Nv 7800 GS AGP Volt Mods

[Kin Hell]

Anyone?

[persivore]

Can you post a picture of the back of the card please.

[Kin Hell]

Can you post a picture of the back of the card please.

No web space.
See PM m8.
Thanks

[alexio]

No web space.
See PM m8.
Thanks

Use www.rapidshare.de in the future

[$oul]

there are some pictures in this thread (scroll down):
http://www.xtremesystems.org/forums/...ghlight=7800gs

I can lookup the names of the ic chips if that's needed.

[Maviryk]

Hi, I am very interested in volt modding the GS also. I have the GS on water, I know the chip has a lot of potential for more clockspeed, but it's starving for voltage. Please help if anyone can. I am currently stuck at 500/1600. Load @ 42C, Idle @ 37C.

[persivore]

I've uploaded pics of the card from Kin Hell here and here

Can someone tell me whats written on the chip I've circled in the attachment please.

[Kin Hell]

Okay, as in Pic going from top to bottom:
--------
ISL6568
CRZ
N535SL
--------

---------
6549CBZ
P548CMM
---------

-----------
5BHF31K G4 (the G4 bit is Underscored & Uppermost on Text line)
ACT08
-----------

---------
6549CBZ
P548CMM
---------

[persivore]

Can you check the voltages at the 2 points I've marked in green please.

[mnewxcv]

no offense but those cards are ugly!

[DTU_XaVier]

#10 I think it's called "getting used to"
Most of us are just used to seeing the capacitors and all the other components infront of the HS...

Best Regards

[Maviryk]

The top one reads 1.83v, the bottom one reads 1.41v.

*holy crap it's unnerving taking readings off of those chips....*

Edit: Actually, let me get a different DMM, I don't trust this one, just tested it out on some known voltages and it's way off.

[Kin Hell]

My MM is real prehistoric. Its got 1 of those Needle pointer thingys. At about 25 years of age, the old MM really isnt suitable for accurate readings needed here. I just use it to check for resistance. So I cant oblige. Also:
Since posing this Thread, I have had a bit of a situation develop.
I got power issues atm on me Rig...maybe these 5 hard drives beat the hell outta my Antec True Control 550. Summat was affecting my 7800 GS so bad that CS Source Bench dropped from 162 fps to 23. I was franticly swapping Molex plugs and removing hardware to try & free up things. Eventually, the AGP card started buzzing a rather pathetic beeping sound when ever I ran ANYTHING 3D. Stop the App, the buzzer stopped, start the 3D App, Buzzer goes off!?!?!? Did this for about 4 or 5 Graphic bench starts and then stopped doing it on the 6th run. At this point, I shut down the PC, Pulled 7800 GS, replugged several times, booted, benched & Im now 1200 3DMark03 points down from my normal score of 15,200 @ 492/1430.
Since starting to resolve this 3 days ago, I gave up re-installing Micro$0ft Winblows ExPee Pro to resolve the issue at attempt number 3.
Something on my PC has or is dying since I first fitted the card & I'm not so sure if its the PSU, the Mobo or the GS. At one point, the Auto detect Speeds returned 650 CORE & 1500 MEM. Now where have I seen this before Nvidia?
Does the 7800 GS OC use more power than an ATI 9800XT? Cos thats what came out & everything was fine there! Cant retry the 9800XT as its not available.
EDIT*
When I run this Bench:
http://www.daionet.gr.jp/~masa/rthdribl/
The Fps starts at high 60's & then hits 118 a second or so after the GPU Fan has increased its speed.
Gotta set default clks here & re-Bench, I have reference benchies to confirm.
Gah!

[Maviryk]

Antec True 550 is a Dual Rail PSU, with a 12v1 and a 12v2 if I recall correctly. The first rail goes directly into your motherboard and the other rail goes to the rest of your components, which is a very bad case for you. Antecs are good for the average joes, you should upgrade.

I would recommend getting a new power supply, since you are overloading it. The buzzer you hear is you 7800GS telling you it's not getting enough power from the molex. Try booting up without anything plugged in and you will see what I'm talking about. The GS definetly uses more power than an ATI 9800 (If I recall, that's about the equivelant of the 5900/6600 nvidia cards)

I would recommend you upgrade your PSU to either

Silverstone Zeus SST-ST56ZF
http://www.newegg.com/Product/Produc...82E16817163111

OCZ Powerstream 520
http://www.newegg.com/Product/Produc...82E16817711002

Or the cream of the crop PC P&C Turbo Cool 510 Deluxe
http://www.pcpowercooling.com/produc...php?show=T51DX

I would trust any of these PSUs to run my rig, which includes watercooling, overclocked prescott/ram, six hard drives, and 6800ultra.

[Kin Hell]

http://www.antec.com/specs/TPII550_spe.html

Mine isn't version 2 of this PSU, But this SPec sheet is for Version 2 of my PSU & is SLI ready quoting 19Amps Max load on the 12V rail?

bbl8r

[Kin Hell]

K!......it apears I have some fluctuation on the 12V or 5V from what Ive seen & heard here, yet Asus Probe constantly shows the following with Prime95 Running Torture mode.

