Page 4 of 9 FirstFirst 1234567 ... LastLast
Results 76 to 100 of 211

Thread: 512mb 7900GT vmod development thread

  1. #76
    Champion
    Join Date
    Jun 2003
    Location
    cape town
    Posts
    1,172
    i should have mentioned that i have tried both a 50k and 20k vr on pin#2 of that LM358.
    i did mention that i varied every output i could find though.
    it supposedly varies the output on pin#1 according to the datasheet.
    anyway i twisted the vr's both to 0kohm resistance eventually but saw no change on the output of pin#1 or for that matter the +- 1.74v reading found in a couple of places.
    maybe that output is not being used
    anyway maybe you have another way?

    also i loaded a 7900gt bios i found with nibitor and discovered timingset7 actually looser on the 512mb cards.
    (not sure if 7 is the one used by 256mb card. someone please confirm)
    also testing the tighter timings now.


  2. #77
    Banned
    Join Date
    Jul 2004
    Location
    Scotland
    Posts
    926
    Good work
    Last edited by humeyboy; 05-13-2006 at 09:09 AM.

  3. #78
    Xtreme Vmod mod
    Join Date
    Aug 2004
    Location
    Cornwall/Newcastle uni, UK
    Posts
    1,234
    Can someone check the things I've marked in yellow in the attachment please.

    @flytek
    I think that pins 1, 2 and 3 on the LM358 are controlling the 1.24v voltage. The voltage might not have adjusted when you added a VR if there was no resistor connected between the output of the regulator and pin 2 on the LM358 (ie if the 2 points are directly connected together)
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	lnreg1.JPG 
Views:	956 
Size:	67.6 KB 
ID:	47105  

  4. #79
    Champion
    Join Date
    Jun 2003
    Location
    cape town
    Posts
    1,172
    well i did some connectivity readings on pin1&2 on LM358.
    from pin1 the resistor below is connected to the horizontal one below it which is in turn connected to the one below pin2 > which is connected to pin2.
    you can almost follow the traces in our nice high rez shots
    and yes if i remember right pin2 is also connected to the right mosfet leg as indicated above.
    if someone else would confirm i'd appreciate it as you have no idea how difficult it is to measure the underside of a mounted card while it is running.
    i've been using a mirror to see what i'm doing like a dentist would
    Last edited by s e t h; 05-13-2006 at 12:34 PM.

  5. #80
    Banned
    Join Date
    Jul 2004
    Location
    Scotland
    Posts
    926
    Are they pics my high res agp pics I got hosted ? (I rec the caps)

  6. #81
    Registered User
    Join Date
    Apr 2003
    Posts
    27
    Quote Originally Posted by persivore
    Can someone check the things I've marked in yellow in the attachment please.
    Resistance: connected
    Voltage: impossible to measure for me, since the part is directly under the ramslots...

  7. #82
    Champion
    Join Date
    Jun 2003
    Location
    cape town
    Posts
    1,172
    Quote Originally Posted by humeyboy
    Are they pics my high res agp pics I got hosted ? (I rec the caps)
    yes they are. thank you very much

  8. #83
    Banned
    Join Date
    Jul 2004
    Location
    Scotland
    Posts
    926
    Well least I did my little part by stripping my card and scanning LOL.

  9. #84
    Xtreme Vmod mod
    Join Date
    Aug 2004
    Location
    Cornwall/Newcastle uni, UK
    Posts
    1,234
    Can someone check if the point I've marked is connected to ground, and the value of the resistor marked please.
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	lnregref.JPG 
Views:	924 
Size:	38.7 KB 
ID:	47107  

  10. #85
    Registered User
    Join Date
    Apr 2003
    Posts
    27

  11. #86
    Xtreme Vmod mod
    Join Date
    Aug 2004
    Location
    Cornwall/Newcastle uni, UK
    Posts
    1,234
    To increase the 1.24v voltage (I still have no idea what it controls or what effect it has - increase at your own risk), either connect a 20k VR over the 667ohm resistor in the above posts, or try penciling it slightly to decrease its resistance.

  12. #87
    Registered User
    Join Date
    Apr 2003
    Posts
    27
    Where can that ominous 1,24V be measured?
    What gives you reason to think it is mem-related at all?

  13. #88
    Xtreme Vmod mod
    Join Date
    Aug 2004
    Location
    Cornwall/Newcastle uni, UK
    Posts
    1,234
    Quote Originally Posted by resonatorman
    Where can that ominous 1,24V be measured?
    What gives you reason to think it is mem-related at all?
    I've attached a picture of where you can measure the 1.24v voltage. Nothing makes me think that its related to the memory at all, but a couple of people in the thread thought that increasing it might help.
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	persivore1.jpg 
Views:	914 
Size:	67.1 KB 
ID:	47109  

  14. #89
    Banned
    Join Date
    Jul 2004
    Location
    Scotland
    Posts
    926
    1.24V sound like default core volts which is 1.2V

  15. #90
    Xtreme Vmod mod
    Join Date
    Aug 2004
    Location
    Cornwall/Newcastle uni, UK
    Posts
    1,234
    Quote Originally Posted by humeyboy
    1.24V sound like default core volts which is 1.2V
    That was checked earlier and was found not to be related to Vgpu. I have no idea what the voltage is used for.

  16. #91
    Banned
    Join Date
    Jul 2004
    Location
    Scotland
    Posts
    926
    The 1.74V sounds like memory going by rumours of it being starved to 1.7v.

  17. #92
    Xtreme Vmod mod
    Join Date
    Aug 2004
    Location
    Cornwall/Newcastle uni, UK
    Posts
    1,234
    Quote Originally Posted by humeyboy
    The 1.74V sounds like memory going by rumours of it being starved to 1.7v.
    The 1.74v looks like its the PWM signal for the Vdd supply before it goes into an LC filter to produce the final Vdd supply. Its just a voltage which is switching between 12? and 0v, which gets averaged out by the multimeter to give 1.74v. I don't think its a seperate voltage supply on the card. This could be proved by seeing if the voltage increases as Vdd is increased.

  18. #93
    Banned
    Join Date
    Jul 2004
    Location
    Scotland
    Posts
    926
    Lets hope it is less than 2v, then at least we know thats why ram cant run at rated specs then in time we get around it, I think Shamino will sort it eventually. (very hard when he dont have the card to work on).

  19. #94
    Champion
    Join Date
    Jun 2003
    Location
    cape town
    Posts
    1,172
    Quote Originally Posted by persivore
    The 1.74v looks like its the PWM signal for the Vdd supply before it goes into an LC filter to produce the final Vdd supply. Its just a voltage which is switching between 12? and 0v, which gets averaged out by the multimeter to give 1.74v. I don't think its a seperate voltage supply on the card. This could be proved by seeing if the voltage increases as Vdd is increased.
    it doesn't afaik. i would have noticed when i had my 2.10vdd up at 2.25

    btw a measure 642 ohms on that resistor you requested.
    some more details below too


  20. #95
    Xtreme Vmod mod
    Join Date
    Aug 2004
    Location
    Cornwall/Newcastle uni, UK
    Posts
    1,234
    I'm 99% sure that the 1.74v voltage is just a PWM signal for Vdd. If you want to check it tho, check the voltages at the points on the inductor I've marked. From what I can tell, it looks like thats what you have done in the above post anyway, but on the back of the card .
    Attached Thumbnails Attached Thumbnails Click image for larger version. 

Name:	pwm1.JPG 
Views:	918 
Size:	84.1 KB 
ID:	47111  

  21. #96
    Champion
    Join Date
    Jun 2003
    Location
    cape town
    Posts
    1,172
    top would measure 1.75v and bottom 2.10v as indicated on my last pic

  22. #97
    Banned
    Join Date
    Jul 2004
    Location
    Scotland
    Posts
    926
    Hmm, wonder if the ram is getting the 1.75v instead of 2.1v to intentually limit it.

  23. #98
    Xtreme Vmod mod
    Join Date
    Aug 2004
    Location
    Cornwall/Newcastle uni, UK
    Posts
    1,234
    Quote Originally Posted by humeyboy
    Hmm, wonder if the ram is getting the 1.75v instead of 2.1v to intentually limit it.
    Nothing will be supplied by the 1.75v voltage. Its a PWM signal, not a DC voltage. The DC voltage for the RAM supply is produced after the PWM signal (the 1.75v voltage) is smothed out to a DC voltage (The 2.11v voltage).

  24. #99
    Champion
    Join Date
    Jun 2003
    Location
    cape town
    Posts
    1,172
    pin#7 on LM358 outputs 5.5v at stock. i had that 10% over to +- 6volts without any change.

    when playing with pin#2 of LM358 i should have checked the effect on the 1.24 and 3.7v measurements on the lowest mosfet.
    maybe the arrangement of capacitors around pin#1&2 of LM358 prevented me from having the usual effect when lowering resistence on pin#2.

    anyway my entire stash of these cards i sold so i'm out of this for a while.

    good luck to everyone trying to get these babies working

  25. #100
    Xtreme Member
    Join Date
    Jan 2006
    Location
    Italy (Pisa)
    Posts
    101
    hi, i'm a italian boy... and is the first time that I write on this forum..

    I bought from new egg a gainward bliss 7900gt 512mb but the ram do not ascend beyond 750mhz, how can do it??

    exists a way to increase the frequency of the ram?

    Ps: sorry for my poor english

    tnx

Page 4 of 9 FirstFirst 1234567 ... 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
  •