TPS5430: Fluctuation in voltage even at low current

Hello Marco, 

Can you please provide a better image of the layout including all copper shapes and also oscilloscope waveform of the PH node. 

We can review in more detail. 

Regards, 
Denislav

Hello Denislav,

Of course. I'm out of the lab for the next three days, here are more PCB details. I'll send the waveform later, OK?

Hello Marco, 

In addition to the PH node measurement, could you check the voltage at the C4 capacitor?

It looks like the feedback  is tied close to the C4 cap. Please check if the voltage there is solid over the load range. 

Regards, 

Denislav

Hello Denislav,

I got the voltage after L2 at LDO input pin, maybe that makes a difference?
Do you mean maybe I should push C4 away or I should lengthen the feed back track as in the TPS5430EVM-173 design bellow?

Hello Denislav,

Unfortunately it will take me longer to be in the lab, so in a hurry I looked up older threads (12 years old) and made some changes. Please offer your opinion. Thanks in advance.

C48 is a bulk capacitor (there wasnt one in previous layout) for the 24V power supply unit is far away

L3 changed for currunt 1A

C4 and others are tantalum only for price drop at fab

Feedback taken from R2 (zero Ohm) close to L2 (12V output) to R22 (27K) and R1 (3K). These values came from modules available at chinese shields. I´ve calculated them using datasheet formula and they are correct. 

C41 to C44 and R33 / R34 made a cutoff filter, I dont know whether this will make any differece so put a jumper to bypass them (JP2) just in case.

Hi Marco,

Denislav is currently OOO, but the additions seem ok from my POV. What spec is the customer asking to improve upon and what are the input/output operating ranges?

Regards,

Richard

Hi Richard,

Input is 24V regulated from a shelf smps. Its location varies, I didnt know that until recently, so made a revision with the bulk capacitor.

Output seems to be close to 500mA / 12V so changed coil. Put a FB after LDO too.

He is asking for low ripple, but it seems to be around 30mV so I think its pretty good, but who knows.

As I stated above all features are working fine anyway, so if the revisions are fine I think this topic may be closed.

Regards and thank you both.

Marco

Hi all,

Client send me PH waveform at 89mA / 12V before changes made above (old pcb, new batch with modifications will come in 20 days)

Hello everybody.
PCBs have arrived. All the modifications described above worked as expected. Thank you all.

Marco