check using flowtherm.  IMHO.  (do not count on solder via fill  
100%.  50% or less might be more realistic - assume 100% interface  
coupling - SAC is not that soft compare to copper, be careful about  
thermal shock if there is a requirement).  my 1.45 cents.
jk
On Aug 30, 2019, at 4:36 PM, Robert Kondner wrote:

> Dave,
>
> Thanks. I was comparing the via copper walls to the area of the via  
> center. I was assuming vias would be solder filled. Solder has a  
> conductivity value of about 50.
>
> For a 12mil via with a 1 mil wall the ratio of wall cross section  
> to open via is about 3. If the epoxy fill is 10% of the thermal  
> conductivity of copper then the wall is the major contributor of  
> heat transfer. The smaller and denser the placement of the vias the  
> copper wall get even more important. I recall figuring that a 40mil  
> via had equal heat conducted by wall and solder. Below 40 mils the  
> wall transfer more heat. So I am wondering if filling the vias  
> (small vias) really does much to transfer heat..
>
> Please check my math!
>
> When I measured 1C/W for my DPAK I called it mission accomplished  
> and I moved on. I am just curious if anyone actually measured things.
>
> Thanks,
> Bob K.
>
> -----Original Message-----
> From: TechNet <[log in to unmask]> On Behalf Of Dave Schaefer
> Sent: Friday, August 30, 2019 10:15 AM
> To: [log in to unmask]
> Subject: Re: [TN] [External] Re: [TN] TYPE III vs TYPE V vias
>
> Robert,
>
> From Amitron website:
>
> " Non-Conductive fill has a thermal conductivity of 0.25 W/mK  
> whereas Conductive pastes have a thermal conductivity anywhere from  
> 3.5-15 W/mK. In contrast, electroplated copper has a thermal  
> conductivity of more than 250W/mK.
>
> So while Conductive via fill can offer needed conductivity in some  
> applications more often than not it is possible to use non- 
> conductive paste and add additional vias. Often this results in  
> superior thermal and electrical conductivity with minimal cost  
> impact. "
>
> Recall reading a more complete paper on this somewhere but can't  
> locate it right now.
>
> More ammunition for the idea of using many smaller VII non- 
> conductively filled vias and calling out increased minimum hole  
> plating (Class 3 hole plating).