The New Guy - 27 June 2007
Has anyone heard of someone changing the heat sink on their Mini? My PPC 1.25 ghz model would need a heat sink not exceeding 47 mm x 118 mm. (47 mm is about 1.5 inches.) A Sonic Tower might work if one ground the 2 of the sides but it would be great to get a good one that might fit (with some retrofitting of course to mount it).
Michelle Steiner - 27 June 2007
Previously, The New Guy wrote:
Has anyone heard of someone changing the heat sink on their Mini?
No one in their right mind.
Support the troops: Bring them home ASAP.
John Byrns - 27 June 2007
What would be the purpose of doing this? Would this reduce the fan noise from my PPC 1.25 ghz Mac Mini?
Regards,
John Byrns
Surf my web pages at, http://fmamradios.com/
Jolly Roger - 27 June 2007
Disregard The New Guy. The guy is obsessed with heat sinks, and is under the mistaken belief that all Macs are insufficiently cooled.
JR
The New Guy - 27 June 2007
I never implied I was in my right mind.
So Micheele, did you look into the transferring costs with your current cell provider to extricate yourself out of your contract? I seem to remember a website that does nothing but find people to take over contracts of other people that don't need them anymore. You'd have to Google that. I should have included that in the original reply but forgot.
The New Guy - 27 June 2007
Yes John. It would dramatically reduce it. The same way good heat sinks dramatically reduce the noise of fans in any computer.
But for now, try what another poster did: place it vertically so the bottom can dissipate heat better. The ultimate would be to place it on a cold surface. The slot loading optical drive should have no problems operating in a vertical position, as far as I know. Which isn't very far. The Mini uses a small fan, similar to what you find on some video cards so when its going hard its really irritating. That's what I'm really trying to get away from; the pitch of the whine more than anything.
Jeffrey Goldberg - 27 June 2007
John,
As Jolly Roger has said, be wary of what "The New Guy" has to say about cooling in general and heat sinks in particular. It's not that TNG is absolutely wrong about absolutely everything, but he is remarkably confused about fundamentals in such a way that a little knowledge really is dangerous.
I most certainly don't want to repeat the discussion that has already taken place, but you can go back a few weeks and look the discussion he started about iMac cooling.
-j
The New Guy - 27 June 2007
How about not trying to sabotage a thread?
I simply asked if anyone has heard of the Mini heat sink being upgraded. The answer is a Yes or a no-answer. If its a Yes, hopefully an URL will follow so I can learn about it. Surely this hardware deficient newsgroup can grapple with that simple task.
Steven Fisher - 27 June 2007
Previously, The New Guy wrote:
How about not trying to sabotage a thread?
I simply asked if anyone has heard of the Mini heat sink being upgraded. The answer is a Yes or a no-answer. If its a Yes, hopefully an URL will follow so I can learn about it. Surely this hardware deficient newsgroup can grapple with that simple task.
Because you asserted in a subsequent post that the change make a hugely positive impact, and you are woefully under-qualified to make a statement like that.
If all you were interested in was an answer to your question, you wouldn't have attempted to shove bad advice down someone else's throat. Leaving that as is would be extremely irresponsible.
In short, you sabotaged your own thread. Next time if all you want is an answer, don't offer bad advice in a follow-up.
Jeffrey Goldberg - 27 June 2007
Is that all you asked?
I responded to message-id
<replytogroup-250C73.18342527062007@news.lga.highwinds-media.com>
in which you stated:
[putting in a larger heat sink] would dramatically reduce [fan noise]. The same way good heat sinks dramatically reduce the noise of fans in any computer.
I did not respond to your question that started this thread.
Anyway, I do hope that you succeed in your project to upgrade your heat sink. I hope that you keep a record of CPU and other critical component temperatures as well as fan noise both before and after the upgrade. When you do, please post your results.
And for the record, I do agree that a good heat sink versus a bad heat sink can have the effect that you describe here. Much of what you say is correct. Although I have no reason to believe that the heat sink in the mini is inadequate. But even when you are right, you've shot your credibility when it comes to talking about cooling. So go ahead and do your heat sink project, but if you recommend it to others before you have results, I (and apparently I'm not the only one) will advise people to take your recommendations with a large grain of salt.
-j
John Byrns - 27 June 2007
The Mac mini fan is only a minor annoyance, what really needs better heat management is my "AirPortExtreme 802.11n" base station whose top surface gets nearly hot enough to fry eggs on.
Regards,
John Byrns
Surf my web pages at, http://fmamradios.com/
Daniel Packman - 27 June 2007
Previously, John Byrns <byrnsj@sbcglobal.net> wrote:
....
The Mac mini fan is only a minor annoyance, what really needs better heat management is my "AirPortExtreme 802.11n" base station whose top surface gets nearly hot enough to fry eggs on.
He who controls the air flow controls how the
eggs are cooked.
E Z Peaces - 28 June 2007
Jeffrey Goldberg wrote:
Anyway, I do hope that you succeed in your project to upgrade your heat sink. I hope that you keep a record of CPU and other critical component temperatures as well as fan noise both before and after the upgrade. When you do, please post your results.
For my 1.25gHz Mini, the HDD temperature is the only one Temperature Monitor displays. Can this Mini keep show other temperatures?
SparkyGuy - 28 June 2007
Apparently replacing the Mini's processor with a Core 2 Duo (2.16GHz Merom T740) results in lower temperatures:
<http://macenstein.com/default/archives/323> <http://homepage.mac.com/macaholicg5/PhotoAlbum12.html>
Maybe you don't need to change your heatsink at all (c:
The New Guy - 28 June 2007
Are you running it with the top on? If so, how do you know its so hot?
If not, try pointing a fan on it. If it works, you can incorporate the fan in a future install. I don't have an Airport in mine so I can't comment.
This is another example though, of how one component can heat up others. Perhaps one of the reasons I find the Mini's fan kind of bothersome is that I'm running it without the top on. If I had the top on, surely the noise would be far less aggravating. But then it would just run that much hotter! A better heat sink would solve all these problems. In fact its quite likely that with a good heat sink I might not need a fan at all. I'd prefer to have a silent 5 volt 120 mm fan on it just to make sure though.
The New Guy - 28 June 2007
I'm using a Western Digital 250 gb IDE HD and I don't get any temperature monitoring off it at all or anywhere else on the computer. I have several programs that supposedly would detect temperature sensors. A while back, before I looked into this, I was running a different hard drive. It may have had a temperature monitor in it so it may be hard drive specific. Does anyone know if the Intel Minis have temp sensors in them? Judging from Jolly Roger's temp postings the iMacs have quite a few. I've got a Seagate 200 gb IDE hard drive lying around. I'll hook it up later to see if it has a sensor built in.
The New Guy - 28 June 2007
I would propose that if a statement is true or false, the credibility of the person making that statement is irrelevant. All the matters is the truth of the statement. Some of you are well educated in science yet lack real world experience so you don't recognize where the priorities lie in the computer cooling world. I don't want to dig up what was amply discussed before though.
If all you were interested in was an answer to your question, you wouldn't have attempted to shove bad advice down someone else's throat. Leaving that as is would be extremely irresponsible.
What bad advice are you talking about?
Michelle Steiner - 28 June 2007
That is correct; however, when the truth of the statement is not self evident, the credibility of the person making it comes into question. And when the person making the statement has demonstrated repeatedly that he doesn't know much, if anything, about the topic, his credibility is very much an issue.
