• State TI Thinks Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 711 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Original question:

[FAQ] TDC1000: Everything You Need to Know for Ultrasonic ToF (Liquid-Coupled) [Collateral, Tools, Designs, FAQ]

TDC1000: Measuring the thickness of Steel or Aluminum blocks

George Gendy1
Prodigy 145 points
Part Number: TDC1000
Other Parts Discussed in Thread: TDC7200

Hello,

I am using a custom board that is intended to measure the thickness of very thick steel block (up to 360mm). Please see the attached pictures for an example (the example shows the sensor attached to an aluminum block though as this is used for initial testing) 

The board uses TDC1000 and TDC7200 for capturing the time of flight. TX signal is raised to 24V. 

The original thread, that this one is linked to, has nice steps to help figuring the setting of the above chips but I appreciate if I can get some feedback on the below points as the trials that has been done so far does not seem to provide good results.

1. Looking at the speed of sound in steel or aluminum, I see that normally the speed is listed in many websites mainly for  3 kind of waves, Longitudinal, Shear, and Extensional. I am not sure which one should I use in my calculations for these blocks. Please advice.

2. For some reason, I am getting the same waveforms regardless of the block thickness to which the US sensor is attached (please see attached waveform: Channel A is Start signal, Channel B is Stop, Channel C is the TX signal). One picture shows the whole sequence and and the other one is zoomed-in to see one of the echo and stop signals. 

3. The sensor that I am using is 1MHz (SMD15T21R111WL). It is currently attached to the block using Scotch double sided tap in both sides (top and bottom). We will use appropriate epoxy in final application. 

4. The settings of the TDC1000 are currently set to:

* Click divided by 8

* Num_TX =3

* NUM_RX=3

*  ECHO_QUAL_THLD_125

* RECEIVE_ECHO_MULTI (for distance/thickness measurements)

* High bit of the TIMING_REG is set to 0x03 to enable standard mode. Force short time of flight is not set.

Your feedback is much appreciated.

Thanks!

George

 

  • 0
    TI__Mastermind 21250 points

    George,

    The best speed to select for your steel block will be for the longitudinal wave. 

    The TX pin isn't the best location to monitor for the return echo.  If you could additionally monitor at the COMPIN pin it will be easier to distinguish what is happening.  Can you monitor here and verify your two thicknesses appear identical? 

    For the connection to the material, it is best to limit airgaps as much as possible.  Instead of tape you might try super glue which would be easier to work with than the epoxy.  I do understand you are trying to prove the concept with a temporary setup, but this coupling can impact performance.

    Thanks,

    Scott

  • 0 in reply to Scott Bryson
    Prodigy 145 points

    Hi Scott

    Thanks for the feedback!

    Please find attached the waveforms after attaching the sensor to the aluminum blocks using super glue. Away from the first row, results of TOF for both the blocks are close. Also I don't see a specific setting that can get good results for both. I hope the format that I put below appears nicely here as I don't see it clearly while editing here. 

    Column 1 shows the test settings. Column 2 shows the results for a 51mm block.  Column 3 shows the results for a 90mm block.

    3 tx, 3 stop, standard TOF (timing register=0x30), PGA_GAIN_15

    TOF (micro sec) Aluminum distance (mm), 3 tx, 3 stop, standard TOF (timing register=0x30), PGA_GAIN_15  
    (6420*TOF)/(2*1000) Notes
    28.8 92.448  
    43.6 139.956  
    51.3 164.673  

    TOF (micro sec) Aluminum distance (mm), 3 tx, 3 stop, standard TOF (timing register=0x30), PGA_GAIN_15  
    (6420*TOF)/(2*1000) Notes
    30 96.3  
    60 192.6  
    68 218.28  
     3 tx, 3 stop, short TOF (timing register=0x00), PGA_GAIN_15, SHORT_TOF_BLANK_64

    TOF (micro sec) Aluminum distance (mm), 3 tx, 3 stop, short TOF (timing register=0x00), PGA_GAIN_15, SHORT_TOF_BLANK_64  
    (6420*TOF)/(2*1000) Notes
    12 38.52 only 2 stop pulses are received
    20 64.2 only 2 stop pulses are received
      0  

    TOF (micro sec) Aluminum distance (mm), 3 tx, 3 stop, short TOF (timing register=0x00), PGA_GAIN_15, SHORT_TOF_BLANK_64  
    (6420*TOF)/(2*1000) Notes
    12 38.52 error 6 in error flag but still 3 stop shown in scope
    17 54.57 error 6 in error flag but still 3 stop shown in scope
    26 83.46 error 6 in error flag but still 3 stop shown in scope
     2 tx, 2 stop, short TOF (timing register=0x00), PGA_GAIN_15, SHORT_TOF_BLANK_64

    TOF (micro sec) Aluminum distance (mm), 2 tx, 2 stop, short TOF (timing register=0x00), PGA_GAIN_15, SHORT_TOF_BLANK_64  
    (6420*TOF)/(2*1000) Notes
    12 38.52  
    17 54.57  
      0  

    TOF (micro sec) Aluminum distance (mm), 2 tx, 2 stop, short TOF (timing register=0x00), PGA_GAIN_15, SHORT_TOF_BLANK_64  
    (6420*TOF)/(2*1000) Notes
    12 38.52  
    16 51.36  
      0  

    TOF (micro sec) Aluminum distance (mm), 3 tx, 3 stop, standard TOF (timing register=0x30), PGA_GAIN_15  
    (6420*TOF)/(2*1000) Notes
    28.8 92.448  
    43.6 139.956  
    51.3 164.673  
      0  
      0  
  • 0 in reply to George Gendy1
    TI__Mastermind 21250 points

    George,

    Thank you for the detailed captures.  The forum does format a little funny, but I think I am able to follow. We will review these and see if we can't spot what's causing the trouble. 

