Flir C2 thermal imaging of SA818 radio node

Today I did some thermal measurements with my Flir C2. These values were taken on Tx during a radio dialogue between two hams. The node was continuously transmitting for several minutes in a row. To me it seems it runs a bit too hot so I have to imagine a cooling solution for the node. What is interesting is that it seems that the most heat is dissipated by the processor not by the SA818 module. This design has a ground pad that should be thermally connected with the corresponding pad of SA818 (low thermal resistance) with some thermal paste. I did not do that, so I wonder if this will improve the thermal profile. Time will tell, but definitely all these will be included in my next project.