Page 63 of 73 FirstFirst ... 135360616263646566 ... LastLast
Results 1,551 to 1,575 of 1814

Thread: ATI 4850/4870 voltmod thread

  1. #1551
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    Thanks for reply , i will measure the resistance later today when i get back home .
    I have tried this mod since the author confirmed that it work on palit sonic video card same design as mine (check post #1164) , but the card was the HD 4850 sonic.
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  2. #1552
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    celemine1Gig thanks for your support .
    messured resistance between
    pin 1 - 19 = 10.06 k
    pin 14 -19 = 9.75 k
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  3. #1553
    IT Engineer in the making
    Join Date
    Jun 2003
    Location
    Former Kingdom of Bavaria
    Posts
    2,094
    I would recommend either 2 200k potentiometers connected in series (resulting resistance of 400k), or alternatively a single 500k (or at least! a 200k) potentiometer. Keep in mind that the latter will have less precision in lower resistance (=higher voltage) ranges.
    Anything less, concerning the total resistance, would give too much of a default voltage bump IMHO.

    Set (all) the potentiometer(s) to their/its max. resistance and connect them/it bewteen pin#19 and pin#14. That should work. Maybe you got a black screen because you hit the OVP voltage limit. Then you could consider yourself lucky.

    BTW: What the heck does that mean "50 h vr"?

    And another thing:

    Quote Originally Posted by makaka View Post
    celemine1Gig ...messured resistance between
    pin 1 - 19 = 10.06 k...
    I'm sorry, if this sounds rude, but do you even remotely know what you are doing?
    Why would you want to measure resistancee between pin#1 and pin#19, which correspond to "EN" and "Vfb"? What do you think has the "Enable" pin to do with the feedback in terms of a vmod.

    Quote Originally Posted by NCP5388 datasheet
    EN Pull this pin high to enable controller. Pull this pin low to disable controller. Either an open−collector output
    (with a pull−up resistor) or a logic gate (CMOS or totem−pole output) may be used to drive this pin. A Low to
    High transition on this pin will initiate a soft start. If the Enable function is not required, this pin should be tied
    directly to VREF.
    Last edited by celemine1Gig; 05-11-2009 at 02:54 AM.
    Quote from one of our professors:
    "Reality is hiding in the imaginary part."

  4. #1554
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    HI there ;
    i will try with 2 x 200k or 500 k potentiometers ,

    BTW: What the heck does that mean "50 h vr"?
    i meant 50 k potentiometers
    I'm sorry, if this sounds rude, but do you even remotely know what you are doing?
    No i don't know :-) , i just saw a post above asking for the resistance between those so i measured it , may be it will be useful for u .

    Also can i solder to other points since the pins are pretty hard for my low soldering skills ,
    If yes please show them in the picture above .

    Also were can i check the vcore , (i tried those)

    they show 1.28v 2d and 1.33v (3d) , note with video card turbo button on .
    never checked with the normal mod
    Last edited by makaka; 05-11-2009 at 06:27 AM.
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  5. #1555
    IT Engineer in the making
    Join Date
    Jun 2003
    Location
    Former Kingdom of Bavaria
    Posts
    2,094
    Try using the empty solder pads to the right, that would normally be occupied by some SMD capacitors. That's a better place to measure the real VGPU, because it is behind the inductors and not before them, like the points you marked.

    And concerning VMem: One of the two AWP7065 probably regulates the memory voltage. Question's just which one (I would guess that it's the upper one). Some esaurements should clear that up.
    Quote from one of our professors:
    "Reality is hiding in the imaginary part."

  6. #1556
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    Quote Originally Posted by celemine1Gig View Post
    Try using the empty solder pads to the right, that would normally be occupied by some SMD capacitors. That's a better place to measure the real VGPU, because it is behind the inductors and not before them, like the points you marked.

    And concerning VMem: One of the two AWP7065 probably regulates the memory voltage. Question's just which one (I would guess that it's the upper one). Some esaurements should clear that up.
    HI there , Thanks for reply .
    The vmem is not required , the card can do 1100 stable .

    Also can i solder (vgpu mod) to other points since the pins are pretty hard for my low soldering skills ,
    If yes please show them in the picture above .
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  7. #1557
    IT Engineer in the making
    Join Date
    Jun 2003
    Location
    Former Kingdom of Bavaria
    Posts
    2,094
    Of course you can use different solder points, as long as they are directly connected to the named pins. And I won't mark them as I can't test the electrical connection. You can do that. It really isn't very complicated.
    Quote from one of our professors:
    "Reality is hiding in the imaginary part."

