PDA

View Full Version : My fan does not like 6200RPM




ayeying
Dec 3, 2009, 11:29 AM
Okay, so a little background story.

I applied a Copper Mod, basically sandwiching a Copper in between the Diode and Heatsink, effectively allowing more heat to be drawn away from the diode, stored in the copper while allowing the aluminum enough time to cool itself without being overloaded by the diode's heat creation.

Since some users expressed that I should get some data with before and after, I've removed the copper mod and reapplied the system's heatsink stock.

So far, my temperatures are higher, approx. 10-15 deg C higher than with the copper mod. However, the fans are still... well, a bit lazy.

Currently, after surfing online for about 40 minutes now, my fans are only at 4600-4800 RPM. Keep in mind, I do grid computing, therefore my CPU is always pinned at 100%. My temperatures are jumping around 82-98 deg C, but mostly holding steady around 85-87 deg C most of the time.

Right now, I'm questioning if the Copper Mod has anything to do with the fan being overactive. I have removed the battery when I performed these procedures, so the SMC/PMU is reset.

I have seen the fans jump to 6200RPM, but only for a very short time, even running a virtual machine along with grid. The fans slow down to 3000-5000 RPM and tend to stay there unless theres considerable load which... i dunno how much higher I can get with a 100% cpu...

Any opinions?



alphaod
Dec 3, 2009, 03:21 PM
IF your CPU is jumping around 82-98 then that's within the threshold of the CPU max temperature threshold. With it even usually around the 80 mark is fine since again the CPU is designed to run at that temperature for extended periods of time without any issues.

My problem is that if I run anything on the computer @ maximum power, my termperatures fluctuate between 95 all the way to 125, which is about 25 about the 100C maximum thermal threshold for safe computing. I just hope the thermal module is reporting the wrong temperatures because if I continue like that, the CPU will fry (not to mention my leg if I'm using it on the go).

Furthermore, during the hot session, I never see my fans go about 4800RPM; quite worrying; this brings me to suspect that Apple's power management has changed the way the way the fans respond to higher temperatures. Apple has been focusing on less noise and such and when they force lower fans, the machine is indeed more quiet, but at the expense of component life; they probably want you to buy a new computer in few years anyways (but here I'm making an assumption, not really based on any facts).

FWIW, to reset your SMC, [plug in the AC adapter] just press Command+Shift+Option (held) and then press the power button (just once), and release everything (wait 5 seconds and boot the machine with the power button)

MacModMachine
Dec 3, 2009, 04:12 PM
removing the battery resets the smc?

i would be curious if the diode reacts to the heat buildup in the copper, registering a much higher temperature then the aluminum, causing the fan to run overtime, just a thought.

my ducting project is a complete success on the air i have reduced my overall temps by 12-15c under load, i will be finalizing the guide for it shortly.

ps, i built the duct into the removable back plate of the air, completely reversable.

ayeying
Dec 4, 2009, 11:00 AM
IF your CPU is jumping around 82-98 then that's within the threshold of the CPU max temperature threshold. With it even usually around the 80 mark is fine since again the CPU is designed to run at that temperature for extended periods of time without any issues.

My problem is that if I run anything on the computer @ maximum power, my termperatures fluctuate between 95 all the way to 125, which is about 25 about the 100C maximum thermal threshold for safe computing. I just hope the thermal module is reporting the wrong temperatures because if I continue like that, the CPU will fry (not to mention my leg if I'm using it on the go).

Furthermore, during the hot session, I never see my fans go about 4800RPM; quite worrying; this brings me to suspect that Apple's power management has changed the way the way the fans respond to higher temperatures. Apple has been focusing on less noise and such and when they force lower fans, the machine is indeed more quiet, but at the expense of component life; they probably want you to buy a new computer in few years anyways (but here I'm making an assumption, not really based on any facts).

FWIW, to reset your SMC, [plug in the AC adapter] just press Command+Shift+Option (held) and then press the power button (just once), and release everything (wait 5 seconds and boot the machine with the power button)

I understand my temperatures are fine... but the fans are just not being as trigger happy as before. I'm not sure why is that. So far, the downclocking still happens, but fans just refuse to go to 6200RPM majority of the time. I have to manually force it via smcFanControl.

removing the battery resets the smc?

i would be curious if the diode reacts to the heat buildup in the copper, registering a much higher temperature then the aluminum, causing the fan to run overtime, just a thought.

my ducting project is a complete success on the air i have reduced my overall temps by 12-15c under load, i will be finalizing the guide for it shortly.

ps, i built the duct into the removable back plate of the air, completely reversable.

The Air's internal battery stores all the SMC information. We don't have a separate battery for it so it's shared with the internal battery. Once removed, all SMC data is reset.

The idea behind the copper was hoping that it'll create a bridge of flowing heat. I'm pretty sure the diode would react to the copper, but I'm hoping that the aluminum would be fast enough to draw the heat away from the copper while at the same time the copper is drawing heat away from the diode.

However, the weirdness here is that the Fan isn't going overtime in either situations. Copper or no copper, the fans seems to be less active :confused:. Honestly, I miss the revving 6200RPM fan right now

I'm curious to see this duct you're creating though.