Some of you are well educated in science yet lack real world experience so you don't recognize where the priorities lie in the computer cooling world.
You have demonstrated lack of education in the field as well as lack of real-world experience, nor have you shown any cognizance of what the priorities are.
In other words, you don't know what you're talking about again.
Support the troops: Bring them home ASAP.
The New Guy - 28 June 2007
Nothing like killing too birds with one stone. :) Many thanks.
The New Guy - 28 June 2007
If people were knowledgeable on a subject, the truth of the statement would be evident. So it all comes down to real world experience.
Your opinion, Michelle. Like I've said, you know software very well. In hardware your knowledge just isn't there. In real world experience, less so. Maybe you just can't stand it when someone disagrees with you?
Tim Streater - 28 June 2007
No, just the opinion of everyone of this NG who has two braincells to rub together, a department in which you were sadly short-changed.
RubyTuesday - 28 June 2007
Ya know, you keep saying that. But looking back at that thread about heat sinks, I can't find one single person that agrees with you. Why does that not tell you something? If you're so right, how can everyone else be so wrong?
Maybe you just can't stand it when _everyone_ disagrees with you.
Ruby.
"Rarely is the question asked: is our children learning?" -- George W. Bush
James Glidewell - 28 June 2007
I'd say it's because the New Guy is a textbook example of the Dunning-Kruger effect...
The New Guy - 28 June 2007
Not at all. Like I said, the hardware experience here is not strong. The software experience here is very strong. Actually there were several people that agreed with several of my points. Once Michelle and others starting slinging insults than others jumped on the bandwagon. It happens. No biggee. And its not about someone "agreeing with you" on everything. Its about specific points. My experience gave me knowledge and now I know why certain designs are lacking. If someone else doesn't have that experience, they may not agree with me. Perfectly understandable. But we're getting off track.
This thread is simply about if anyone has read or heard of the Mini heat sink being replaced. I Googled but didn't have much luck.
Grandpa - 28 June 2007
Do you read for comprehension or just blather? Its his Airport Base Station, not the Mini that he says is getting hot.
Sheesh!
Grandpa
The New Guy - 28 June 2007
Right you are. My mistake. Thought it was the internal Airport.
Jolly Roger - 28 June 2007
As is the hardware experience.
My
experience gave me knowledge and now I know why certain designs are lacking. If someone else doesn't have that experience, they may not agree with me. Perfectly understandable.
Exactly what experience do you have? What degrees do you hold?
JR
Jolly Roger - 28 June 2007
On 2007-06-28 16:50:45 -0500, James Glidewell said:
New Guy is a textbook example of the Dunning-Kruger effect...
From that page:
"Meanwhile, people with true knowledge tended to underestimate their competence."
Word. It's quite typical for engineers in most engineering teams in which I've worked to create performance estimates (with little outside (marketing etc.) influence) that lean heavily to the pessimistic side. Ans the sure-tell sign of a person without much experience is that they tend to bite off more than they can chew. ; )
JR
The New Guy - 28 June 2007
If you want to believe that, fine.
Real world experience. Trying different things and seeing their effect. The only thing that matters is results. And I get results in my own equipment.
Look at the medical profession. Here you have highly educated professionals that are almost useless for anything but trauma (accidents). In disease, they are useful for little more than diagnosis. They are highly trained, and by society's definition, highly educated. Yet they largely responsible for our society's disgraceful health condition. If you told them "I ate something different and noticed these health benefits", they wouldn't listen. Or 99% wouldn't. They are simply not willing to step out of the mainstream and risk ridicule by their peers by noticing something different. You will almost never hear a medical doctor ever recommend fasting yet that has helped untold numbers of people cure terminal health conditions. In some ways education can hinder people. It traps them in established ways of thinking.
Results are the only thing that matters.
Before I gave results showing very effective and silent cooling options. Temperature results. The only thing that matters.
A lot of you think that exhausted hot air is not a sign of poor cooling. I guess that could be true if you believe that components and their heat sinks can run at far higher temperatures than typical room temperature with no long term problems. Most electronics experts will attest to the fact that you want to run electronics at as cool a temperature as possible or as close to room temperature as possible. This is indeed possible and can be tested simply by holding your hand and feeling the exhausted air. If it feels warm, that means its a lot higher then room temperature.
Cooling can be quite easy. Some techniques may seem a little unorthodox, but they are not costly and they really work.
The New Guy - 28 June 2007
Don't you mean the optimistic side? Like their performance goals may be a little unrealistic and not sustainable by market prices? Or were they trying to scare the others by warning that if they didn't do things their way, repercussions would follow, like product failure?
Ans the sure-tell sign of a person without much experience is that they tend to bite off more than they can chew. ; )
We all do that once in a while. But if you want to reach your goals, sometimes its better to aim a little beyond them.
Michelle Steiner - 28 June 2007
Previously, James Glidewell wrote:
I'd say it's because the New Guy is a textbook example of the Dunning-Kruger effect...
I think you hit the nail on the head.
Support the troops: Bring them home ASAP.
Michelle Steiner - 28 June 2007
I software and hardware, your knowledge isn't there.
In real world experience, less so. Maybe you just can't stand it when someone disagrees with you?
Oh, I can stand it alright. What I can't stand is when someone who knows less than I do insists that I'm wrong and he's right, especially when it is obvious that he doesn't know what he is talking about.
I'll stack up my hardware knowledge and real-world experience against yours any day.
Support the troops: Bring them home ASAP.
Grandpa - 28 June 2007
Just how long did the engineers say that the Mars rovers were going to last? And how long have they lasted? Would you say that they were "optimistic?" Most of the design engineers that I worked with were always "hedging their bets" when it came to design life. They'd guarantee 500 hours MTBF, but we would typically actually experience 2000 hours or more. No electronic engineer that I worked with would consider pushing the environmental envelope to get a little more performance. Everything was always derated. But you probably wouldn't know about that criteria.
D-K indeed.
Grandpa
Michelle Steiner - 28 June 2007
Previously, The New Guy wrote:
This thread is simply about if anyone has read or heard of the Mini heat sink being replaced. I Googled but didn't have much luck.
That should give you a hint right there.
You should apply to Apple for this job; it fits your self-described expertise:
<http://jobs.apple.com/index.ajs?BID=1&method=mExternal.showJob&RID=6501& CurrentPage=2>
Requisition Number
2946588
Job title
Engineering Project Mgr: Enclosures
Location
Santa Clara Valley
Country
United States
City
Cupertino
State/Province
California
Job type
Full Time
Job description
As an Engineering Project Manager within the Product Design organization you will be responsible for managing the development and implementation process for one of our Product Design teams. The position will support multiple product lines from our Mac engineering organization. In this role you will drive the day to day program activities to meet the overall program objectives with a heavy focus on the mechanical design engineering cycle and deployment planning. Key skills to include: the ability to create and maintain program schedules; the ability to predict pitfalls and develop a better course of action; the ability to identify and direct all resources required to execute successful programs; and the ability to collaborate across engineering, operations and our OEM partners
to meet very aggressive cost, schedule, quality and TTV goals.
The ability to work closely with the respective Eng Managers, Operations managers, Marketing managers and Product Design Engineers, as well as our vendors and OEM partners is crucial to success. The ideal person is a Project/Program Manager with a mechanical engineering or related technical background.
