• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

When is a Component Too Hot?

Status
Not open for further replies.

wuchy143

Member
Hi All,

I am using Netburner's MOD5234 PCB that mounts onto my board which has some transistor logic, optical transceiver and that's about it. My PCB(the one the Netburner mounts to) also has a regulator on it(LM1117) to power the Netburner PCB at 3.3V. The input to the regulator is +5V.

I"m seeing temperatures in the high 90's F on the Netburner and on the optical transceiver. The regulator is a little warm at about the same temperature but that's somewhat expected.(It's a TO-220 package which I goofed and didn't provide a heatsink...DUMMY!) So, I'm confident that with a proper heat sink the regulator should be at a more ambient temperature. I'm still a little confused about the netburner and optical transceiver though.

Does anyone have experience with this netburner PCB and knows if these temperatures are expected? Also, in an electrical system what is the temperature rise from ambient that is the max so you don't degrade the part too quick or do something even worse...like create a fire. :(

Thanks!!

-wuchy143
 

Hero999

Banned

MikeMl

Well-Known Member
Most Helpful Member
Another data point: You can spit on you finger, touch the IC. If it burns your finger, that is not sufficient to say it is too hot. If the spit sizzles and turns to steam, then the IC is Probably too hot. If the spit stays there, then the IC is well within its spec (even though it burned your finger);)
 

wuchy143

Member
all great responses. thx

I think my next step is to call an applications engineer and see if the tempature rise on the Netburner Module and Optical Transceiver are what's expected. I don't mean to be overly cautious but this is going into a very critical type of application and cannot afford to have this bite me.

Regrards!

-wuchy143
 

bountyhunter

Well-Known Member
Another data point: You can spit on you finger, touch the IC. If it burns your finger, that is not sufficient to say it is too hot. If the spit sizzles and turns to steam, then the IC is Probably too hot. If the spit stays there, then the IC is well within its spec (even though it burned your finger);)
We found that the case temp where it is too hot to be able to hold your finger on continuously is about 60C. If it is hot enough to make you yank your hand away, it's obviously significantly hotter than that. Approaching 125C, you will really get a burn on your finger...... I remember getting the "M" logo on a motorola can transistor seared into my fingertip once....

In general, if it is uncomfortably hot but not making you yell, it's OK.


The problem is you have to know the power dissipation and thermal resistance of the package to know how much hotter the junction is than the package surface. With power packages like a T0-220, the temp rise from the surface of the case to the junction is maybe 4C/Watt. In tiny surface mount packages it is WAY higher.
 
Last edited:

Hero999

Banned
Note that the 125C maximum for IC's is the JUNCTION temp,
You're certainly right about that.

not the case temp. In reality, IC's don't usually shut down until the junction reaches 160 - 165C.
Surely it'll be cooked to death before then?
 

bountyhunter

Well-Known Member
You're certainly right about that.


Surely it'll be cooked to death before then?
NO, IC's don't cook at 165C. Even 185 for brief operation is survivable. Some power transistors designed for very high power use allow continuous junction temp up to 200C (those are not common). I know from testing parts with no thermal shutdown that the commercial IC's explode at a junction temp in the 220 - 230C range.... at least ours did.
 

wuchy143

Member
problem is you have to know the power dissipation and thermal resistance of the package to know how much hotter the junction is than the package surface. With power packages like a T0-220, the temp rise from the surface of the case to the junction is maybe 4C/Watt. In tiny surface mount packages it is WAY higher.
great rule of thumb thanks!

I have an idea but what exactly is the junction. Is that just referring to the silicon chip inside or...what?
 
Last edited:

Hero999

Banned
Yes, it's the silicon chip in side.

Technically the junction only applies to discrete components, IC have a die.
 

bountyhunter

Well-Known Member
Junction temperature is the die temperature. We call it "junction temperature" because of how it's measured. Every IC has a substrate diode that is "calibrated" by putting the IC in an oven with no voltage and raising the whole thing to a specific temp (like 125C) and then applying a small current through the substrate diode and measuring it's VBE, ergo we use that P-N junction to measure die temperature. When the device is under power dissipating a specific amount of power, you measure the case temperature. Then the IC power is abruptly switched off and the current is sent through the substrate diode and a "snapshot" VBE tells the temp it is by comparing that VBE to the calibration VBE. Knowing the temp of the case and die at a specific power dissipation allows you to calculate the thermal resistance from junction to case.
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top