• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 30 subscribers
  • Views 165 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TDC1000: TDC1000

Damien Giron
Prodigy 10 points
Part Number: TDC1000
Other Parts Discussed in Thread: TDC7200, EVM430-FR6047, MSP430FR6047

Hello,

We are a team that has developed a contactless water speed measurement system in a pipe, using Audiowell ultrasonic probes arranged in a V configuration, and Texas Instruments components TDC1000 and TDC7200.

Our device is designed to measure speeds between 0.1 and 1 m/s, requiring a precision of 5-10% in the range of 0.1 to 0.2 m/s for alarm triggering. The system must operate stably between 35°C and 80°C. However, we are encountering difficulties in achieving and maintaining accuracy and stability of measurements, especially at high temperatures.

Our electronic contractor has implemented specific modifications, including sending 14 pulses to only retrieve 4 stop bits to compensate for the loss of the first bits by the TDC1000. A software average of 50 valid values out of 100 attempts has been adopted in place of the hardware average, which did not provide consistent results. To counter the effects of the expansion of gel-containing joints between the probes and the pipe, which affect the speed measurement during temperature changes, a mechanical bridging of the probes has been performed. Although the system is beginning to stabilize with an accuracy of ±0.03 m/s, we still observe inconsistent speed variations for several minutes, particularly during temperature changes.

We would like to seek your advice on the following points:

  • Is the choice of TDC1000/TDC7200 well-suited for our application, despite being at the end of the development phase and unable to consider changing technology?
  • Is the practice of sending a high number of pulses to ensure the capture of a few stop bits, as suggested by our contractor, common? Does this technique pose any risks or side effects?
  • Do you have any other suggestions or recommendations to improve the stability and precision of our system?

We thank you in advance for your help and look forward to your recommendations.

Sincerely,

  • 0
    TI__Mastermind 20730 points

    Hello Damien,

    Thanks for posting to the sensors forum!

    • Is the choice of TDC1000/TDC7200 well-suited for our application, despite being at the end of the development phase and unable to consider changing technology?
      • The TDC1000 is not really recommended for this type of application since it technically a flow measurement. There is a bug that misses the generation of a STOP pulse at the first zero crossing which reduces the accuracy of measurements. It seems like you are doing external averaging of several measurements so this should help to remove the influence this missed pulse would have on your measurements.
    • Is the practice of sending a high number of pulses to ensure the capture of a few stop bits, as suggested by our contractor, common? Does this technique pose any risks or side effects?
      • This is not uncommon, the purpose is usually not to generate more STOP pulses but typically the goal is to ensure that there is a healthy margin of return signal which can be done two ways by increasing the sound pressure level (SPL) during a measurement. Typically, you can increase the amount of pulses or others opt into increase the voltage driving the transducer. If you are using two transducers one dedicated to RX and the other to TX then there isn't much of an impact. If you are using one transducer for both this essentially means you will have a longer decay time during your TX sequence which means you will have a longer blind zone.
    • Do you have any other suggestions or recommendations to improve the stability and precision of our system?
      • Are you measuring the temperature in your system? The speed of sound in most mediums changes depending on the temperature so if you are seeing a huge deviation at various temperature points my thought is that you might not be compensating for the drift at temperature.

    Best,

    Isaac

  • 0 in reply to Isaac Lara
    Prodigy 10 points

    Hello Isaac,

    Thank you very much for your detailed response and valuable advice.

    Indeed, we have tested the MSP430FR6047 with the EVM430-FR6047 evaluation board and observed better results. However, the hardware development is already completed, and we had chosen the TDC primarily for its lower power consumption since we already have a microcontroller dedicated to sending results via BLE.

    We are using separate transducers for reception (RX) and transmission (TX), which should mitigate the issues you mentioned. We have also implemented a linear regression to correct measurement errors, as we have noticed that results at room temperature vary with speed.

    Regarding temperature compensation, we use an NTC thermistor in direct contact with the pipe, but it has significant inertia (5 to 15 minutes). What type of correction or approach would you recommend to manage this inertia and improve measurement accuracy?

    We are also reconsidering the idea of mechanical bridging, as we have observed that the measured speed can vary significantly depending on the tightness, ranging from x0.1 to x10. We are thinking of reverting to a simple mechanical fixation to avoid these variations.

    Thank you again for your help, and I look forward to any further suggestions you might have.

    Best regards,

    Damien

  • 0 in reply to Damien Giron
    TI__Mastermind 20730 points

    Hello Damien,

    I did not get a chance to review your reply today. Please allow me to review and provide feedback on your question tomorrow. 

    Thank you!

    Best,

    Isaac

  • 0 in reply to Isaac Lara
    TI__Mastermind 20730 points

    Hello Damien,

    Thanks for the patience here! Thanks for the feedback regarding on the usage for the TDC in this application and for confirming the usage of the seperate TX and RX channels.

    Regarding the NTC thermistor this is an issue that we have seen with various customers in the past. The way other customers have gotten around this issue is by using PTC1000/500 temperature sensors that are directly coupled with the liquid through the pipe. This provides a better accuracy of what the temperature is of the liquid being measured over the temperature of the pipe since there will be a discrepancy between the two. Typically, this is more expensive so its not an option for all customers.

    Damien Giron said:
    We are also reconsidering the idea of mechanical bridging, as we have observed that the measured speed can vary significantly depending on the tightness, ranging from x0.1 to x10. We are thinking of reverting to a simple mechanical fixation to avoid these variations.

    I am not sure I understand what you mean exactly by mechanical bridging here. If you could explain what you mean by this concept that would be helpful.

    Best,

    Isaac