8800GT voltage mods

[Demo]

Demo,
Yes. Need something to mark them on, like say, a pic of the card for example.

I'll take pics. gimme a few mins.



Lemme know if you need a closer view of a specific area, though my camera is a stone age point and shoot DSC-P200 sony.

[largon]

Demo,
Red probe to one of the marked points, black to any ground:

[Demo]

Im getting 1.11 on idle, and 1.15 on loaded that means i already have the 1.1v in my "Extra" settings right ?

[largon]

Appears so.

[Demo]

My card overclocks at an ultra crappy rate, i get a pink screen at 675core/1687shader mem wont move past stock ( 900 )

altho it doesnt lock up but shows a pink screen overlay, what could be causing this ?

[jason4207]

Try the BIOS mod anyway. Perhaps you are getting a bad read w/ your DMM. I get readings of up to a 0.05v difference depending on where I put my ground probe.

[I34z1k]

Anyone found a hardmod for the Gigabyte non-ref card?

[Demo]

Try the BIOS mod anyway. Perhaps you are getting a bad read w/ your DMM. I get readings of up to a 0.05v difference depending on where I put my ground probe.

But seeing as how it overclocks now at a very low level, its not worth the hassle.

[Renegade5399]

OK, so I finally bit the bullet and performed the vGPU mod. My RAM already does 1000MHz so the vRAM mods were not needed. I still need to do the OCP mod. With it as it is right now I can do 864/2000/2000 and all benches will complete. Without the OCP mod, you really have to "sweet spot" the voltage setting so not to trip the OCP while keeping stability. AM3 and 3dMark01 (yes 01) are helpful for this. Once 3dm01 hits the multi texturing test, voltage will spike due to the load. So right now I am running 1.295v idle and it goes up to 1.337v - 1.355v under load. 1.4v on the meter will trip the OCP and I get the black screen and 0 volts on the meter. I have to say though that while I am seeing better bench scores, the scores didn't scale like I was expecting. Still, this thing performs AWESOME! Thanks for all the tips. Now it's time to sell this and get a GTS, LOL!

[nvrmind]

If default vmem voltage is 2.0v, and you use a 20K resistor for the vmem mod what is the lowest and maximum voltage you will be able to achieve?

[r4st4m4n]

Seeing all the nice number people were getting, I finally decided to jump into the train. I bought a solder iron, DMM, all the hardware needed for the Vmod and a HR 03-GT to replace the stock cooler.

Everything wasn't smooth but at the end I finally managed to get it working.

Here is my max AT stable OC before the volt mod: 712/1728 @ 1.1V load temp 78*C


Here is my current max AT stable OC after the volt mod and the installation of the HR 03-GT 783/1890 @ 1.2V load temp 51*C


Picture of the moded card: (not very pretty but it does the job :p)


Turns out pretty good for a 1st volt mod and considering that I had never soldered anything before.

A big thanks to the extremesystems forum members for all the useful information that gave me the guts to do it and particularly to Largon for the volt mod and to Thrilla to post this awesome soldering guide: http://www.xtremesystems.org/forums/...d.php?t=115387

[Renegade5399]

LOL wow, I mounted my VR on the same corner! And the Molex connector is a MUST. Good job on your first volt mod!

[r4st4m4n]

LOL wow, I mounted my VR on the same corner! And the Molex connector is a MUST. Good job on your first volt mod!

Thanks

I actually took both ideas (molex and VR placement) from pictures posted on this forum (maybe yours ).
This is the advantage to have many skilled people here: it avoids beginners to make too much mistake and allows to know great tips

I've continued my testing and succeeds to run 10 loops of Crysis GPU benchmark and 2 loops of 3dmark06 at 799/1998/975 @ 1.2V idle.
This card definitely needed better cooling and more volts to shine
After finding the max OC at 1.2V idle, I'll try up to 1.3V

[Mr Alfabet]

Pencil voltmod 8800gt Xpert Vision(Palit)(pink)
907/2322/2246 with 1,47v core
24/7 with 1,34v core
My is Xpertvision 8800GT 512MB DDR3 PCI-E
Non-Reference | Non-Sonic model (That's the card )

does that vGpu mod also work for the Gigabyte Turboforce cards? the ones that go to 1.2V with software...

[Kabauterman]

Hey Guys,

I have some havy problemes with the Mod on the First Page.
I have a Point of View 88GT with Reference design Board.

I solder a 500kOhm Resistor to the left Alternativ Point (I was much easyer than the Point in the Middle.

The Mods Works good ... so long...
I can tune up and down the Voltage
and I can raise the Clock BUT
If I get more than 1,3 V the Card freezes i everey Benchmark.

Under 1.3 V all is ok and I can Bench but the advantage with the Clock is not good 770 without mod 780 with 1,3 V more is not possible.

Can someone Help me?
Is it possible that the Mod works better on the middle Soldering Point?

[quake6]

natedog420,
I got the buzzing too, it's quite common in 8800-serie cards it seems.
btw, I would advise against removing the resistors as this will completely prevent the phase controller from maintaining balance between the two phases = not good, at all. This type of OCP mod may be dangerous...
The safe(r) way is to do it like described in G80GTS OCP -thread:

I was wondering can I use 1.2k fixed resistor something like this?

http://www.queenmao.com.tw/image_pro...007-Carbon.jpg

[largon]

I solder a 500kOhm Resistor to the left Alternativ Point (I was much easyer than the Point in the Middle.

