LM5071: LM5071

Hi SUNDARAMOORTHI,

Thanks for your reaching out.

Could you share the schematic of your design so we can look into it?

AN1430 (https://www.ti.com/lit/ug/snva141b/snva141b.pdf) is an reference design using LM5071, which you can refer and check the components selection. 

Best regards,

Diang

Hi Diang Xing,

please refer the schematics and let me know if you need anything else.

Hi SUNDARAMOORTHI,

Thanks for your reply.

I will take a look and get back to you by 23rd/Feb.

Best regards,

Diang

Hi SUNDARAMOORTHI,

One thing I noticed is that you did not add the decoupling caps in your design. They could be critical for dc/dc converter, since the input side ESL ESR could be large. 

I would recommend you to add ceramic cap as closing to the highlighted circuit as better in your layout.

Besides, could you share the datasheet of D9? D9 should be a large package Zener or TVS diode that can absorb the energy that stored in stray / leakage inductance.

Best regards,

Diang

Hi Diang,

Received your suggestions, and the D9 data sheet is attached.  DATA SHEET

Both Development board and our design D9 packages match (SMA).

Hi SUNDARAMOORTHI,

Thanks for the datasheet, the diode looks good.

Best regards,

Diang

Hi SUNDARAMOORTHI,

Did you see any improvement after adding the decoupling cap? If you have any waveforms of that flyback converter it will be helpful too.

Best regards,

Diang

It appears we're still grappling with issues related to POE IC failures. I've provided the schematics for the section preceding the POE, where the signal transformer also intermittently fails or becomes damaged. Could overvoltage or overcurrent from the POE switch be contributing to these failures? Additionally, while some devices function properly via the Aux input, they fail to operate with the POE input. share your suggestions.

Hi SUNDARAMOORTHI,

Thanks for your updates. 

For the signal transformer, what is the data rate you will need? And if the secondary side of the signal transformer is powered by the PoE, then the PoE failure may also compromise the signal transformer.

If you have any waveforms of that flyback converter it will be very helpful for us to look into this issue. Besides, the two circuit schematic may be helpful too.

PoE Powered Devices Debug Guidelines (https://www.ti.com/lit/an/slvaf74/slvaf74.pdf)

PoE PD Schematic Review Guidelines (www.ti.com/.../slvaf59.pdf) 

Best regards,

Diang

Hi Diang Xing,

we are using the Data rate of 100 Mbps. I have attached flyback converter waveforms for your reference.           

Hi Sundaramoorthi,

Diang, our PD Apps engineer is currently out of office. Please expect a delay in his response!

Regards,

Brandon

Hi Sundaramoorthi,

Thanks for your patience. From the waveforms I could see the converter can work at least in a basic way. 

I am wondering if you did any thermal test like thermal image of the top/bottom of this board. One thing I could notice the secondary side diode and RC snubber resistor packages are not large enough. 

RB080L-30DDTE25 with SMA package thermal resistance usually is very high for power rectification. You may need a diode with a package such as PDS360. The snubber resistor is also better to be 0805 or larger.

Best regards,

Diang

Hi Diang, thanks or your reply,

I will check and Update the Diode and Snubber resistor.

In most of the devices on our board, there are shorts between the following IC pins:

Pins 1 and 2,

Pins 2 and 3,

Pins 8 to 9 and 10. Could it be due to voltage spikes from the PSE switch, ultimately leading to IC failure? Any other specific reason is there.

Hi Sundaramoorthi,

Thanks for your updates.

Based on your information:

Pins 1 (VIN), 2 (RSIG), 3 (RCLASS) are shorted;

Pins 8 (RTN), 9 (OUT), 10 (VCC) are shorted.

Are these pin shorts measured on the board or standalone chip? Is it measured when the board unpowered? 

If it is on the board: VIN - RISG could be measured short considering RISG could be tied to VEE and there could be a big capacitance between VEE and VIN. RLCASS may not be shorted to VIN normally.

Between OUT and VCC there normally a push-pull switch. And if the measurement is the forward bias of the push-pull switch's body diode, you may see a short measurement. For VCC to RTN, there is also a large capacitor which may cause a short measurement.

Best regards,

Diang 

Hi Sundaramoorthi,

Since we did not get back from you for a long time. I will close this thread for now. Please reply or open a new thread if you have further questions.

Best regards,

Diang