* At least 8 years of combined experience in project management and mechanical engineering delivering products comprised of multiple materials and high esthetic quality.
* Detailed knowledge of the product development processes as well as a strong understanding of manufacturing
processes.
*Experience interfacing with Operations, Mechanical Engineering/Design and Marketing
Preferred Experience:
* Masters in Engineering preferred.
* Multi-national development experiences also a plus.
* Understands CAD/Mechanical design
Support the troops: Bring them home ASAP.
Michelle Steiner - 28 June 2007
In other words, nothing worthwhile.
Before I gave results showing very effective and silent cooling options. Temperature results. The only thing that matters.
You didn't measure the right things, and your interpretation of your results is screwy.
Support the troops: Bring them home ASAP.
Jolly Roger - 28 June 2007
No, I don't mean the optimistic side. Experienced engineers don't set themselves up for failure. You haven't ever worked in engineering, have you?
See above.
JR
Bob Carter - 28 June 2007
The New Guy wrote:
Does anyone know if the Intel Minis have temp sensors in them?
Yes they do - that is how I was able to provide you with temperatures in an earlier thread. Here is a screenshot showing the sensors probed by Temperature Monitor:
http://freespace.virgin.net/george.coward/other/temp-mon.png
The Mini is the latest model (just a few days' old). I've suspended SETI@home for the time being, so those temperatures are with the Mini idle and a room temperature of 18 C.
The New Guy - 28 June 2007
Well if you call cooling at near ambient levels nothing, so be it. Its the pinnacle of simple, air cooled, cooling, nevertheless. Nobody wants to rely on liquid cooling if they don't have to. The air being exhausted out of my machines feel the same as the ambient temperature. That's cooling that works. And its near silent almost all the time. And the only thing that isn't silent is that tiny Mini fan. That's why I'm searching for a better heat sink.
I didn't measure anything at all. You don't even remember. If you go back to that thread that would never die, you will see that I gave an example of a siltentpcreview article that also showed near ambient levels of cooling that were far lower than the ones supplied by Jolly Roger's iMac and G5 Tower. It wasn't my figures. Remember, I don't have temperature sensors in my Mini. But any idiot knows that the cooling is working if the exhausted air is cool.
Michelle, you seem exceedingly angry lately. Maybe some more cardio would relieve some of the angst?
The New Guy - 28 June 2007
I've never had to answer to anyone but the market, thankfully. And the market treats me well, so far. My results are good and people appreciate them when they open their eyes to new ideas. Sadly, few people do.
If you're worried about your job, then I guess you may design pessimistically. If you strive for high goals, you may risk more. It all depends on the job.
Well I guess nobody has replaced the Mini's heat sink so far........:)
The New Guy - 28 June 2007
The fact that you have the latest model is the reason you have those precious sensors. The PPC models, at least mine, didn't. The only sensor was in the hard drive, if it had one that is.
Please post the same figures if you ever pop off the top. I presume that was taken when idling?
And to those familiar with the Apple temperature sensors: If I used an infrared or laser thermometer on the surface of say, the heat sink (by the say, its interesting that they spell heat sink as one word on that screenshot), would I get the same reading as the Apple sensor would give me? I should hunt someone down nearby that has a newer Mac to try it. I'm going to get one of those nifty thermometers but it would be a drag if my readings were not consistent with what an onboard sensor would give.
The New Guy - 28 June 2007
I guess it depends on the consequences of failure. Sometimes its not a serious consideration. When you're dealing with something millions of miles away, it most definitely is. Ironically, the best cooling will help all those components last longer.......lol. :)
Jolly Roger - 28 June 2007
On 2007-06-28 20:23:09 -0500, The New Guy said:
any idiot knows that the
cooling is working if the exhausted air is cool.
No, cooling is working if the temperature of the components being cooled is low. The temperature of the exiting air is insignificant.
You have a lot to learn, my friend.
JR
Jolly Roger - 28 June 2007
This is why people here tend to think you are a know-it-all teenager. You have zero knowledge of engineering, you hold no degree, yet you claim to know why engineers design things the way they do.
JR
The New Guy - 28 June 2007
Well, that was polite at least.
Now would someone please enlighten me on how a component can be cooled at near ambient levels and the exhausting air is hot. This is supposing that we're in a normal enclosure that doesn't trap air of course.
The New Guy - 28 June 2007
Why not just respond in specifics. If you don't agree with something, stay on point. General insults don't do anyone any good.
Daniel Packman - 28 June 2007
Previously, The New Guy wrote:
...... Its about specific points. My experience gave me knowledge ...
Experience isn't getting an empirical idea of how hot exhaust air is. You need to measure the temperature of component of interest (eg, cpu) and find the operating range of said component. Then if make a modification (heat sink, fan, baffles, etc), then you need to report the resulting temperature of the component. Finally, if both the before and after temperature are within the stated operating range, you need to find some information to suggest that there is some value to your modification - perhaps the component life is lengthened. But you need to get some solid information suggesting that this increase in component life is significant in this applicatin. If the component normally lasts 10 years and your modification makes it last 20, it is probably a waste of time.
Jolly Roger - 28 June 2007
If you really want specifics, you'll take a course in thermodynamics.
JR
Michelle Steiner - 28 June 2007
Previously, The New Guy wrote:
Michelle, you seem exceedingly angry lately. Maybe some more cardio would relieve some of the angst?
You seem exceedingly smarmy and insulting lately. Maybe knowing what you're talking about would let you stop making a complete ass of yourself.
Support the troops: Bring them home ASAP.
Michelle Steiner - 28 June 2007
Previously, The New Guy wrote:
Now would someone please enlighten me on how a component can be cooled at near ambient levels and the exhausting air is hot.
The heat transfers from the component, heating the air, and the air then leaves the enclosure. That is a concept that you refuse to understand.
Support the troops: Bring them home ASAP.
John Byrns - 28 June 2007
I agree that it is the "temperature of the components being cooled" that is important, but I don't agree that the "temperature of the exiting air is insignificant", although "The New Guy's" conclusions about the exhausted air don't necessarily follow.
Regards,
John Byrns
Surf my web pages at, http://fmamradios.com/
Jolly Roger - 28 June 2007
On 2007-06-28 23:30:37 -0500, John Byrns said:
I don't agree that the "temperature of the exiting air is insignificant"
It's insignificant when you consider that the air flow rate is variable.
JR
Steven Fisher - 28 June 2007
If a statement can be proven true or false, sure. However, you've done nothing of the kind. For all you know, a larger heat sink will lower air circulation sufficiently to completely fry the CPU. You do know the heat has to go somewhere after the heatsink draws it away, right?
Replacing the heatsink. *YOU* go ahead risk your mini by doing it if you're brave - just don't assert that it will make things better when it isn't proven. I mean, you haven't even TRIED it on ONE computer yet, where the hell do you get off telling people to do it because it will have a huge, positive impact? At a minimum, it will void the warranty, so it's automatically bad advice on those grounds. Honestly, it's difficult to imagine someone who gives as little thought to things as you.
Andy - 28 June 2007
It really begs the question (and excuse my language if that sort of thing if frowned upon here) of why the fuck you think you're in a position to second guess the Apple ENGINEERS that designed and built the fucking things in the first place?
Personally, I can't remember reading anything about the mass-recall of Mac Minis that had died due to overheating issues. Can anyone point me towards a source that shows *any* Mac has overheating issues under normal conditions?
Sheesh...
Andy.
Gregory Weston - 29 June 2007
You may be worried about something other than your job, or indeed anything about _your_ life specifically. Engineers and architects (whether we're talking about computer systems or physical structures) are very often financially liable for the expenses incurred when the things they've created don't match what someone - often not them - has promised. Sometimes those "expenses" manifest as injury or death, which can be a crippling burden for anyone with a conscience.
If you strive for high goals, you may risk more. It all depends on the job.
True. If nobody's safety depends on you getting it right you can go pretty wild.
G
Gregory Weston - 29 June 2007
Yes, any idiot probably would assume that without bothering to verify.
G
Jolly Roger - 29 June 2007
It's also possible that the air flow was way faster than needed, which begs the question "How much power was the machine consuming at the time?".
JR
Michelle Steiner - 29 June 2007
Previously, Jolly Roger wrote:
It's also possible that the air flow was way faster than needed,
But was it faster than a unladen swallow?
Support the troops: Bring them home ASAP.
Jolly Roger - 29 June 2007
Would that be a "gulp"?
JR
Gregory Weston - 29 June 2007
African or European?
Jolly Roger - 29 June 2007
Does one swallow faster than the other?
JR
isw - 29 June 2007
It could also mean that the heat sink is packed with dust, or that it has become thermally disconnected from the chip it's supposed to cool.
The only *proper* test, is to measure the temperature of the chip you're trying to cool -- not the exiting air, not even the heat sink (which doesn't actually "sink" any heat at all, but rather just moves it from one place to another).
I don't play an engineer on TV, but I am one 8^}
Isaac
isw - 29 June 2007
It is impossible for the heat sink to be at a lower temperature than the exiting air that is removing its thermal energy.
Isaac
Jolly Roger - 29 June 2007
On 2007-06-29 11:19:51 -0500, isw said:
The only *proper* test, is to measure the temperature of the chip you're trying to cool
Exactly.
JR
Jolly Roger - 29 June 2007
True in the strict sense, but you're not taking into account that some systems funnel heated air from multiple components into a single air channel before final exit.
JR
The New Guy - 29 June 2007
My point all along is if the heat sink is properly designed for the task, it will hardly get warm. So it really can't heat the air as long as their is reasonable airflow aided by a fan. It all boils down to the job the heat sink is doing since we never operate in sealed boxes - there is always airflow.
Now if the heat sink is barely getting warm, it doesn't heat the air so the exiting air is near ambient temperature. So exhausted air temperatures are very indicative of cooling efficiency, hard as that may be to accept for some of you.
The New Guy - 29 June 2007
I think we all agree on that.
but I don't agree that the "temperature of the exiting air is insignificant", although "The New Guy's" conclusions about the exhausted air don't necessarily follow.
Explain please. What I'm curious about I guess is how something can be cooled to near ambient levels with the exhausted air being hot. Now you may not agree that cooling to near ambient levels is important. But that is not what we're discussing here.
The New Guy - 29 June 2007
Here again airflow is coming into play. Well, unless you don't care about noise, airflow is simply not a factor unless noise is not a factor. Its rather hard, in any design, to have high airflow without significant noise. So if you care about noise, you cannot depend on fans to take care of cooling that the heat sink is not doing.
William Mitchell - 29 June 2007
It seems to me that, to a first approximation, the temperature of the exausted air would only depend on the amount of heat being generated and the rate of air flow, and would be independent of the condition of the heat sink. If you do something to make the heat sink work poorly, then the temperature of the chip would increase until the temperature differential is large enough to transfer the same amount of heat as before.
Bill Mitchell
Dept of Mathematics, The University of Florida
PO Box 118105, Gainesville, FL 32611--8105 mitchell@math.ufl.edu (352) 392-0281 x284
The New Guy - 29 June 2007
So? Do it after your warranty expires. Or buy a used one. I would imagine that a sizable percentage of the readers here are using used computers. And anyone, how would it void the warranty? How would they even know? Have you ever dismantled a Mini? Its not very complicated. Also, the CPU heat sink is under a plastic housing so you could replace the plastic holders with metal ones like I did and they would never know. I doubt that they would dismantle the plastic housing when they're working on it if they don't have to. I mean, its not as if a lot of people are fooling with Mini heat sinks. In fact, its obvious that its rarely been done. Otherwise someone would have posted a link.
so it's automatically bad advice on those grounds. Honestly, it's difficult to imagine someone who gives as little thought to things as you.
Its difficult to imagine how someone can be so fearful of change.....:)
The New Guy - 29 June 2007
Have you ever heard of any product that could not be improved? I'm trying to improve one area of the Mini that bugs me. That's all. In fact, most inventions and improvements came about from precisely that reason.
Personally, I can't remember reading anything about the mass-recall of Mac Minis that had died due to overheating issues. Can anyone point me towards a source that shows *any* Mac has overheating issues under normal conditions?
That's not the point. The point is the machine uses a small fan and any small fan is noisy when spinning fast. If I could somehow funnel air from a large fan down to the heat sink that would be an improvement too. But I don't know how to do it. That, of course, wouldn't be nearly as good as just using a better heat sink in the first place.
The New Guy - 29 June 2007
Good points. But we're hardly talking about maiming and death here! But you do show why an engineer may design pessimistically as Jolly mentioned above. And really engineering is always a mixture of both. You want to push the envelope but have to refrain from risk. The engineering balance.
The New Guy - 29 June 2007
But that never happens! How could any case trap air inside? Did someone wrap a plastic bag around it? Airflow - air movement which is the opposite of "trapping the hot air inside".
Well this idiot has components running at near ambient levels in a near silent system. And it works.
The New Guy - 29 June 2007
Well that's quite impossible as I just mentioned in the previous thread.
It's also possible that the air flow was way faster than needed,
So its noisy. Here we go again: You cannot depend on airflow if you care about noise.
The New Guy - 29 June 2007
So then the person that applied the heat sink didn't know what they were doing and used deficient thermal compound. Once again, if you don't know how to apply thermal compound, you shouldn't be messing with your heat sink.
The only *proper* test, is to measure the temperature of the chip you're trying to cool -- not the exiting air, not even the heat sink (which doesn't actually "sink" any heat at all, but rather just moves it from one place to another).
I guess you haven't worked with heat sinks. If a heat sink is working well and is up for the job needed, it will hardly get warm to the touch. I guess a lot of you only have experience with really lousy heat sinks. Think of the example I gave many moons ago. 1: You hold a small piece of metal that warms to your touch. 2: You hold a large piece of metal that doesn't warm to your touch. #2 is the better heat sink. It doesn't get as warm. The better the heat sink, the cooler it runs. The cooler it runs, the lower the temperature of the exhausted air. Which is exactly what I've been reiterating from the beginning. Hopefully some of your eyes are being opened.
The New Guy - 29 June 2007
Yes, that's kind of obvious. Nowhere did I intimate that that was not true. But it IS very much possible for the heat sink to be NEAR ambient temperature levels. VERY near.
But this bring up an interesting question that you engineers can answer. When we go outside, the higher the wind, the lower the temperatures feels. Is it ever possible that a similar thing happens with inanimate objects? You take a heat sink with nothing heating it, then you blast it with, say, 70 F room temperature air. Can the temperature of the heat sink ever get lower than 70 F? It FEELS colder in a wind to us, but what about the actual temperature solid structures? I think someone mentioned before that we feel cool because of the rate of evaporation from our skin. I'm not sure if that applies here but of course solid structures have no evaporation rate.
The New Guy - 29 June 2007
And that design is going to be relegated to the scrap heap more and more, the Mac Pro case design being a good start.
The trouble is the shape of the typical motherboard. Instead of about 12" x 9" (108 square inches), it could be maybe 18" x 6 " or whatever the maximum depth needed for the longest PCI card. A large fan (250 mm or about 10") on the bottom would push air up and ductwork would conduct the air into each area. You could have separate channels for the CPU heat sink, Video card(s), ram and chipset(s). Then the hard drives (2 fixed for Raid 0 for the OS and Programs and others in removable hard drive drawers for data) could be mounted on the other side of the motherboard with a channel of air for them. But to reshape motherboard design......that would take a lot of convincing. Sort of like reinventing the wheel. If they just oriented the ram so it was in the same direction as the PCI slots, that would help a lot. The CPU channel would be still be at the far right, then maybe the ram and chipsets, then the PCI slots. With ram getting hotter and hotter, (note the heat sinks on most all high end ram), it just makes sense to cool it with some airflow. But not like the Mac Pro using the hot CPU heat sink air to cool the FB-Dimm ram heat sinks.
We may see some very different case design come about once 250 mm or larger fans become more common.
Grandpa - 29 June 2007
No they will not cool lower than 70 F. There is no evaporative action, which is what is happening to your stupid hand when you feel the exit air. This whole thing is how you FEEL about it. Check out some thermodynamic laws about heat transfer. Do you know "easy" it is to transfer heat from one object at 26C to ambient air of 25C? If the incoming air is 25C and the exit air is 25C just how much heat is being removed from the system?
Grandpa
Tom Stiller - 29 June 2007
No; the cooling effect you observe under those conditions is mostly due to evaporation of the moisture on the skin. Inanimate object do not perspire. The heat of vaporization of water is approximately 539 cal/gram.
Even if such objects did perspire, no amount of airflow would cool them below the temperature of the air passing over them; the moisture would just not vaporize. You would do well learn a little about partial pressures and thermodynamic equilibrium.
Jolly Roger - 29 June 2007
On 2007-06-29 12:25:17 -0500, The New Guy said:
Have you ever heard of any product that could not be improved? I'm trying to improve one area of the Mini that bugs me. That's all.
What area would that be? I have yet to see a shred of proof from you that the Mac mini runs too hot.
JR
The New Guy - 29 June 2007
I'll admit I'm much more sensitive to noise than most people. Its not the heat issue that prompted this original post though. It was the noise. Of course that is related to the heat.
Jolly Roger - 29 June 2007
I barely hear any of my three Mac mini fans, even when I'm pushing them. Then again I live in a big city.
JR
The New Guy - 29 June 2007
When my CPU use is above about 75% they ramp up. Are you using Intel Minis? Maybe they are quieter?
Methinks an Intel Mini is something I should get soon. So many advantages: 667mhz ram speed, double the ram capacity, full size Sata hard drive use without an adaptor, the ability to upgrade the CPU, the ability to run a 22" widescreen (1680 x 1050) which is currently the most bang for the buck.....the list just grows. Especially attractive with the recent crash in ram prices. But it looks like you can't add a high gain receiving antenna to the built in wireless. Is that correct? With the portability factor, that is especially interesting.
Bob Carter - 29 June 2007
Yes, will do. As a matter of fact I plan to pop the hood in the very near future in order to add more RAM.
I presume that was taken when idling?
Yes; the temperature doubles when pegged at 100%. I'd like to resume SETI@home on it, but 81-88 C cannot be good for the CPU. I will be looking into "non-invasive" cooling solutions in order to avoid voiding the warranty...
John Byrns - 29 June 2007
The fan in my Mac mini is on and making noise any time the computer is not sleeping, I would be happier if the fan only came on when the CPU is under stress as with my iBook book.
Regards,
John Byrns
Surf my web pages at, http://fmamradios.com/
Matthew T. Russotto - 29 June 2007
Previously, Grandpa wrote:
Just how long did the engineers say that the Mars rovers were going to last? And how long have they lasted? Would you say that they were "optimistic?"
Without knowing the confidence values and the distribution of failure, I don't know. If the numbers usually bandied about for the "design lifetime" were actually, e.g., the expected minimum lifetime with a confidence of 95%, and the chance of failure follows some distribution characterized by high infant mortality (quite likely!), then the usual case would be lifetimes much longer than the design lifetime.
However, this is off the topic of whether "The New Guy" is suffering from Dunning-Kruger... I'd give way better than 95% confidence for a YES on that one :-)
There's no such thing as a free lunch, but certain accounting practices can result in a fully-depreciated one.
Matthew T. Russotto - 29 June 2007
Free clue: No one cools a component to "near ambient levels". It's both impractical and unnecessary.
There's no such thing as a free lunch, but certain accounting practices can result in a fully-depreciated one.
Matthew T. Russotto - 29 June 2007
You seem to be under the misconception that the heat sink is doing any cooling at all. It's not. The heat sink is being passively heated by the component. The airflow is effectively doing ALL the cooling (radiation to the environment is not significant); the heat sink itself is just a middleman.
There's no such thing as a free lunch, but certain accounting practices can result in a fully-depreciated one.
Andy - 29 June 2007
The Mac Minis I've seen have been near enough to silent for my tastes.
Then again, I own a Dual G4 MDD "Wind-tunnel" model _without_ the PS/Fan upgrades and don't really find it a problem.
Andy.
Daniel Packman - 29 June 2007
Previously, Andy <nospam@no.no> wrote:
.....
The Mac Minis I've seen have been near enough to silent for my tastes.
Then again, I own a Dual G4 MDD "Wind-tunnel" model _without_ the PS/Fan upgrades and don't really find it a problem.
HELLO? WHAT? COULD YOU SAY THAT AGAIN? :-)
Michelle Steiner - 29 June 2007
Well, when you become competent in the area, then perhaps you will have something constructive to add.
Support the troops: Bring them home ASAP.
Michelle Steiner - 29 June 2007
Previously, The New Guy wrote:
Methinks an Intel Mini is something I should get soon.
Methinks a Dell, Gateway, HP, or Compaq is something you should get soon, so you would bug their newsgroups instead of this one.
Support the troops: Bring them home ASAP.
Michelle Steiner - 29 June 2007
Previously, The New Guy wrote:
When we go outside, the higher the wind, the lower the temperatures feels.
Not always. Try going out in the 110 (+/- 5) degree temperature here in Phoenix on a windy day. The temperature feels hotter than if there were no wind.
Support the troops: Bring them home ASAP.
Michelle Steiner - 29 June 2007
Previously, The New Guy wrote:
I guess you haven't worked with heat sinks.
That's your stock bullshit answer, if someone disagrees with you, you tell them that they have no hands-on, real-world, etc., experience.
Support the troops: Bring them home ASAP.
Michelle Steiner - 29 June 2007
Or does it spit?
Support the troops: Bring them home ASAP.
Dave Balderstone - 29 June 2007
I recall Orlando at 100F, 98% humidity, and smoke from forest fires covering the city.
No amount of wind could have made a difference.