12.281/12.342
5.134/5.160
3.296/3.328 - this one never use to drop below 3.312 under load - prolly causing the CPU back off I guess!?

Over the last month or so, I have had to reduce my overclock slightly from 250fsb to 244fsb. It plays 24/7 at 244FSB & 3DMark 06 cr^ps out at the CPU test. I objected to de-clocking further to get that prog to run CPU tests & if I rememeber rightly, the 9800XT bombed out on the CPU test in 06 @ 244.
So if the 7800GS needs more current, its making my rig crash randomly for the extra current draw over the 9800XT....maybe!
I dont feel at all embarresed about this cos Asus Probe is fine with Prime95 Torture Test and rthdribl demo in a 640 x 480 window makes absolutely NO differences to the Voltages in Asus Probe?
Motherboard temp is UP though, by about 6/7 degrees....at 30
Surely I'd be seeing some sub - acceptable levels in Asus Probe if I had voltage issues?
Will do some more stressing at various FSBs to see if there are any differences. I got a feeling as soon as I up the VCore, I'm going to see AGP issues again & maybe I should be running some Pixel 3 Shader stuff to really stress the card.

This is all a bit confusing but it looks like the PSU is gunno get what for!

*edit* I suppose the PSU getting it is better than the Kitten getting it eh!

[Maviryk]

It doesn't matter what the ratings say on the 12v1 or 12v2 rails, because as more current is drawn in one rail, there will be less available on the other rail.

Do NOT trust your onboard monitor. Using a reliable DMM is the only accurate way. A reliable PSU will get you cleaner power, less fluctuaction, and therefore higher Overclocks. I would take a single 38A PSU over a combined (12v1 + 12v2) 38A anyday, and you would be wise to do the same.

For more information on PSUs read this article
http://www.silentpcreview.com/article28-page1.html

What does all this mean? Essentially, the only potential benefit of dual 12V lines is improved safety, and this is questioned by the engineers I spoke with. There are many downsides to dual 12V lines, including higher cost and the extra headache of ensuring adequate 12V current for all the components in complex, high power systems. For the consumer who is trying to make a choice among the myriad of PSUs available on the retail market today, the most practical approach regarding dual 12V lines and power capacity is to consider only the combined 12V current capacity. (It will certainly be interesting to see how Intel handles this issue in the next version of ATX12V.)

[$oul]

The top one reads 1.83v, the bottom one reads 1.41v.

*holy crap it's unnerving taking readings off of those chips....*

Edit: Actually, let me get a different DMM, I don't trust this one, just tested it out on some known voltages and it's way off.

I geuss the top one is memory voltage (since GDDR3 runs at 1.8v-2v i think)

anyone tried this already ?:
http://www.vr-zone.com/?i=3271&s=8

[Maviryk]

... How the hell do they come up with this stuff this fast... Well, I'm at work right now, will let ya'll know in a later tonight if I have a dead card with a voided warranty.

[Maviryk]

Persivore, I borrowed a calibrated DMM from the lab, I hope you see this before I go back to work tomorrow. Top reads 5.220v, bottom reads 1.437v.

[Kin Hell]

Persivore, I borrowed a calibrated DMM from the lab, I hope you see this before I go back to work tomorrow. Top reads 5.220v, bottom reads 1.437v.

Thanks for your support RE my PSU:thumbsup:
I noticed whilst running Prime95, my PSU was making like an Electronic squeaking sound. Wondered if it was a fan, but stopping the PSU's fans momentarily didnt stop the digital-squeak!
Dropped the fsb to 220 and VCore to Stock, then ran Prime95 squeak free. At this poinit, I heard another squeaking sound & then realised it was my wallet groaning again!
Another thing happened 7 weeks ago,.... a failing ATA133 Drive on my raid Array...it's all adding up now.
Moving on.....

...I'm guessing here, the bottom reading of 1.437v is controlled by the Cards bios!?
Alexio at this thread: http://www.xtremesystems.org/forums/...ghlight=7800gs
has a modded bios that appears to work, changing the cards 3D voltage from 1.4v to 1.5v

[Maviryk]

Well, I'm at work now, so no more taking accurate readings unless I can borrow the DMM again. Zombie has done the Vcore voltage limitor mod, I'm waiting for him to post on its results, and for Persivore to give some more feedback before I do any more mods to the card. According to Zombie the BIOS volt mod does not work. The 7800GTs only have voltage hard mods, I would think the GS would be the same.

I'm gonna wait for all the info to come in before I put the mods on my card, due to the fact that it has a water loop attached to it.

[persivore]

Can you check the resistance between the pin I've marked in red and ground please. Check resistance with the power off.

[$oul]

Can you check the resistance between the pin I've marked in red and ground please. Check resistance with the power off.

I got 0.865 KOhm on my DMM, so 865 Ohm

[$oul]

... How the hell do they come up with this stuff this fast... Well, I'm at work right now, will let ya'll know in a later tonight if I have a dead card with a voided warranty.

by reading pdf's :p

here is the one for that chip:
http://www.intersil.com/data/fn/FN9187.pdf

they remove the resistors from the ISEN1 and ISEN2 pins.