does the the cpu affect the voltage drop from idle to load?

veryhumid · 08-10-04, 08:00 PM

I asked batboy this earlier, but I am just curious to know if anyone has an explanation. Is the motherboard and power supply the only things that maintain cpu voltage and voltage drop (droop) from idle to load? I find my system has a pretty big drop, I was wondering if it may be my chip. Is it possible some chips would have higher voltage drops than others? Thanks for the input.

batboy

VisEtVirtus · 08-10-04, 11:52 PM

Hmmm... I'm a little curious about this as well.

Styyn · 08-11-04, 05:17 AM

I'd give a qualified yes to that. I say qualified because I'm basing this solely on my limited observations. Under full load I was getting .04-.06v droop with my 2.8c, whereas my 3.2E D0 usually droops .06-.08, with a very rare and very brief .1v droop. This is with a PCP&C 510 psu and a fully sinked and wc'd MAX3.

aNTiChRisT · 08-11-04, 06:27 AM

The power consumtion of the chips will affect the amount of the droop, but only as a % -- if you have no droop, it wont for any chip.

~t0m

veryhumid · 08-11-04, 08:36 PM

Quote:

Originally Posted by Styyn

I'd give a qualified yes to that. I say qualified because I'm basing this solely on my limited observations. Under full load I was getting .04-.06v droop with my 2.8c, whereas my 3.2E D0 usually droops .06-.08, with a very rare and very brief .1v droop. This is with a PCP&C 510 psu and a fully sinked and wc'd MAX3.

yeah i have like a .09 droop when set to certain voltages, .05 for some. what are the rails looking like on your psu when at full load?

thanks for the input everyone!

NookieN · 08-12-04, 12:40 AM

The voltage differential between idle and load is proportional to the current that the CPU draws. The voltage drop on a 2.8C will be less than with a 3.2E because the E chip (a Prescott) draws much more current than a Northwood.

On a side note, the voltage drop is normal; in fact it's part of the operating spec. If a motherboard provides a consistent voltage from idle to load, that's not necessarily a good thing; it can be detrimental to the long-term life of a chip.

batboy · 08-12-04, 08:23 AM

Come to think of it, I have noticed a greater amount of undervolting with a Prescott compared to a Northwood, so I think Nookie is right about some CPUs pull more power. The thing is, his IC7 Max3 is dropping from like 1.55v idle to maybe 1.5v load (this is with a 1.575v BIOS setting). This is with a Northwood CPU. I've never seen an Abit mobo drop that much before.

veryhumid · 08-12-04, 09:20 AM

yeah i am going to kind of rebuild my computer. Put in the old true 430W and try an IC7-G. I'm really asking this because I wonder if my chip maybe damaged.

Styyn · 08-13-04, 03:41 PM

Sorry about the late response. Anyway, as displayed by MBM5, my 12v rail idles at 12.04 and droops to 11.92 under full load. I realize MBM isn't terribly accurate when it comes to voltage readings, so if anyone could tell me the best way and locations for voltimeter readings I can give some more accurate results

08-10-04, 08:00 PM	Thread Starter #1
veryhumid Join Date: Apr 2004 Location: New Hampshire	does the the cpu affect the voltage drop from idle to load? I asked batboy this earlier, but I am just curious to know if anyone has an explanation. Is the motherboard and power supply the only things that maintain cpu voltage and voltage drop (droop) from idle to load? I find my system has a pretty big drop, I was wondering if it may be my chip. Is it possible some chips would have higher voltage drops than others? Thanks for the input. batboy __________________ < FOLD! < Heatware < EVGA X58, 920, TRUE, 12GB Corsair, 2 x EVGA GTX 285, Enermax Revolution, Velociraptor, 7 64-bit
	QUOTE Thanks

08-10-04, 11:52 PM	#2
VisEtVirtus Member Join Date: Apr 2004 Location: Denver, Colorado	Hmmm... I'm a little curious about this as well. __________________ Phenom II X6 1100T @ 3.8ghz 8gb Gskill Ripjaw series 2x 120gb OCZ Vertex3 SSD 2x 1TB WD Black HDD Radeon 6950 @ stock
	QUOTE Thanks

08-11-04, 05:17 AM	#3
Styyn Member Join Date: May 2004 Location: St. Paul, MN	I'd give a qualified yes to that. I say qualified because I'm basing this solely on my limited observations. Under full load I was getting .04-.06v droop with my 2.8c, whereas my 3.2E D0 usually droops .06-.08, with a very rare and very brief .1v droop. This is with a PCP&C 510 psu and a fully sinked and wc'd MAX3.
	QUOTE Thanks

08-11-04, 06:27 AM	#4
aNTiChRisT Member Join Date: Feb 2004 Location: England	The power consumtion of the chips will affect the amount of the droop, but only as a % -- if you have no droop, it wont for any chip. ~t0m
	QUOTE Thanks

08-12-04, 12:40 AM	#6
NookieN Member Join Date: Jul 2002 Location: The West	The voltage differential between idle and load is proportional to the current that the CPU draws. The voltage drop on a 2.8C will be less than with a 3.2E because the E chip (a Prescott) draws much more current than a Northwood. On a side note, the voltage drop is normal; in fact it's part of the operating spec. If a motherboard provides a consistent voltage from idle to load, that's not necessarily a good thing; it can be detrimental to the long-term life of a chip.
	QUOTE Thanks

does the the cpu affect the voltage drop from idle to load?

Internet Services

Website Development

Tech Hardware

08-12-04, 08:23 AM	#7
batboy Senior Moment Join Date: Jan 2001 Location: Kansas, USA	Come to think of it, I have noticed a greater amount of undervolting with a Prescott compared to a Northwood, so I think Nookie is right about some CPUs pull more power. The thing is, his IC7 Max3 is dropping from like 1.55v idle to maybe 1.5v load (this is with a 1.575v BIOS setting). This is with a Northwood CPU. I've never seen an Abit mobo drop that much before. __________________ 2600K, Asus P8P67 Deluxe, 2X2GB G.Skill Pi PC3-2133, PC & Cooling 750W watercooled: Eheim 1048 pump, Swiftech Apogee GTX and 2X120mm rad FS020 case, OCZ SSD, Sapphire HD 6950, Acer X241W 24" widescreen
	QUOTE Thanks

08-12-04, 09:20 AM	Thread Starter #8
veryhumid Join Date: Apr 2004 Location: New Hampshire	yeah i am going to kind of rebuild my computer. Put in the old true 430W and try an IC7-G. I'm really asking this because I wonder if my chip maybe damaged. __________________ < FOLD! < Heatware < EVGA X58, 920, TRUE, 12GB Corsair, 2 x EVGA GTX 285, Enermax Revolution, Velociraptor, 7 64-bit
	QUOTE Thanks

08-13-04, 03:41 PM	#9
Styyn Member Join Date: May 2004 Location: St. Paul, MN	Sorry about the late response. Anyway, as displayed by MBM5, my 12v rail idles at 12.04 and droops to 11.92 under full load. I realize MBM isn't terribly accurate when it comes to voltage readings, so if anyone could tell me the best way and locations for voltimeter readings I can give some more accurate results
	QUOTE Thanks

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search