TPS23754: Extra inrush current on Ipi when PD power on

Hello, 

The multiple current startups probably happen if the design is not able to startup the first time. The high, narrow spikes is more likely a scope feature rather than something happening on the board itself. Even still, they are not very high. The peak spike is ~400mA, which is not able to damage the device. 

As for the Vout waveform, this is probably due to the high amount of output capacitance and the very high secondary soft-start capacitor (C89). The max secondary soft start cap size we use is 10uF, and typically use 1uF. A 22uF is going to give a different behavior. 

For the output cap, I don't know the load transient performance you need but the amount of output cap is very high. The higher the output voltage, the lower the output current for the same power level. With that in mind, this design has 5x what our 12V reference design has, where it should have equivalent or lower. 

So to me, I don't see this behavior as an "issue" per say, rather a design choice. My advice would be reduce C89 to 10uF or lower and you'll probably see a much smoother startup. You could additionally lower the output cap if you want, but I think it's probably fine just a slower startup.

https://www.ti.com/lit/df/tidrbh4/tidrbh4.pdf?ts=1650987096512&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FTPS23754 

If this post answers your question, please indicate so by marking this thread as resolved. Thank you.

Regards, 

Michael P.

Applications Engineer

Texas Instruments 

Hey Wendy, 

Yes a large amount of output capacitance will take time to charge up, especially with the higher output voltage of 15V. So it could be a combination of these two. Additionally, there is a lot of capacitance on VC, which is typically needed for a higher voltage ACF. So it could just be part of it. 

The input current limit has two limit phases ---- first there is the in-rush phase, which is when the PSE first applies power and the input voltage is ramping up. The current in this stage is limited internally, and this limit is set by the PoE standard. this is the Inrush current limit

The second phase is once the power is fully applied, and it is related to the PoE Class level. So the total power must be limited to 25W. So it depends on the input voltage, but the device itself can officially handle 970mA (recommended). Once it goes beyond this the internal current limit should take over and turn the internal pass FET off.  

Looking at the spikes, they appear to be in the ~300mA range. Therefore, I think it should be okay. If you want a final check, you could run the device through a Sifos tester to confirm compliance. 

If this post answers your question, please indicate so by marking this thread as resolved. Thank you.

Regards, 

Michael P.

Applications Engineer

Texas Instruments