- Joined
- Jun 3, 2008
- Messages
- 385 (0.07/day)
- Location
- Pacific Coast
| System Name | Z77 Rev. 1 |
|---|---|
| Processor | Intel Core i7 3770K |
| Motherboard | ASRock Z77 Extreme4 |
| Cooling | Water Cooling |
| Memory | 2x G.Skill F3-2400C10D-16GTX |
| Video Card(s) | EVGA GTX 1080 |
| Storage | Samsung 850 Pro |
| Display(s) | Samsung 28" UE590 UHD |
| Case | Silverstone TJ07 |
| Audio Device(s) | Onboard |
| Power Supply | Seasonic PRIME 600W Titanium |
| Mouse | EVGA TORQ X10 |
| Keyboard | Leopold Tenkeyless |
| Software | Windows 10 Pro 64-bit |
| Benchmark Scores | 3DMark Time Spy: 7695 |
If you have a 1180 GTX 285 like me, here is a vmod guide. I didn't see one already posted on this site; but I didn't look too hard either. 
First of all, I have to say that the 1180 is so much worse than the earlier 1280 models.
I think NVIDIA went to the 1180 when they introduced the 2GB version since the 1180 could support either 1 or 2GB. The earlier 1280 model series is much easier to volt mod; you can find volt mod guides for them anywhere. If you can find an earlier 1280 series card, get that instead.
A volt mod guide for the 1180 is much hard to find.
So, there goes:
The Theory:
In essence, what we are doing is adjusting the voltage to the Vsense (pin 19) of the ISL6327 chip. Under load, the card is designed to output approximately 1.15v to the GPU core at full load. Lower amounts of 3d load and 2d loads will cause the core voltage to automatically drop to a lower output since less is required. It is important to understand that the voltage to pin 19 will always be ~1.15v at full GPU load. In other words, even after the voltage mod, the voltage at pin 19 will still be around 1.15v at full load. What we are doing is tricking the ISL6327 controller; we cause the controller to sense a lower voltage than is actually present.
How do we trick the controller into providing more core voltage to the GPU core? We drop some of the voltage at Vsense to ground.
Let's look at a schematic. The following schematic was provided by an author at a Czech website.
http://phpbb3.ocmag.eu/viewtopic.php?f=155&t=147
Here is the schematic:
See what is going on? The voltage output from the inductors passes onto the GPU core which isn't pictured but would be shown off the bottom of the schematic. Some of that same output is also fed back to the Vsense pin (pin 19). It is important to know that this diagram is not correct; it is only a drawing of a typical application of the controller. It doesn't show the actual schematic of the circuit used for the 285.
The 285 actually has some component between the inductor output and the ISL6327 Vsense pin. I am not sure what component it is because I had a hard time tracing it. I spent over an hour trying but had to give up. The coating put on the card really covers up the traces well. Regardless, there is some resistance present which we will use to our advantage. We will call this resistance X.
Alright, in order to do this mod properly, we will want to do just as the Czech author stated. We would physically cut pin 19 from the PCB in order to isolate it from the circuit. We would then route pin 19 through a potentiometer and then back to its original source. What this does is allows you to bias the voltage to pin 19. Vsense would be on the middle pin of the potentiometer and ground would be on one end and original input to vsense would be on the other end of the potentiometer. This is beneficial because it will cause the voltage drop over resistance X to stay constant. We would just be picking the voltage potential to pin 19; choosing between inductor output or ground or anywhere in between.
The problem with this is that pin 19 is VERY tiny and cutting it is very dangerous without screwing up and cutting something you shouldn't. I elected to leave pin 19 intact. Instead, I soldered onto a point at the same potential and put a potentiometer between it and ground. I chose not to solder directly to pin 19 due to the size and number of items in close proximity to it. The output of pin 19 extends out a little ways and then passes through the board to the other side of the PCB. I soldered a wire onto a capacitor on the back side of the board in connection with pin 19. This capacitor is in series with the ISL6327 controller; I soldered onto the side with the same potential as pin 19.
So, we have voltage output from the inductors which then passes onto the GPU core. We are trying to increase that voltage. In order to do so, we need to trick the ISL6327 controller into thinking that the voltage output is lower than it actually is. As I said, it would be best to bias the voltage mod accross a potentiometer such as the Czech author instructed. But technical issues make that difficult. As the resistance between pin 19 to ground accross the potentiometer is lowered, voltage will drop in that portion of the circuit. This voltage drop is absorbed accross resistance X. If we want the output to the GPU to be 1.2v, we must drop the voltage to pin 19 across resistance X so that the ISL6327 controller will sense a shortage of voltage and will raise the voltage up until 1.15v is present at pin 19.
As I said earlier, it is important to know that voltage to pin 19 will always be 1.15v under full GPU load since the controller will boost voltage output until 1.15v is present at pin 19. We dropped the voltage at pin 19 by dropping it across resistance X to ground. The ISL6327 controller sensed this voltage drop and raised the voltage output until voltage at pin 19 equalled 1.15v. This effectively increased the actual core voltage since ISL6327 is now sensing a false voltage.
So, what potentiometer do you use? A 1K Ohm potentiometer with at least 10 turns would be great. That would provide you with really good sensetivity; resistance would only change by 100 Ohms for every full turn. Be careful while turning the potentiometer though. For the first 6-7 turns, the core voltage will barely change. When the core voltage starts to change, it will change slowly at first. The higher you increase the core voltage, the faster it will rise. The adjustment is not linear, it is exponential. So turn it slowly.
Pictures:
Enough text already! We want pictures!
The wire is soldered to a capacitor in series with pin 19 on the side of the capacitor which has the same potential as pin 19. The black circles with nothing going to them are alternate ground points.
As shown, I suggest adding a small resistance in series with the potentiometer to prevent you from accidentally lowering the resistance too much. I also suggest adding a resistance accross the potentiometer to improve potentiometer sensitivity. Theoretically, you can just use two wires and a potentiometer to do this mod, but I suggest that you add the resistors such as I have shown just to protect the core from accidentally increasing it too high and to increase potentiometer sensetivity.
In the picture, you see the 4.7 Ohm resistor in series with the potentiometer. The red dots on either end of the 4.7 Ohm resistor are to show that the wire is soldered to the resistor. DO NOT THINK that the resistor is supposed to be soldered accross a circuit of the PCB as it may appear in my crappy paint drawing.
Before soldering the potentiometer to the circuit, I would set it to around 20 Ohms. For 1.25v core voltage, the final resistance of the mod with be around ~10 Ohm, but set the potentiometer higher and then dial the core voltage up when the GPU is running.
GPU voltage measuring point:
The above picture isn't of the 285 and the voltage regulation circuitry is layed out a little bit different. Ignore the differences in circuit layout in the picture versus the circuit layout of your 1180 285. I chose the picture because it was a good quality shot although not quite correct. It should serve the purpose though.
The side of the inductor clostest to the GPU core is the side you want to get your voltage indication from.
Either one of the inductors will suffice as well as either of the grounds.
Final pictures:
Potentiometer mounted to heatsink:
Resistors attached in series and in parallel to the potentiometer. I used a 10 Ohm resistor in series with the potentiometer since that is the lowest I could find at RadioShack. I recommend a lower value though since 10 Ohm will limit your core voltage to about 1.25v. 6 Ohm will limit your core voltage to ~1.3v. etc.
Wire soldered to inductor for voltage measurement:
Wires soldered to ground:
In the video card BIOS, I DO NOT recommend that you change the number of active performance levels or the VID of 3d and 2d performance settings. Change the voltage with the hard mod only. Changing the VID of 2d and 3d performance levels causes stability issues. Only modify the BIOS in order to change the core, shader, and memory clock and to modify the fan profile.
The above image is just to show you that the video card does sense the hard vmod. As I said, do not change these values.
If you increase core voltage much higher than 1.3v with just this mod only, you will probably start running into issues. The screen will probably go blank black or display some other solid color and the GPU fan may go to 100%. This is an indication that you have exceeded the video card's built-in overcurrent protection (OCP). You have caused the current going through the voltage regulator to exceed the OCP limit and the card will shut the core down to prevent it from damage. If you want to increase the core voltage more, you will have to do the OCP mod in order to go higher.
First of all, I have to say that the 1180 is so much worse than the earlier 1280 models.
I think NVIDIA went to the 1180 when they introduced the 2GB version since the 1180 could support either 1 or 2GB. The earlier 1280 model series is much easier to volt mod; you can find volt mod guides for them anywhere. If you can find an earlier 1280 series card, get that instead.A volt mod guide for the 1180 is much hard to find.
So, there goes:
The Theory:
In essence, what we are doing is adjusting the voltage to the Vsense (pin 19) of the ISL6327 chip. Under load, the card is designed to output approximately 1.15v to the GPU core at full load. Lower amounts of 3d load and 2d loads will cause the core voltage to automatically drop to a lower output since less is required. It is important to understand that the voltage to pin 19 will always be ~1.15v at full GPU load. In other words, even after the voltage mod, the voltage at pin 19 will still be around 1.15v at full load. What we are doing is tricking the ISL6327 controller; we cause the controller to sense a lower voltage than is actually present.
How do we trick the controller into providing more core voltage to the GPU core? We drop some of the voltage at Vsense to ground.
Let's look at a schematic. The following schematic was provided by an author at a Czech website.
http://phpbb3.ocmag.eu/viewtopic.php?f=155&t=147
Here is the schematic:
See what is going on? The voltage output from the inductors passes onto the GPU core which isn't pictured but would be shown off the bottom of the schematic. Some of that same output is also fed back to the Vsense pin (pin 19). It is important to know that this diagram is not correct; it is only a drawing of a typical application of the controller. It doesn't show the actual schematic of the circuit used for the 285.
The 285 actually has some component between the inductor output and the ISL6327 Vsense pin. I am not sure what component it is because I had a hard time tracing it. I spent over an hour trying but had to give up. The coating put on the card really covers up the traces well. Regardless, there is some resistance present which we will use to our advantage. We will call this resistance X.
Alright, in order to do this mod properly, we will want to do just as the Czech author stated. We would physically cut pin 19 from the PCB in order to isolate it from the circuit. We would then route pin 19 through a potentiometer and then back to its original source. What this does is allows you to bias the voltage to pin 19. Vsense would be on the middle pin of the potentiometer and ground would be on one end and original input to vsense would be on the other end of the potentiometer. This is beneficial because it will cause the voltage drop over resistance X to stay constant. We would just be picking the voltage potential to pin 19; choosing between inductor output or ground or anywhere in between.
The problem with this is that pin 19 is VERY tiny and cutting it is very dangerous without screwing up and cutting something you shouldn't. I elected to leave pin 19 intact. Instead, I soldered onto a point at the same potential and put a potentiometer between it and ground. I chose not to solder directly to pin 19 due to the size and number of items in close proximity to it. The output of pin 19 extends out a little ways and then passes through the board to the other side of the PCB. I soldered a wire onto a capacitor on the back side of the board in connection with pin 19. This capacitor is in series with the ISL6327 controller; I soldered onto the side with the same potential as pin 19.
So, we have voltage output from the inductors which then passes onto the GPU core. We are trying to increase that voltage. In order to do so, we need to trick the ISL6327 controller into thinking that the voltage output is lower than it actually is. As I said, it would be best to bias the voltage mod accross a potentiometer such as the Czech author instructed. But technical issues make that difficult. As the resistance between pin 19 to ground accross the potentiometer is lowered, voltage will drop in that portion of the circuit. This voltage drop is absorbed accross resistance X. If we want the output to the GPU to be 1.2v, we must drop the voltage to pin 19 across resistance X so that the ISL6327 controller will sense a shortage of voltage and will raise the voltage up until 1.15v is present at pin 19.
As I said earlier, it is important to know that voltage to pin 19 will always be 1.15v under full GPU load since the controller will boost voltage output until 1.15v is present at pin 19. We dropped the voltage at pin 19 by dropping it across resistance X to ground. The ISL6327 controller sensed this voltage drop and raised the voltage output until voltage at pin 19 equalled 1.15v. This effectively increased the actual core voltage since ISL6327 is now sensing a false voltage.
So, what potentiometer do you use? A 1K Ohm potentiometer with at least 10 turns would be great. That would provide you with really good sensetivity; resistance would only change by 100 Ohms for every full turn. Be careful while turning the potentiometer though. For the first 6-7 turns, the core voltage will barely change. When the core voltage starts to change, it will change slowly at first. The higher you increase the core voltage, the faster it will rise. The adjustment is not linear, it is exponential. So turn it slowly.
Pictures:
Enough text already! We want pictures!
The wire is soldered to a capacitor in series with pin 19 on the side of the capacitor which has the same potential as pin 19. The black circles with nothing going to them are alternate ground points.
As shown, I suggest adding a small resistance in series with the potentiometer to prevent you from accidentally lowering the resistance too much. I also suggest adding a resistance accross the potentiometer to improve potentiometer sensitivity. Theoretically, you can just use two wires and a potentiometer to do this mod, but I suggest that you add the resistors such as I have shown just to protect the core from accidentally increasing it too high and to increase potentiometer sensetivity.
In the picture, you see the 4.7 Ohm resistor in series with the potentiometer. The red dots on either end of the 4.7 Ohm resistor are to show that the wire is soldered to the resistor. DO NOT THINK that the resistor is supposed to be soldered accross a circuit of the PCB as it may appear in my crappy paint drawing.
Before soldering the potentiometer to the circuit, I would set it to around 20 Ohms. For 1.25v core voltage, the final resistance of the mod with be around ~10 Ohm, but set the potentiometer higher and then dial the core voltage up when the GPU is running.
GPU voltage measuring point:
The above picture isn't of the 285 and the voltage regulation circuitry is layed out a little bit different. Ignore the differences in circuit layout in the picture versus the circuit layout of your 1180 285. I chose the picture because it was a good quality shot although not quite correct. It should serve the purpose though.
The side of the inductor clostest to the GPU core is the side you want to get your voltage indication from.
Either one of the inductors will suffice as well as either of the grounds.
Final pictures:
Potentiometer mounted to heatsink:
Resistors attached in series and in parallel to the potentiometer. I used a 10 Ohm resistor in series with the potentiometer since that is the lowest I could find at RadioShack. I recommend a lower value though since 10 Ohm will limit your core voltage to about 1.25v. 6 Ohm will limit your core voltage to ~1.3v. etc.
Wire soldered to inductor for voltage measurement:
Wires soldered to ground:
In the video card BIOS, I DO NOT recommend that you change the number of active performance levels or the VID of 3d and 2d performance settings. Change the voltage with the hard mod only. Changing the VID of 2d and 3d performance levels causes stability issues. Only modify the BIOS in order to change the core, shader, and memory clock and to modify the fan profile.
The above image is just to show you that the video card does sense the hard vmod. As I said, do not change these values.
If you increase core voltage much higher than 1.3v with just this mod only, you will probably start running into issues. The screen will probably go blank black or display some other solid color and the GPU fan may go to 100%. This is an indication that you have exceeded the video card's built-in overcurrent protection (OCP). You have caused the current going through the voltage regulator to exceed the OCP limit and the card will shut the core down to prevent it from damage. If you want to increase the core voltage more, you will have to do the OCP mod in order to go higher.
Last edited:
