• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 78 subscribers
  • Views 947 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.

MSP430FR6047: Technical comparison between MSP430FR6047 and other TI's Ultrasonic ICs such as TUSS4470.

Hossein Forouzani
Genius 3030 points
Part Number: MSP430FR6047
Other Parts Discussed in Thread: TDC1000, PGA460, TUSS4470, , TDC7200

Hello,

I well know TI's ultrasonic sensor ICs, such as PGA460 and TDC1000. Those ICs can be programmed via SPI port with the help of an MSP430 or a C2000 (or any other MCU). Some are direct drive, and some others are the transformer drive types.

Recently I got familiar with MSP430 ultrasonic sensing mictrocontrollers, which are designed for ultrasonic application-specific purposes.
I want to know:
1-if they have any extra benefits in comparison with ICs like  PGA460, TDC1000, and TUSS4470?
2-I think these MCUs have not any output driver submodule and are appropriate for transducers using standard MCU voltage levels (3v).  I need to drive a 100 v transducer so that I'm curious to know whether any of MSP430 Ultrasonic MCUs provides an output driver submodule or not? If no, we should consider an independent circuit for driving a 100v transducer (MOSFET gate driver + Transformer).
Therefore taking into account the BOM, I think using a TUSS4470 + a low price MCU to program it could be more cost effective because TUSS4470 has an output driver submodule.
Does the MSP430 Ultrasonic MCU Design team have any other idea?

Regards,
Hossein

  • 0
    TI__Intellectual 2100 points

    Hi Hossein,

    The 'FR6047 is tailored for flow metering. Some system advantages are...

    - resolution in picoseconds for difference time of flight (acoustic delay)

    - energy/measurement  <3nAs

    - HW supported sequencer for 1 transducer pair

    - high RX sensitivity to be able to measure without high energetic ultrasound pulses

    - very even running Oscillator/PLL with 5ps error

    … those performance values cannot be achieved with the other devices. The other devices are suitable to analyze ultrasonic echoes, and measure larger, not quite that fine delays.

    Energy consumption is one of the major concerns in battery operated flow metering; therefore we are not driving high energy pulses of into transducers, but rather clean and ultra symmetric ones => this gives you much better results than brute physical force.

    The target applications are utility flowmeters for gas, water and heat starting from DN15...DN40/50 .  Our customers are using the 'FR604x devices also with larger diameters, here with external drivers to overcome the then higher channel attenuation. This then with external driver-stages to keep the distortion of the time base as small as possible. 

    In other words the devices you listed are meant for different uses... of course can they be used for other things.

    have a nice day

        Johann

     

  • 0 in reply to Johann Zipperer
    Genius 3030 points

    Hi Johann,

    Thank you for your reply.

    Although the items you mentioned are positive points of MSP430 Ultrasonic Application-Specific MCUs, some of them are not distinctive features and can be found in Ultrasonic ICs such as TDC1000, For example, picoseconds resolution is crucial to low flow measuring and this feature of MSP430FR6047 could be highlighted as one of the two most significant specifications of it. However, it is not an exclusive feature of MSP430FR6047. In order to achieve picoseconds resolution, a precise time-to-digital converter such as TDC7200 can be used to measure the time of flight duration. Therefore in my point of view, this is not encouraging enough to persuade me to shift from TDC1000 to MSP430FR6047.

    I think the main advantage of MSP430FR6047 is its integrity which leads to BOM reduction. There should be no more need for external ADC, SPI communication between an Ultrasonic IC and a Microcontroller.

    In addition to this, 'Optimized ultra-low-power standby mode' is another important feature of MSP430FR6047 (maybe unique in comparison with other devices used in ultrasonic measurement)

     

  • 0 in reply to Johann Zipperer
    Genius 3030 points

    Johann Zipperer said:

    The target applications are utility flowmeters for gas, water and heat starting from DN15...DN40/50 .  Our customers are using the 'FR604x devices also with larger diameters, here with external drivers to overcome the then higher channel attenuation. 

     Based on the technical feedback from your customers, what is the largest diameter could everyone have used MSP430FR604x devices to measure the flow rate?

