Author Topic: Thermal transfer for a TO-92  (Read 2791 times)

0 Members and 1 Guest are viewing this topic.

Offline abitTopic starter

  • Contributor
  • Posts: 32
Thermal transfer for a TO-92
« on: May 02, 2015, 01:04:25 am »
If I solder it in-place with the leads bent so that the flat face of the IC is in contact with an exposed pad to the ground-plane - is this going to provide significant heat relief? How should I make this reliable or better, with some thermal compound, or this is not a good solution or not worth the effort? Lets assume the IC might going to run hot for a short while only, or switching, and in case it might not have thermal shutdown built-in as most transistors do not. Should I design a PCB with this option in mind?

Update:

Thanks, I suspected this, not worth the effort designing a PCB that way..
« Last Edit: May 02, 2015, 01:41:29 am by abit »
 

Offline radioFlash

  • Regular Contributor
  • *
  • Posts: 163
  • Country: us
Re: Thermal transfer for a TO-92
« Reply #1 on: May 02, 2015, 01:19:33 am »
To get good thermal transfer, you need to have some pressure applied, and bending the leads won't provide much force and won't be reliable, and TO-92s have a very large thermal resistance anywy, so you won't gain much. There are TO-92 heatsinks, but I don't think they're worth it. If you're concerned about heat, consider a TO-220 or other package.
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21909
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Thermal transfer for a TO-92
« Reply #2 on: May 02, 2015, 01:33:09 am »
As long as you have something filling the gap (you don't really need clamping pressure, but without it, the gap can creep open is the thing), you can get slightly better, but the plastic stinks for conduction, so it's not going to be much.

If you can do SMT, SOT-89 or 223 will be much better, up to a watt or two.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline jay

  • Regular Contributor
  • *
  • Posts: 62
  • Country: ee
  • Coder
Re: Thermal transfer for a TO-92
« Reply #3 on: May 02, 2015, 02:44:42 pm »
Thermal adhesive might do the trick.
SW engineer trying to design HW because it's more fun.
 

Offline digsys

  • Supporter
  • ****
  • Posts: 2209
  • Country: au
    • DIGSYS
Re: Thermal transfer for a TO-92
« Reply #4 on: May 02, 2015, 02:52:51 pm »
Quote from: jay
  Thermal adhesive might do the trick. 
Arctic cement / Arctic silver will give you the best thermal dissapation, but as others have said, the package isn't designed to get rid of heat.
Hello <tap> <tap> .. is this thing on?
 

Offline abitTopic starter

  • Contributor
  • Posts: 32
Re: Thermal transfer for a TO-92
« Reply #5 on: May 09, 2015, 03:03:33 am »
"Arctic cement / Arctic silver"

Thanks, I never hear of these, thanks!
 

Online Gyro

  • Super Contributor
  • ***
  • Posts: 9692
  • Country: gb
Re: Thermal transfer for a TO-92
« Reply #6 on: May 09, 2015, 10:23:57 am »
I suspect that, given the high thermal resistance of the case, any minor difference in thermal resistance of exotic heat transfer compounds isn't going to make any significant difference.

I'd stick with bog standard heatsink compound, maybe even superglue if the mating surfaces are in close flat contact (and secure them that way), Datron used to use it all the time to maintain matched components in close thermal contact in their instruments.

The most effective thing you can do is maximise the area of copper attached to the center pin, that's the one the die is attached to.
« Last Edit: May 09, 2015, 10:45:37 am by Gyro »
Best Regards, Chris
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf