I'd like to know it this behaviour is caused by the 1.39f5 SMC version or by the Base_XX blob in the NVRAM, what I call hardware descriptor. Apple continuously upgraded the Base_XX, from Base_17 in February 2009 to Base_21 in the end of 2011.
If you wanna troubleshot this at the lowest level possible, I can upgrade your Base_XX to the same one as the mid-2012 MP51 (Base_21). Just do a dump of your BootROM, compress it and send me by PM.
@h9826790 did you had this bug too? Resolved with Base_21?
A quick test shows that my cMP's fan now are all normal with the Sapphire PULSE RX580 8GB (ROM modded as per
this post, including the
part number mod).
This is the capture about 1min after I boot to desktop
Both PCIe and PSU fans are in auto mode, and their min speed are normal.
However, I sill can't 100% sure if they are back to normal yet. As I said, this high PCIe / PSU fan bug is intermittent, and we still has no idea what's the real complete condition to trigger it. Especially I used my own version of fan control software to automatically manage the PCIe / PSU fan for years already (I turn on force fan mode for these two fans and manage the target fan speed, which will completely override all native system fan setting and protection). Not sure if that leave anything inside the SMC / PRAM / OS which can affect the fan behaviour
What I am 100% sure is my way to manage the fan does leave something in the SMC which can affect the fan's native behaviour. e.g. As you can see, my min booster fan speed is above the normal idle RPM, and this number will stay in SMC, carry to Windows. So, if I really want to make sure my fans are back to normal. I will need
1) Make a clean installation
2) reset SMC
3) run this clean OS for few week with lots of cold / warm reboot
This is quite a impossible test for me at this moment.
But anyway, I will occasionally disable my own fan control, and see how the fans react to the native fan management in the next few weeks.
[doublepost=1536022193][/doublepost]Anyway, when I looks at the SMC key / value, I found some interesting number that I never pay attention to
1) The CPU average target temperature is 80C.
This fit the observation so far. With native fan setting, CPU's fan basically stay at idle until hit 80C, then the fan spin up to keep it no more than few degree C above 80C.
OF course, this is the CPU diode temperature, not the T-case. This is why Intel says my CPU's max temperature is 68C (T-case), but Apple still let it run to above 80C (T-diode).
2) Memory target temperature is 75C
There is another thread about memory throttling. I haven't really do any test in this area yet. May look into it more later. But so far, my Micron memory seems very stable in performance, not that temperature sensitive.
3) PSU target temperature is 90C!
This is so far the most ridiculous value. And now we know why the SMC can allow our PSU run at above 60C but the fan still at idle. And the PSU temperature (TpPS) is identical to the PSU component temperature (Tp1C), which make sense. IMO, if anyone setup their own fan profile in MacsFanControl. It's better to base on PSU 2 temperature (which is the component temperature), but not the PSU 1 (Tp0C), which is the PSU ambient temperature.
4) North bridge target temperature is just 65C
65C is reasonable target value on the cMP. However, as many threads discuss before, lots of users' cMP NB can run into 80C but the SMC still doing nothing. I set booster fan react to both CPU and NB temperature. And once NB is above 60C, the booster fan will spin up. This is why my NB value (TN0D) is just above 60C, and as you can see from my last post, my booster fan speed is above normal idle. Anyway, it seems I pick a good number, my NB always stay below 65C.
