IR kitchen thermometers are cheap and remarkably useful to have on your bench.
Has anyone done any testing on the cheap FLIR units. Low res but given you can get the CAT S60 phone with one built in seems a boon.
Rugged Phones & Accessories | Cat phones US
Rugged Phones & Accessories | Cat phones US
Thanks. The derating curve is based on case temp, not junction temp, so just need to read the case temp to a reasonable degree of accuracy.
The devices will be dissipating about 3.75 watts each so according to the derating curve, I can get by with just a bit over 100°C but I'm going put the limit at 100°C for just a bit more margin.
I'm well within that at normal room temperature (25°C) but I'm wanting to make sure they don't exceed 100°C when the ambient temperature is 50°C. That's not a likely scenario in the home, but just in case.
se
What you have been saying and depicting is totally incorrect as an engineering model as well as a physics model. You have to get the concept correctly first, to neglect the correct concept means that any "engineering model" you try to build with an erroneous concept will also be erroneous.
Instead of abandoning all physics to try to create a new "model", why not just use the exponential skin depth approximation equation?? At audio frequencies with a 1 mm diameter wire, it is only a factor of 3 in error at 20 Khz. Given the cross section and resistivity of a 1mm diameter wire, the actual error is of no consequence.
That is why the exponential skin depth approximation equation was developed and used. It gets away from the vessels, is easy to use, and normal people can understand the concept. The fact that some, like Hawksford, try to use the approximation equation in an application which requires more accuracy, is par for the course but must be viewed with skepticism. Using an approximation carries with it the chance of ruining the analysis due to errors.
John
The outside case temperature will not be very good. Does the tc have to be isolated from the collector? The collector terminal right where it exits will probably be the closest reading to the junction temperature. I don't know how much the lateral spread of heat of the case encapsulation will pull the temp down at the surface. In addition, if there is any glob top on the chip surface to protect the wire bonds from the encapsulation TCE, that will act as a thermal insulator.
The collector terminal is a part of the lead frame, and the chip is soldered directly to it. I would try there first. If it shows the highest temperature there, bingo.
If you really can only go to the top, use a thermal epoxy to bond the tc halfway between the screw mount and the lead edge of the plastic, then put a dab of foam over that to isolate it from the air. That will be about the closest to the die center as you're going to get.
John
Maximum junction temp is 150°C, but derating is based on case temp. The junction to case temp derates at 10°C/W. So I just need the case temp, and I don't think the case temp will vary by a huge amount from one point to another.
se
I am familiar with it. It works fine.
We also need models which we can use to explain technical/scientific concepts to the general public in as few words as possible to get a concept to them without being misleading in any substantial way. In other words.... marketing. I read in forums here many time about the marketing prose to non-scientific people and how crazed it sounds to more learned people.
How would you explain skin effect to a consumer with no science background.... in your words? In one or two sentences. [without equations]
But also oxidation and why copper purity and gold plating and tight connnections helps them get and keep better results. All sorts of other things actually. Its no wonder many audio/video equipment/cable marketing seems so off-beat.... I mean excluding outright lies.
THank you,
Richard
Last edited:
In one or two sentences? 'Don't believe the snake oil salesmen, you really don't need to worry about it in your hifi' would seem to work?
Gold plating generally requires a nickel layer which add iron, which we all know is bad, unless its good. Unplug and replug now and again and will either work as well as anything else, or fail totally in which case time to get the pliers out. Or abandon the second worst connector ever designed?
Gold plating generally requires a nickel layer which add iron, which we all know is bad, unless its good. Unplug and replug now and again and will either work as well as anything else, or fail totally in which case time to get the pliers out. Or abandon the second worst connector ever designed?
- Status
- Not open for further replies.