Asus M3A79-T Deluxe

[Chicken Patty]

http://en.wikipedia.org/wiki/Phase-locked_loop

read that, the thing is that how does it come into play with our computers. What does it do?

Thanks by the way.

[Oliverda]

read that, the thing is that how does it come into play with our computers. What does it do?

Thanks by the way.

Have you heard something about the clock generation?

[Chicken Patty]

Have you heard something about the clock generation?

no sir

[Oliverda]

no sir

http://en.wikipedia.org/wiki/Clock_generator

[Oliverda]

Some of my major issues with the board: http://img221.imageshack.us/my.php?image=clipboard02dd7.jpg

[Chicken Patty]

thank you oliverda

[gOtVoltage]

Some of my major issues with the board: http://img221.imageshack.us/my.php?image=clipboard02dd7.jpg

1.40v in Bios = 1.46v can be from what ever the VDD is set to if on Auto..(2.5v vdd up to 2.8v vdd..)The Higher you go with VDD the Higher it raises the Vcore for both CPU/NB


Example :
2.5vdd + 1.40vcore = 1.45v
2.6vdd + 1.40vcore = 1.465v
2.7vdd + 1.40vcore = 1.482v
2.8vdd + 1.40vcore = 1.50v

Those are not exact but thats how the Auto VDD function works.. i usually start with 2.5vdd and set the Vcore manually and even then it will ADD Xtra vcore...Just how this Mobo is...

As for the prime with 1.030v load it probally a glitch with reading the Diode..

I use (CPU-ID Hardware monitor) and (CorTemp).. seems the best compared to Speedfan which dosnt read and AMD Monitor which can also be buggy...

Im still testing my Max for above 3.52ghz ,, Waiting for Replace ment Cooler to arrive...Since i sold one of my BIX !!!'s i decided to add a 120.3 Feser,, I know i got the Cheap 17mm one$99bucks ...i just had to try it...I have a PA 120.3 somewhere in the Attic but am too lazy to check.. Everything went into attic which wasnt attached to a running rig .

P.S. Everest reads things fine..The ASUS mobo does read 10c higher overall... ..But then again full load on 3.2ghz and Air @ 63c dosnt surprize me if you subtract 8 - 10c it would yield 53c with 1.46v seems kinda Kosher to me if Prime is 100% and no crashes after a few hours.

[TiTON]

Great job Gotvolt!!

i was curiosu what the VDD did. I am currently running mine on auto. I didn't know what it did, and adding more and less didn't appear to give me any extra boost.

Is VDD only used to "bump" vCore.. or is the voltage actually used for something else, and the vcore Bump is just a side effect?

Ton

[Oliverda]

1.40v in Bios = 1.46v can be from what ever the VDD is set to if on Auto..(2.5v vdd up to 2.8v vdd..)The Higher you go with VDD the Higher it raises the Vcore for both CPU/NB


Example :
2.5vdd + 1.40vcore = 1.45v
2.6vdd + 1.40vcore = 1.465v
2.7vdd + 1.40vcore = 1.482v
2.8vdd + 1.40vcore = 1.50v

Those are not exact but thats how the Auto VDD function works.. i usually start with 2.5vdd and set the Vcore manually and even then it will ADD Xtra vcore...Just how this Mobo is...

I want 1.40V at load and at idle as well. It overvolts the CPU at load like hell. I think it is not normal.

As for the prime with 1.030v load it probally a glitch with reading the Diode..

I use (CPU-ID Hardware monitor) and (CorTemp).. seems the best compared to Speedfan which dosnt read and AMD Monitor which can also be buggy...

I know that the 1.030v is a fake value. I would like to show that when the CPU temp (which is incorrectly reported by the mainboard) reaches the
approx. 68-70C degrees the system automatically switches back the CPU's multiplier to 6.5x.

[justapost]

http://en.wikipedia.org/wiki/Phase-locked_loop

The oscillator generates a periodic output signal. Assume that initially the oscillator is at nearly the same frequency as the reference signal. Then, if the phase from the oscillator falls behind that of the reference, the phase detector changes the control voltage of the oscillator, so that it speeds up. Likewise, if the phase creeps ahead of the reference, the phase detector changes the control voltage to slow down the oscillator. A low-pass filter smooths out abrupt changes in the control voltage; it can be demonstrated that some filtering is required for a stable system. Since initially the oscillator may be far from the reference frequency, practical phase detectors may also respond to frequency differences, so as to increase the lock-in range of allowable inputs.