Same in Nawlins... And Atlanta...
isw - 29 June 2007
Well, no. Actually, the heat generated by the underlying device is constant. It is coupled by the heatsink to the passing air no matter what the heatsink's temperature is -- a smaller sink will necessitate a greater temperature above that of the air, however. The temperature rise of the air is determined by how much airflow there is, and how much power (watts) it is carrying away, and has nothing at all to do with how large the heatsink is, or its temperature.
Bottom line: the rise in temperature of the air in passing through the unit is dependent *solely* on the total amount of heat generated inside (in watts, say), and is *not related* directly to the temperature of any particular heatsink which may be inside. Applied to a given device, a small, hot heatsink or a large, warm one, would result in exactly the same exit air temperature.
Which is exactly what I've been reiterating from the beginning.
I'd suggest you "guess" again (to quote something I read recently).
Isaac
Jolly Roger - 29 June 2007
On 2007-06-30 00:08:26 -0500, isw said:
The temperature rise
of the air is determined by how much airflow there is
This is what I've been telling The New Guy all along. Good luck with getting him to believe it...
JR
Jolly Roger - 29 June 2007
On 2007-06-29 18:56:01 -0500, The New Guy said:
When my CPU use is above about 75% they ramp up. Are you using Intel Minis? Maybe they are quieter?
I have a G4 mini and two Intel minis. None of them seem excessively loud. Then again my office environment is not so quiet I can hear a pin drop anyway.
JR
Steven Fisher - 29 June 2007
Previously, The New Guy wrote:
If you review this thread, I never told anyone to do it. I asked if it had been done.
On the contrary: You recommended it to someone in the third post on it. Thus, the storm.
Steven Fisher - 29 June 2007
Previously, John Byrns wrote:
The fan in my Mac mini is on and making noise any time the computer is not sleeping, I would be happier if the fan only came on when the CPU is under stress as with my iBook book.
There's a small cable in your mini that you or whoever worked on your mini forgot to plug back in. Without that cable, the fans run at full.
Jolly Roger - 30 June 2007
Same here. In fact the ATI Radeon X850 XT in my G5 tower makes *way* more racket with it's stupid noisy little fan than *any* of Apple's fans.
JR
Bjarne Bäckström - 30 June 2007
"Gutta cavat lapidem..." doesn't work on this stone, as it seems.
Gregory Weston - 30 June 2007
I didn't see the previous thread, but it's certainly not impossible in the general case. What makes it impossible in yours?
And consuming extra power and, ironically, generating excess heat.
G
Gregory Weston - 30 June 2007
You do know that last thing wasn't a sentence, right? It didn't express a complete thought, so I'm not really sure what you intended to say.
And you know they're near ambient how? I thought you said you didn't measure anything.
Michelle Steiner - 30 June 2007
Why? It doesn't substitute for "real world experience". If you don't believe me, just ask New Guy; he'll tell you.
He knows that they're near ambient because the exhaust air is near ambient.
Support the troops: Bring them home ASAP.
John Byrns - 30 June 2007
Well then blame Steve Jobs and the Apple factory for that, my Mac mini has never been worked on by me or anyone else since it first left the Apple store, I don't even have a clue how to open the box. Speaking of that how do you open the box, I assume it takes some kind of special tool? I would like to add more memory and an airport card as I did with my iBook.
Regards,
John Byrns
Surf my web pages at, http://fmamradios.com/
Steven Fisher - 30 June 2007
<shrug> It happens sometimes. Don't worry about it.
If you get a local dealer to install the memory, they'll be able to reconnect the cable at the same time. Getting a mini's case open is only the first problem in working on it.
Jolly Roger - 30 June 2007
Technically, he *feels* the exhaust air is near ambient.
JR
Jolly Roger - 30 June 2007
On 2007-06-30 09:46:04 -0500, John Byrns said:
Speaking of
that how do you open the box, I assume it takes some kind of special tool? I would like to add more memory and an airport card as I did with my iBook.
You can find various web sites on the web with detailed instructions showing how to take Mac minis apart. Google is your friend. Here's one:
<http://www.applefritter.com/Mac_Mini_Take_Apart_Guide>
JR
Jolly Roger - 30 June 2007
There's a video here as well:
<http://www.smashsworld.com/2005/01/taking-apart-mac-mini-how-to.php>
JR
Michelle Steiner - 30 June 2007
You getting new agey on JR?
Support the troops: Bring them home ASAP.
Jolly Roger - 1 July 2007
Hey I can be sensitive.
JR
The New Guy - 1 July 2007
Great.
I wonder if there is a water block that is small enough to fit on that CPU? For the 1.25 ghz PPC it would need to be 47 mm x 118 mm. The 118 shouldn't be a problem. The 47 might very well be.
The New Guy - 1 July 2007
The fan in my Mac mini is on and making noise any time the computer is not sleeping, I would be happier if the fan only came on when the CPU is under stress as with my iBook book.
For that to happen you would need a massive heat sink. Maybe almost as large as the entire Mini!....:)
The New Guy - 1 July 2007
Free clue: I do. Its done quite easily using exterior air and venting it immediately so the heat doesn't rise in the enclosure. Now whether you believe its necessary, that up to you. But its very practical indeed.
The New Guy - 1 July 2007
I think we're a little beyond these semantics. The point is you need a large, well designed heat sink to get that heat away from the heat generator. The better the heat sink, the less airflow you need. Now it just happens that 120 mm and larger fans fit nicely on larger heat sinks. So now many users are finding they can run that 120 mm fan at inaudible speeds and still get good cooling.
The New Guy - 1 July 2007
Reminds me of when I was driving through the desert in the evening and the temperature was over 100 F. You're right, it did feel like it was heating my skin. I guess that was because it was above body temperature. It sure felt eerie the first time.
The New Guy - 1 July 2007
Hence almost all high performance are very large.
The temperature rise
of the air is determined by how much airflow there is, and how much power (watts) it is carrying away, and has nothing at all to do with how large the heatsink is, or its temperature.
Well just by chance, all high performance heat sinks are very large. Coincidence?
The New Guy - 1 July 2007
But when the Mini runs hard doing some intensive task, isn't that fan whine a little grating on the ears? And wouldn't it be nice if the fan whine was at a lower frequency? That's sure what I would love. There are so many advantages to the Intel Mini, so if I could solve this little pesky issue, it would be grand.
The only negative to the Intel Mini that I can think of so far is that I can't hook up a second hard drive like I can on the PPC Mini because my 2.5" > 3.5" IDE adaptor enables 2 devices on that channel. Unless someone has rigged up a 2.5" optical IDE to 3.5" IDE adaptor? Now if the Mini used a Sata optical drive, it might be easier. But unfortunately its still IDE.
The New Guy - 1 July 2007
Same here. In fact the ATI Radeon X850 XT in my G5 tower makes *way* more racket with it's stupid noisy little fan than *any* of Apple's fans.
And ironically the Mini uses that same size noisy fan. If the heat sink was larger, one could use a larger fan. Same airflow with less rpm's. Noise problem solved.
You may notice on some high end video card heat sinks, they are positioning the heat sink "on top" of the card so its not in between the PCI cards. Then you can use a 120 mm fan and keep the noise levels down.
Jolly Roger - 1 July 2007
The Mac mini uses the same exact fan as the Radeon X850 XT? Citation please? If that's true, the mini's is somehow *way* quieter, because I never hear it.
