- Thread Starter
- #81
slightly less, 1.14
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
GIGABYTE Radeon HD 7870 Windforce3 Voltage Mod?
- Thread starter N0ble
- Start date
slightly less, 1.14
- Thread Starter
- #83
Yes, if i do around 3 turns it will go down by 1mv.... it should be more than that?? yet i see the temp increase by 3c when going from 1144mv to 1148mv.. this is all done under load.
Using multimeter to check.
Just to add, got my multimeter set to DCV 2000m and the current readout under load is 1147.. unless im mad thats 1.147v??
Something not right here and i think its more along the lines of the voltage read out being wrong as the mod is deffo doing something as i can run the clock a lot higher....
Last edited:
- Joined
- May 10, 2009
You're correct on mV to V.
That sounds about right for the mod point. The far side of the resistor (681) should be core voltage and that should be going up. The mod point voltage should stay more or less the same.
Temps going up indicate that *something* is certainly happening.
What does GPUz say the core voltage is under load?
That sounds about right for the mod point. The far side of the resistor (681) should be core voltage and that should be going up. The mod point voltage should stay more or less the same.
Temps going up indicate that *something* is certainly happening.
What does GPUz say the core voltage is under load?
- Thread Starter
- #85
1.168.
As an example, i have just rigged up my stock card to test comparison.
You see the voltage increas on the multimeter inline with what voltage is displayed on MSI Afterburner.
So as an exmple in MSI afterburner its set to 1.250 and my multimeter reads 1.225. The card seems to be getting a lot hotter than the modded card by 10c and as soon as i try and if i set higher then 1260mhz it crashes the display driver.
This is really odd.... im not sure what is happening here.
- Joined
- Mar 27, 2011
- Location
- Soviet Mexico
I would guess that the .025mv difference is just a difference in calibration between the two, or even polling times. So not really a big deal. Low end multimeters are not known for their accuracy at low voltage ranges.
I'd probably be checking Temperatures at the core, temperatures at the vrms, and if that looked go i would give it more juice. 1.25volts isnt really that much that is actually the starting voltage required for alot the circuitry on the card.
I'd probably be checking Temperatures at the core, temperatures at the vrms, and if that looked go i would give it more juice. 1.25volts isnt really that much that is actually the starting voltage required for alot the circuitry on the card.
- Thread Starter
- #87
I'm worried that by doing the mod i have done... its some how messed up the readings on the card?
I just ran my stock card and the multimeter showed me 1280mv at full load.... but only at 1220mhz core... yet my modded card would show no higher than 1147mv yet i magaed to have that running at 1333mhz...
this is freaky... i will run checks on VRMs next to see what happens there..
- Thread Starter
- #89
think i will try that next.
- Thread Starter
- #92
http://www.3dmark.com/3dm11/4130448
http://www.techpowerup.com/gpuz/drms7/
Just got this. 1320mhz core and 1361mhz memory, if i go higher on memory it makes the bench worse... so hopefully a little more voltage will resolve that issue...
Voltage still reads funny on core though
http://www.techpowerup.com/gpuz/drms7/
Just got this. 1320mhz core and 1361mhz memory, if i go higher on memory it makes the bench worse... so hopefully a little more voltage will resolve that issue...
Voltage still reads funny on core though
- Thread Starter
- #95
well it would appear that way, but i ran with the above settings on 3dmarks and every time it bombed the graphics drivers when under load.
So it seems with the mod the card somehow works better yet the voltages are showing lower..... makes no sense..
- Joined
- Mar 27, 2011
- Location
- Soviet Mexico
is that F or C? if is 90c thats quite excessive, and they're gonna need some extra airflow, On alot of gigabytes cards I have found they tend to gimp on VRM cooling, and i have had it cause stability issues in the past. So i usually stick some extra heatsinks on their vrm's if they are accessible enough, or o go cut some some special.
- Joined
- May 10, 2009
I'm going to guess that it's an issue with the voltage read point. I would try soldering a new read point onto one of the large square pads under the VRM section. One side will be ground, the other should be vcore. There may be something else that also uses the current read point.
- Thread Starter
- #98
Orginal resistance form mod point was 3.2ohms.
With the trimpot attached and me finding a sweet spot, the final resistance was 2.4ohms.
When detaching the mod point and measuring the trimpot i got a reading of 12.8ohms, i did start with it at max which read 53 but moved it down until it actually made a difference in a benchmark.
With the trimpot attached and me finding a sweet spot, the final resistance was 2.4ohms.
When detaching the mod point and measuring the trimpot i got a reading of 12.8ohms, i did start with it at max which read 53 but moved it down until it actually made a difference in a benchmark.
- Thread Starter
- #100
