Author Topic: SMD heat sink positioning  (Read 2299 times)

0 Members and 1 Guest are viewing this topic.

Offline SimonTopic starter

  • Global Moderator
  • *****
  • Posts: 18038
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
SMD heat sink positioning
« on: November 17, 2017, 03:55:54 pm »
So I am using some nice little SMD mosfets: http://www.farnell.com/datasheets/1648177.pdf?_ga=2.10961076.1206128083.1510930412-1136071739.1509783694&_gac=1.45419536.1510340349.EAIaIQobChMI0LKqmti01wIVVZ0bCh0-igWzEAQYASABEgL7UfD_BwE

I am thinking that maybe I should for good measure I put a heat sink in. Does gluing a heat sink to the top of the plastic case help that much? or is it better to have a surface mounted heat sink soldered on right next to the parts on  2oz copper ?
 

Offline phil from seattle

  • Super Contributor
  • ***
  • Posts: 1044
  • Country: us
Re: SMD heat sink positioning
« Reply #1 on: November 17, 2017, 05:47:20 pm »
The data sheet talks about PCB area - page 1, note 2 and page 2 "thermal characteristics".  It's characterized for a PCB, vertical in still air. I wonder if adding a heat sink on top would disrupt convection.

To your point about heatsink on plastic package. I've not seen any studies but lots of anecdotes say that it's only somewhat effective. I'd think forced air on the PCB surface would be more effective but that's just a guess. I use lots of solder filled vias to thermally bridge to the bottom side of the PCB.  It seems to help though I've only done casual measurements. Your MOSFET package doesn't seem to have a slug so it's probably not the best for that.

That MOSFET has a pretty low Rdson so you may not need to worry about it.
 

Offline SimonTopic starter

  • Global Moderator
  • *****
  • Posts: 18038
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: SMD heat sink positioning
« Reply #2 on: November 17, 2017, 05:52:46 pm »


That MOSFET has a pretty low Rdson so you may not need to worry about it.


Yes that was my hope. The PCB will be potted, I'll go for some decent potting but I'm thinking a little heat sink would at least put a cooling surface outside the potting. Indeed i'm nervous about using the plastic as a contact surface. So I'm thinking a heat sink soldered to the PCB track close by with plenty of solder will help take any heat out.
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 9992
  • Country: gb
Re: SMD heat sink positioning
« Reply #3 on: November 17, 2017, 05:53:26 pm »
If it comes to it, you can use this style of solderable smd heatshink. Various sizes available:

http://uk.rs-online.com/web/p/heatsinks/7124217/

Edit: Just saw the potting bit
Best Regards, Chris
 

Offline SimonTopic starter

  • Global Moderator
  • *****
  • Posts: 18038
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: SMD heat sink positioning
« Reply #4 on: November 17, 2017, 05:56:13 pm »
sort of what I'm after although my mosfet is smaller than a TO263, I was thinking something to solder to the board on the wide track that comes out of the drain
 

Offline cowana

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: gb
Re: SMD heat sink positioning
« Reply #5 on: November 17, 2017, 06:08:30 pm »
Soldering a heatsink to the other side of the board (and thermally connecting the two sides with plenty of vias) is a technique often used on high power SMD devices in low-volume projects where the extra assembly step is acceptable.
 

Offline Twoflower

  • Frequent Contributor
  • **
  • Posts: 742
  • Country: de
Re: SMD heat sink positioning
« Reply #6 on: November 17, 2017, 06:52:00 pm »
First you should check if this is actually needed. And if this is needed: Is there a way to reduce the losses. See if it is possible to increase the GS voltage.

And you could place the MOSFETs on the back-side of the PCB. That would allow to use the metal package to work as head-spreader. For this you might need some thick thermal conductive material (depending on how much height variations are present) and a metal spacer-plate to bridge the space between the case and the trannies. This might work also for top-side power circuits, but the spacer-design might be more complicated.
 

Offline SimonTopic starter

  • Global Moderator
  • *****
  • Posts: 18038
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: SMD heat sink positioning
« Reply #7 on: November 17, 2017, 07:33:12 pm »
First you should check if this is actually needed. And if this is needed: Is there a way to reduce the losses. See if it is possible to increase the GS voltage.




I will but I won't want to make prototype board twice over it i rather design them in and then remove them from the BOM is not required.

These look good: http://uk.rs-online.com/web/p/heatsinks/1263115/ assuming they are available.
 

Offline ocset

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Re: SMD heat sink positioning
« Reply #8 on: November 18, 2017, 01:06:47 am »
we used those type of fets when i worked at an rf place in london.
It is about 20 degc/watt through the palstic case. Much better through the tab on the bottom.
They need thermal vias in pad...to bottom copper, then to heatsink.
Ill send some stuff on thermal vias.
Always make them <0.3mm diameter. -to stop wicking of solder.

I must admit though, at first they just gap padded them to heatsink above the board...ie going through  the plastic case.....but then they changed to thermal vias
« Last Edit: November 18, 2017, 01:10:47 am by treez »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf