Apple 2010 mac mini fan always on?

[kenreau]

I'm having some difficulty determining if my new 2010 mac mini has issues or not.

 

I changed out the o.e.m. 320 GB HD for a SSD and upgraded the ram. It boots up and appears to be working fine, but the fan comes on and runs full speed. I left it running for about 5 minutes and the fan continued to run full speed. I wondered if maybe I didn't get a temperature sensor re-installed or plugged back in correctly, so I've pulled it apart twice now to trouble shoot, unplugged the temperature sensor (molex?) connectors and it continues to run full speed.

 

Can anyone comment if this is typical with the 2010 mac mini's? do they eventually cycle down to a slow or no fan? My notion is with the SSD their should be little or no heat and thus no fan, especially from a cold start up. My old 2009 fan almost never comes on. I'm trying to sort out if I got a got a bad unit or may have botched something in the swap out of parts.

 

Thanks

Kenreau

 

[wgscott]

I haven't done this yet, but I think what may have happened is that you de-attached the temperature probe from the old disk and haven't correctly attached it to the new one. I think I read about this on another 2010 mini thread somewhere here. What is the temp readout?

 

[Bob Stern]

I expect William Scott's diagnosis is correct.

 

These instructions should be useful:

 

[kenreau]

Thanks guys. Problem solved.

 

One of the little gold connectors on the temp. sensor lead wire that snaps into the main pcb came out of its plastic (molex?) housing and was not making contact. The plastic housing actually broke and fell apart - it is ultra minimal construction and may only be good for a couple of installs. I surgically snapped it in to place and sealed it in place with a little dab of hot melt glue. Should be good until the second coming...

 

For the original work, I found a youtube "2010 mac mini hard drive" video swap out by a 'Travis' that was very helpful.

 

Thanks again

Kenreau

 

[wgscott]

Can you hear any improvement (apart from the fan not running)?

 

[kenreau]

Not yet....I didn't even insert it in my system yet. From what I've read, it is the best sounding mac mini already. I just shipped it off to the Mach2 Music guys to scrub it with their optimization for music server mods.

 

http://www.mach2music.com/index.html

 

[wgscott]

OPERATING SYSTEM

We want the music server applications to have as much of the computers attention as possible. This greatly affects the sound quality, and after all this is a Dedicated Music Server. So we optimize the Snow Leopard Operating System for playing music by removing anything we can that doesn't support that goal. Our special combination of modifications removes or turns off numerous non-music related applications and processes. We strip out almost 1.5GB of unnecessary code.

 

99% of what I assume they are stripping out (eg unix subsystem) shouldn't get in the way, and may be needed in order to do operating system updates. OS X is a very efficient operating system that really doesn't need to be further optimized. It is much more efficient than 10.5, for example. Since it is a unix operating system, the various parts are only used on an as-needed basis, and don't hog system resources.

 

Turning off Dashboard, hourly Time-machine backups, and Spotlight/metadata indexing, and not running inefficient apps during playback like Safari, Spaces and fast user switching should be more than adequate.

 

[kenreau]

Could be... not to get too far off on another subject tangent, but check out some of the RMAF room blogs at AudioCircle. A number of rooms (dB Audio labs, Emperical Audio, GR Research) used these Mach2 Music 2010 mac minis to reportedly great effect.

 

Kenreau

 

[wgscott]

But from what I read on your link, it sounds like their main deal is that they install memory and a SSD, but you just did that (at least the SSD), right?

 

I often listen to music at work while doing heavy-duty computational stuff that maxes out the processors, has lots of disk io and hogs lots of memory, and I never detected even a blip of a degradation in sound quality.

 

Anyway, I would still like to hear about what differences you perceive. I've been holding off on the mods, but if it would really improve the sound quality to have an SSD, I'd go for it. (I put one in my laptop when my standard drive failed, and can't detect any difference in sound quality, but that is my Macbook Air, which isn't really suited to be a music server).

 

[kenreau]

I'm anxious myself to give it a listen. You may not be able to perceive the differences if listening in a typically noisey work environement and with computer speakers(?). My notion is, like so many other fine tuning enhancements, it is only really audible in a dedicated listening enviroment. I should know in a couple of weeks.

 

Kenreau

 

[dmccombs]

Wgscott,

 

If you do some tests, you will find turning off certain apps and processes provide huge sonic gains. I used to agree with your theory that none of this SHOULD matter, but some simple testing shows it surely does.

 

If you take the time to do some testing and you have a fairly resolving system, you will be shocked. ;-)

 

[dmccombs]

Ken,

 

Your Mach2 is on its way back to you. You already had your hardware in place so all we did was the software modifications. Your feedback should be very relevant to wgscott's questions.

 

Yes, RMAF was blast. We got to A/B the Mach2 against some MacMinis that were already prepped and being used in the show. 2 rooms asked if they could use our Mach2 for the rest of the show, after hearing the A/B tests.

 

Steven Stone highlighted the 2 rooms we were in (among others). http://www.avguide.com/blog/tas-rmaf-steven-stone-digital-and-new-technologies

 

The two rooms we were in won Best of Show - Cost No Object, and Best of Show - For the Lowest Cost. Of course these were terrific rooms before the Mach2 was used, but in hearing the A/B tests, I know we made them better.

 

The bottom line is that both software mods and hardware mods make the MacMini an even better Music Server. Way Better...

 

[wgscott]

Hopefully not electrically.

 

It might be that I don't have a good enough system or hearing to tell, but I'll give it a try.

 

A slightly easier test would be to compile something with both processors maxing out. Would that degrade the sound? That one is real easy for me to do.

 

[dmccombs]

I prefer to test the actual proposed changes I am considering, but I think playing your music, then maxing out your CPU would be a good test.

 

My guess is that even if the music doesn't skip, it will sound relatively sloppy (i.e. cymbals not as crisp, bass not as tight, soundstage seems smaller, etc.)

 

BTW at Mach2, we do more in our mods than just kill processes that free up CPU cycles. I think Ken's feedback will be interesting since we only did software updates on his system. He already had the hardware mods. Ken may hear more differences than your test since you are just dealing with the CPU load.

 

Nonetheless, your test should be interesting, if you system is pretty resolving.

 

 

 

[wgscott]

So we need an independent way of deciding whether or not my system is sufficiently resolving.

 

With B&W P5 headphones plugged into my 2009 mini at work, I can't tell any difference, but I will concede that this is probably not good enough to hear a difference. The problem is, if I don't hear a difference on my stereo system, how can I know it isn't something that can just be explained away?

 

If I look at what consumes the most CPU, memory, disk i/o, etc, under standard operating conditions, I find that the main culprits are spotlight and the associated md- processes, time machine hourly incremental backups, screen saver, and dashboard (a memory hog, once activated). Generally not a good idea to leave Safari (or any web browser) open, especially with a javascript or flash applet running. Auto software updates are often a hidden problem as well -- I just noticed a "google software update" consuming CPU and presumably hanging on my work 2009 mini.

 

For this reason, I find Menu Meters invaluable, as it allows me to keep an eye on CPU and memory consumption, disk i/o, temperatures, upload/download speeds, etc from the menu bar, while introducing very little overhead.

 

I'm happy to test any proposed changes you have, but I don't know what they are; I can only guess. As far as I am aware, most of what is in /System/Library/LaunchDaemons that isn't needed for normal system operation is off by default.

 

grep Disabled /System/Library/LaunchDaemons/* | cut -d : -f 1 gives a listing. Most of those load only on demand anyway. The stuff in the canonical unix file system directories (/usr, /bin, /sbin, /etc, /var) doesn't make demands on the system unless it is running, but on the other hand is assumed to be present for operating system incremental updates and security updates. It would worry me to delete any of that (and since I use OS X primarily as a unix machine anyway, probably wouldn't want to, even if I could). All of this of course is pure speculation, since I don't know what you have in mind.

 

 

 

[dmccombs]

Thanks for the offer for testing, but I run my tests as I think of things to try. I let my ears be the judge.

 

I thought you wanted to test out how CPU load affects sound. If so, go for it however you deem best. You obviously have a good grasp of Unix/Linux.

 

[wgscott]

I wasn't clear what you were asking of me. I assumed I wouldn't be able to test what you are doing if it is a proprietary trade secret.

 

I haven't had a chance to test this on my home stereo system, but with decent headphones plugged into the headphone jack on my 2009 mini at work, maxing everything out, there is no change from when the system is idling.