EDAboard.com | EDAboard.de | EDAboard.co.uk | WTWH Media

removing heat with thermal tape and small heatsinks

Ask a question - edaboard.com

elektroda.net NewsGroups Forum Index - Electronics Design - removing heat with thermal tape and small heatsinks

Goto page 1, 2, 3, 4, 5, 6, 7  Next

Winfield Hill
Guest

Sat Feb 02, 2019 7:45 pm   



I'm using Vishay's CRCW1210-HP 1210 resistors, rated at 0.75
watts each. I've got three of them dissipating 2.2 watts,
plus three SOT-23 transistors right next to them, rated at
0.5 watts each, dissipating 1.1 watts, or 3.3 watts total.
A Flir measurement shows a surface temperature of 150 C.

Shocking! I resisted a temptation to check with my finger.

The heat-dissipating region is 10mm x 20mm, and I'm thinking
of adding a 10x20mm heatsink with double-sided thermal tape.
According to the Alibaba seller's specs, the dT rise should
be 2.2 C/watt (seems too good). I'll probably also add a
40mm fan to blast 6W from an adjacent 25 x 50mm heatsink,
and its (heated) side-flow should help cool the resistors.


--
Thanks,
- Win

John Larkin
Guest

Sat Feb 02, 2019 7:45 pm   



On 2 Feb 2019 10:15:54 -0800, Winfield Hill <hill_at_rowland.harvard.edu>
wrote:

Quote:
I'm using Vishay's CRCW1210-HP 1210 resistors, rated at 0.75
watts each. I've got three of them dissipating 2.2 watts,
plus three SOT-23 transistors right next to them, rated at
0.5 watts each, dissipating 1.1 watts, or 3.3 watts total.
A Flir measurement shows a surface temperature of 150 C.

Shocking! I resisted a temptation to check with my finger.

The heat-dissipating region is 10mm x 20mm, and I'm thinking
of adding a 10x20mm heatsink with double-sided thermal tape.
According to the Alibaba seller's specs, the dT rise should
be 2.2 C/watt (seems too good).


By a factor of at least 5.

Would the heat sink go on the bottom of the board? That would benefit
from strategic use of copper pours and vias to transport heat from the
parts.

Thin tape may not conform to the surfaces, and leave air gaps.
Compressible gap-pad material would be better.


I'll probably also add a
Quote:
40mm fan to blast 6W from an adjacent 25 x 50mm heatsink,
and its (heated) side-flow should help cool the resistors.


At 5.5 watts, you could just blast air down onto the board, sort of
like this:

https://www.dropbox.com/s/yn52khzb0jz9dy3/Uzed_Fan_Top.JPG?dl=0

Again, copper pours make a huge difference in getting heat out of
small parts. 0.75 watts isn't bad for a 1206 with decent end-cap
cooling. I'm using some AlN 1206 resistors rated for 11 watts!

The SOT23s, with the usual tiny pads, might be 300 K/w. That sounds
grim at 1.1 watts.

Internal copper layers, a ground plane and some copper pours, can make
a small board almost isothermal, so it wouldn't matter too much where
you blow the air.









--

John Larkin Highland Technology, Inc

lunatic fringe electronics

John Larkin
Guest

Sat Feb 02, 2019 8:45 pm   



On 2 Feb 2019 10:15:54 -0800, Winfield Hill <hill_at_rowland.harvard.edu>
wrote:

Quote:
I'm using Vishay's CRCW1210-HP 1210 resistors, rated at 0.75
watts each. I've got three of them dissipating 2.2 watts,
plus three SOT-23 transistors right next to them, rated at
0.5 watts each, dissipating 1.1 watts, or 3.3 watts total.
A Flir measurement shows a surface temperature of 150 C.

Shocking! I resisted a temptation to check with my finger.

The heat-dissipating region is 10mm x 20mm, and I'm thinking
of adding a 10x20mm heatsink with double-sided thermal tape.
According to the Alibaba seller's specs, the dT rise should
be 2.2 C/watt (seems too good). I'll probably also add a
40mm fan to blast 6W from an adjacent 25 x 50mm heatsink,
and its (heated) side-flow should help cool the resistors.


Hey, Win, please post some pics of the top and bottom of the board.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

Klaus Kragelund
Guest

Sat Feb 02, 2019 8:45 pm   



If you have room for a fan, then why not just spread the components out?

Cheers

Klaus

Winfield Hill
Guest

Sat Feb 02, 2019 10:45 pm   



John Larkin wrote...
Quote:

Hey, Win, please post some pics of the top and bottom of the board.


https://www.dropbox.com/sh/oezklu2uy7eiywd/AADvVvJvAn2XJrYcIYxyJHb8a?dl=0

Things will move around a bit to improve the power situation.


--
Thanks,
- Win

Winfield Hill
Guest

Sat Feb 02, 2019 10:45 pm   



John Larkin wrote...
Quote:

On 2 Feb 2019 10:15:54 -0800, Winfield Hill <hill_at_rowland.harvard.edu
wrote:

I'm using Vishay's CRCW1210-HP 1210 resistors, rated at 0.75
watts each. I've got three of them dissipating 2.2 watts,
plus three SOT-23 transistors right next to them, rated at
0.5 watts each, dissipating 1.1 watts, or 3.3 watts total.
A Flir measurement shows a surface temperature of 150 C.

Shocking! I resisted a temptation to check with my finger.

The heat-dissipating region is 10mm x 20mm, and I'm thinking
of adding a 10x20mm heatsink with double-sided thermal tape.
According to the Alibaba seller's specs, the dT rise should
be 2.2 C/watt (seems too good). I'll probably also add a
40mm fan to blast 6W from an adjacent 25 x 50mm heatsink,
and its (heated) side-flow should help cool the resistors.

Hey, Win, please post some pics of the top and bottom of the board.




--
Thanks,
- Win

Jon Elson
Guest

Sun Feb 03, 2019 12:45 am   



John Larkin wrote:


Quote:

Thin tape may not conform to the surfaces, and leave air gaps.
Compressible gap-pad material would be better.

Right. Bergquist makes stuff called "Gap Pad" in several thicknesses.
The stuff is a bit sticky, but not good enough to hold a heat sink onto the
board. So, you'd need some sort of screws and spacers to anchor the heat
sink in place with the right crush and pressure to maintain thermal
conductivity. And, the Gap Pad stuff is not real high thermal conductivity,
either. We've used it to keep some medium-power chips cool even when run in
vacuum. But, it might not be good enough for higher power densities.

Jon

John Larkin
Guest

Sun Feb 03, 2019 1:45 am   



On 2 Feb 2019 13:30:56 -0800, Winfield Hill <hill_at_rowland.harvard.edu>
wrote:

Quote:
John Larkin wrote...

Hey, Win, please post some pics of the top and bottom of the board.

https://www.dropbox.com/sh/oezklu2uy7eiywd/AADvVvJvAn2XJrYcIYxyJHb8a?dl=0

Things will move around a bit to improve the power situation.


Interesting trace pattern on the bottom side, near R30.

If you're going to spin the layout, you might add more copper and
thermal vias to move heat to bottom side pours, where there's more
room. Especially if capacitances don't matter much.

When I do a guard trace that dead-ends, like your bottom leftside, I
usually get a call from the fab house telling me that I have a trace
that dead-ends.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics


Guest

Sun Feb 03, 2019 1:45 am   



On Sunday, February 3, 2019 at 11:16:03 AM UTC+11, John Larkin wrote:
Quote:
On Sat, 02 Feb 2019 17:40:16 -0600, Jon Elson <elson_at_pico-systems.com
wrote:

John Larkin wrote:



Thin tape may not conform to the surfaces, and leave air gaps.
Compressible gap-pad material would be better.
Right. Bergquist makes stuff called "Gap Pad" in several thicknesses.
The stuff is a bit sticky, but not good enough to hold a heat sink onto the
board. So, you'd need some sort of screws and spacers to anchor the heat
sink in place with the right crush and pressure to maintain thermal
conductivity. And, the Gap Pad stuff is not real high thermal conductivity,
either. We've used it to keep some medium-power chips cool even when run in
vacuum. But, it might not be good enough for higher power densities.

Here is (as is usual with EETimes) a very bad article about gap pads.

https://www.eetimes.com/document.asp?doc_id=1334245


It doesn't mention graphite cloth, which is electrically as well as thermally conductive, but worked fine for me (as is mentioned in my 1996 millidegree thermostat paper).

--
Bill Sloman, Sydney

George Herold
Guest

Sun Feb 03, 2019 1:45 am   



On Saturday, February 2, 2019 at 7:01:17 PM UTC-5, John Larkin wrote:
Quote:
On Sat, 2 Feb 2019 15:50:34 -0800 (PST), George Herold
gherold_at_teachspin.com> wrote:

On Saturday, February 2, 2019 at 4:31:08 PM UTC-5, Winfield Hill wrote:
John Larkin wrote...

Hey, Win, please post some pics of the top and bottom of the board.

https://www.dropbox.com/sh/oezklu2uy7eiywd/AADvVvJvAn2XJrYcIYxyJHb8a?dl=0

Things will move around a bit to improve the power situation.


--
Thanks,
- Win

What John said, more copper. (Some power R's spec thermal resistance
for different pad sizes.)
Can you use transistors in a package with a thermal pad?

George H.

I like SOT89s for a couple of watts, but there aren't a lot of parts
in that package.
Right, those.


GH
Quote:


--

John Larkin Highland Technology, Inc

lunatic fringe electronics


John Larkin
Guest

Sun Feb 03, 2019 1:45 am   



On Sat, 02 Feb 2019 17:40:16 -0600, Jon Elson <elson_at_pico-systems.com>
wrote:

Quote:
John Larkin wrote:



Thin tape may not conform to the surfaces, and leave air gaps.
Compressible gap-pad material would be better.
Right. Bergquist makes stuff called "Gap Pad" in several thicknesses.
The stuff is a bit sticky, but not good enough to hold a heat sink onto the
board. So, you'd need some sort of screws and spacers to anchor the heat
sink in place with the right crush and pressure to maintain thermal
conductivity. And, the Gap Pad stuff is not real high thermal conductivity,
either. We've used it to keep some medium-power chips cool even when run in
vacuum. But, it might not be good enough for higher power densities.

Jon


Here is (as is usual with EETimes) a very bad article about gap pads.

https://www.eetimes.com/document.asp?doc_id=1334245



--

John Larkin Highland Technology, Inc

lunatic fringe electronics

George Herold
Guest

Sun Feb 03, 2019 1:45 am   



On Saturday, February 2, 2019 at 4:31:08 PM UTC-5, Winfield Hill wrote:
Quote:
John Larkin wrote...

Hey, Win, please post some pics of the top and bottom of the board.

https://www.dropbox.com/sh/oezklu2uy7eiywd/AADvVvJvAn2XJrYcIYxyJHb8a?dl=0

Things will move around a bit to improve the power situation.


--
Thanks,
- Win


What John said, more copper. (Some power R's spec thermal resistance
for different pad sizes.)
Can you use transistors in a package with a thermal pad?

George H.

John Larkin
Guest

Sun Feb 03, 2019 1:45 am   



On Sat, 02 Feb 2019 17:40:16 -0600, Jon Elson <elson_at_pico-systems.com>
wrote:

Quote:
John Larkin wrote:



Thin tape may not conform to the surfaces, and leave air gaps.
Compressible gap-pad material would be better.
Right. Bergquist makes stuff called "Gap Pad" in several thicknesses.
The stuff is a bit sticky, but not good enough to hold a heat sink onto the
board. So, you'd need some sort of screws and spacers to anchor the heat
sink in place with the right crush and pressure to maintain thermal
conductivity. And, the Gap Pad stuff is not real high thermal conductivity,
either. We've used it to keep some medium-power chips cool even when run in
vacuum. But, it might not be good enough for higher power densities.