Thank you for the link oliverda. So if i increase VDDA the circuit might not fall behind more frequent with higher clocks.

[demonkevy666]

I want 1.40V at load and at idle as well. It overvolts the CPU at load like hell. I think it is not normal.



I know that the 1.030v is a fake value. I would like to show that when the CPU temp (which is incorrectly reported by the mainboard) reaches the
approx. 68-70C degrees the system automatically switches back the CPU's multiplier to 6.5x.

thermal throttling ?> maybe

that was VDDA right ? I don't think it's bumping mine up :/ me temp was 45C at 1.35

[gOtVoltage]

Great job Gotvolt!!

i was curiosu what the VDD did. I am currently running mine on auto. I didn't know what it did, and adding more and less didn't appear to give me any extra boost.

Is VDD only used to "bump" vCore.. or is the voltage actually used for something else, and the vcore Bump is just a side effect?

Ton

Its like the how VID works just adds more to the Base Voltage and helps keep the Voltage supply up when go high CPUclocks..I stumbled accross it by shear luck one day

I want 1.40V at load and at idle as well. It overvolts the CPU at load like hell. I think it is not normal.



I know that the 1.030v is a fake value. I would like to show that when the CPU temp (which is incorrectly reported by the mainboard) reaches the
approx. 68-70C degrees the system automatically switches back the CPU's multiplier to 6.5x.

I even experianced the Thermal shut down ,, i tested with High than Normal Vcore and Max VID..i mean Above 1.75volts total and i could make it happen..This Mobo reads 10c higher than Normal so any temp range below 65c can easily read 70c+ when its not suppose too..

As for the VID or Volts being 1.40v consistan youll have to play with each VDD setting and adjust the VCORE for CPU untill you have 1.40-1.42v under load ... Using CPUID hardware monitor and Coretemp will read VID and VCORE close to accurate..

Its like a game in its self just to get what you want Volt wise with the SB750 chip set .. SB600 was easier to set i believe..Then again AMI is little different from settings in AWARD...Thus i give props to AWARD bioses..

Thank you for the link oliverda. So if i increase VDDA the circuit might not fall behind more frequent with higher clocks.

Yup,, it does help alot when High Oc is involved especially when running High NB too.

thermal throttling ?> maybe

that was VDDA right ? I don't think it's bumping mine up :/ me temp was 45C at 1.35

youll need to adjust each setting till you find the sweetspot..Trust me its not as easy as it looks and can be time consuming ,,Just like finding good Timing on ram i do one setting at a time and write down what it did then try another untill i get what i want

NEW UPDATE 11/2/2008

ASUS website has 504 BIOS ready AGESA update...

I just downloaded it ...Trying to look up info if anyone else has tested/tried it yet....

[XRogerX]

Well mine will be here tomorrow and i will post some overclock with a 5400+ BE

[m411b]

NEW UPDATE 11/2/2008

ASUS website has 504 BIOS ready AGESA update...

I just downloaded it ...Trying to look up info if anyone else has tested/tried it yet....

Read post 265 and there after!

[gOtVoltage]

Hahaa, partyfoul im bad,,its just a reminder

[Jazzman]

I have done some research on this bios option and discovered some things. First, the majority of web descriptins of CPU Tweak say it is part of a performance solution for the TLB bug fix. Supposedly, selecting this option without the TLB fix option enabled causes increased cpu heat under load.

It was enabled by default in the ASUS M3A79-T bios so I left it there prior to my research. Under heavy load, while under water, I was getting about 47.5C - 49C in core temp. I have been thinking this was a little high for only 3.2 ghz.

After disabling CPU Tweak in the bios, I noticed that the red circle around the green button in the upper-right corner of AOD was gone. I then decided to test cpu temps while running OCCT-a great high load stability test program. My temps dropped on average 4C !!!

YMMV, but disable this bios option and run a high load test program to see the impact on temps. Some posters here have seen increased performance with it enabled but I have tested it all day and gotten better performance with it off.

[merkk1]

I was testing the red ring on the green botton in AOD to day and on my Foxconn A79A-s running 3dmark-06 it pick me up 200-300 points on my score by having it on . And it dose make the cpu run hotter but i'am using water cooling so it was no big deal .

[Daveburt714]

Thank you for the link oliverda. So if i increase VDDA the circuit might not fall behind more frequent with higher clocks.

I'm curious about this effect too. It's pretty obvious that raising CPU VDD will increase Vcore as well, but what would happen if you raised VDD to 2.8v and lowered Vcore to 1.3. Would that result in a 1.4v Vcore while improving Clock Gen syncronization?

And what would be the effect of that on the PWM section of the Mobo?

Excuse me if I'm a little skiddish, but I killed my last board with the AOD.ini mod to allow higher CPUvid, and I actually have to buy my own parts...

[soundood]

now in the uk, (has been for a week or so)

http://www.microdirect.co.uk/Product...ource=googleps

this internet site is sound, i have ordered from them lots of times, never a problem

[RASTAVIPER]

I'm curious about this effect too. It's pretty obvious that raising CPU VDD will increase Vcore as well, but what would happen if you raised VDD to 2.8v and lowered Vcore to 1.3. Would that result in a 1.4v Vcore while improving Clock Gen syncronization?

And what would be the effect of that on the PWM section of the Mobo?

Excuse me if I'm a little skiddish, but I killed my last board with the AOD.ini mod to allow higher CPUvid, and I actually have to buy my own parts...

Let me see if I got it right.
For example you want to put 1,475 in vcore to keep stable a good OC.
You believe it's better to have this vcore by combining VDD +lower vcore, then just increasing the vcore?
The combination of VDD and Vcore helps OCed systems??

[Daveburt714]

Let me see if I got it right.
For example you want to put 1,475 in vcore to keep stable a good OC.
You believe it's better to have this vcore by combining VDD +lower vcore, then just increasing the vcore?
The combination of VDD and Vcore helps OCed systems??

Well you have it pretty much right RASTAVIPER..

I would actually rather keep my Vcore around 1.4 though. Justapost is a really bright guy with lots of experience using X4 systems, so it was more of a question for him than an option for people to try....

[justapost]

I'm curious about this effect too. It's pretty obvious that raising CPU VDD will increase Vcore as well, but what would happen if you raised VDD to 2.8v and lowered Vcore to 1.3. Would that result in a 1.4v Vcore while improving Clock Gen syncronization?

And what would be the effect of that on the PWM section of the Mobo?

Excuse me if I'm a little skiddish, but I killed my last board with the AOD.ini mod to allow higher CPUvid, and I actually have to buy my own parts...

I don't understand this eighter dave. Have no clue why a higher vdda results in an higher vcore. Maybe it's just the sensor reading. Noone prooved it with a DMM yet. As far as i understood the wiki page an higher voltage results in an higher frequency and the circuit has to increase the voltage if the frequency is too low. I assume the stock 2.5V are good for stock clocks but with higher clocks that pll will have to do more correction work.

[RASTAVIPER]

Well you have it pretty much right RASTAVIPER..

I would actually rather keep my Vcore around 1.4 though.

Vcore at 1,4 and vdd at which number??
Something like this?
2.5vdd + 1.40vcore = 1.45v

[gOtVoltage]

Vcore at 1,4 and vdd at which number??
Something like this?
2.5vdd + 1.40vcore = 1.45v

thats pretty much how it is,, use less or more Vcore for your target volt range that you will use...

Ive been testing High Vcore for two weeks @ 2.8vdd + 1.65v = 1.70v for both NB and CPU...This mobo can handle just about anything...Also the Same Voltages were used when i pushed Mobo@ HTT 292 x12 = 3.5ghz while having PCIE/core and HT @ 2.8v while keeping Southbridge @auto ...

This mobo has let me achieve the highest HTT @ 3.5ghz vs other Mobos ive tried..Yes Mobo can be buggy and does need lots of work to tweak. Updated Bioses will definatly make it even better.

The stock heatpipe works great. I had pull it off just to see if it had good flat contact.. After a small tweak of the pipe(not the funnyPipe) i installed the same thermal tape on pwms and used MX-2 on NB/SBchipset ,,thats all i did,, Oh yeah i flashed to newest BIos yesterday and it gives me the same results as the 40x vs 504.

[Oliverda]

The VDDA voltage doesn't have any impact to the VCORE on my board. It overvolts same the CPU at 2.5V VDDA like at 2.8V.

Overvolt mania by ASUS.....yeah...