http://i4.photobucket.com/albums/y14...porX20mods.jpg
With a FB -> GND resitance of 4.62K Ohms I get 1.343v core voltage load.
Printable View
http://i4.photobucket.com/albums/y14...porX20mods.jpg
With a FB -> GND resitance of 4.62K Ohms I get 1.343v core voltage load.
It works on my HD5850 XFX model 585X-ZAFC...??? she use the same HD5870 Vapor-x PCB...Quote:
Originally Posted by Meaker
Here a Pict
[img]http://img687.imageshack.us/i/1pcp.jpg/[/img]
[img]http://img688.imageshack.us/i/4voltagechip.jpg/[/img]
Conductive paint works? Or it raise voltage too much???
If its the same PCB it should, not sure about the conductive ink, would be great if it did work. Might be too conductive hwoever.
Thats pic got a poor quality. Is that the "NCP5395T" voltage regulator? Does anyone know if it's software controllable too?
Hello,
This is my first voltmod, I would not make mistakess.
This connection is correct? to increase VGPU to about 1.35v ?
http://img443.imageshack.us/img443/4129/vmod.png
Thank you for your help and sorry for my English :rolleyes:
I found out that the NCP5395T voltage regulator, that is on the Sapphire Rev. 2 Card (non-vaporx and non-toxic) is used on mobo's to regulate cpu voltage too. Couldn't it be possible that this regulator can be controlled via software?? Does anyone know a bit more about that?
Well I dont know if they changed it but i've been up to 1110mhz core on this mod.
Walbed it needs to be around 50k ohms as far as I can tell.
46.5K ohms should do it.
This is because there is already around 5.5k ohms to ground, this mod is adding another in parallel.
thank you for your response, I start!
Thanks for this.
vGPU Confirmed working on my Sapphire 5830. I hardmodded using 50k resistor giving 1250mW at full resistance.
Can reach 1100 Mhz on core, but it locks at 1150Mhz on core even with 1350mW.
Only have the vGPU mod done - could it be OCP kicking in?
The core is under water at 38'C full load at 1250 mW and 1100 Mhz.
This bodes well for software voltage adjustments when RBE supports 5830s properly :up:
This might sound silly, but I can't get any normal programs to work for general OCing on these boards. I'm always stuck at the CCC max of 900 in every program. What does everyone use?
The simplest option is to use MSI Afterburner and edit the config file to allow unofficial overclocking.
http://www.guru3d.com/article/radeon-5830-review/26
However, even with that I ran into a limit of 1040 Mhz on the core (and 1300Mhz on the memory, but I'm less concerned with the memory limit).
I ended up using RBE to enable Rivatuner support, then using Rivatuner and changing the MaxClockLimit under Power User, RivaTuner\Overclocking\Global to 100.
Now I feel I am either at the card limit or OCP @ 1150Mhz.
EDIT - Having read the 5850/5870 thread it seems OCP kicks in somewhere after 1.35v, so I guess I'm at the limit with ~1150Mhz, which seems strange as it clocks to 1000Mhz on default cooling and voltage.
EDIT - I'm now using the latest ATI Tray tools Beta for overclocking as this means I simply use one program for both overclocking and graphics settings (i.e. I only install the Display Driver from ATI).
I sent a email to BAGZZlash. If voltage control is possible hes the guy who can make it happen :cool:
Just a FYI the Gigabyte HD5830 BIOS pumps 1.15v to the GPU instead of the 1.068v Sapphire feed the 5830. From what I can tell (theres not many good Gigabyte 5830 pictures around) The PCB is identical.
ED - I flashed the Gigabyte BIOS to the Sapphire HD5830 I have, theres deffinately a improvement. I found 24/7 stable GPU clocks with the Sapphire BIOS to be 935MHz, with the Gigabyte BIOS I'm currently testing (very successfully I might add) 970MHz in Unigine Heaven. Gigabyte vBIOS FTW on Sapphire cards :up:
Just to let you know that my Sapphire 5830 was running 1.15 V since day one. That's measured under load with a multimeter though.
Maybe theres different Sapphire vBIOSes with different voltages set? Could you open your Sapphire vBIOS with RBE and confirm if 1.068v or 1.15v is set? Thanks :up:
All 3D Voltages are set to 1.063 V in this BIOS.
Maybe it's just a discrepancy between the software voltage setting and the measured voltage?
Thank you Ket for drawing my attention to this interesting thread. :up:
Unfortunately I'm currently stuck on 5830 vMod research for RBE. Those BIOSes don't contain the voltage registers I'm used to use for modding (since the cards aren't equipped with a voltage controller I know how to program such as VT1165MF) or they don't kick in.
Sorry for being unable to announce better news... :shrug:
Maybe, or your card might overvolt a bit, mine doesn't which is why I only got to 935MHz before flashing to the Gigabyte vBIOS. The Gigabyte vBIOS actually lets me go all the way to 1000MHz on the GPU, way over what the stock vBIOS could do, but it just needs a bit more voltage to get it stable.Quote:
Originally Posted by Jimbo Mahoney
I don't know how much merit there is in it but from research I've done it would seem there is a software controllable voltage controller on some cards (5830 I have has one) but only one voltage is defined in the vBIOS and not several. It would seem you could replace the voltage registers with new values, theres just no table. This would explain the difference of 1.068v and 1.15v in the Sapphire and Gigabyte vBIOSes where the Gigabyte vBIOS deffinately gave the card I have a boost, letting me grab another 35MHz from the core. I'll try just entering 1.3v and reflashing the vBIOS and let you all know how it works out, but I don't think it'll work with the current way RBE handles 5830 vBIOSes :(Quote:
Originally Posted by BAGZZlash
That's not too unlikely, but still I'd have to find out where the value(s) is/are located and - even more importantly - how it/they is/are encoded. Usually there's a formula like the one which is well known for the VT1165. As long as I don't know that formula...Quote:
Originally Posted by Ket
Anyone know about doing this to the ASUS EAH5830?
May be ending up with one of these in my hands, want to know if the boards are similar or if I'm going to have to dig around myself.
I'll post up details of the chip on my card when I've checked it again :up: I've also done a bit of experimenting, multiple voltages can be set depending on the card power state so a funky formula seems like the likely solution.Quote:
Originally Posted by BAGZZlash
The PCB on the Asus isn't quite reference it uses a 6 and 8 pin PCI-E connections instead of 2x6 PCI-E connections, but you might be ok using the Gigabyte vBIOS. Just make a backup of the vBIOS on the Asus in case you need to blind flash back.Quote:
Originally Posted by Dolk
I wish you good fortunes on fully identifying this voltage regulator BAGZZlash, a couple of the markings have been smudged at the factory it seems :down:
ED - I'm pretty sure I figured the chip markings out they are; P6213AJ VCF99X :up: Google doesn't come up with any results though. Must be a very new chip or something. I'd dare say its the very same chip as whats on pedrovicz XFX card, the only difference is his chip marks are P6213AJ VCF99XP