If the heat
sink was larger, one could use a larger fan. Same airflow with less rpm's. Noise problem solved.
There is no noise problem here - and that's with three different minis. Even when I put the CPU through its paces there's no noticeable noise.
JR
John Byrns - 1 July 2007
Then why does the Book fan only run when the CPU is stressed? Are you saying the iBook has "a massive heat sink" maybe even as large as an "entire Mini!"?
Regards,
John Byrns
Surf my web pages at, http://fmamradios.com/
isw - 1 July 2007
That, or any other sort of "heat pipe", doesn't get rid of any heat at all; it just moves it from one place to another with good efficiency. Heat pipes have been used in laptops for a long time.
If you really want to stick a large radiating element on your mini, a surplus heat pipe might be a way to go.
Isaac
Tim Streater - 1 July 2007
This is untrue and underpins your lack of understanding of this subject.
Now
it just happens that 120 mm and larger fans fit nicely on larger heat sinks. So now many users are finding they can run that 120 mm fan at inaudible speeds and still get good cooling.
Tim Streater - 1 July 2007
Correct. The closer to ambient, the harder the cooling process becomes. Exponentially harder.
Tim Streater - 1 July 2007
I'd just say he's just a thick cunt who is unaware that he is. Unless he's MC in disguise.
Tim Streater - 1 July 2007
[snip]
This is one of the bits Mr NG doesn't get.
Michelle Steiner - 1 July 2007
It's not semantics; it's basic concept. Until you understand basic concepts, you will not be able to have a meaningful discussion on the topic, and will continue to post the nonsense that you continually write.
Support the troops: Bring them home ASAP.
The New Guy - 1 July 2007
It runs from the motherboard to the daughtercard. But I'm guessing his machine is running normally and like me, is sensitive to noise, plus he's probably in a quiet environment so he hears it more easily.
I've been experimenting with putting noise generators in a column. And it does work well. Remember high frequencies don't go around corners like bass does - directional is a more accurate term. If you have some milk crates you can try this. It won't win any Architectual Digest design awards but it does seem to work. Place your Mini vertically with the bottom facing the interior, and then stack 2 or more milk crates above and below it. Then wrap something soft like thick blankets or a soft lined sleeping bag around the column. Leave a few inches at the bottom for air to come in. You should hear a dramatic reduction in noise. You'll need longer wires for this to work or you could use 2 blankets with the wires coming out. This is not practical but its easy to do if you have milk crates or something else that allows air to pass vertically upwards. The downside to this is there are still hard surfaces for sound to reflect off of inside the milk crates, some more than others depending on the design. Ideally the inside of the milk crates should be lined with sound absorbing material. Closed cell foam might be good. I was trying to research what commonly available materials absorb higher frequencies but its not easy to find. If anyone has any tips, that would be great.
The New Guy - 1 July 2007
Some people use putty knives but most are not the full width to encompass all the clips. Anything 4" or more wide should do it. I used what looks like a putty knife but really was designed for mortar work I think. Work slow and be very patient. Lots of tutorials online relating to upgrading the Mini's memory.
The New Guy - 1 July 2007
Once you get it open, you can compare your layout with the pictures online. You'll see 2 black wires wrapped together that go to the daughtercard (where the 2.5" hard drive and optical drive plugs into) from the motherboard. This controls the fans otherwise they'll be going full bore all the time.
The New Guy - 1 July 2007
No. I said same size. But that could have been clearer on my part.
I wonder if there's much difference between the newest Intel Mini, the first Intel Mini, and the PPC Mini. Sounds like there is. The CPU on the Mini - is that a notebook style CPU? Or a desktop style? Or is there even much of a difference anymore?
The New Guy - 1 July 2007
I know nothing about notebooks. Somebody else can answer that I'm sure. But we all know how hot some notebooks get. Its hard cooling something like that.
The New Guy - 1 July 2007
Yes, I was looking at some heat sinks and realized that the heat pipe design, because it elevates the heat sink off the motherboard, might allow a mammoth heat sink to be installed if only the mounting area is small enough. Often the manufacturer never mentions what those dimensions are. Its just not something that people need. They look for Socket 775, AMD, etc.
The New Guy - 1 July 2007
Tim, what exactly was untrue about what I said?
Jolly Roger - 1 July 2007
I have the PPC mini and two Intel minis. None of them are very loud.
The CPU on
the Mini - is that a notebook style CPU? Or a desktop style? Or is there even much of a difference anymore?
The G4 mini has a 1.42 GHz PowerPC 7447a (G4) processor:
JR
Daniel Packman - 1 July 2007
Previously, The New Guy <replytogroup@here.thanks> wrote:
Post your measurements of the component and temperature
of exhaust air. Measure this for several values of airflow.
Matthew T. Russotto - 1 July 2007
Q = -hA(Ts - T)
There's no such thing as a free lunch, but certain accounting practices can result in a fully-depreciated one.
The New Guy - 2 July 2007
But is the CPU on the Intel Mini a notebook style?
If it is, the cost should be higher to upgrade or replace than if it was a desktop style. That is if there is even a difference. Hopefully someone knows this stuff. I sure don't.
The New Guy - 2 July 2007
I'm shopping around for an infrared type thermometer.
But I'll have to test it with a machine that has built in sensors to make sure its calibrated correctly. Unless you can look it up, the sensors are not exactly obvious on the motherboard, are they? And would the temp of the motherboard itself influence the ambient air reading? Sort of like mounting a thermometer outside on a surface that gets hot or cold. Its probably going to influence the reading. I hope to avoid inaccuracies like that.
Jolly Roger - 2 July 2007
Huh? There's no "notebook style" 7448. It is what it is.
JR
The New Guy - 2 July 2007
That brought me here:
http://biocab.org/Heat_Transfer.html A nice summary of correct terminology to help express myself better. Thanks.
isw - 2 July 2007
I can't say about the mini and its CPU in particular, but it's common for the temperature sensor to be built into the CPU chip itself. It's just another transistor, wired in a special way.
You can find little apps that can give you a readout from it.
And remember, it will be measuring the temperature of the active surface of the processor directly, so it'll be considerably warmer than any external surface you can stick a thermometer on.
Isaac
The New Guy - 2 July 2007
Thanks Isaac.
Tim Streater - 2 July 2007
Isaac said, correctly, that the heat sink is providing an impedance matching between the chip (which is small) and the air (which ultimately carries the heat away). As with any impedance matching device, it has an optimum size. For a given chip producing a certain amount of heat, and a given airflow, there will be more or less one size of heat sink that is correct. Size in this case may simply be nothing more than the surface area due to it having lots of fins.
You vary one of the parameters, such as heat produced, and if you want to remain at equilibrium (i.e. constant chip temperature), you better also vary the other parameter you have control over, namely airflow. So we see what is observed, namely the fan speeds up. You optimise the heat-sink size to maximise the heat transfer from the chip to the heat sink, and then to the air. It doesn't necessarily lead automatically to the conclusion that the heat sink is large.
Jolly Roger - 2 July 2007
Many of us have tried, in vein, to explain this to The New Guy repeatedly. He refuses to believe it. Good luck with that.
JR
Michelle Steiner - 2 July 2007
Previously, Jolly Roger wrote:
Many of us have tried, in vein, to explain this to The New Guy repeatedly.