  8. #1558
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    Quote Originally Posted by celemine1Gig View Post
    Of course you can use different solder points, as long as they are directly connected to the named pins. And I won't mark them as I can't test the electrical connection. You can do that. It really isn't very complicated.
    thanks i will do it
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  9. #1559
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    HI there ;
    After checking i found that the pin 14 is not directly connected to any other point . this is clear from the PCB pic ,

    so even if not directly connected can i solder to it ?
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  10. #1560
    Xtreme Enthusiast
    Join Date
    Dec 2005
    Location
    Estonia, Tartu
    Posts
    889
    pin14 probably "dives" into PCB where it's connected to GND. So You have to use some other GND point near that IC. celemine1Gig might know it better...
    Vmods - what can possibly go wrong? ©

  11. #1561
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    Quote Originally Posted by ceemic View Post
    pin14 probably "dives" into PCB where it's connected to GND. So You have to use some other GND point near that IC. celemine1Gig might know it better...
    Thanks for reply , will any GND point work ? or it has to be near the IC , In theory any GND will work but go figure , this is why i am asking
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  12. #1562
    IT Engineer in the making
    Join Date
    Jun 2003
    Location
    Former Kingdom of Bavaria
    Posts
    2,094
    Yep, normally any GND point should work. The best option would be to use a point directly connected to AGND, as this is the real reference for the feedback line, but as AGND and GND shouldn't differ by much (AGND should be less prone to interference), it should theoretically not matter at all if you used just GND instead. The difference between GND and AGND as reference should max. be in the mV range.
    Quote from one of our professors:
    "Reality is hiding in the imaginary part."

  13. #1563
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    Quote Originally Posted by celemine1Gig View Post
    Yep, normally any GND point should work. The best option would be to use a point directly connected to AGND, as this is the real reference for the feedback line, but as AGND and GND shouldn't differ by much (AGND should be less prone to interference), it should theoretically not matter at all if you used just GND instead. The difference between GND and AGND as reference should max. be in the mV range.
    Thanks just done the Vmod , 1.48v and 900 mhz from the gpu stable (under water 40c load )
    Thanks again for your help
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  14. #1564
    IT Engineer in the making
    Join Date
    Jun 2003
    Location
    Former Kingdom of Bavaria
    Posts
    2,094
    And now post a highres picture with your exact solder points, for other poeple with the same card, as reference. I mean, I know how to work out mods, but others obviously don't, so this way you can give something back to the community.
    Maybe you can even put all the needed info + pic in one final posting, so one can link to that one posting in case someone's gonna ask for the mod in the future.
    Quote from one of our professors:
    "Reality is hiding in the imaginary part."

  15. #1565
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    Quote Originally Posted by celemine1Gig View Post
    And now post a highres picture with your exact solder points, for other poeple with the same card, as reference. I mean, I know how to work out mods, but others obviously don't, so this way you can give something back to the community.
    Maybe you can even put all the needed info + pic in one final posting, so one can link to that one posting in case someone's gonna ask for the mod in the future.
    will do it when i will get back home
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  16. #1566
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    First i want to thanks celemine1Gig and ceemic for their help finding the right vmod for the HD 4870 Palit soic edition (dual bios and dual fan) ,
    here is the video card picture :

    The vmod for the card and the vcore measure points :


    Here is some result using 2 Gb ram (470 mhz) , E8500 @ 4.2 Ghz (470x9) , p5q , 1000w psu , raid 0 :
    Note : stock volt according to point above was 1.335 v (maybe palit over volted the card ). the card max stable was 800 mhz GPU and 1050 GDDR5
    @ 1.44 v (load) 900 Mhz stable

    @ 1.44 v (load) 920 Mhz

    @ 1.44 (load) 940 Mhz

    @1.5v (load) 950 Mhz

    @ 1.55v (load) 960 Mhz

    @ 1.6v (load) XXX??? dark screen with restart GPU voltage show -0.03 v maybe a rest or OVP ,
    975 also passed the test @ 1.55v but there was a lot of artifact , so i decided to not publish the result .
    Any OVP mod for my card , i am looking for 1000 Mhz benchable
    Last edited by makaka; 05-16-2009 at 06:51 AM.
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  17. #1567
    Xtreme Enthusiast
    Join Date
    Dec 2005
    Location
    Estonia, Tartu
    Posts
    889
    makaka, 960MHz is very good OC
    Vmods - what can possibly go wrong? ©

  18. #1568
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    Quote Originally Posted by ceemic View Post
    makaka, 960MHz is very good OC
    Still need more , i looked for 1000 Mhz , any idea about OVP
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  19. #1569
    IT Engineer in the making
    Join Date
    Jun 2003
    Location
    Former Kingdom of Bavaria
    Posts
    2,094
    The OVP mod would probably include cutting one trace. To decrease the 1.3V limit at pin#17 (DIFFOUT), one would have to increase resistance between Vfb (pin#19) and DIFFOUT, in order to decrease the voltage at diffout. Probably not worth the effort as ~1.6V is very high already. I'd rather say that perhaps some other mods. Better filtering VGPU could help get another extra MHz. Sometimes, voltage alone isn't really that helpful.
    Quote from one of our professors:
    "Reality is hiding in the imaginary part."