    I saw in one of TI's documents that MSP430FR604x MCUs support up to DN1000. Is this genuine?

  • 0 in reply to Hossein Forouzani
    TI__Intellectual 2100 points

    Hi Hossein,

    yes, you are right on telling the PS resolution can be achieved other ways.... ,   and yes you are right again to tell it is the integration that counts on the 'FR604x devices.... . There are several ways to measure transit times; one is to use a stopwatch; the other is correlation. The problem with stopwatches are the max slew rate of the sensor signals. Any small distortions on power supply will lead to a time distortion. For separate dies this is not a problem, you use one silicon process for the MCU, another one for the sensor frontend, but having a stopwatch on the same die with an microcontroller it is. You would need to use a process that allows large scale integration. Those have even flatter slew rates... You would need to hide the noise of CPU and the peripheral tasks that are running in background from the ultrasound frontend. This is difficult with a stopwatch.... since same dies and just 1mm away!

    A correlation approach removes all that shortcomings, that's why. Correlation has a lowpass property; if you concept is laid out smart you could exploit coherency effects that allows your own generated noise to be invisible, and so on.

    This were some of the original ideas why we've chosen this approach...

    Large diameters are tricky...  you've got the channel attenuation and the "BlowOff" The flow coming from the side changes the direction of the sonic club. The main beam might not hit the RX-transducer anymore...

    e.g. see the planar simulation below...

    At zero flow everything I still ok, but as soon flow is added...

    The effect is remarkable the narrower the club is (the larger and higher sensitive the transducer disk)

    Below some 2D simulations.

      have a nice day

          Johann

  • 0 in reply to Johann Zipperer
    Genius 3030 points

    Johann,

    Johann Zipperer said:
    There are several ways to measure transit times; one is to use a stopwatch; the other is correlation. The problem with stopwatches is the max slew rate of the sensor signals. Any small distortions on the power supply will lead to time distortion. For separate dies, this is not a problem, you use one silicon process for the MCU, another one for the sensor frontend, but having a stopwatch on the same die with a microcontroller it is. You would need to use a process that allows large scale integration. Those have even flatter slew rates... You would need to hide the noise of CPU and the peripheral tasks that are running in the background from the ultrasound frontend. This is difficult with a stopwatch.... since the same dies and just 1mm away!


    1-Therefore now we can draw another conclusion:
     'FR604x' MCUs due to their integrity of peripherals (Analog Modules) and CPU, use 'correlation' method to measure the TOF (Time Of Flight), so that the possibility of time distortion in 'Stopwatch' method, caused by e.g., power supply possibly distortion could be eliminated; while Ultrasonic ICs such as TDC1000 are designed to work with 'Stopwatch' method to measure the TOF' because there is no concern with parameters such as slew rate of sensor signals, time distortion, and noise; this comes from using Ultrasonic IC, Time-to-digital IC, and MCU separately.(e.g TDC1000+TDC7200+a microcontroller like an MSP430)


    Thank you for your detailed post.
    2-However I'd like to know if there is any referable experimental fact showing what has been the maximum diameter of a pipe that one of 'FR604x' MCUs has measured its flow? Is there any valid data? The image below is taken from slau720b.

    3-What are the documents discussing 2D simulations? I'm looking for more details and further readings.

    BR
    Hossein

  • 0 in reply to Hossein Forouzani
    TI__Intellectual 2100 points

    Hello Hossein,

    there are plenty of free simulation tools out there e.g. "OpenFoam" using the native geometry entry, or via "FreeCAD".

    Those allow also volumetric simulations...   To explain some simple effects I use an old simulator than can 2D only.

    My simulator is available in German only with an rather exotic way to describe the geometry.

    The 'FR604x devices allow additional tricks; you can tune the excitation frequencies to the optimum, you can sense the aging of the transducers by observing the channel attenuation, eigen-resonance of the excited transducer (this be able to observe deposits building up). You can listen if other meters are in acoustic neighborhood etc....

    Hossein have fun

          Johann

  • 0 in reply to Johann Zipperer
    Genius 3030 points

    Hi Johann,

    Thank you for your time.

    Have a nice day.

    Hossein

**Attention** This is a public forum