smcFanControl — Cool Your MacBook Pro

Clodas writes, "smcFanControl 1.1 is a simple GUI that lets you control how fast each fan spins on your MacBook, MacBook Pro, or Mac Mini. The temperature of my MBP when idle averaged around 63 degrees celsius. After running smcFanControl 1.1, my temperature dropped to 43 celsius within 10 minutes of use. This now allows me to sit my MBP on my lap, something I was unable to do previously since the machine got so hot. I have my fans set to spin at a minimum of 3000 RPM and I still don't hear the fans spinning. Apple by default has them set to 1000 RPM. I really recommend smcFanControl 1.1 for any that feel their MB, MBP, or Mini are too hot to handle."

Comments

Finally!

[TPIRman]

I really recommend smcFanControl 1.1 for any that feel their MB, MBP, or Mini are too hot to handle.

Finally! I can use my Mac mini on my lap again.

Re:Finally!

[JimDaGeek]

Amen to that! I just got my first Mac (macbook) and really love it. However I have really hated how hot it gets. I just purchased a Lapinator and hope that will help when it arrives.

I am glad that this code is under the GPL. Instead of having to have a GUI app always running to make sure the fans stay at a certain RPM, maybe I or another can rip out the guts and make it a cron job that runs every 5 mins or so?

OT:
P.S. Does anyone know of a good Mac usenet/email group for learning all I can about the Mac? Professionally, I have been programming Windows and Linux for 10 years so I do not want my hand held or told "you don't need to know that". A lot of the tips/solutions/etc I have been finding for Mac releated questions have been more for the "I want it to just work and don't care how" crowd. I am more of the "I love Linux and want to get dirty with my new Mac" crowd.

Uh huh

[nodwick]

I have my fans set to spin at a minimum of 3000 RPM and I still don't hear the fans spinning. Apple by default has them set to 1000 RPM.

And how long does your battery last between recharges now?

Re:Uh huh

[4D6963]

Interesting question tho, I wonder what's the extra power drain. If someone could give an estimate.

Re:Uh huh

[dal20402]

Been using smc and then smcFanControl since they showed up... battery life on my MBP doesn't seem appreciably different, not that it was any good to begin with, with the 7200rpm HD.

The best way to save battery is to dim the screen. At less than half brightness I can get nearly 3.5 hours in normal usage. At full brightness it's more like 2.5+.

Re:Uh huh

[kloppe]

A fan's power draw is a couple of watts at most, and no, your fan won't die 'orribly from it using speed control software. :) Higher speed probably means the bearings will wear out a bit faster though. For more info, check out http://en.wikipedia.org/wiki/Pulse-width_modulatio n

Re:Uh huh

[The Lerneaen Hydra]

Seeing as how most small sized fans (40mm-ish) dont usually take more than 0.1A @ 12V => 1.2 watts (at full speed), compared to the 40-50-ish watts that the rest of the system uses the difference is negligable.

Re:Uh huh

[dal20402]

Yes.

If your screen buzzes when dim you have a bad inverter board. Unlike the famous CPU whine, Apple was willing and able to fix this problem from the very beginning. If your machine has the bad inverter, send or take it back to Apple for repair.

Re:Uh huh

[Firehed]

Well, desktop fans have yet to rip themselves apart. The software won't throttle the fan faster than it was designed to go, and all fans are designed to be able to sustain their maximum rotation speed until they reach their MTBF. Some people who use their laptops as DTR machines will have them under a constant heavy load (and then run folding in the background, just to make sure their batteries never stand a chance), and their fans aren't tearing themselves apart.

Basically, fans can be fed anywhere from 0 to 12 volts (maybe laptops are 0-5v, but it's pretty standard). It usually takes 5-6v to get them spinning and overcome inertia, but after that you can drop it down to 4v or so to keep it going. The app just overrides the software control telling it to throttle the voltage until it hits near a certain RPM. The voltage control almost certainly works on a percentage, but even if not, there's only so much available to give it - you can't just pull random extra voltage in from somewhere to overvolt the fan.

As to the relation to a hard drive... couldn't say. I've seen numerous desktop-sized fans that use quite a bit more power than a notebook hard drive, which (in my experience) draw 2.5w or less (ie, you can power them from a USB port with no extra plugs). There are fans out there that draw 12w and up, but those tend to be the high-speed 120mm fans that can do serious damage to objects that happen to get in their way. Notebooks, on the other hand, tend to use very small fans such as 40mm units, which have a power draw in the half-watt range at full tilt. Varies by fan of course, but this probably won't drain your battery any faster than plugging in a flash drive and pulling a few files from it.

I'm just a bit irked that Slashdot posted this today. Not 36 hours ago, I left my MBP at the Apple store for them to fix the heat problems.

Re:Uh huh

[flyingfsck]

I am presently trying to quiet down a MiniITX machine. It has a pair of howling fans in it. I found that a 47 Ohm resistor in series with the fan will reduce the noice to a whisper and still run the machine nice and cool. So, I had to do some measurements.

A 1 inch miniature fan uses about 60mA at 12V. So running at full tilt, it consumes about 0.72W of power. The minimum setting where it will still start reliably, is about 6V at 50mA, so the power is about 0.3W. So, from that you can see that the power consumption of miniature fans is negligible - almost nothing.

The performance of the little fans vary a lot, but the power consumption of even the worst fan I measured, is still negligible. I am actually typing away on this little machine right now, and I think one fan needs a bigger resitor - still too noisy - need to turn the soldering iron back on...

Well, it works ...

[Sonic McTails]

I reset my fan seed to 6000 to see how cool my laptop could get, and its amazing getting cooler then most PCs I owned, whose fans are 10x as load. I would pay for this app if it was a commerial program!

I wonder how it works, I'd love to see the source code for it ...

Re:Well, it works ...

[dal20402]

There's no benefit in setting fan speed to 6000 at idle. Here are idle CPU1 temps for my MBP (after it's been running for at least 20 mins) at various speeds. Each MBP tends to get different results, so YMMV.

Default (1000rpm): 59-62 C
2000rpm: 49-51 C
2500rpm: 46-48 C
3000rpm: 42-44 C
3500-6000rpm: no change: 38-41 C

Note that the faster speeds DO make a difference when the MBP is doing intensive work, as it appears that setting the minimum speed to higher also causes the fan to ramp up more quickly. At sustained 100% CPU load the machine is always hot but the lowest temperature was reached when I set the minimum to 5000rpm: about 78-81 C.

On the outside, the machine is MUCH cooler when using any setting over 2500rpm. It really is a "laptop" now. And below 3000rpm the fans are barely audible. I don't know what Apple was thinking when they chose such a low default.

Re:Well, it works ...

[donaldGuy]

As posted .. the source code is included in the distribution.

It is inside the .app itself... Just right click, click 'Show Contents' .. then go to Contents/Resources/Sources.. all the C++ source and xcode project files are there ..

Also .. I created a more standard source distro .. If anyone cares they can mirror it somewhere better, but here it is... http://rapidshare.de/files/36736578/smcFanControl- 1.1.src.tar.gz.html

Peace,
Donald

Re:Well, it works ...

[mustafap]

>I would pay for this app if it was a commerial program!

The source is GPL, so you can buy it as a commercial program. Please feel free to post your credit card details, and I'll sell it to you for the bargain price of $49. I'll even ship you the source code!

Re:My pref...

[Sonic McTails]

The fans will automatically rev up to 6000RPM max if necessary to keep the machine cool, and as a failsafe, the machine will automatically turn off if it gets too hot*

* At least, I believe it will, a lot of other PCs do, and my old mac did ...

macbook pro redesign

[posterlogo]

I think this heat problem was generally regarded as a major source of delay for the merom macbook pro's as a simple update wasn't going to cut it this time -- the whole interior had to be redesigned to allow for better cooling. Let's hope they got the problem fixed. I also wonder why the fan speed wasn't considered an adequate fix -- is there something inherently unstable about this? Don't fan's in other laptops run that fast?

Re:Well, look at it this way.

[Mooga]

Does fertility really matter? I mean, we're posting on SLASHDOT. If we had a life, we wouldn't need to post about cooling our laptops on slashdot...

Undervolting

[Aardpig]

I think it would be even better if you could unvervolt the MBP. My 2.26GHz Pentium M Sonoma system used to run very hot (95C) under full load (e.g., mprime); by undervolting from 1.35V to 1.18V, I've cut that down to 75C. Not only does this solve a heat problem; my fans are also quieter (since they are under less stress), and I have a substantial power saving to boot (recall, power consumption scales as voltage squared).

1000 posts. Hmmm, maybe I should get out more...

fan failure, not battery life, the issue

[SuperBanana]

And how long does your battery last between recharges now?

The question is whether the fans will be run past their expected/rated lifetime before the computer has.

As we all know- small fans (CPU fans, chipset fans anyone?) don't last very long. That's precisely why they're only run when necessary. Given the MB/MBP's thermal output, Bad Things will happen if those fans fail- probably no worse than it shutting itself down or crashing. Still won't be good for it.

That said, keeping the fans on a very low speed to maintain a cooler temperature will improve general component life.

Re:fan failure, not battery life, the issue

[TubeSteak]

The question is whether the fans will be run past their expected/rated lifetime before the computer has.

As we all know- small fans (CPU fans, chipset fans anyone?) don't last very long.

Computer fans generally die out because they use shiatty oil impregnated sintered bronze bushings. Bushings soaked in oil are vastly inferior to a good sealed bearing.

Though heat is still an issue either way, since it'll degrade the lubricant, bearings will have a longer lifespan without maintanence.

You can revive a noisy computer fan if you peel back the sticker & put a drop of oil into the hole, but no promises on how long that'll last.

This might also resolve the "it won't spin up" problem, though sometimes that's just the motor dying & not the bushing sticking.

Re:fan failure, not battery life, the issue

[ben there...]

Yeah, but who really cares about the fan? When it's dying, you'll know it. All sorts of freaky noises will spew from your system. $20 to replace a fan vs. up to $300 to replace a component that failed because of heat.

Re:fan failure, not battery life, the issue

[Matey-O]

This is innacurate. My first MBP's fans died within two weeks. While diagnosing the problem, I intentionally ran it as hard as I could with one, and later, no fans.

With all fans dead, temp reached 92C, but the system DID NOT FAIL. It _did_ clock cycle to keep running, but I could not get it to freeze up.

Pshaw!

[RealGrouchy]

Come on!

MacBook Pros will never be cool! :P

- RG>

Re:Well, look at it this way.

[elmarkitse]

The 'decrese in fertility' is totally unrelated to the temporary discomfort of burning your wanker by having something too hot sitting on it.

Do a google search for "intra scrotal temperature fertility" and you'll find all kinds of interesting tests that basically say that heat messes with how the dna in your sperm is formed. An increase in heat doesn't damage what you've got, but your newly forming sperm goes 'all out of whack'

Also, wearing tight underpants is worse the boxers or commando, and sitting is worse that walking.

http://www.reproduction-online.org/cgi/content/ful l/129/4/505

http://humrep.oxfordjournals.org/cgi/content/abstr act/20/4/1022

Only fix, I suppose, is to get your computer off your lap, strip naked, stand up, walk around, and eject any damanged sperm. You can just tell whoever walks in that you're busy saving the human race.

Thermal Management For Laptops

[loose electron]

I designed the chip for something called TAFI (Temperature and Fan IC) a way back. This widget (or a variation on it) now sits inside every PC/Mac/Laptop/Desktop box on the planet that has variable speed fans, that cool on demand.

How it works (the simplified 2 mile high view)- Sitting over inside the microprocessor is a diode, that is at the same temperature as the microprocessor chip. The forward biased voltage of a diode changes with temperature. With some signal processing, you can turn that into a temperature number.

The temperature is available for readback over a serial bus. (SMA,SMB, I2C, the original was SMA if I remember correctly) A software routine reads the temperature and makes the call "cool me off" or "at desired maximum temperature" which gets turned into a number that gets loaded over the bus back to the TAFI chip. That number gets dumped into a DAC, which becomes the voltage for powering the DC motor fan.

Presto! Variable speed fans dependent on how hot the microprocessor is!

Before that, all the PC's had fans that ran full blast 24-7-365.

Whoever did the software better realize that they are messing with the thermal management system and could seriosly fry their computer, if they set things up to not cool enough. So like any hardware hack, YMMV and you are taking a chance of doing permanent damage to the machine.

The fan motor, in comparision to the processor, does not suck that much juice, so I expect that it won't change battery run time in a big major way. A little, but not gobs.

Re:Thermal Management For Laptops

[Khuffie]

This only changes the 'minimum' speed your fan runs at; if the cpu gets too hot, the machine will automatically increase the fan speed above that as needed. It's just that the minimum speed on my MacBook is at 1500, and it lets me adjust that (it doesnt allow me to go below that).

obligatory..

[Awod]

Get em' while their hot..

You think it would work correctly from the factory

[klubar]

What a hack... You think after you've dropped a couple of bills on a brand new computer it would be designed to work correctly.

Seriously guys...why can't Apple make a laptop that doesn't double as a Friallator? It seems that with all that computer power available in a laptop, the system could do a better job at adjusting fan speed... perhaps on the power control panel there should be various settings... simmer, roast, boil and flame (Apple with Sony batteries only). Alternatively, Apple could come out with a line of cookware design to work with your laptop.

Re:Well, look at it this way.

[BKX]

The proper term is "receiver." (Yes, it's safe for work, and much funnier than the original.)

Apple sits back and relaxes

[Pliep]

So there you have it. Apple can sit back and relax while others fixe their "hot MacBook" problems. I think anyone who uses this app should return their machine to the store and claim a cooler machine or their money back. If we simply install a tool and then be happy, when is Apple ever gonna fix problems like these?

Too bad that is just a hack...

[not already in use]

I'm one of the brave souls who took apart their MBP and reapplied the thermal paste in reasonable quantities. Problem was, with the stock application heat wasn't transferring properly to the copper tubing from the processors. Unfortunatly, the thermal sensors are on this copper tubing so the problem becomes two-fold. Heat isn't being properly dissipated from the processors, and the computer thinks it is cooler than it really is, so the fans don't kick in when they should. Once the thermal paste was reapplied and everything was put back togother, I noticed my fans would come on more often or even come on full speed at times, which is a noticable sound. As a result, the MBP noticibly cooler to the touch and CoreDuoTemp reports a much cooler cpu.

Re:You think it would work correctly from the fact

[arodland]

Well let's see. An ordinary sort of DC fan such as you might find in a laptop might draw 50mA. At 12V, this is 600mW. The battery in the MBP 15" has a capacity of 60Wh, meaning that it could run the fan for 100 hours. Alternately, suppose that using your lappy, you draw 20W with no fan activity, so the battery lasts 3 hours. Now run the fan 100% of the time, you're pulling 20.6W, and the battery lasts... 5 minutes less. I wouldn't worry too much. :)

I was just looking at it...

[Ayanami Rei]

It uses IOKit to walk a device tree looking for fan-like things. It enumerates them, and lets you set or read a series of RPM thresholds (min, max, target). The GUI is built on top of the CLI program that does all that.

Re:Yes: PowerBook

[Macthorpe]

Am I the only one here that objects to the term 'lap-safe'?

It's a LAPTOP. If you can't put it on your lap, surely it ceases to be described as one!

Re:Yes: PowerBook

[FLEB]

They rarely, if at all, are anymore. The only terminology you'll hear used by manufacturers is "notebook", for the burning-lap sterility reasons mentioned.