![](https://community.hwbot.org/uploads/set_resources_1/84c1e40ea0e759e3f1505eb1788ddf3c_pattern.png)
nachtfalke
-
Posts
860 -
Joined
-
Days Won
8
Content Type
Profiles
Forums
Events
Blogs
Posts posted by nachtfalke
-
-
waterchiller -it´s a PHASE CHANGE unit ! use a refrigerant compressor , condenser , expansion valve ( cap tube or TEV) , and vaporizator( platte HX - heat exchanger) so this is a EXTREME cooling 100000000%% , that water is cooled by refrigerant unit ..make temperature exchange refrigerant / chilledwater!!! it´s NOT dependent of ambiental temperature like "normal" water.
so my question here is: IF i use a LN2 pot, mount a waterblock on his BASE , and this water block is allready introduced in my waterloop ( don´t need anymore the radiator and fans
, just 2 or more waterblock´s), and then i cooled the pot with LN2/ DICE , that will make heat exchange POT vs waterblock , it is entuziast leaque aswell?? =))) because i use water...like people said above !! BUT this water is NOT anymore dependent of ambiental temperature!! like should be in enthuziast leaque!!
if i had a winter with -20-30 * and bench outside with my waterloop, use proper cooling liquid ( antifreeze) in this case is NORMAL for enthuziast leaque -because the water is cooled natural form nature and like i said above -cooling dependent of ambiental temperature!!
-
i shaw , anyway good job from far
-
good one!!! did you done OCP mod on RT pwm controller?
-
nice man , really good work!
-
i know , but improvment to 9.12whql is small actually i´m limited form cpu freq anyway..i did that 2k1 with 11.11whql and i not changed it anymore:) i will return with cpu on cascade:Dand ofcourse with proper driver!
thanks!
-
-
-
i told you , Qimonda no scalling with volts ..not only on RV670 , almost all cards that it had( nvidia aswell)..only isolated case , so i´m in same problem like you with my HD3870TOP , so i will give a try with another card , this time with reference ofcourse:D
-
i had same problem with Asus HD3870 TOP , i have no PLL bios ..and card are limited at 1100mhz and ram 1250mhz, even this vard has Samsung 0.8ns ram ic´s ...i´m dissipointed
-
mine also Qimonda,( the old infineon) don´t required Vmod ...because no scalling with volts , need reverse vmod:D
-
:)these cards are really hard to bench ,if you get a card with good ram´s/mem controller then you will have good efficiency! will see that:)
actually cpu freq limitted some my submision like 2k6 , 2k5 and aqua ..but i will give a try with cpu under my cascade , actually work in progress!!!
-
with 8600GT i had issue at lower ambient ( -5*C -last winter) , so i think is the same problem depends from gpu to gpu, prehaps earlyer issue on mine:D
actually i moded 3x 8800GT ( G92 chips 65nm) card all insulated for sub0 benching and i give a try under my rotary phase change cooling( -65* ) 2 of them has cold slow ...i have poor performance in specialy heavy load ( line nature 2k1 , 2k3 etc..), and the last ,work very well on gpu 1080mhz , but if i rise the mem( hynix FP1) clock even 5mhz from stock freq sistem freeze instantly ...so the mem controller inside the GPU has cold issue or at these temps(-60) , i modded the bios directly at 1050mhz( because on air cooling this card can do 1250mhz with 2,25v) on ram and updated the bios card but no luck
the sistem frezze even in bios ...so i have no solution for this issue ..
from 4X 9600GT only one like -55* without shutdown the gpu power ( specify for CB issue)with vrm onboard..and 2 of them has same CS issue like your!
try to put a thermal pad or alu/cooper plated( 1mm-2mm -3mm tall) gpu die/ evap base...only way is to rise the gpu temp ...if you want necessarily to bench that card:D
-
ok ,then for sure you heard of "cold slow"
some chip have this issue like your card.
for CB issue on these chips can only solve it with a external gpu power ( like epower for example), but for CS issue we need bios support and we haven´t yet
-
the gpu has cold slow bug at that temp ,normal for G94 and G92 chip´s:D
-
-58 -60 on heavy load , 28ccm rotary compressor:D
-
-
good job man , really nice work:D
-
thanks, my actually backup is at 6 point´s difference to first score:( no WR today
GPU thermal protection send a signal to pwm when gpu temp are ~-50*C( depens from gpu to gpu) and shutdown the card, so i think only way to solve CB at those cards( G92 , G94) need to be zombified , because CB shutdown the gpu power
-
nice clock for water, keep pushing:)
have you also problem with OCP trigger over 1.33v?
-
i change the cpu SS ( that -68* ) now cpu ran 5740mhz 2k1 stable but card is dead:(
-
thanks guys:) , when i finish that cascade i can bench o cpu ~ 6000mhz -6050:D with this exemplar, so 2k1 on old cards will be very funny:D
-
no, i did that SS
, and this chiller ( also is a dual rotary unit), now i have a cascade work in progress ...will be very very strong:D
here the build: http://forum.lab501.ro/showthread.php?p=385152#post385152
-
depends lot of gpu mem controller , this card is imposible to run on my SS because it has CB at -40:( , my ss can do -65*
-
thanks budy:D i do my best!
card ran 900mhz but think i need change the chocke inductors with another 6 ( 3+3) more powerfull, because the org is weak, i was limited to 1.26 -1,28v , 1.3v random freze in N.
Andrix85 - Radeon HD 4870 @ 905/1110MHz - 134662 marks 3DMark2001 SE
in Result Discussions
Posted
that phase change from dimastech are very weak !!!