    Scott

  • 0 in reply to Scott Bryson
    Prodigy 145 points

    Scott,

    Thanks a lot!

    Looking forward for your feedback (hopefully soon)

    Have a great weekend!

    George

  • 0 in reply to George Gendy1
    Prodigy 145 points

    Hi Scott,

    Just  a quick clarification while you look at the waveforms. Eventually when we get sorted out, measuring the TOF should be based on using the same stop signal for different block sizes, right?

    For example, if we came to the conclusion that stop signal # 2 is the one that gives accurate results, we should use that specific stop signal when we measure the TOF for different blocks. Please confirm.

    Thanks,

    George 

  • 0 in reply to George Gendy1
    TI__Expert 3450 points

    Hi George, 

    The STOP that should be used to determine the ToF should always correspond to the received echo on the COMPIN pin. I see in your plots that the received echo appears at the expected time (on the TX waveform), measurement reliability will come with fine tuning. 

    Configuring the device in short ToF mode is required for ToF measurements this small. Circled below is the received echo waveform appearing on TX:

    51mm:

    90mm:

    Please send a list of your device register settings and we will have a look. Meanwhile, experiment with different PGA gain and SHORT_TOF_BLANK_PERIOD values to see if you can get a better response on COMPIN. Also ensure that the Echo/Listen period is set properly for both blocks. If a bandpass filter is not in use, consider installing one for better noise rejection, see the section 8.3.5 of the data sheet. 

    Regards,

    Gabriel

  • 0 in reply to Gabriel Lake1
    Prodigy 145 points

    Hi Gabriel,

    Thank you for the quick feedback! This was spot on :)

    After using the below settings I got better wavefroms which is much better than the previous ones. However, fine tuning still needed as the the numbers are still a bit off.  Please see below waveforms and register settings and advise on the tuning process/register values.  I will also need to repeat the process though when I test the steel block. 

    Settings:

    Num_TX = 3

    Num_RX = 3

    ECHO_QUAL_THLD_125

    RECEIVE_ECHO_MULTI

    Timing register = 0

    PGA_GAIN_0 (also PGA_GAIN_3 works fine)

    TOF_TIMEOUT_1024, SHORT_TOF_BLANK_128

    CONFIG_0(0):0x43
    CONFIG_1(1):0x43
    CONFIG_2(2):0x00
    CONFIG_3(3):0x03
    CONFIG_4(4):0x5f
    TOF_1(5):0x00
    TOF_0(6):0x00
    ERROR(7):0x00
    TIMEOUT(8):0x23
    RATE(9):0x00

    TOF (micro sec), first stop (6420*TOF)/(2*1000) Notes
    20 64.2 51mm block
    30 96.3 90mm block
  • 0 in reply to George Gendy1
    TI__Mastermind 21250 points

    George,

    I found two sources for the speed of sound in Aluminum. One matches the 6420 m/s  (which you are using) and another listing at 6320 m/s.  Looking at the pulse you captured, it looks like you probably should have triggered a little sooner.  If we use the 6320 m/s value for Al and calculate backward the TOF expected for 51 mm, then we should see a echo time of 16.14 us. 

    With the SHORT_TOF_BLANK_128 setting, the wait time will be 16 us, which seems to be the likely cause for your delayed trigger.  We recommend that you try to set the blank period to the 64 setting.  You may then need to adjust your threshold and/or gain to accommodate any changes in the noise preceding your echo.

    Can you give this a try and let us know if it helps you get a better result? 

    Thanks,

    Scott

  • 0 in reply to Scott Bryson
    Prodigy 145 points

    Hi Scott,

    Once more, thank you for the great feedback!

    I could get better results with the below settings (please note that I also have to change the NUM_RX to 1 and leave Num_TX to 3).

    However, during the trials to get these working settings I saw some weird behavior: in some of those trials I set the NUM_RX to 3 and Num_TX to 3 and I saw 3 stop signals on the scope but the flag register still reported error 0x02 or 0x04. Same thing happened when I set both the NUM_TX and NUM_RX to 1. I saw the stop one signal but the flag register still report the above error. Any idea why.

    Working settings:

    Num_TX = 3

    Num_RX = 1

    ECHO_QUAL_THLD_410

    RECEIVE_ECHO_MULTI

    Timing register = 0

    PGA_GAIN_0 (also PGA_GAIN_3 works fine)

    TOF_TIMEOUT_1024, SHORT_TOF_BLANK_64

    Waveforms from working settings:

    Waveforms from the settings that resulted in error flags (but correct number of stops signals showed up on the scope):

    Thanks,

    George

  • 0 in reply to George Gendy1
    TI__Mastermind 21250 points

    George,

    Increasing the TX pulses will increase the excitation of the transducer and should result in a larger echo signal up to a point.  The returning echo must be large enough to trigger the STOP event.  The number of RX events is how many possible STOP events you want to capture. 

    If the echo is not large enough to trigger all the STOP events you are trying to capture, then the device may return ERR_SIG_WEAK or ERR_NO_SIG depending on how many trigger events occur.  This flag is only reset when read, so it might be possible that the register was still set after trying a previous setting.

    It looks like you reported your setup is working now, but if you would like to troubleshoot the other settings further I'd recommend that you run several iterations on your setup and ensuring that you are reading these error flags after every TOF measurement.  If any variations are observed from test to test, we could then try to pin point the source of the flag.  

    Thanks,

    Scott