• State TI Thinks Resolved
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 133 views
  • Users 0 members are here
Support feedback
Options
Options
Related

TDC1000: Gain difference between production batch

selim yönet
Prodigy 10 points
Part Number: TDC1000


Tool/software:

We have been using the TDC1000 in our products for over seven years, and recently we have observed differences in the LNA or PGA gain between certain batches of the component.

Observation:

  • Last year, we produced 400 devices and tested all of them. The results were as expected.

  • This year, our manufacturer produced 50 devices as a pilot batch for mass production. After testing, we noticed that this batch showed an average gain approximately 10% lower than the previous one.

  • To verify, we swapped the TDC1000 between two devices (from different batches), and the difference in gain was confirmed to originate from the TDC1000 itself.

Our questions:

  1. What is the maximum gain variation that should be expected between different batches of the TDC1000?

  2. If a 10% difference is not within the expected range, how can we verify that the TDC1000 parts are genuine and not counterfeit?

Best regards,

  • 0
    TI__Mastermind 44410 points

    Hello Selim,

    We specify the PGA gain error in the datasheet at a typical 5%.  We do not have a specification for the LNA error.  Can you please provide your test data from the different batches.  Also, are you seeing this gain difference in the PGA or LNA?

    As far as checking for genuine parts, we do not verify parts, but if they were purchased from a authorized distributor or direct from TI, they are authentic.  You can find our authorized distributors below.

    https://www.ti.com/ordering-resources/faqs/purchasing-online/authorized-distributors.html   

  • 0 in reply to Eddie LaCost
    Prodigy 10 points

    Hello Eddie,

    The PGA error is defined for 0 dB gain, but in our case the PGA gain is set to 21 dB. Therefore, we cannot directly interpret the gain error from the datasheet.

    Normally, our device signal gain is measured and calibrated by adjusting the drive voltage of the ultrasonic sensor. However, instead of sharing bulk test data (which may not be meaningful), I redesigned the calibration software so that the ultrasonic sensor is driven with a fixed voltage across all devices. Then, I tested one device from an older batch and one from a newer batch, both running the same firmware.

    Old Batch Device LNA output: (First negative swing voltage is ~55mV)

     

    Old Batch Device PGA output: (First negative swing voltage is ~585mV)

    New Batch Devices LNA output: (First negative swing voltage is ~56mV)

    New Batch Device PGA output: (First negative swing voltage is ~556mV)

    From this comparison, the difference clearly originates from the PGA stage.

    We expect up to ~5% variation due to component tolerances, which is why we implemented a calibration step. However, within the same batch of TDC1000 devices, the PGA gain drift is approximately 10%, which is beyond our design expectation.

    Since our manufacturer sources the TDC1000 directly, we are not aware of the exact supply chain and therefore cannot comment on its origin.

    Could you please advise whether this variation between batches is within the expected performance range for the TDC1000, or if further investigation is needed?

  • 0 in reply to selim yönet
    TI__Mastermind 44410 points

    Selim,

    I am looking into this further.  What is the input voltage that was used for these tests?

  • 0 in reply to Eddie LaCost
    Prodigy 10 points

    The TDC1000 is driven by the receiver ultrasonic sensor, while the transmitter sensor is driven with 26 V. Therefore, we do not drive the TDC input directly with a voltage.

  • 0 in reply to selim yönet
    TI__Mastermind 44410 points

    As a test, you can use a function generator to create a sine wave with a specific amplitude that you connect to the input.  Using a method like this could be useful to determine the exact gain by comparing input and output.

  • 0 in reply to Eddie LaCost
    Prodigy 10 points

    We know both the PGA input (the LNA output signal) and the PGA output signal, so we can calculate the PGA gain, right? Based on the test above, the old TDC1000’s PGA gain is approximately 10.63, while the new one is about 9.92. If a test with a signal generator is required, I will perform it.

  • 0 in reply to selim yönet
    TI__Mastermind 44410 points

    Thanks.  I think this is good enough.  You should start the customer return process for these devices where we can perform failure analysis.  If you purchased through a distributor, you can also work with them.  Also, if you have a dedicated FAE(field application engineer) or sales person, you can also work with them.

    https://www.ti.com/support-quality/additional-information/customer-returns.html