• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

AV8 - Vcore fluctuation (AV8 owners input wanted!)

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Gnerma

Senior Member
Joined
Aug 9, 2002
Location
Bakersfield, CA - USA
AV8 Vcore fluctuation

Hello my fellow AV8 owners, lets get serious about our boards shall we? I understand my cutesy new loli avatar might get in the way of taking me seriously but please try to get past it if possible.

I'd like a few opinions and experiences of AV8 users regarding Vcore stability. Help me out and we can get to the bottom of this. I'm sure we'll all benefit in the end.

As you might expect by the fact that I started this thread I've observed large & baffling Vcore fluctuations with my setup.

  • Winchester 3200+ @ 2.5 Ghz, 1.625 Vcore in BIOS.
  • Fortron 530w PSU
  • Abit AV8 @ 250 fsb
  • 2 x 512mb Kingston Hyper-X PC3500 @ 2.65v, 204 Mhz, 2-3-2-6 timings
  • X800XT 533 gpu & 528 mem (only clocks up when a 3d game is launched.)
  • 1 WD Raptor 74gb, 1 Maxtor DM10 300gb 16mb cache
  • Audigy 1 & Revolution 7.1
  • 1 Lite-on DVD burner
  • 2 x 120mm case fans, 1 x 90mm cpu fan all running undervolted to ~7v through my fan controller.

At full load, Abit EQ reads Vcore between 1.55v and 1.6v although it usually hangs around 1.58v. I checked the voltages with a multimeter through a molex and the 5v aux connector a while back. They tested to 12.08v, 5.09v & 3.43v respectively. The readings showed zero fluctuation in these numbers. I know I should perhaps test through the ATX connector though.

All of my other voltages do not fluctuate at all according to Abit EQ. It reads 5v at 5.12v and the 3.3v at a perfect 3.3v.

Also, observing the voltage in BIOS after a cold boot shows spot on voltages with very little fluctuation at first, but it slowly jitters down to be somewhere near to what I see in windows. This suggests to me that the power reg components (MOSFETs etc) are warming up causing the undervolting.

So I'd like to hear from other AV8 + Winchester CPU owners. Does your Vcore fluctuate as much as mine? What PSU are you running? If you have installed MOSFET cooling has it improved the fluctuation and/or overall Vcore level? If you reply please post your system details as I did because sigs change.


Regarding AV8 MOSFET cooling

I've thought about trying it to see if it helps the fluctuations, but upon looking at the mess that that area of the board is I was a bit put off. The AV8 uses 12-13 of the 'small' type MOSFETs, and they are hardly lined up into a neat little row.

(power regs - full board pic)

So have any of you taken a shot at taming this wild mess? If so please post some pics. Thanks a lot.
 
Last edited:
Well I sometimes see a fluctuation of .05v's in bios, but nothing in MBM5. I'm running a brand new antec truepower 430, with both motherboard power connectors in.

winchester 3000+
av8 3rd eye
truepower 430
512mb kingston hyper x pc3500
512mb corsair xms pc3200
200gb WD sata hdd
60gb ide WD hdd
2 optical drives, floppy drive
9700 pro stock clock
stock amd heatsink with 80mm tornado
2 exhaust fans, 1 120mm intake fan.

right now i'm doing some stability testing so i'm running 2.3ghz 250httx9, 209fsb, 2.8 memory voltage, 1.4 vcore, 4x LDT.
 
Tyranos said:
... I'm running a brand new antec truepower 430 ...
What are the voltage ratings on that PSU if you don't mind me asking? Also, how does your Vcore behave if you crank it up to 1.5 or 1.6? Do you see that 'heating up' phemomenon in BIOS I described after booting up your rig? (from a cold boot of course)
 
*Pardon the intrusion*

I own an ASUS A8V (same chipset) and my board undervolts exactly like yours. I tweaked the pots on my FSP530 and things are a lot better now.

