TPS54218: Clarification on Figure 36 & Figure 37

PY (Pengyu) Song

Part Number: TPS54218

Hi team,

I would like to check on Figure 36 & Figure 37, what is the "Power Dissipation" referring to? Does it refer to the power loss on the chip itself OR the total output power? For example in application, 1.8Vout/2A with 90% efficiency, total output power is 3.6W, power loss is 0.36W (assuming all power loss on the device for simplicity), should I use 3.6W OR 0.36W as Y-axis for below 2 curves?

Thanks!

over 5 years ago

0 JohnTucker over 5 years ago

TI__Guru**** 191445 points

Hi PY,

Those curves are generally following the correct trend. For Ta = 25 C and Tj = 25 C power dissipation must be 0 (see fig 36...there is some offset). For Tj = 150 C and Ta 150 C, power dissipation is 0 W. Essentially, Ta + Pdiss * 37 C/W = Tj (for the EVM). There appears to be some unintended offset in those graphs. Since they were generated 10 years ago, it may be difficult to track down the source of the difference. I will update this post tomorrow.

0 PY (Pengyu) Song over 5 years ago in reply to JohnTucker

TI__Expert 5860 points

Dear John,

Thanks for the comments.

The key question here is: Pdiss = Vout*Iout, OR, Pdiss = Vout*Iout*(1-efficiency)?

I suppose it should be the 2nd one, i.e. the power loss dissipated on the TPS54218 itself. Please help confirm.

0 JohnTucker over 5 years ago in reply to PY (Pengyu) Song

TI__Guru**** 191445 points

PY,

eff = Pout / Pout + Pdiss. Solving:

Pdiss = ( (1/Eff) - 1) * Pout

0 JohnTucker over 5 years ago in reply to JohnTucker

TI__Guru**** 191445 points

PY,

Just to close the loop on this. These are simple estimates based on the EVM Theta J-A of 37 C/W. Looking first at Figure 37, which plots Pd vs Ta with Tj constant, Pd = (150 - Ta) / Theta J-A. So for Ta = 20, Pd = (150 - 20) / 37 = 3.51 W, and for Ta = 150, Pd = (150 -150) / 37 = 0 W, with a linear interpolation in between. Figure 36 is the same data with Tj variable and Ta constant. It should be either offset by 5 deg. or changed to Ta = 20 C. In any case it is just an estimate. We now have more sophisticated methods to measure SOA.

Power management

Power management forum

TPS54218: Clarification on Figure 36 & Figure 37