Junction to Board Thermal Resistance for CY7C2665KV18

[Jmar_12]

In a question I posted on 4/21/2021 (Thermal resistance values for CY7C) it was stated " As per our packaging team we have simulated Theta JB which is 13.45 C/W using an 8-Layer (101.5 x 114.4 x 1.6mm) PCB."

However, this value is greater than the  data sheet Theta JA (12.23 °C/W) which does not make too much sense. For a typical FBGA package without an exposed heat sink or heat spreader, majority of the heat is removed from the base of the package. 

Can you provide some insite on this?

[PradiptaB_11]

Hi,

We will need to check this with our team internally. Will keep you updated on this. 

Thanks,

Pradipta.

[PradiptaB_11 · Accepted solution]

Hi,

I think there was a typo from our side in the previous post. Please refer to the below values that we have with us both from simulation and measurements. A device soldered on the PCB should observe a lower Theta JB value. We believe the PCB in the data below plays a bigger role in the lower value than the socket.

Type	Simulation	Measurement
Board Mount	Soldered	Socketed
test Condition	102x114x1.6mm, 8-layer PCB	170x220×2.35mm, 8-layer PCB
Theta JA	16.11⁰C/W	12.23⁰C/W
Theta JB	11.72⁰C/W	9.34⁰C/W
Theta JC	2.1⁰C/W	2.1⁰C/W

Thanks,

Pradipta.

[Jmar_12]

Pradipta, 

Thanks for the response. You stated "A device soldered on the PCB should observe a lower Theta JB value. We believe the PCB in the data below plays a bigger role in the lower value than the socket." I agree with your statement, but the soldered value you have listed (11.72C/W) is higher than the socket value (9.34C/W) which contradicts what you said above. Can you clarify? 

[PradiptaB_11]

Hi

Yes, even though both PCBs are 8 layer, the socketed PCB is much larger and thicker, which is why we believe the Theta JB value is lower on the socketed PCB. So, we should not consider this  as a contradiction. 

Thanks,

Pradipta.