6 CASE FANS + 1 TORNADO
2 x HDD's
1 Sony DVDRW
Radeon 9600
Newcastle 3500
2x512 PDP XBLK (TCCD)
1 SB live 5.1

VCore 1.65V (Windows shows it as 1.628V)

I'm searching for a MB monitor for linux ATM. I'll post my results when I find one.
VDimm 2.8V
V AGP=1.6V
V Vlink=2.6V
 
Super Nade said:
*Pardon the intrusion* I own an ASUS A8V...
GET OUT! (kidding :) )



Super Nade said:
... my board undervolts exactly like yours. I tweaked the pots on my FSP530 and things are a lot better now.
Thats something I've also been considering doing, mostly because I can do that without popping any warranties but I would probably not go over 12.20 - 12.30 max (reading with my multimeter through the ATX connector of course). Exactly how much did you turn your pots up? Do you have any specific before and after Vcore numbers? Thanks for the reply.
 
Gnerma said:
What are the voltage ratings on that PSU if you don't mind me asking? Also, how does your Vcore behave if you crank it up to 1.5 or 1.6? Do you see that 'heating up' phemomenon in BIOS I described after booting up your rig? (from a cold boot of course)


(+5, +12, +3.3 MAX at 410w)
+5v= 36a
+12= 26a
+3.3= 28a
-5v= .5a
-12= 1.0a
+5SB= 2.0a

Vcore acts the same at 1.4volts and 1.6v. I did notice on a cold boot, the vcore would fluctuate +.014v higher, then after a few minutes, -.014v lower and stay that way. So, 1.5 looks like 1.500, 1.514, 1.500, 1.514.... and after a while it looks like 1.5, 1.486, 1.5, 1.486.
 
Thanks a lot Tyranos. You have me wondering if the 18A rating on the Fortron 12v is the root of the problem here. After all 22A - 24A seems to be a standard issue recommendation when it comes to decently overclocked A64 setups. Of course knowing how information travels in internet hardware circles, I take things that everybody "knows" like this with a grain of salt. I usually am quicker to put my trust in first hand information from obviously experienced users; hence I listened to Larve and own a Fortron 530.

I figure somebody with my problem has a few options. One, I could tweak the pots in my PSU. These units don't have a warranty blowing sticker or any such thing so this can be done at the cost of only effort. Two, I could add some passive cooling to relevant components. This would probably help a lot with the fluctuation but is almost a sure warranty killer on the board. Three, I could try different PSUs which would cost $20 - $40 per unit swap after I sell & ship my previous one.

Before I do anything I'd like to hear from more people about their setups though. Also we're still waiting for Super Nade's reply to my questions :burn:
 
-A64 3500+ Newcastle @240x10 1.55 vcore
-Fortron 530w
-Mushkin PC3200 2x512
-Radeon X800XT PE stock
-Creative Audigy

Abit EQ reads my vcore @1.58-1.6. I think I read on xtremesystems that the AV8 overvolts the vdimm by .05. Maybe it does this to the vcore also?
 
Oh i forgot to add...
-1 80mm chassis fan and a retail AMD heatsink/fan
-2 WD 80Gb IDE hardrives
-and a LG DVD player
 
12 V before : 11.286~11.484 V
12 V after : 11.8 ~ 12.3 V
A real PITA as the pots are very very sensitive. I would much prefer to do this without powering up my rig, shorting appropriate 12V PSU pins instead. Took me forever as its better being slow than have a dead MoBo.
VCore before 1.65 shows as 1.624
VCore after 1.65 shows as 1.64, sometimes as 1.63 but I'm scared to mess with it further ;)

{I did this a while ago. I can check over the weekend and update this post)


I think the problem is with the voltage regulators on board. I sinked all I could find, but the ones hidden by the HS seem to be the troublemakers. There is nothing much I can do about them. I may try to lap a penny or two and stick it in there. I'm using big BGA RAM sinks so that's stabilized my VCore fluctuations a bit.

