AD5735 DCDC Converters problem

Hi, 

This thread will be move to its appropriate forum (+) Q&A - Precision DACs - EngineerZone (analog.com). Someone here may be able to support you.

However, I also noticed that this part is no longer recommended for new design. Is this a redesign or for new design if it is for a new design the recommended replacement is https://www.analog.com/en/products/ad5755-1.html 

Thanks,

Jellenie

Hi, JellenieR

Thanks for your reply

I think about AD5755 for replacement, but it's also not recommended too.

and about AD5755-1, I think it's need to be change the circuit design ( am i right? )

thanks

Hi, 

To help you better, could you share your schematic and circuit details (op. votlages, load voltage & current connected to the outputs). If it's not convenient to share, you can send it to my email. 

Before we could recommend replacement parts, we' need more info in your application. Is this for a new design or repairing existing designs?

Best regards,

Ian

Hi, iandal

Thanks for your reply

but, i'm not here for searching for replacement parts.

I just want to know what caused the problems.

if you have something to advise me to solve problems.

i will send you more details.

Best regard

Lee

Hi Lee,

This is an interesting problem you have here. Here are some questions I have to help with the identification of the problem:

1) Does the problem occur on freshly built/assembled units?

2) What's the total % of failing units? Can the failure be correlated with specific lot numbers?

3) Is the problem easily replicated? What are the steps required to get a failing result?

4) Are all channels (A, B, C, D) exposed to the same load?

5) What mode is the AD5735 operating at?

6) Can you measure the duty cycle of SWA? If not, can you provide zoomed in waveforms which will only show 2-3 cycles so we can measure the duty cycle?

7) If the load conditions are changed on channel A, do the Vboost_A voltage and SWA waveform stay the same?

Cheers,

Ivan

Hi, isantos

Thanks for your reply

I answering your questions.

1) Does the problem occur on freshly built/assembled units?

- We've been using AD5735 chips for over 10 years and it's only on recently assembled  boards that the problem arises.

2) What's the total % of failing units? Can the failure be correlated with specific lot numbers?

- Recently we made 120 PCB boards and 80% of them had issues.

3) Is the problem easily replicated? What are the steps required to get a failing result?

4) Are all channels (A, B, C, D) exposed to the same load?

- Any conditions ( Load or No Load ) 

5) What mode is the AD5735 operating at?

- I'm using default setup except " Status Read "

   Status register is show the Vout_A Fault

6) Can you measure the duty cycle of SWA? If not, can you provide zoomed in waveforms which will only show 2-3 cycles so we can measure the duty cycle?

- We make boards in our factory in Shanghai, China.

  i got three samples that  problem occurred, and now all of them had been changed AD5735 chips that we had a few in stock. and it's work fine.( about 3 days in testing)

I'll attach a few pictures that after change the chips

can you see that difference between this and i attached one first times. 

7) If the load conditions are changed on channel A, do the Vboost_A voltage and SWA waveform stay the same?

- yes

Best regard

Lee

Hi Lee,

Thanks for sharing the information.

1) Have you checked if all the failing AD5735 units belong to a single lot?

2) Let's say we have a newly assembled unit, what steps are required to get a failing board?

3) With regards to the zoomed in SWA waveform, can you take similar SWA waveforms on a failing board where the VboostA is ~46V? 

4) Since you are operating on VOUT mode, what is the setpoint voltage for ChA output and what is the measured voltage on the failing board?

Cheers,

Ivan

Hi, isantos

thanks a lot, your concern

I got no failing board now,  so i ask our shanghai factory manager to check your request.

my answer is,

1) Have you checked if all the failing AD5735 units belong to a single lot?

- I check three of them ( #1650, 3693079.1 )

   I will do ask to check for rest of failing board at shanghai factory

2) Let's say we have a newly assembled unit, what steps are required to get a failing board?

- No special steps were required to obtain the fault condition.

3) With regards to the zoomed in SWA waveform, can you take similar SWA waveforms on a failing board where the VboostA is ~46V? 

4) Since you are operating on VOUT mode, what is the setpoint voltage for ChA output and what is the measured voltage on the failing board?

- also ask to check to shanghai factory ( It will probably take some time )

Best regard

Lee

Hi,

I am from shanghai factory.

Here are my test results:

Recently we made 120 PCB boards and about 85% of them had issues.

1) Have you checked if all the failing AD5735 units belong to a single lot?

-I have checked all of the AD5735 units,they belong to a single lot ( #1650, 3693079.1 ).

2) Let's say we have a newly assembled unit, what steps are required to get a failing board?

- No special steps were required to obtain the fault condition.

3) With regards to the zoomed in SWA waveform, can you take similar SWA waveforms on a failing board where the VboostA is ~46V? 

-First,I tested the good one:

normal: channel A's SW output ( output voltage is about 15V)

normal: channel A's VboostA output ( output voltage is about 14.3V)

then,I tested the failing one:

abnormal: channel A's SW output ( output voltage is about 50V)

abnormal: channel A's VboostA output ( output voltage is about 45.6V)

4) Since you are operating on VOUT mode, what is the setpoint voltage for ChA output and what is the measured voltage on the failing board?

-I changed the setpoint voltage for ChA output,but the measured voltage on the failing board with no changes,which is always 0V.

Best regard

Michael

Hi michael-shanghai ,

Apologies for the late response, are you still encountering the same problem?

Can you do a readback of the status register (refer to page 38 and 39 of data sheet) and get back to me the readback values?

Cheers,

Ivan