Abit IT7: CPU Vcore undervolting help!

[CMatty]

Hi,

I just built a PC for a friend with the following: Abit IT7 MAX, p4 2.53 ghz, corsair XMS 3200. All is well apart from the cpu vcore is undervolting. When set to 1.5v in the bios it is reported as 1.45v in PC Health and the CPU at stock speed and voltage wont boot into windows. I dont think it is his power supply because he has got a 550W Enermax. Should I return the it7 and get the it7 max 2 or do they undervolt too???

Thanks

Matty

[MuFf|n M@N]

check the psu connector to the mobo it might not be all the way in

[OPPAINTER]

Matty,

Why don't you just set the voltage up a tad in the bios??

OPP

[CMatty]

Yeah I have upped the voltage in the BIOS and the PSU connector is all the way in. Its just that the cpu is at 2.8 ghz at the momment with the swiftech P4 heatsink and delta fan. I reckon it could eaily do 3.0ghz+ but when I run prime 95 torture test the cpu vcore goes from 1.66V on idle to 1.62V. This is when the voltage is set to 1.7V in the bios. I have seen it as low as 1.60V thats 0.1V of under voltage.

Matty

[OPPAINTER]

Well from 1.66 idle to 1.60 under load is about the average loss I get these days with most Mobos. Just may have to live with it, if you want some higher voltage you can allways do a v-mod to the mobo.

OPP

[Charles Wirth]

This is correct operation from the IT7, volt mod it.

[Creative]

Originally posted by OPPAINTER
Well from 1.66 idle to 1.60 under load is about the average loss I get these days with most Mobos. Just may have to live with it, if you want some higher voltage you can allways do a v-mod to the mobo.

OPP

I dont think we should have to live with it Opp. Why cant mobo manufacturers release proper products that when you set it on something, it actually means that setting. Im sick of setting 1.85 in my bios to only boot up at 1.78 and then on load drop to 1.73....thats .12 volts that its losing...very unacceptable imo....making us do mods to our own boards to make them work like they were supposed to:mad:

I wish companies would fix this problem and not upset us end users....:( :)

[sysfailur]

I've actually noticed my IT7 gives me alot less voltage than what I'm asking for it as well. When set @ 1.7 I get 1.66, and when set @ 3.2v for ddr, I get like 3.04 :| Weaksauce.

On the other hand, on my Epox 8k5a3+ I get MORE voltage than whatever I set it at in the bios. It's nuts. I get 2.24v @ 2.2 and 3.22 @ 3.2.

[Charles Wirth]

Creative it is the way the circuit is designed and it works as intended.

[Creative]

Cant they use better quality parts tha wont do this though?

Or am I in over my head talking bout hardware that I know nothing about how it works.....Im more of a commerce student that a Engineering student :D
hehehehe

[Charles Wirth]

Well, we know it works perfect as is. You are overclocking and need more than it gives so mod is the only option. I am not sure if you chimed in with your suggestion in the "abit suggestion" thread. Most people want more vcore.

[Creative]

Ill do it now fugger, never really noticed that thread up the top...:)

Thats what I liked about Epox and thier bios vcore options.

[GreenBeret]

Originally posted by sysfailur
On the other hand, on my Epox 8k5a3+ I get MORE voltage than whatever I set it at in the bios. It's nuts. I get 2.24v @ 2.2 and 3.22 @ 3.2.

sysfailur, if I were you, I wouldn't trust what the 8K5A3+ reports. I had an Epox 4G4A+ (i845G board). With the stock BIOS, I got higher voltages than what I set or saw in the BIOS. However, when I updated to newer BIOSes, all the voltages dropped. Not only the ones that I could change in the BIOS, but the ones that I couldn't change also dropped: 5V, 12V, 3.3V etc.

[sysfailur]

Originally posted by Niko


sysfailur, if I were you, I wouldn't trust what the 8K5A3+ reports. I had an Epox 4G4A+ (i845G board). With the stock BIOS, I got higher voltages than what I set or saw in the BIOS. However, when I updated to newer BIOSes, all the voltages dropped. Not only the ones that I could change in the BIOS, but the ones that I couldn't change also dropped: 5V, 12V, 3.3V etc.