Jon


I use some stuff from 3G, 2mm thick, that is rated 6 w/m-K. It's soft
and fairly compressible and a little tacky on both sides. A square
inch, compressed down to 1 mm thickness (with modest force) is under
0.3 k/w. It's a good dielectric, too.

https://www.dropbox.com/s/6g80f5yw695e8ah/Baseplate.jpg?dl=0

https://www.dropbox.com/s/fev8skbaaws5yib/L1.JPG?dl=0


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

John Larkin
Guest

Sun Feb 03, 2019 1:45 am   



On Sat, 2 Feb 2019 15:50:34 -0800 (PST), George Herold
<gherold_at_teachspin.com> wrote:

Quote:
On Saturday, February 2, 2019 at 4:31:08 PM UTC-5, Winfield Hill wrote:
John Larkin wrote...

Hey, Win, please post some pics of the top and bottom of the board.

https://www.dropbox.com/sh/oezklu2uy7eiywd/AADvVvJvAn2XJrYcIYxyJHb8a?dl=0

Things will move around a bit to improve the power situation.


--
Thanks,
- Win

What John said, more copper. (Some power R's spec thermal resistance
for different pad sizes.)
Can you use transistors in a package with a thermal pad?

George H.


I like SOT89s for a couple of watts, but there aren't a lot of parts
in that package.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics


Guest

Sun Feb 03, 2019 12:45 pm   



Winfield Hill <hill_at_rowland.harvard.edu> wrote in
news:q34moq01jom_at_drn.newsguy.com:

Quote:
I'm using Vishay's CRCW1210-HP 1210 resistors, rated at 0.75
watts each. I've got three of them dissipating 2.2 watts,
plus three SOT-23 transistors right next to them, rated at
0.5 watts each, dissipating 1.1 watts, or 3.3 watts total.
A Flir measurement shows a surface temperature of 150 C.

Shocking! I resisted a temptation to check with my finger.

The heat-dissipating region is 10mm x 20mm, and I'm thinking
of adding a 10x20mm heatsink with double-sided thermal tape.
According to the Alibaba seller's specs, the dT rise should
be 2.2 C/watt (seems too good). I'll probably also add a
40mm fan to blast 6W from an adjacent 25 x 50mm heatsink,
and its (heated) side-flow should help cool the resistors.



Skip the tape and use thermal epoxy. Need to retain serviceability?

Place those elements on a daughterboard with the heat sink, and
stand it off the main board on pins. Make more than one and trade
out the entire unit upon failure. Service the failed units.

Looks like you are pushing them pretty hard.

I was amazed that the Euro radial leaded resistor form factor has
the ratings they do.

Our US (read Vishay, most of the industry) 1/8 W units get pretty
hot at their rated value. The Euro (read British) resistor in the
same sized package is rated at 1W. Not sure I believe them.

Anyway, the daughterboard route allows you to modify those
elements as well and create the improved set up all properly
integrated onto the sinking element.

Goto page 1, 2, 3, 4, 5, 6, 7  Next

elektroda.net NewsGroups Forum Index - Electronics Design - removing heat with thermal tape and small heatsinks

Ask a question - edaboard.com

Arabic version Bulgarian version Catalan version Czech version Danish version German version Greek version English version Spanish version Finnish version French version Hindi version Croatian version Indonesian version Italian version Hebrew version Japanese version Korean version Lithuanian version Latvian version Dutch version Norwegian version Polish version Portuguese version Romanian version Russian version Slovak version Slovenian version Serbian version Swedish version Tagalog version Ukrainian version Vietnamese version Chinese version Turkish version
EDAboard.com map