TPS2378: Heat issues occured.

Pang

Part Number: TPS2378
Other Parts Discussed in Thread: TPS54560,

Hi Ti,

I am using TPS2378 pairing with the TPS54560 to achieve output 5V@4A. Currently I had tried using the TPS2378EVM-105 to pair with my TPS5460 as per below schematic and found some heat issues.

Input from PSE to supply to TPS54560.

During this connection when I started a load with 5V@250mA the heat at TPS2378 increased to 100degC in few seconds.

1) Could you help to advise what should I rectify with my current TPS2378 design?

2) What is the reasons that cause the TPS2378 temperature going up just in few seconds?

3) What is the Iout(max) able to support by TPS2378? Because it hits up to >100degC and my load current is at 250mA only.

over 6 years ago

0 Darwin Fernandez over 6 years ago

TI__Mastermind 40100 points

Hi Incog,

The 100uF bulk cap might be too large especially if the DCDC is pulling some current away from it charging. Note, per the standard, the PSE must try to charge this bulk cap by a specific time. This is why the PD's load (DCDC converter) shouldn't pull current during the first 80ms when the PSE first applies power.

To test this out, you can try disabling the DCDC when you plug in. Then you can try to start decreasing the PD's output bulk cap.

0 Pang over 6 years ago in reply to Darwin Fernandez

Expert 2385 points

Hi Darwin,

I would like to check how do I determine the bulk capacitance value? Tried to check from the datasheet but couldn't find the recommended one.

0 Darwin Fernandez over 6 years ago in reply to Pang

TI__Mastermind 40100 points

Typically, for this power level it'll be 47uF. I have seen however, 100uF used in the past. You must make sure however, that there is no load current (like the DCDC converter) taking away current from the bulk cap that will cause the bulk cap to charge up slower during startup. This can heat up the IC while the PD is in inrush. This is why per the IEEE802.3bt standard, the PD shouldn't add a specific amount of load current during startup (~80ms).

0 Pang over 6 years ago in reply to Darwin Fernandez

Expert 2385 points

Hi Darwin,

Thanks.

I have changed the Cbulk to 47uF after 10s of current loading the temperature hits >90degC. It's like a slight improvement only. Could you help to advise what's the next step?

0 Darwin Fernandez over 6 years ago in reply to Pang

TI__Mastermind 40100 points

Hi Incog,

It sounds like there is something loading the PD output that the PD cannot fully charge the bulk cap. Is there a load on the output of the converter? If so, can you try with no load? If not, can you try removing the converter IC?

0 Darwin Fernandez over 6 years ago in reply to Darwin Fernandez

TI__Mastermind 40100 points

This is just to see if the PD IC can get out of inrush. And point to whether we should be focusing at the load or the PD IC design.

0 Darwin Fernandez over 6 years ago in reply to Darwin Fernandez

TI__Mastermind 40100 points

Hi Incog,

It looks like this answer was rejected, did you check if the PD is always in inrush (due to short or loading between VDD-RTN)? Does RTN-VSS ever reach ~0V?

Regards,

Darwin

0 Thomas Lewis over 6 years ago in reply to Darwin Fernandez

TI__Expert 5615 points

Incog,

I wanted to check in as we have not heard back from you in a week.

Please let us know if there is anything else we can do to assist...otherwise, we are going to assume this thread is resolved.

Thank you for your interest in PoE products from TI,

Power management

Power management forum

TPS2378: Heat issues occured.