Well... maybe, but... I can tell you that running 3.2v into my ram on my 8k5a3+, I can run higher mhz @ the max timings than on my IT7 @ "3.2" supposedly... The reason for it I believe is b/c the drop in voltage. On the IT7 I max @ 222 @ 2-6-3-2 where-as on the 8k5a3+ I can do 225.8 @ 2-5-2-2-1.

[chainbolt]

Originally posted by FUGGER
Creative it is the way the circuit is designed and it works as intended.

I don't think so, it is certainly not intended to work this way. The IT7 and the IT7-II as well as the BD7-II are substantially undervolting in default settings when the CPU is in idle. This is - despite all the good things we can say about these boards - BAD and a sign for either insufficient timings are bad components/weak circuitry. Other boards, in particular Asus, are NOT undervolting. On top of this you have the built in "real" overvoltage on most Asus boards.

The missing 0.05 volt might be excalty what is missing to get the full oc potential out of CPU - not nice.

[Charles Wirth]

Chainbolt, it is. I know for a fact as it was explained to me by Abit.

The CPU is allowed to draw what it needs.

[Creative]

So when you set it to 1.50 in bios and in windows it undervolts to 1.45.....this is because the cpu isnt calling all 1.50 volts, it just needs that 1.45 to run? What about when I crank it up to 1.85 and the cpu is at 3.1ghz and I prime it and she goes to 1.78 load.....is that cos the cpu is only calling 1.78 volts?

I know that sounds a bit rude in my tone, but FUGGER no way man am i being rude :) Im just making observations about this issue.....cos if that is the case, the cpu doesnt call the maximum vcore than you have set.....as chainy said, that could be the difference between a good OC and a :banana::banana::banana::banana: OC :(

Like my BD7-II undervolted on the ram to 2.6vdimm or something so I could only get 160 in 3:4 mode but with the IT7 and 2.9vdimm(grr it should be frikken 3.2 like it says in the bios, again undervolting :mad:) I can get it up to 176 for a sandra bench....just proving that the loss in volts is hindering my overall OC....:(


Did that just make sense? :p
hehehehe

[Charles Wirth]

I understand your point with overclocking. This is a limitation must of us deal with. they are all the same.

[chainbolt]

Originally posted by FUGGER
I understand your point with overclocking. This is a limitation must of us deal with. they are all the same.

That's the point: they are NOT. Only recent ABIT boards allwow the CPU to draw so much that even the idle voltage is below the Intel specified default. That is what is surprising me.

Intel is indeed allowing certain voltage fluctuation, in particular under load. But no other manufacturer than Abit allows the CPU voltage to fall so much under the default.

I don't think it's a big deal. The BD7 and the IT7 ( I have both boards) are the best overclocking P4 platforms. But fact is fact: they undervolt, and would probably do even when better when reaching the specified voltages.

[Charles Wirth]

You see is as undervolting, but it is the way it the way the circuit is designed. I asked this question to Abit about year ago wondering the same with a TH7-II.

P4T533 is the one with the vcore problem, 1.56v +-.03v jumping about 80 times per minute. Im glad I didn't screw up a CPU.

[chainbolt]

Originally posted by FUGGER
You see is as undervolting, but it is the way it the way the circuit is designed. I asked this question to Abit about year ago wondering the same with a TH7-II.

P4T533 is the one with the vcore problem, 1.56v +-.03v jumping about 80 times per minute. Im glad I didn't screw up a CPU.

I simply don't believe in what Abit is telling here. It like getting a car which has a high oil consumption, and then the manufacturer is telling you: that is the way it is engineered. Sure, we call it bad engineering, when all other cars need less oil.

That is how I see it. When all other boards I know off, don't have such a fluctuation (the P4T533 maybe an exception, my board was fault free by the way), there is absolutely no reason for ABIT not to provide the same voltage stability.