overclocking an e2140 for beginner

str · 10 Sep 2007 at 20:23

That's a great CPU with only 1.2v default vcore and an excellent overclocking result. :cool:

Everest shows your M0 has a 9c delta between motherboard reported CPU temp and the core temps whereas mine is only 5c and from what I've read there is a bigger margin of error on the motherboard reported CPU temp compared to the core temps so I think we can assume they're accurate.

str · 10 Sep 2007 at 20:25

Robbie G said:
You don't even need that. My 2140 does 3.2 easily on the stock cooler, load temps never go over 55C.

Can you do the same test and screenshot as dread and I have done please mate?

dread · 10 Sep 2007 at 20:45

str said:
That's a great CPU with only 1.2v default vcore and an excellent overclocking result.

Everest shows your M0 has a 9c delta between motherboard reported CPU temp and the core temps whereas mine is only 5c and from what I've read there is a bigger margin of error on the motherboard reported CPU temp compared to the core temps so I think we can assume they're accurate.

I can't moan for £40

I only bought it as a stop gap, but i think iam gonna keep it for quite a while !!

dread · 10 Sep 2007 at 23:07

3.3

http://img72.imageshack.us/img72/7223/33vs4.jpg

yet to get it stable at 3.4

mosfet · 10 Sep 2007 at 23:14

str said:
Got my E2180 today:

The default vcore is nice and low at 1.2375v and I've got it as above running at 2.66GHz (266 x 10 with 1.2375v) being cooled by AS5 and a Tuniq Tower. Not sure about temps as there is only a ~5c delta between Tcase and Tjunction and the TjunctionMax is only 85c but maybe that's how it is with M0 stepping CPUs.

Also notice the vcore at 1.184v in CPU-Z meaning this motherboard suffers vdroop quite badly.

This is a good start, but I'm sure they're capable of much, much more

I've got a 2180 in the post from a competitor, OCUK still haven't got them on the site for some unknown reason. They look like fantastic chips.

Sorry to the OP for going off topic!

str · 10 Sep 2007 at 23:32

At stock 2GHz 1.2375v and at 2.66GHz also with 1.2375v while running Orthos core temps were ~52c and at 3GHz with 1.325v it was at ~60c so the core temps don't go up too much at all and also my case cooling is configured to run in silent mode.

It needed 1.375v to run Orthos at 3.2GHz but it failed after about 15 mins and an even bigger increase to 1.425v to run Orthos at 3.33GHz which again failed after about 5 mins so I doubt I'll go for any higher than 3GHz.

Still 3GHz at only 1.325v and decent temps is fine for such a cheap CPU.

str · 11 Sep 2007 at 12:20

Changed my mind about clocking it higher as it's a bit cooler today.

Screenshot of 3.33GHz (333 x 10 with 1.4625v):

Will let it run to see if it can do 24 hours.

mosfet · 11 Sep 2007 at 13:58

That's still very hot, too hot for my liking, especially with such a good cooler. I'm about to order a Zerotherm BTF-95 after seeing one (Wayne?) in the E2140/E2160 thread.

str · 11 Sep 2007 at 14:09

It's only 10c higher compared to running at 3GHz but yeah it's running hotter than I'd like. Once I know it's stable at that speed I'll be lapping my Tuniq Tower and I'll also lap the E2180's IHS and hopefully get the core temps down to a warm 65c or so. Also at 3.33GHz it idles at 36c and goes up to around 45c during normal light use so isn't so bad really.

Bolerus · 11 Sep 2007 at 14:43

i keep reading these threads and get far too tmepted to overclock my E6400

(on an asrock dual vista, but only got ddr1 400 memmory)

str · 11 Sep 2007 at 14:53

Yeah I know what you mean. Seeing dread at 3.3GHz made me revisit my setup and try 3.33GHz again.

DavidB · 11 Sep 2007 at 19:24

Yeah I know what you mean. Seeing dread at 3.3GHz made me revisit my setup and try 3.33GHz again.

lol, ill be trying later too, damn my competetiveness !

str · 11 Sep 2007 at 21:55

Orthos failed at 3.33GHz after 6 hours:

I suspect it could be the northbridge voltage needed an increase so I've also fitted 40mm quiet fans to the northbridge and southbridge heatsinks and I'm testing again with Orthos.

dread · 12 Sep 2007 at 13:00

stick to 3.2 on the e2140 all the extra temps etc etc aren't worth the extra 100mhz

str · 12 Sep 2007 at 13:18

Yeah I agree.

I've gone back to 3GHz and my system is in silent mode with the fans running at a very low rpm.

Participant · 12 Sep 2007 at 20:24

p4radox said:
What program is that measured with?

The one that came with the Gigabyte board and speedfan (both read the same).

str said:
Can you do the same test and screenshot as dread and I have done please mate?

Maybe if I get time but I honestly can't be bothered. Take that to mean whatever you like

Biffa · 12 Sep 2007 at 20:35

Robbie G said:
The one that came with the Gigabyte board and speedfan (both read the same).

Neither of which measure core temperatures. Its a good idea to try Coretemp or TAT or add 15 degrees to Speedfan.

Robbie G said:
Maybe if I get time but I honestly can't be bothered. Take that to mean whatever you like

Usually means you have tried it and it failed Orthos or Prime

What people don't seem to understand is that by standardising the tests for stability by using a select type of load testing software means that we can all learn what the best way to get a stable overclock is for a particular set of components. Also helps us to judge what components work best in a given situation to get the best stable overclock.

I can understand people who are happy with their Orthos failing, gaming stable clock on stock fans but please understand why we ask for orthos/Prime screenshots.

Participant · 12 Sep 2007 at 21:09

Biffa said:
Neither of which measure core temperatures. Its a good idea to try Coretemp or TAT or add 15 degrees to Speedfan.

Usually means you have tried it and it failed Orthos or Prime

What people don't seem to understand is that by standardising the tests for stability by using a select type of load testing software means that we can all learn what the best way to get a stable overclock is for a particular set of components. Also helps us to judge what components work best in a given situation to get the best stable overclock.

I can understand people who are happy with their Orthos failing, gaming stable clock on stock fans but please understand why we ask for orthos/Prime screenshots.

Who said I don't understand it? It's hardly a tricky concept. I simply can't be arsed. And I can promise you I have never even downloaded "Orthos" let alone run it. I'm sure if I did it would fail and read temperatures of 3483425 degrees but as long as it's stable and ok for my use it's not a concern to me. The BIOS also reads the same as the above by the way but I suppose you are going to tell me it's a useless measurement and to add 30 deg C.

str · 12 Sep 2007 at 21:53

Robbie G said:
The one that came with the Gigabyte board and speedfan (both read the same).

Maybe if I get time but I honestly can't be bothered. Take that to mean whatever you like

I take that to mean as you say that you can't be bothered which I can understand as I know not everyone has the time to make sure an overclock is stable or don't care about stress testing. The stability isn't important anyway more I was interested in seeing your setup with the stock cooler (CoreTemp default VID and CPU-Z to see vdroop and stepping) and the delta between Tcase and Tjunction in Everest. Orthos is just to put a load on the CPU to compare similar load.