Thermal impedance thetaJA EPC2088 vs EPC2367

Designing with GaN FETs and ICs Thermal Management
1

I’m trying to compare the thetaJA of the EPC2088 to the EPC2367. However, the PCB mounting conditions given in the datasheet seem to be different, or at least not clearly the same. Is there any data to make an apples-to-apples comparison between the two? I assume the EPC2367 would be better, but I’d like to know by how much.

2

Hi Lloyd,

Unfortunately there is no perfect way to make an apples-to-apples comparison for the RthetaJA for the two parts, since those parameters in the datasheet are defined using different boards. RthetaJA will always be very board-dependent. If you want to compare the thermal properties of the parts, you can start with RthetaJB and RthetaJC, since these properties are intrinsic to the FET itself.

Two options that could be helpful:

  1. Use our thermal calculator tool which will give a simulation for the overall thermal resistance experienced with a given part, given specific physical and electrical parameters (that you can specify). LINK: https://epc-co.com/epc/design-support/gan-power-bench/gan-fet-thermal-calculator
  2. Look at thermal camera images in the eval board user guides for both parts. The evaluation board used for these parts is very similar, so it should have very similar thermal properties. The thermal images (Figure 16 in both user guides) show the temperature of the top of the FETs on the board. So it’s not exactly the junction temperature, but it is still useful. The operating parameters for both conditions are almost identical. Only differences are output current and airflow. EPC2088 eval board user guide: epc-co.com/epc/portals/0/epc/documents/guides/EPC90154_qsg.pdf
    EPC2367 eval board user guide: https://epc-co.com/epc/portals/0/epc/documents/guides/EPC90164_qsg.pdf

Thanks,

Daniel Carlson

3

Thanks, Daniel,

From the thermal images, it looks like the larger, lower RDSon FETs are performing worse than the smaller FETs. Higher switching losses? I’ll check out the thermal calculator tool, appreciate the pointers!

4

Hi Lloyd,

Yes, the EPC2367 would experience higher switching losses than the EPC2088 in those operating conditions.

Also keep in mind that the test done on the EPC2367 uses 28A, while the test done on the EPC2088 used 25A. And the air flow on the 2367 was 400LFM, while the airflow on the 2088 was 1000-1500LFM. Both of those factors also contributed to the 2367 getter hotter than the 2088, so it’s not just the difference in switching losses.

Thanks,

Daniel

5

Thanks again, Daniel, well noted.