I made a pic of what you have found out:
Ground is on the back of the card, where the casing of DVI connector is soldered to the PCB.
EDIT: STEvil you beat me to itbut my pic is prettier!
I made a pic of what you have found out:
Ground is on the back of the card, where the casing of DVI connector is soldered to the PCB.
EDIT: STEvil you beat me to it
www.teampclab.pl
MOA 2009 Poland #2, AMD Black Ops 2010, MOA 2011 Poland #1, MOA 2011 EMEA #12
Test bench: empty
Found the datasheet for our ram I think: http://www.qimonda.com/download.jsp?...rev143_www.pdf
How do I get nibitor to read timings?
Edit - Lists 2.3v as maximum relative to Vss (ground, ~0v).... so 2.2 should be fine for benching then crank it up for final run lol
Last edited by STEvil; 02-10-2008 at 01:41 AM.
All along the watchtower the watchmen watch the eternal return.
Use your digital multimeter and make sure those are both grounded, but they should be.. I used the on on right.
EDIT
Data sheet compared to stock timings you showed. I'm not near my bench machine to get its timings but I suspect they are the same... unless thats why yours gets 20mhz better than me at stock lol
Last edited by STEvil; 02-10-2008 at 02:08 AM.
All along the watchtower the watchmen watch the eternal return.
Top two caps
All along the watchtower the watchmen watch the eternal return.
810 core 1900 shader @ 1.5v
2000 shader probably would have gone as well, was being conservative trying to find a stable setting lol
All along the watchtower the watchmen watch the eternal return.
2.2v seems to be the max stable for the qimonda on my 8400.. 475mhz 3D03, 477mhz 3D01. Any more volts and it gets unstable.
A note on Nibitor guys.. I figured out how to work it and there is a "test timings" option so you dont have to flash the bios to play with them.. this is a good thing as I was testing some timings and a lot of the time ended up with screen corruption and etc...
Last edited by STEvil; 02-11-2008 at 04:37 PM.
All along the watchtower the watchmen watch the eternal return.
Maybe they dont like more volt because they have the same problem as the 8800 GTīs with the Qxxxxxxxx chips.
Dont even want to try and spell it :p
When giving to much volt Vtt or was it Vddq increases to much.
on 8800 GT they did a mod to lower Vmem and that allowed for around 40-50 Mhz more on Mem.
Of those three legs on the 431L, which leg takes power from the card in general, which leg sends power to the memory, and what is the third leg doing? Is that maybe a three terminal, third leg as offset, the one alone on a side as incoming power, and the one with the red marked dot to send power to the memory chips?
Should be a supply, ground, and output. I havent looked really close.
I got my card to run the mem at 2.4x volts stable but my solder job fell off so i gotta redo it.
That mosfet to the right of the 431L next to the two caps gets rediculously hot.
All along the watchtower the watchmen watch the eternal return.
The 431 is the drive for the memory power mosfet's, if the mosfet just next to it get's really hot (like... you can't touch it), then please add some extra sinking on your card, otherwise it mights just burn or crack.
Things got hot enough with sustained running at 2.48v to unsolder my wire to the 431L transistor
All along the watchtower the watchmen watch the eternal return.
Auwh, think you can measure actual surface temp of the mosfet via IR or something?
If I could find my IR thermometer yeah.. dunno where it is though.
All along the watchtower the watchmen watch the eternal return.
I'm running into the overcurrent protection at 1.95v and 904mhz+ lol
Anyone got an idea on how to disable/increase it?
All along the watchtower the watchmen watch the eternal return.
Pin 13 to ? should be overcurrent: http://pdf1.alldatasheet.co.kr/datas...C/APW7068.html
All along the watchtower the watchmen watch the eternal return.
There is a resistor between the OCSET pin and ground. Measure it, remove it and solder in a resistor with ~2-3 times more resistance. (I cant see the pictures in the pdf but going by the description that should be right
Q9550 || DFI P45 Jr || 4x 2G generic ram || 4870X2 || Aerocool M40 case || 3TB storage
Looks like its the "44B" resistor (hard to read it).
http://www.xtremesystems.org/forums/...75&postcount=7
An SMD capacitor across the SMR should increase OCP as well correct?
All along the watchtower the watchmen watch the eternal return.
Confirmed.
Remove the 44B resistor and OCP is done for
Reached 1000/2500 desktop @ 2.3v ati tray tool artifact tester. Couldnt bench over 905/2430 due to heat though.
Ran up to 2.5v testing, nothing over 2.05 gave me any benchable gains on my water setup.
All along the watchtower the watchmen watch the eternal return.
Question ? Are you guys using rivatuner to overclock ? because I use rivatuner to overclock my eVGA 8400GS and it will only allow me to oc to 702 Max ? is it the drivers or rivatuner ? How do you guys get it higher than 702 ?
Sorry for the noobie question. Any help appreciated.
chispy.
You need a volt mod to go above 700 usually.
All along the watchtower the watchmen watch the eternal return.
Originally Posted by STEvil
Thank You for the response , the card its volt moded already but all i can get it to its to 702/512. The core its holding me back or the drivers or rivatuner ??? dont know why but in rivatuner the max allow overclclock on the core that shows its 702 turning the slider all the way tru the right side to the end. Any help or guidance as to why its this will be appreciated.
chispy.
Posting Permissions
Reply With Quote
Bookmarks