Doctor: You shall take all the pills I prescribe. The red one in the morning with a glass of water, the green one in the noon with a glass of water, the blue one late in the afternoon with a glass of water.
Patient: Three different pills each day! What I am suffering from Doctor?
Doctor: You don’t drink enough water.
George
Your point is certainly not missed by me.
Jan, is there any interest in a thermal analysis paper?
jn
Quote from page 342 of Vision of the Brain: Semir Zeki: 9780632030545: Amazon.com: Books (1993)
“We have already alluded to the visual normality of most dreams and hallucinations, to the fact that they are integrated, centrally generated, images which must somehow be re-entered into the cortex as if they were coming from the outside. If dreams and hallucinations depend upon re-entry into V1 or V2, the above assumption can be tested, at least in patients with a damaged V1 and V2, or so one would imagine. Unfortunatelly the task is not that simple and direct proof of the supposition is in fact lacking.
…
That hallucinations occur in subjects who are completely blind tells us, of course, that these are internally generated images which are more or less normal and depend upon the integrity of at least some of the visual areas….”
George
http://www.aavid.com/sites/default/files/news/Aavid-Liquid-Cooling-Advances-CIPS-2012.pdf
>Edit
http://www.google.gr/url?q=http://escher.elis.ugent.be/_p_/doc.php%3Ffile%3DP107_031.pdf&sa=U&ei=lnhsUfuxMM3MPfb9gdgB&ved=0CB0QFjAB&usg=AFQjCNH04AfbxTBU9YP7Vy5N0kQ_XCyukQ
(Fig.2 is what you were explaining to me some time ago)
George
Last edited:
+1
(thanks to Mr Eddy, I now have visions of waking up every morning, due to Çher's body parts falling down on me)
Nice.
I've pretty much used all that stuff. Recently, I used the copper press paks for cooling 76 mm diameter diodes for a burn in rack. I'm trying to remember, I think I used 2000 amps for that rack, burned in 20 diodes at a time in series.. I've also done the power module thingy, mostly from Powerex. Mil power hybrids were my teething rings, so to speak. Lots of bare dice, with either gold/silicon eutectic scrub attach, or 50/50 lead indium. On moly tabs, alumina, beryllia, or direct to a copper package. The DBC stuff is really interesting, direct bond copper to alumina..copper oxide melts 2 degrees C below pure copper, man that process must be a bear to control..
That aavid thing's a nice synopsis, but doesn't delve into what DIY'ers would need to know.
I'd think a theoretical analysis showing how die size, package thickness, attach techniques, etc would alter dissipation and derating. And typical issue surrounding methods, such as flatness of surfaces etc..
Tis simple enough to put a spreadsheet together where the geometry is the defining inputs, the output being either temp rise or theta jc.
Also, some kind of analysis for leaded devices..most people believe axial resistors and diodes dissipate via air instead of the leads.
jn
edit...I don't know, that second paper looks kinda complex..
Last edited:
Yes jn, I wish you write an article.
edit...I don't know, that second paper looks kinda complex..
The second paper was for to show the heat spread angle and to help JC on the calculations you asked him to do.
>Edit. Eq. (7), (8), (9) is what is essential to look into.
alfa and gamma are given by eq.(3) and (6) respectively.
George
Last edited:
Us humans are lucky enough that our hearing mechanism is smarter and more adept than the acoustical people give credit to -- if you give the ear/brain thing enough clues it can work what it's supposed to hear, separate the wheat from the chaff. However, this can only hapen if the low level information is not clogged with spurious muck, subtle distortion that intermodulates with key "inner" detail and makes it too difficult for the mind to work out what's going on. The majority of audio systems don't work well enough to give the fine discerning capabilities of our hearing a chance to do this ... hence, "hifi" sound ...
I support you on that. E.g., why does the copper tube have to be polished?
edit: I see it is on thermal management, the paper, but the question still stands.
Last edited:
Something on magnetic field theory as applied to practical applications like hi-fi for mortals like me would also not go amiss.
Hi everyone, I hate to say it again, but I was told once, many decades ago, that at least some semiconductor manufacturers used freon baths to test and determine temperature sensitive characteristics of at least some semiconductors. I never saw this in action, but it made sense. A cooled conductive fluid, continuously bathed over a particular device could approximate an 'infinite' heatsink.
Now, what can we do with to TO-92? I need to dissipate 50ma at 30V or 1.5W per unit, to make an effective line stage. Can I do this with device? If not, what is the alternative?
Now, what can we do with to TO-92? I need to dissipate 50ma at 30V or 1.5W per unit, to make an effective line stage. Can I do this with device? If not, what is the alternative?
Joachim and Kevin, I know that you mean well, and in a realistic situation, your suggestions would be very useful, but not in this case. You see, if you try to get 1 1/2 W from a 1W device, that is specified as 1W with an INFINITE HEATSINK AT 25 degrees C, then you are going to have a big problem just KEEPING the outside case of a TO-92 device at perhaps 0 degrees C, in order to keep the internal temperature that the chip operates at, within worst case limits. In fact, even IF you could, you would soon lose the device after a limited time.
Sorry everyone, that I can't do a better show, but please look at this set of pages from a 1983 (you know, when the schematic was dated) datasheet from the EXACT device originally specified for the JC-80, some 30 years ago. Now look at one particular graph which shows the MAXIMUM dissipation with different packages, with temperature, it is called: 'Power Dissipation Vs Case Temperature'.
Now, please realize that although the T0-220 package was used by the JC-80, the TO-92 device is about the only one left for sale from Supertex Inc, today.
Now everybody get a cup of coffee or tea in front of you, then study the graph carefully, and come to realize that it is almost impossible to make the T0-92 package do the job without overheating within the first few seconds.
Now, PLEASE, just read the graphs and come to understand what they state. It is a good engineering skill to learn, if you haven't already.
Now, please realize that although the T0-220 package was used by the JC-80, the TO-92 device is about the only one left for sale from Supertex Inc, today.
Now everybody get a cup of coffee or tea in front of you, then study the graph carefully, and come to realize that it is almost impossible to make the T0-92 package do the job without overheating within the first few seconds.
Now, PLEASE, just read the graphs and come to understand what they state. It is a good engineering skill to learn, if you haven't already.
- Status
- Not open for further replies.