Moderator: Steven

 
djfilms
Topic Author
Posts: 3
Joined: 10 Aug 2017, 18:43

Best sensor for custom?

10 Aug 2017, 18:47

Just found your excellent app! One question: I have a 2012 Mac Pro upgraded to 2x3.46 GHZ, 6-core and 96 GB RAM. The custom control allows you to choose the sensor or ambient for control. Is there a recommended sensor for best avoidance of overheating for this Mac Pro?

Any info gratefully received.
 
PerryM
Posts: 3
Joined: 25 Aug 2017, 16:40

Re: Best sensor for custom?

25 Aug 2017, 17:28

Maybe I can answer with some logic. I hope somebody with more knowledge will come in.
I have a Mac Pro 4,1 hacked to 5,1 with twin processors and a couple of GPUs. I am currently running some scientific calculations which usually max out the processors and sometimes the GPU (GTX 980) so lots of heat! It seems to me that we basically have a heat energy balance - the processing puts in a certain amount of heat and there is nothing we can do to reduce that. It is ultimately no good changing heatsink grease etc as long as the heat sinks are getting hot. The only thing that will get the heat away is moving air through the heat sinks and out of the case.
Now it is simple physics to understand that the amount of heat energy that can be transferred will be proportional to the temperature difference. In other words, the hotter the heat sinks the better the transfer.
Having accepted this fact - I then did some experimenting. The most effective fans for controlling the processor heatsink temp is the relevant 'BOOST' fan. This is mounted in the heat sink so not surprising! I have thus set my Control to be based upon the relevant heat sink temperature. I have found that for continuous load, if I set the range 50 - 60degC then I keep the core below 100C. Dependant on the exact job in hand, they usually settle about 90C (A gets more abuse than B).
All the other fans are basically shifting air through the case so I have set them all to be based upon 'Ambient'. The coolest I ever see with the computer at rest is 30C and after a long period of max processing I can see it rise to 37-38C. I have thus set the range to 30-40C. This gives nice quiet fans for normal computer use but enough air flow to keep the balance for max processing.
The screen shot below shows the status quo whist the computer is full blast (after several hours) on a warmish day in UK. The room temp is about 28C. (edit - it refuses to paste the screen shot but the tech data is below)
I await any comments with interest.

Macs Fan Control 1.4.8.1
MacPro5,1
CPU: Intel(R) Xeon(R) CPU X5570 @ 2.93GHz [100]
GPU: NVIDIA GeForce GT 640 [active], NVIDIA GeForce GTX 980
macOS 10.12.6

Fans:
PCI-min{800}curr{3021}max{4500}mode{2,TA0P,30,40}
PS-min{600}curr{1425}max{2800}mode{2,TA0P,32,40}
EXHAUST-min{600}curr{1919}max{2800}mode{2,TA0P,30,40}
INTAKE-min{600}curr{1919}max{2800}mode{2,TA0P,30,40}
BOOSTA-min{800}curr{1838}max{5200}mode{2,TCAH,50,60}
BOOSTB-min{800}curr{1778}max{5200}mode{2,TCBH,50,60}