I followed the procedure in here:-
http://www.ocforums.com/showthread.php?t=352933
 
hm...... i got a abit av8 and all i can say its a total pos, doesn't overclock worth crap, best i can get is 240 x 10 with my ram divided, i have tccd pd memory and when i get it at 240 x 10 its soo unstable it shuts down every 2-3 mins
 
Saud said:
hm...... i got a abit av8 and all i can say its a total pos, doesn't overclock worth crap, best i can get is 240 x 10 with my ram divided, i have tccd pd memory and when i get it at 240 x 10 its soo unstable it shuts down every 2-3 mins
This is totally off topic. If you want to talk about your overclocking situation, please start a new thread. Thanks :)
 
Super Nade said:
12 V before : 11.286 ~ 11.484 V
12 V after : 11.8 ~ 12.3 V

I think the problem is with the voltage regulators on board. I sinked all I could find, but the ones hidden by the HS seem to be the troublemakers. There is nothing much I can do about them. I may try to lap a penny or two and stick it in there. I'm using big BGA RAM sinks so that's stabilized my VCore fluctuations a bit.
Wow looks like you had a really big problem with your PSU initially. Perhaps the pots were just set wrong out of the factory. Although I don't see those massive voltage swings on my unit, it never even moves. I assume that you mean "Idle ~ Load" and not "Load low ~ Load high"?

In reading through all of that Fortron 530 literature, it seems like your situation is more typical though. It has also made me think that adjusting the pots would not help me much if at all since my unit seems to be at pretty good settings already.

Anyway, how stable is your Vcore? Pretty much just 1.64 with occasional drop to 1.63?
 
Alright well here is the plan. As soon as I am sure that these boards will take Revision E chips without issue, I'll sink up my MOSFETs and hope for the best. I took some time to really look at other AGP s939 boards and I didn't see any compelling reasons to switch. Not that there are many other boards available. In fact, the only other boards I'd probably consider buying is the Asus A8V.
 
My VCore is pretty stable 1.65+-0.02 since I sinked a few FET (the ones hiding under/near my HS bother me though. No problems since.

I was leaning towards the DFI board, but I'm really happy with my current setup. The only problem with the current crop of boards is the BIOS. I don't know why every maker struggles with the BIOS. They have a big team, they have the CPU before we get it, so what's the big deal?
 
Isnt 0.05V fluctuation harmless? Vcore on my AV8, when set to 1.425V, various between 1.4V (under load) out to 1.45V.
 
Dexx said:
Isnt 0.05V fluctuation harmless? Vcore on my AV8, when set to 1.425V, various between 1.4V (under load) out to 1.45V.
It depends on what you mean by harmless. If you are asking if I think the fluctuation will damage my hardware in any way no I do not. Think of it this way though, this is a classic "chain is only as strong as it's weakest link scenario." What if my processor needed a steady supply of 1.55v to be perfectly stable at 2.5 ghz? If that were the case, and assuming my Vcore fluctuates between 1.55 & 1.6v, what is all that extra voltage above 1.55 for? Nothing. It is a load of wasted power and unwelcome heat. The same could be true for your CPU. I live in the desert and summer is right around the corner. The less heat I have to deal with the better.

That is the theory behind the fluctuation being a bad thing anyway. Although a recent PM discussion with larve has me looking at it from a more neutral perspective.

EDIT - BTW Dexx how about some Spexx?


Super Nade said:
The only problem with the current crop of boards is the BIOS.
Current crop of AV8s or DFI boards? Do you mean the Nforce3 or 4 DFI? And what is wrong with the BIOS' you think they should get right?
 
Last edited:
Both VIA and NF4 boards are/were having massive problems working with the Winnys. They have been attributed to the change in default working VCore from 1.5 to 1.4 between the earlier NC and the current Winnys. Expect more teething troubles with Venice.
 
Massive problems working with Winchesters? Sure I agree that not posting without a BIOS update as was the case with some boards (including the AV8) is a massive problem but I'm not sure what else you mean. Give me some examples. Perhaps these nebulous problems come from the fact that we push our systems to the breaking point and the makers don't exactly test for that. And I don't think they should really have to.
 
mine fluctuates from 1.57-1.61 when i set 1.575 in the bios, and 1.68-1.71 when i set 1.7v in the bios.

That said i never use it over 1.6v for extended periods of time (1hr+) since it crashes in 3d apps over 255htt~ so its 10*250 @ 1.575v for 24/7 use.

I was wondering if any of you get fluctiations on the NBVDD? Mine jumps all over the damn place.. between 1.6 and 1.75v when i set 1.65v in the bios...
 
Last edited:
Back