Macs Fan Control 1.2.1 sometimes uses inordinate CPU

[Schmye Bubbula]

Macs Fan Control 1.2.1 sometimes uses inordinately high CPU: 15-25%, when normally uses 0.5-2%.

Console log attached below, reflecting period with high CPU, then quit & relaunch.

Code: Select all

Macs Fan Control 1.2.1
MacBookPro5,2
Intel(R) Core(TM)2 Duo CPU T9600 @ 2.80GHz[100]
Mac OS X 10.6

Fans:
Left side-min{2000}curr{2003}max{5700}mode{2,TC0D,66,77}
Right side-min{2000}curr{2003}max{5700}mode{2,TC0D,66,77}


TempSensors:
TB0T (Battery TS_MAX) - 33.0977, 33.0977
TB1T (Battery Point 1) - 33.0977, 33.0977
TB2T (Battery Point 2) - 30.1992, 30.1992
TB3T () - 34.875, 34.875
TC0D (CPU Diode) - 56.375, 56.375
TC0F () - 47.1289, 47.1289
TC0P (CPU Proximity) - 50.125, 50.125
TG0D (GPU Diode) - -1, 129
TG0F () - 49.7812, 49.7812
TG0H (GPU Heatsink) - 46.375, 46.375
TG0P () - 49.125, 49.125
TG0T () - -1, 128
TG1H () - 52.75, 52.75
TN0D (Northbridge Chip) - 56.375, 56.375
TN0P (Northbridge Point 1) - 52.625, 52.625
TTF0 () - 49, 49
Th2H () - 45.75, 45.75
Tm0P (Mem Controller) - 41.5, 41.5
Ts+S () - -1, 0
Ts0P (Left Palm Rest) - 31.3125, 31.3125
Ts0S () - 41.25, 41.25


Disk drives:
CRUCIAL_CT128M225 - 0

[Steven]

Schmye Bubbula
Try not to set both fans "Based on CPU Diode".
CPU optimizations are coming in the next updates.

[Schmye Bubbula]

Oh, OK.... So I'm still kind of interested academically: inasmuch as the inordinately high CPU usage by Macs Fan Control is only intermittent — and does not seem to be associated with other CPU-intensive tasks I'm doing (although I suppose it could reflect a lingering effect from prior CPU-intensive tasks) — is this a bug that will get fixed, or are you saying that it's against best practices in principle to set both fans "Based on CPU Diode"? Curious minds want to know. Thanks!

[Schmye Bubbula]

For almost two weeks I've run Macs Fan Control without both fans "Based on CPU Diode." (I set one fan to "Based on Northbridge Chip.") I can't tell a difference. So could I get a clarification?—Is having both fans set to the same sensor causing occasional, intermittent high CPU usage a known bug that will get fixed, or are you saying that it's against best practices in principle to set both fans "Based on CPU Diode"?

[Steven]

high CPU usage lasts for ... (time)?

[Schmye Bubbula]

Well, I suppose my data points would not be good ones, sorry, because the few times I notice it happening, I don't let it play out—I just quit and relaunch Macs Fan Control, and things return to normal. I can say that I have LoadInDock running at all times, and I don't have anything else CPU-intensive running at the same time.

But it sounds like you haven't gotten any other reports of this. Like I said, it is infrequent and intermittent.

I guess what I'm really interested in is getting an answer to the question of whether it's a possible bug, or simply a matter of best practices not to set both fans to "Based on CPU Diode"? Why did you say that?

[Steven]

Yes it's a bug, better to say not optimized well. This will be done soon.

And it's logically NOT correct to set both fans to be related to the same sensor. Normally one is for CPU and the other - GPU

[Schmye Bubbula]

Oh, OK, thanks for the clarifications, Steven! Looking forward to the next update.

Yes, it makes sense for the norm to be one fan for CPU, and the other for GPU. In my case, however, I'm not a gamer or such, and never use my high falutin' NVIDIA GeForce 9600M GT; instead, I always keep my Energy Saver System Preference Graphics setting on "Better Battery Life" instead of "Higher Performance," and therefore use merely my other NVIDIA GeForce 9400M (which hijacks some of my regular system RAM), and I noticed that it hardly ever gets hot, the way I use my computer. So my rationale was that since the CPU Diode seems to be the most volatile (even more than the Northbridge chip), I would get the most overall cooling if I let it control both the left & right fans. Moreover, the left fan also seems to cool the AC adapter power supply section of the board—which heats my case more than anything else—and I noticed that if I set the other sensor to something else, the left fan often wouldn't speed up to cool that side in as timely a manner. So would you now concur with my reasons for running both fans off the CPU Diode sensor, all things considered?... Thanks!