<grammar/spelling nazi>
"in vain"
<end nazi>
Support the troops: Bring them home ASAP.
Kurt Ullman - 2 July 2007
Maybe that meant they got so frustrated they cut themselves..
Jolly Roger - 2 July 2007
Good catch. Thanks.
JR
Michelle Steiner - 2 July 2007
I can see how you made that error; he is, after all, a blood sucker--metaphorically speaking, of course.
Support the troops: Bring them home ASAP.
The New Guy - 2 July 2007
But the reason the manufacturer wants to use smaller heat sinks is to save material costs. Aluminum and copper are very expensive.
You vary one of the parameters, such as heat produced, and if you want to remain at equilibrium (i.e. constant chip temperature), you better also vary the other parameter you have control over, namely airflow. So we see what is observed, namely the fan speeds up. You optimise the heat-sink size to maximise the heat transfer from the chip to the heat sink, and then to the air. It doesn't necessarily lead automatically to the conclusion that the heat sink is large.
Name one high end heat sink that is small then.
You cannot speed a fan up much and retain a quiet system. Its all in the heat sink - not in the fan. Unless you don't mind a loud system. Few people do though. A few years ago people had lower standards. Those noise standards are getting a little higher now that people realize they don't have to put up with that anymore.
Tim Streater - 2 July 2007
No, the optimum size is that which, for the given conditions, transfers the maximum heat to the cooling stream.
The New Guy - 3 July 2007
The maximum heat transferred is always going to come about using a large heat sink though, right? I noticed you didn't bother answering a question that I asked you twice......namely, "Name a high end heat sink that is small."
Some of you think I'm talking without thinking, yet when it comes to specifics, you are silent.
Name a high end heat sink that is small!
Anyway, its obvious that very few if any people have ever replaced the Mini heat sink. Back to some intensive Googling. And thanks to those of you that responded with specifics.
Tim Streater - 3 July 2007
Not necessarily, as we all keep pointing out. Probably you would have to use calculus or numerical methods to calculate the optimum size.
I noticed you didn't bother answering
a question that I asked you twice......namely, "Name a high end heat sink that is small."
Of course not. Why should I know anything about specific heat sinks made by particular manufacturers? Or care, more to the point? And who says that any such heat sink bought pre-designed, is going to be the best for a particular application? I'd rather have the one designed for the situation. What these and previous threads have been trying, with little evident success, to dun into your head, are the principles involved.
Another of the constraints is bound to be cost, something else you seem to poo-poo. One would make the bulk of the heat sink out of aluminium, to reduce cost and weight, and have a copper plug in it where it meets the chip. The plug itself would need to be designed, again to provide a good impedance match with the aluminium. The impedance mismatch is due to the different thermal conductivities of copper and aluminium.
Gregory Weston - 3 July 2007
Not necessarily. Other factors come into play, and "large" is ambiguous. All else being equal, increased surface area will generally help.
G
Tom Stiller - 3 July 2007
Right, but with properly designed fins. It does no good to have large fins with insufficient cross section area to accommodate the heat flow at any point along the its height.
isw - 3 July 2007
Nope; the amount of heat transferred is always the same; only the temperature is different.
I noticed you didn't bother answering
a question that I asked you twice......namely, "Name a high end heat sink that is small."
"High end" heat sinks are never small because they are after-market items, bought mostly by individuals for whom the clue-meter reads very low. IOW, they're big because that's what people who know nothing about proper device cooling will pay for. That, and the "mine's bigger than yours" strategy of marketing.
Compare to "high-end" audio cables at a few dozen dollars a meter -- they're not better because they're bigger or more expensive.
--
There are some fairly subtle reasons why very large heat sinks might be useful to folks who want to overclock processors "to the max", but those don't apply to *properly designed* commercial gear.
Isaac
isw - 3 July 2007
Actually, thermal calculations are similar to the ones used to calculate series resistors. IOW, really simple.
1) From the chip's data sheet, determine the "thermal resistance" from the chip (it's called the "junction" for historical reasons) to the package (called the "case") -- it's called "theta j-c", and it's expressed in degrees per watt.
2) Use the data sheet for the heat sink grease (or pad) to determine theta c-s (case to sink, again in degrees per watt).
3) Use the heat sink data sheet to determine theta s-a (sink to air, degrees per watt, and it's given for still air and for a range of flows). For a custom sink, you can easily measure theta s-a.
4) Simply *add* theta j-c, theta c-s, and theta s-a. That gives you theta j-a, the temperature difference per watt transported between the junction (the chip) and the air, for that specific heat sink, installation method, and air flow.
5) From the chip data sheet, look up the power consumed by the chip (in watts), or measure it.
6) Multiply that number by theta j-a. That gives you the actual temperature difference between the chip and the air when transporting the specified number of watts.
7) Take the maximum air temperature you expect (you *do* know where this thing is going to be used, don't you?), and add to it the temperature difference determined in (6). That tells you the chip (junction) temperature for that air temperature.
8) Use the device data sheet to determine the maximum allowable operating temperature for the chip. Compare to the number you calculated in (7). If (7) is lower, you can go with it, or you can consider a smaller heat sink, a smaller fan, or any similar combination. If the junction temperature is higher than specified, you have a different set of choices, involving a possibly reduced functional lifetime, or a better sink or fan. Or you can change the operating parameters of the device (e.g. reduce the clock rate) to reduce the amount of heat generated.
--
Now you know more about thermal calculations than most of the newly-minted engineers I've worked with during my career.
Isaac
The New Guy - 3 July 2007
Many audiophiles swear that changing the cables was one of the best investments they ever made in their system. I haven't experimented with audio cables, so I have no comment. Let's just keep this in the computer world, ok?
There are some fairly subtle reasons why very large heat sinks might be useful to folks who want to overclock processors "to the max", but those don't apply to *properly designed* commercial gear.
It all boils down to what temperature you will put up with. I've found ways of getting much lower temps with non-standard techniques while keeping the noise floor very low.
The New Guy - 3 July 2007
When something barely fits into some cases, its large. Its rather hard to dramatically increase surface area and keep the overall dimensions the same.
The New Guy - 3 July 2007
Finally the truth comes out. You're simply not familiar with today's heat sinks. See what I mean? This is typical of the lack of real world hardware experience here.
Have a look at the websites of Thermaltake and Scythe, two of the most respected names in the industry. And then show me a high end heat sink that isn't large.
Tim Streater - 3 July 2007
You really don't have a fucking clue, do you?
I know about heat sinks to the same extent that I know about how cars work. As someone with a physics degree, and from Imperial College at that, I can be relied upon to understand the basics of heat sinks. Friend Isaac obviously has a better detail knowledge of the theory, but then I never studied it from an engineering perspective, just from one of heat flow, and what heat is.
Equally, I know the basic principles of how car engines work. But that doesn't mean I know how to design them. When I buy a car, I might well match it to the purpose - a small car for commuting, possibly a slightly larger one if I'm a commercial traveller. But I don't buy a Range Rover unless I'm a farmer.
You give the impression of someone who's heard that you put fuel in one end of the car, and it makes the wheels turn. Then you come along and say "Gee, I've got this fluid called water, a whole lot cheaper, why don't we run the car on that instead?".
That the Mac Pro's heat sinks are large is not at all interesting. They're just the end result of the design.
And by the way, increasing the surface area of something within the same dimensions is easy. You might try googling for "fractal solids s