TempSensors:
TA0P (Ambient) - 36.875, 36.875
TCAC (CPU A core from PCECI) - 88.5781, 11.4219
TCAD (CPU A Diode) - 72.5, 72.5
TCAG () - 57.2812, 57.2812
TCAH (CPU A HeatSink) - 53.875, 53.875
TCAS () - 53.875, 53.875
TCBC (CPU B core from PCECI) - 86.9062, 13.0938
TCBD (CPU B Diode) - 69.75, 69.75
TCBG () - 80, 80
TCBH (CPU B HeatSink) - 52.75, 52.75
TCBS () - 52.75, 52.75
TH1F () - 37.125, 37.125
TH1P (Drive Bay 0) - 37.125, 37.125
TH1V () - 38.2188, 38.2188
TH2F () - 35.9023, 35.9023
TH2P (Drive Bay 1) - 36, 36
TH2V () - 36.875, 36.875
TH3F () - 35.9062, 35.9062
TH3P (Drive Bay 2) - 35.875, 35.875
TH3V () - 36.875, 36.875
TH4F () - 35.7266, 35.7266
TH4P (Drive Bay 3) - 35.75, 35.75
TH4V () - 37.6367, 37.6367
THPS () - 38.2188, 38.2188
THTG () - 58, 58
TM1P (DIMM Proximity 1) - 44, 44
TM2P (DIMM Proximity 2) - 45.375, 45.375
TM2V () - -1, 129
TM3P (DIMM Proximity 3) - 47.375, 47.375
TM3V () - -1, 129
TM4P (DIMM Proximity 4) - 47.375, 47.375
TM5P (DIMM Proximity 5) - 45.5, 45.5
TM6P (DIMM Proximity 6) - 47.875, 47.875
TM6V () - 43.0586, 43.0586
TM7P (DIMM Proximity 7) - 43.75, 43.75
TM7V () - 36.875, 36.875
TM8P (DIMM Proximity 8) - 35.75, 35.75
TM8V () - 48.5898, 48.5898
TM9V () - 50.9141, 50.9141
TMA1 () - 47.75, 47.75
TMA2 () - 48.5, 48.5
TMA3 () - 49.75, 49.75
TMA4 () - 49.75, 49.75
TMB1 () - 38.75, 38.75
TMB2 () - 37.75, 37.75
TMB3 () - 38.25, 38.25
TMB4 () - 38.5, 38.5
TMHS () - 49.75, 49.75
TMLS () - 37.75, 37.75
TMPS () - 49.75, 49.75
TMPV () - 48.5898, 48.5898
TMTG () - 75, 75
TN0D (IOH Diode) - 65.875, 65.875
TN0H (IOH HeatSink) - 53.625, 53.625
TNTG () - 65, 65
Te1F () - -1, 129
Te1P (PCIE Ambient) - 35.875, 35.875
Te1S () - -1, 129
Te2F () - -1, 129
Te2S () - -1, 129
Te3F () - -1, 129
Te3S () - -1, 129
Te4F () - -1, 129
Te4S () - -1, 129
Te5F () - -1, 129
Te5S () - -1, 129
TeGG () - 85, 85
TeGP () - -1, 0
TeRG () - 60, 60
TeRP () - -1, 0
TeRV () - -1, 129
Tp0C (PSMI Supply AC/DC Supply 1) - 39.8125, 39.8125
Tp1C (PSMI Supply AC/DC Supply 2) - 46.4219, 46.4219
TpPS () - 46.4219, 46.4219
TpTG () - 90, 90


Disk drives:
Samsung SSD 850 EVO 500GB - 32
WDC WD6400AAKS-41H2B0 - 41
Hitachi HDS722020ALA330 - 38
Hitachi HDS722020ALA330 - 39
Hitachi HUA722010CLA330 - 38
 
djfilms
Topic Author
Posts: 3
Joined: 10 Aug 2017, 18:43

Re: Best sensor for custom?

25 Aug 2017, 17:53

Perry,

Multitudinous thanks for your detailed and clear reply! Having drunk at the fountain of your knowledge, I'll follow in your path for setting my fan controls.

Again, much appreciated. regards, David
 
PerryM
Posts: 3
Joined: 25 Aug 2017, 16:40

Re: Best sensor for custom?

26 Aug 2017, 06:20

Screen Shot 2017-08-25 at 18.01.20.jpg
Screen Shot 2017-08-25 at 18.01.20.jpg (101.54 KiB) Viewed 342 times
You're welcome. Here is another attempt at adding the window.
 
djfilms
Topic Author
Posts: 3
Joined: 10 Aug 2017, 18:43

Re: Best sensor for custom?

26 Aug 2017, 12:33

Perry, Terrific! Thanks. regards, David
 
PerryM
Posts: 3
Joined: 25 Aug 2017, 16:40

Re: Best sensor for custom?

26 Aug 2017, 13:11

So my question for others:
Is a 17/19degC difference between the CPU Diode and the CPU Heatsink typical (at prolonged full power) or could I reduce this with a new application of thermal grease?
To continue my thoughts from above - if the difference was smaller then the heatsink would be hotter and the heat transfer more efficient. Thus the CPU could be cooler for the same fan speeds.
I realise that I am thus disagreeing with my own previous statement about thermal grease!

Who is online

Users browsing this forum: No registered users and 1 guest