The Mods Works good ... so long...
I can tune up and down the Voltage
and I can raise the Clock BUT
If I get more than 1,3 V the Card freezes i everey Benchmark.
(...)

500kohm? Is that a typing error?
The VR one is supposed to be 500ohm, not 500kohm (kilo-ohm).

Is it possible that the Mod works better on the middle Soldering Point?

All marked mod soldering points are equal.

I was wondering can I use 1.2k fixed resistor something like this?

http://www.queenmao.com.tw/image_pro...007-Carbon.jpg

Yes.

[Kabauterman]

500kohm? Is that a typing error?
The VR one is supposed to be 500ohm, not 500kohm (kilo-ohm). All marked mod soldering points are equal.
Yes.

Ohh ^^ yes I mean 500 ohm ...

I think I have find the Problem

I must do the OCP mod :-(

[EightyDuce]

Ohh ^^ yes I mean 500 ohm ...

I think I have find the Problem

I must do the OCP mod :-(

Yeah thats the OCP kicking in, drop some 3.92k in parallel on the caps next to the X78s and you'll be set. There should be some pictures to help you in this thread.

[r4st4m4n]

Yesterday, I wanted to see how high I could go on the voltage. I went all the way up to 1.35V idle, giving me slightly over 1.4V under Ati tool load, the clock being 848/2106 and the OCP still didn't kick in

Is it possible the OCP is not working on my card ?

What is a 24/7 safe voltage knowing that my card stay cool, always under 55 under load ?

[scottc19]

Yesterday, I wanted to see how high I could go on the voltage. I went all the way up to 1.35V idle, giving me slightly over 1.4V under Ati tool load, the clock being 848/2106 and the OCP still didn't kick in

Is it possible the OCP is not working on my card ?

What is a 24/7 safe voltage knowing that my card stay cool, always under 55 under load ?

I don't think we know for sure what the max is... a few people have killed there cards, but this may be due to some other reasons and the fur benchmark...
I have mine at 1.5v load and am doing fine... but take your risks as I am also on water... ~41C or less on load... Also with these high volts the mosofets get really hot, you have to have heatsinks on all of them and a fan!

[r4st4m4n]

I don't think we know for sure what the max is... a few people have killed there cards, but this may be due to some other reasons and the fur benchmark...
I have mine at 1.5v load and am doing fine... but take your risks as I am also on water... ~41C or less on load... Also with these high volts the mosofets get really hot, you have to have heatsinks on all of them and a fan!

Yes, I have the PWM heatsink budled with HR-03 GT on them.
I have the fan of the HR 03-GT blowing upwards to the ram sink / mofsets sinks, an intake fan in the front of the case, another intake fan on the side of the case at the level of the card centered on the GPU and a 3rd 80 mm exhaust fan zip tied on the empty PCI brackets under the one used by the card. Not really quiet but I believe it should give plenty of airflow to the mosfets.

I probably won't go as high as you. I think 1.35~1.4V is the maximum I'm willing to go. If you run 1.5V 24/7 on water, I guess 1.4V should be ok for air.

[Renegade5399]

It seems some cards scale better than others. I still have to do the OCP mod, but had to wait for my resistors to arrive. That will come shortly. For now I have to work within the limits of the OCP. I am able to set 1.3v idle and under load it goes up to about 1.337-1.35v. When running AM3 or 3dm01, anything over 1.386v will trip the OCP. This is not too bad for me however as I am able to run 864/2106/1062 bench AND CS:S/TF2/COD4/Crysis stable. The card will cut out after about 2 hours of Crysis though due to the OCP. Something must get warm and cause current draw to go up. No biggie though as I will fix that soon.

So here's a shot of my fastest ever Lightsmark 2007 run:

[jason4207]

Yesterday, I wanted to see how high I could go on the voltage. I went all the way up to 1.35V idle, giving me slightly over 1.4V under Ati tool load, the clock being 848/2106 and the OCP still didn't kick in

Is it possible the OCP is not working on my card ?

What is a 24/7 safe voltage knowing that my card stay cool, always under 55 under load ?

Have you run AquaMark3?

[r4st4m4n]

Have you run AquaMark3?

I just downloaded it yesterday and tested the spike created when idle Vcore is 1.2V. But it seems Crysis demo make the voltage going higher than Aquamark 3
I've also tested 3dmark01 and the test 8 or 9 (can't remember exactly the name) is what makes the voltage going higher by far. While any other test including Crysis bench only make the voltage go up to +0.055, this test make the voltage jumps up to +0.08

I'll try tonight to set idle Vcore at 1.35V and run the 3dmark01 test, if it doesn't make the OCP kick in, I think nothing else will and I'm not willing to go higher on voltage anyways.

Yesterday I was a little bored to run benchmarks, so I tested the stability by playing Crysis. So far it seems 784/1944 is stable at 1.2V. At least it is after 1+ hour of Crysis.