ADP2443 output fault

Hi RoyLL,

Your point in item 3 tells me something is not right in soldering. The ADP2443 is not damaged because re-soldering makes performance ok again. Can you share your schematic and remeasure waveform again this time with VIN and VOUT in the same waveform capture?

Hi Fil,

Thanks for your feedback.

Resoldering did make two boards performance ok again. However, it didn't work for other boards. From item 3 above, we also analyzed the soldering quality and probability of open circuit of ADP2443 Pins. We make more efforts to test ADP2443 circuit but cannot determine the root cause of recently occurred failure. Here are some more information for your reference:

1. Schematic is as following:

2. For a board which worked abnormally (resoldering didn't not fix it), we changed the R405(RT Pin/150K) with a 280K Resistor to lower down the switch frequency to 600kHz from 1.2MHz. The board came back to work correcttly again. The input (24V) and output (3.3V) waveform is shown in Fig 1.

Fig1. 24V input and 3.3V output waveform --- performance ok

2. In order to evaluate ADP2443 performance degration, R405 was changed to set switch frequency back to 1.2MHz, and adjusted the circuit component parameters as the datasheet suggestions. 

The Vin and Vout waveform is shown in the same capture Fig2 as following:

Fig2: 24V input and 3.3V output waveform --- performance fault state

You can see the output waveform is not stable and overvoltage. Just as the previous Q&A, the situation happened on 6-8 boards in recent days.

3. During the test, we also found some other situations regarding to the abnormal ADP2443 on the failed boards. For two boards, we can test the outvoltage is 3.3V correctly without load (BD15 was removed). When we attached BD15 to connect the circuit, the 3.3V output changed between 4.2V and 6.5V.

The phenomenon of all the failed ADP2443 are all nearlly the same. And the quantity of failed boards in such short time is really cannnot be neglected. Figuring out the root cause of ADP2443 failure is really urgent for us in order to make better design to prevent such failures.

Waiting for your professional support. Thanks and merry Xmas.

Hi RoyLL,

Is there particular reason why you need such very long soft-start time of 388ms? This is not typical in many situations and may not be necessary. Can you try to reduce this value to 4ms?

For item 3 above (3.3V output changed between 4.2V and 6.5V), on these boards can you follow these circuit values and see if performance improves?

Hi Fil：

I'm sorry to reply to you for such a long time.

We can adjust the soft start time to 4ms (Css= 22nf). We originally thought that extending the soft soft time to 400ms (Css = 2.2uf) could reduce the impact of possible surge current to the load. Later, when testing with the current probe, it was found that the output surge current also increased with the extension of soft start time.

After your analysis and recent tests, we also believe that adjusting the parameters to the recommended value is beneficial to the long-term stable operation of the board, but we can't explain clearly to the customer that the cause of the above board failure must be the result of incorrect DC/DC peripheral parameter configuration, because it needs a long time to verify.

We will continue to test and verify the design according to the parameters recommended in the data manual.

Hi RoyLL,

Ok, please follow recommended parts in data manual. If problem persists, please send me the waveforms I asked. You may send me private message so you can send directly to my email.