TPS54231 5V output issue

While the compensation is not ideal, it should be close enough for the supply to operate.  First thing you need to do is look at the switching node.  Is there any switching when the output is ramping up? Next look at the input voltage at the IC.  is it collapsing?  Try disconnecting the load completely.  Does it start up under no load condition?

hi JohnTucker,

The situation is i probe the input voltage and output voltage before i connect the load. i can get 12V input and 5.09V output. its correct.

but i can't get 5.09V output after i connect a 5ohm 10W load to it. its become the waveform that i send out earlier and the load can't draw 1A. I doubt i had missed something but the circuit quite simple. so i doubt is the compensation problem. 

do you meant switching node is the point (red circle) on the picture?

hi John Tucker,

i finally manage to capture the switching node waveform.

the waveform above is take before i connect load to it.

the waveform above is take after i connect load to it.

the Vsense is 960mV

i also found there is beep.. sound after i connect load. i think some frequency have trigger the inductor to come out the sound.

i also used WEBENCH to simulate my circuit, it work fine in the simulation. i headache now don't know where is the problem.

hi John Tucker,

i attached 2 waveform below. it seems like the waveform no much different but frequency is higher after connected a load.

the picture above shown PH pin waveform before connect load.

the picture above shown PH pin waveform after connect load.

 hi,

i have probe the 2 pins. But not sure what is the meaning of the waveform. Can you advice me?

Picture above shown comp pin before connect load.

Picture  above shown comp pin after connect load.

Picture above shown SS pin before connect load.

Picture above shown SS pin after connect load.

HI Abull Yang,

the waveform seem like SS pin(yellow) drop 1st before the COMP pin(blue).  Should i change the comp capacitor and resistor value? how i calculate the value? Please advice me because i dont understand the formula on the datasheet. Thanks

heng choy wong,

The first thing that you should do is to check the next stage. You know, typically, the current should not reach the current limit. I think the output current should be very high during the startup or there is a huge capacitor at output stage. I don't think it should be caused by the loop stability but the next stage circuit. Please monitor the output current of TPS54231 during the start up. There must be something wrong.

 Thanks

BR.

Abull

hi Abull,

i try connect 15ohm load to it(5v/15 0hm=0.33A . it also can't work. There are  one 47uF and one 10uF only at the Vout. i not sure how to monitor the startup current? Please see the figure below, it reach 1.2V then no more increase to 5V.

Based on the COMP pin voltage you appear to be drawing more current than you think.  I would remove all loading from your circuit to debug your TPS54231.  You can add some external controlled load to check that it can deliver the full 2 A.

Not COMP pin drawing too much current.  Peak switch current is directly proportional to COMP pin voltage.  The peak COMP voltage looks to be at the maximum level which would correspond to current limit on the output.

hi Abull,

How can i apply an EVM board? is it request from TI website?

hi John,

Good news. I found that the output inductor current rating only 0.4A. so I manage to get 1A current at the output after i changed the output inductor to higher rating current inductor. but i facing another problem is the heat dissipation. The TPS54231 is hot very fast with 10second after i connect 1A load( 5 ohm). do you have any idea? my routing have big trace for output and input voltage. so i not sure what cause the IC too hot.

Hi John,

Found the PH pin (switching node) is around 21Khz before load connected.

But it run up to 600Khz after load connected. Is this normal?

Thanks.

hi John,

For switching regulator power dissipation formula is POWER OUT/EFFICIENCY. so for my case is 5V and 1A output so 5W/90%=5.6Watt power dissipation. Do i multiple Roja with 5.6watt (5.6 * 116.3 = 650C)to get the estimate heat the tps will generate? 

What you calculated was power in, not power dissipated.  Pdiss is Pin - Pout or Pdiss = Pout * (1/eff -1) = 5 ( 1.111-1) = 0.555 W.  also there are other losses besides the IC in that calculation.  most are small but usually there will be significant loss in the catch diode and inductor.  You need to know the series resistance and the duty cycle to subtract out those losses.  Then you can multiply the IC power dissipation times theta JA to get the temperature rise of teh IC.  Also bear in mind those thermal parameters are for the standard JEDEC coupon.  It is useful to get an "apples to apples" comparison between different ICs.  Your performance an your PCB may be significantly different.