  20. #1570
    Xtreme Enthusiast
    Join Date
    Jun 2005
    Posts
    601
    well what should be safe VGPU then for 24/7 usage under water , (40- 44 c load) ?
    2600K working in 4.8 GHZ so far

    2600k @(4600 Ghz) 1.42v : (under water)
    Asus B2 mobo
    4 GiG DDR 2400 MHZ
    GTX 570 @ 1.063v ( 910/1820/2001 Mhz) .(under water).
    1020 W PSU

    Hobie's : Overclocking .... Overclocking .... Overclocking

  21. #1571
    Xtreme Enthusiast
    Join Date
    Sep 2006
    Posts
    881
    Is there a pencil vgpu mod for the XFX 4850? I've found the mod thread, but it only has the solder mod. TIA.

  22. #1572
    Xtreme Enthusiast
    Join Date
    Jun 2007
    Location
    Athens
    Posts
    558
    Quote Originally Posted by makaka View Post
    well what should be safe VGPU then for 24/7 usage under water , (40- 44 c load) ?
    1.4 is pretty safe for 24/7 overclock

    more than 1.5 you will degrade the card :/
    i killed one 4850 and one 4870 @ 1.5 ( core artifacts ) and the were at 35*C load
    CPU : Q9550 / Board : Asus P5E64 WS Evolution / Ram : 2x1 OCZ D9GTR DDR3 / Vga : HD 4870 / PSU : PPC&C 750W / SSD Ocz Vertex 30Gb / All under Water & Tec's
    Overclockers Wannabe Athens Dept...

  23. #1573
    Registered User
    Join Date
    Apr 2005
    Location
    Croatia
    Posts
    2
    Hello
    Is there a pencil vgpu mod for the Gigabyte 4850 512mb? Or soldering mod?
    Here are the pictures
    Last edited by John S.; 05-25-2009 at 02:40 PM.
    Intel C2D E5200@3200Mhz 1.35V+Scythe Ninja 2
    Gigabyte EP35-DS3+Thermalright HR-05-SLI
    2x2gb Kingmax 800MHz
    Sapphire 4870 1Gb+Accelero S1
    Seasonic S12II 430W
    Seagate 250GB+Samsung Spinpoint F1 1TB
    Samsung 2233BW

  24. #1574
    Xtreme Member
    Join Date
    Dec 2007
    Location
    Sweden
    Posts
    322
    Quote Originally Posted by Vatos_locos View Post
    pff done it

    Hi guys!

    Have you figured out another vcore pencil-mod for this card as well?

    Please let me know if you do!


    Yours // fng77
    MotherBoard: GigaByte G33M DS2R
    Memory: 2x1 Gb HyperX DDR2-9600, 1200 Mhz
    Cpu: Intel e2160 @ 3,6 Ghz on Air
    Vga: XFX 4870 XXX
    Cooling: Ninja Mini, 2x Ximatek 120mm, 1x Zalman 80mm , Stock Gpu
    Case: Antec Fusion Remote

  25. #1575
    Registered User
    Join Date
    May 2009
    Posts
    3
    Quote Originally Posted by makaka View Post
    First i want to thanks celemine1Gig and ceemic for their help finding the right vmod for the HD 4870 Palit soic edition (dual bios and dual fan) ,
    here is the video card picture :

    The vmod for the card and the vcore measure points :


    Here is some result using 2 Gb ram (470 mhz) , E8500 @ 4.2 Ghz (470x9) , p5q , 1000w psu , raid 0 :
    Note : stock volt according to point above was 1.335 v (maybe palit over volted the card ). the card max stable was 800 mhz GPU and 1050 GDDR5
    @ 1.44 v (load) 900 Mhz stable

    @ 1.44 v (load) 920 Mhz

    @ 1.44 (load) 940 Mhz

    @1.5v (load) 950 Mhz

    @ 1.55v (load) 960 Mhz

    @ 1.6v (load) XXX??? dark screen with restart GPU voltage show -0.03 v maybe a rest or OVP ,
    975 also passed the test @ 1.55v but there was a lot of artifact , so i decided to not publish the result .
    Any OVP mod for my card , i am looking for 1000 Mhz benchable
    I have the same card, but I don't have the proper tools to do the mod. All I can afford to do ATM is a pencil mod.

    Would a pencil mod be possible on this card?

    Thanks!

Page 63 of 73 FirstFirst ... 135360616263646566 ... LastLast

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •