THS4521: THS4521IDR Performance problem

Hello user,

  A few questions:

1. What is the specification are you measuring in dB?
2. What is the total A vs total C vs total Z represent?
3. Would you be able to share a schematic with us?
4. How are you getting abnormal vs normal waveforms? Example: is it with different units but same schematic? 

Thank you,

Sima

1. What is the specification are you measuring in dB?

“dB” is the unit of sound pressure level .

The sound pressure level is a commonly used physical quantity in Acoustic measurement

2. What is the total A vs total C vs total Z represent?

The total A represents the total value of A-weighted sound pressure levels,

the total C represents the total value of C-weighted sound pressure levels,

the total Z represents the total value of Z-weighted sound pressure levels,

three weighted networks are used in different test scenarios.

Ignore A-weighted sound pressure levels and C-weighted sound pressure levels,

Focus on Z-weighted sound pressure levels.

Z-weighted means ZERO weighted namely no weighting is carried out.

3. Schematic

4. How are you getting the abnormal vs normal waveform?

   Test conditions:

   The same PCB,

   The same schematic,

   The only difference is the chip(THS4521).

5. 
6. 

7. Our instruments have been in mass production for five years.

   The waveform of the instrument where abnormal background noise is found during the

   production process is shown in Figure 1.

   This problem was solved by replacing the chip(THS4521).

   Then test again to get a normal waveform, as shown in Figure 2.

   All measurements are tested on the same instrument and only replaced the chip(THS4521).

   The failure rate of this batch of instruments has reached 25%, and the problems are the same.

   The solution is the same, all by replacing the chip(THS4521).

Hello, 

I expect that your circuit is likely unstable. It is quite difficult to get an ideal integration circuit stable with no parallel feedback resistance included in the circuit. I would suggest adding some feedback resistance in parallel with the capacitance and see if that solves the issue. 

Best, 

Jacob 

There are feedback resistors in the circuit, and different magnifications can be switched by the ADG409.

There are feedback resistors in the circuit, and different magnifications can be switched by the ADG409.

Hello, 

In the below schematic image I do not see where there are feedback resistors connected between the THS4521 outputs and inputs. Is there another connection that is now shown in the schematic? 

Hello,

There are feedback resistors in the circuit, and different magnifications can be switched by the

ADG409. One of the magnifications

Hello,

  Is this schematic the one of the multiple THS4521 in use? From our understanding, you are using THS4521 in different gain configurations which in production is being switched via the ADG409. Then does this mean the THS4521 exists in design as configured either as first schematic or second. If so, then are you seeing an issue from the first schematic configuration or the second schematic configuration? And, is this 25% failure rate issue seen for the THS4521 configured the exact same way?

Thank you,

Sima   

Hello,

1. This schematic is the one of the multiple THS4521 in use.
2. The first schematic can be configured with the ADG409 for 3 different feedback networks,

each with a different gain and the second schematic as one of the gains. The second schematic is

the equivalent form of the first schematic when the ADG409 control pins are configured to

00.The second schematic is just to answer the question Jacob Freet asked, indicating that there

are feedback resistors in the first schematic.There is only the first schematic on the PCB.

3. The 25% failure rate issue is seen for the THS4521 configured the exact same way.

Please help confirm the processing result of the above reply

Hello,

1. This schematic is the one of the multiple THS4521 in use.
2. The first schematic can be configured with the ADG409 for 3 different feedback networks,

each with a different gain and the second schematic as one of the gains. The second schematic is

the equivalent form of the first schematic when the ADG409 control pins are configured to

00.The second schematic is just to answer the question Jacob Freet asked, indicating that there

are feedback resistors in the first schematic.There is only the first schematic on the PCB.

3. The 25% failure rate issue is seen for the

Hello,

  Sorry for the delay and thank you for the additional information.

  Does this mean that the first schematic is showing the THS4521 configured when all three feedback resistor switches are open? When switching to other gain configurations using the failed THS4521 units, do they pass or continue failing?

Best Regards,
Sima 

Hello！The first schematic shows three feedback networks.
The ADG409 is controlled via an FPGA and only one feedback network is connected to the THS4521 at a time.

Hello,

  I understand that you have at least one feedback network connected during the test. Does the THS4521 unit fail for each feedback network selection, or only specific feedbacks or magnifications? This way I can run a stability analysis on the failing configuration.

   Also, I would suggest doing an ABA swap which is taking out the failing unit, and soldering it to the passing board. Then, check if the device measurements is failing or passing. If the measurement is passing, then place the unit back on the failing board and test again, and let us know how it performs back in the same board.

Thank you,
Sima 

The faulty THS4521 measures an abnormal waveform under all three feedback networks.
We have already made an ABA swap and come to the following conclusions :
The faulty THS4521 is installed on the passing board and still measures the abnormal waveform mentioned above.

Hello,

  In that case, I will run a simulation to check the stability of the system. If it is close to instability, some units might pass and others might fail due to tolerances in the open loop gain of the device. I will get back to you shortly with an update.

Thank you,
Sima 

Please help to confirm whether there is any new progress
thank you

Please help to confirm whether there is any new progress
thank you

Please help to confirm whether there is any new progress

Hello,

  Sorry for the delay. I ran simulations on the the last attached schematic, and below are the results:

   The gain is correct, and circuit is showing a closed-loop bandwidth of close to 10MHz. 

   The phase margin of the circuit is stable at around 73 degrees of phase margin. For a circuit to be stable, the phase margin has to be around or greater than 45 degrees. 

   Please confirm is simulation is showing the correct schematic and component values as you are seeing during your fault conditions. Also, what is the value of the output capacitance, C132, this could drastically change the stability of the circuit. 

THS4521_Sound.tsc

THS4521_SoundStability.TSC

Thank you,

Sima 

Hello, the file provided cannot be opened, so we have sent it to the testing institution for testing. Please help confirm the specific reason given in the testing report

thank youProduct Test Report.pdf

Hi,

Please allow us time to look over the test report you sent over and we will get back to you if we have any questions or updates.

Best Regards,

Ignacio

Ok, do you have an approximate time? When do you expect the result?

Hi,

We had one of our team members with experience in the failure analysis process look over the document and we are reaching out to the appropriate team to further investigate the findings in the report that can confirm if there are any quality/manufacturing concerns. I apologize for any delay; these processes sometimes take their time to work through. We will reach out with an update as soon as we have anything to share.

Best regards,

Ignacio

Hello

Could you please take a look at this report and confirm the results? thank you

Hello,

  Thank you for conducting and sharing this report. We had our test manager look over the report, and we were advised to conduct further analysis. I sent over the report to our quality and reliability expert. I will follow up to get a response as soon as possible. We apologize for the delay on solving this issue.

Best Regards,

Sima

Hi

Could you please help me confirm the latest progress of this case? thank you

Hello, 

  Our team expert looked over the report, and our next step is following up with the fab lot quality team. The cracks/burn marks mentioned in the report are laser trim resistance. However, we are looking at how this device was trimmed with the fab, and they will conduct further investigation and analysis. Our team will push for a quick response and analysis report. 

Thank you,
Sima 

Hi

According to the information you replied, does the burn mark belong to the product itself? So is this the reason why we use the exception? Can we apply for a return?

Hello,

    Yes correct, these burn marks are intentional via laser trimming.

     Our FAB expert looked over the report. Their comments on the trim quality is normal, and there should not be issues from the trims in the report. 

     We wanted to ask how many units in total are in production, and have these units been in production for 5 years as said in earlier replies? If so, did the failure rate of 25% of the units appear after production or are these units new to the lot production?

Thank you,
Sima 

Why are these cut marks not in the cut path but in the middle? At present, 600 tablets have been used, and the defect rate is about 12%. This problem is very bad, so we hope your company can help find the cause of the problem

Could you please help me reconfirm the questions we raised in the previous article? thank you

Hello,

  That is the path where resistance is trimmed; the trim amount relates to the amount of resistance. That is a large quantity of defects. I talked to our FAB expert, and he explained that these trims should not be an issue. 

   Where you able to perform a Decap test on an OK part? I believe the report only shows a not good part unit 2.

Thank you,
Sima 

So what causes the stratification in our test report?

Hello,

  Are you referring to the delamination? Our expert mentioned that it is normal, but it would be helpful if we can compare an OK unit to a NG unit in both the SAM test and the de-cap test.

Thank you,

Sima