just feel a lot more confident about my P8Z68 than I did about my Striker, and for almost $130 less.
 

wowempire

Member
Jan 12, 2014
140
0
0
Hey, i have a last question, i won't make a new topic for that . I can put a D945 3.4 GHz dual-core OR an e7500 on that motherboard?
 

BonzaiDuck

Lifer
Jun 30, 2004
15,699
1,448
126
Hey, i have a last question, i won't make a new topic for that . I can put a D945 3.4 GHz dual-core OR an e7500 on that motherboard?

That, sir, I wouldn't know -- for sure. It's an LGA-775 motherboard, if I recollect correctly. Check the Biostar(?) web-site and page for the mobo's specs. There should be a list of compatible processors. I think prospects are good, but do the research.

The Gigabyte budget mATX LGA-775 boards we used had a list that "didnt' keep up" with CPU releases. Per its last BIOS release -- a columnar datum that qualifies each item in the CPU list -- our boards would run any "Wolfdale" processor listed as E8400 to E8600. Our E6700 -- replacing entirely the E6700 Conroe (which certainly worked even on initial BIOS releases) -- was a Wolfdale released in spring, 2010. The website didn't list it.

But following the simple logic -- the board used earlier Wolfdale cores -- we updated the BIOS and -- yeah -- fine! -- Ausgezeichnet! Bueno! Superiore! -- the E6700 worked without a hitch, fully recognized in BIOS and Windows.

Use similar prudence, and Via con Dios, compagnero!
 

DigDog

Lifer
Jun 3, 2011
13,444
2,084
126
^ the grammar nazi in me is furious now

because because because
Vaya Con Dios, Companero
!And stop aksin me! It hurtz!
 
Last edited: