PS3 PS3 Phat Cell BE VDDC Voltage question

NightLess

NightLess

Member
Hi everyone, I have a question about the effect of BE VDDC voltage on Cell temperature. I have 4 PS3 40nm Frankies, all of which have been delided and re-added with MX4 thermal paste. Although the temperature performance of the other 3 after delid was very good, the fan speed was 29%-32% at 67°MAX for one hour of playing TLOU, but there is a special one, it behaves like a PS3 equipped with 90NM RSX, and the fan speed at 67°MAX reached 38% (even 40% at the beginning of starting the game)!

Although I re-applied thermal paste, confirmed the flatness of the heat sink and IHS surface, replaced 15-blade or 19-blade fans, and even replaced 3 heat sinks, I couldn't improve this situation. . .

For this situation, I want to know if the voltage has an effect on the temperature of the CELL? I used a multimeter to measure the voltage of CELL VDDC when it was powered on and it was a stable 1.155v. The service manual states that BE VDDC is 1.0v and the resistance of VDDC is 1.2Ohm (8x Felix Tantalizer installed). I got 1.8Ohm on a donor board. I wonder if there is something wrong with CELL?

Although the machine can operate normally, is this output voltage correct? Is the 1.2Ohm of Cell too low? Does it also affect the temperature? Thanks!
 
NightLess said:
Hi everyone, I have a question about the effect of BE VDDC voltage on Cell temperature. I have 4 PS3 40nm Frankies, all of which have been delided and re-added with MX4 thermal paste. Although the temperature performance of the other 3 after delid was very good, the fan speed was 29%-32% at 67°MAX for one hour of playing TLOU, but there is a special one, it behaves like a PS3 equipped with 90NM RSX, and the fan speed at 67°MAX reached 38% (even 40% at the beginning of starting the game)!

Although I re-applied thermal paste, confirmed the flatness of the heat sink and IHS surface, replaced 15-blade or 19-blade fans, and even replaced 3 heat sinks, I couldn't improve this situation. . .

For this situation, I want to know if the voltage has an effect on the temperature of the CELL? I used a multimeter to measure the voltage of CELL VDDC when it was powered on and it was a stable 1.155v. The service manual states that BE VDDC is 1.0v and the resistance of VDDC is 1.2Ohm (8x Felix Tantalizer installed). I got 1.8Ohm on a donor board. I wonder if there is something wrong with CELL?

Although the machine can operate normally, is this output voltage correct? Is the 1.2Ohm of Cell too low? Does it also affect the temperature? Thanks!
Click to expand...

As always for any processor, voltage does have an effect on temperatures.
The equation for the power draw of a processor can be roughly derived from the formula:
Pd=C*f*V^2
Where C is the capacitance of the silicon gate (intrinsic to the processor, can be considered constant for most applications)
F is frequency
V is voltage
This means that your power draw increases by the squared of the voltage.
Undervolts are playing with the silicon lottery, and as always it differs from unit to unit and tends to favour later chips as the manufacturing process matures.
 
villahed94 said:
As always for any processor, voltage does have an effect on temperatures.
The equation for the power draw of a processor can be roughly derived from the formula:
Pd=C*f*V^2
Where C is the capacitance of the silicon gate (intrinsic to the processor, can be considered constant for most applications)
F is frequency
V is voltage
This means that your power draw increases by the squared of the voltage.
Undervolts are playing with the silicon lottery, and as always it differs from unit to unit and tends to favour later chips as the manufacturing process matures.
Click to expand...
That's true. This situation is very common on computer CPUs. As the yield rate increases, the overclocking performance of later chips is often better. Maybe I should buy a socket power meter to see if their operating power consumption is roughly the same? Only this one is special, and I have no way to deal with it.. Thank you for your answer.
 
NightLess said:
That's true. This situation is very common on computer CPUs. As the yield rate increases, the overclocking performance of later chips is often better. Maybe I should buy a socket power meter to see if their operating power consumption is roughly the same? Only this one is special, and I have no way to deal with it.. Thank you for your answer.
Click to expand...
If you want to have an estimated reading yeah. It would help pinpoint your power delta against stock.
 
villahed94 said:
If you want to have an estimated reading yeah. It would help pinpoint your power delta against stock.
Click to expand...
OK, I will buy one. If you can successfully change the VID of the CELL, please make a tutorial. Yesterday I saw that Felix seemed to be working on this issue, but he has not made a tutorial yet due to the uncertainty.
 
I took power measurements on both machines and interestingly they are a bit different?

CECHB00 40nm Frankenstein
XMB:~150w MUSIC:~158w TLOU:~177w PS2 Kingdom Hearts:~176w

CECHA+B 40nm Frankenstein
XMB:~134w MUSIC:~140w TLOU:~160w PS2 Kingdom Hearts:~159w

I only conducted a brief test. They were in the same environment. Although I didn't have a thermometer, I turned on the air conditioner. You can see that my A+B (A motherboard, B case) has a relatively low power consumption, 67 degrees @ 30% (19-blade fan), while B00 is 69 degrees @ 32% (15-blade fan). If it is at 67 degrees, it will be at 34% or even higher, which is relatively noisy, so I chose to increase it by two degrees.

I have also seen other users' silicon lottery, even though it is equipped with 90nm rsx, the temperature is incredibly low.
 
NightLess said:
OK, I will buy one. If you can successfully change the VID of the CELL, please make a tutorial. Yesterday I saw that Felix seemed to be working on this issue, but he has not made a tutorial yet due to the uncertainty.
Click to expand...
I will do it later today, I found the address for CELL VID. It's 0x3110.
 
You must log in or register to reply here.

Similar threads

Back
Top