TPS5450 out of regulation

You cannot use the same circuit to generate 3 different voltages.  At least the feedback resistors must change.  The TPS5450 is internally compensated.  the output filter must be selected to work well with that internal compensation, so usually the L and C will need to change for different output voltages.  Also, the ESR of theoutput cap is important.  It looks like you may be using switcherpro for these designs.  Make sure you are using the exact capacitor that switcherpro specifies (or at least one with the same characteristics).  Can you post waveforms of the 3.3 V while it is working properly then while it is at 4 V?  Also be aware that the TPS5450 is at the limit for power handling in that package.  Check you r thermal performance as well.

Sorry, I didn't explain clearly.

I have used three TPS5450s to generate +3.3V, +5V and +15V separately.

I have tried to reduce the input voltage +24V down to +20V or +18V but +3.3V is still out of regulation.

The load of my +3.3V is only three clock fan out buffer ICS551MLFT.

Many thanks.

When there are no inputs to three ICS551MLFT, the TPS5450 is regulated at +3.3V.

The regulated +3.3V will switch to +4.1V when inputs are fed to ICS551MLFT.

Please help.

You probably have a leakage path that is allowing one of those other devices to let some higher voltage onto your 3.3 V rail.  I'll need to see your full schematic and waveforms to be sure.  You can try disconnecting them one at a time to find out which is causing your issue.

I will let you know the result.

one more question:

if the load current of TPS5450 is variable, Can TPS5450 maintain constant +3.3V output with fixed LC output filter?

Many thanks.

In general yes.  There is a specification for that called "load regulation".  It is the change in voltage with respect to load current and applies to slowly changing loads.  It is a graph in the EVM users guide and datasheet.  The other consideration is "load transient response".  That is the change in output voltage due to a step load change.  there should be an example of that in the datasheet as well.

I have changed +3.3V to +5V since ICS551MLFT also accept +5V power supply.

with +5V power supply, the TPS5450 isn't out of regulation.

It is strange.

I will let you know the new progress.

Best Regards.

If 4 V is leaking onto a 3.3 V rail, tehn it may pull it out of regulation.  The 5 V rail will source current back into the leakage path and maintain regulation.  Now I am fairly certian that you are infact dealing with a leakage path.

Thank you John.

Have a nice weekend.