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.

EVM430-FR6047: Optimizing Parameters Based on Type of Gas

Part Number: EVM430-FR6047
Other Parts Discussed in Thread: EVM430-FR6043,

Hi,

I am currently attempting to get a proper signal for various different gases. I understand that, to account for the sound speed of the gas being used, I need to change the ADC capture such that: ADC Start Capture Time = (Ultrasonic Path Length/Speed of Sound) – 40us. (In my case, my path length is ~160mm).

With a gas such as Argon, I am able to get a clear enough signal such that my ToF matches well with the theoretical sound speed of Argon. Below are my settings for the Argon measurement.

When switching the gas to something with a greater sound speed (Methane, Helium, Hydrogen, etc), I am unable to properly optimize my settings to get a good signal. For example, below is the ADC capture I am getting when using Hydrogen instead of Argon.

I am unsure of what additional parameters (such as "Start PPG Count" or "Turn on ADC Count") need to be changed to capture a proper signal. I know my connections are fine, as I am able to switch back to Argon or Air and get a clean signal again. 

  • Hi Andrew,

    You seem to understand the concept of having to configure for the speed of sound in the particular expected environment. I recommend reading the Quick Start Guide for Gas Flow Meter, section 4.2 configuration parameters. If you're increasing the speed of sound then you will need to lower the gaps, in turn decreasing the speed of sound would require you to increase the gaps. The speed of sound of helium is about 3x faster than Air, so by default I would adjust the timing parameters to reflect the increased speed.

    Based on the signal, I would say that your timing is off between when you start sampling and the speed of the samples.

    Here is the link to the Ultrasonic Sensing Design Center User Guide and the configuration parameters, including the "Start PPG Count" and "Turn on ADC Count".

    Regards,

    Luke

  • I agree that the issue appears to be a discrepancy between when I start start sampling and the speed of the samples. Could you be a little more specific on what timing parameters you're referring to?

    The "Gap between pulse start and ADC capture" does not seem to be the issues as I am able to center the waveform in the ADC capture window.

    The "UPS and DNS Gap" seems long enough to ensure that the DNS signal does not have interference from residual reflections of the UPS signal. Additionally, varying it significantly (100 us vs. 14,000 us) seems to have no affect on my captured waveform. 

    The "UPS0 to UPS1 Gap" will just affect the time between two measurements, not affect the waveform of a single measurement.

  • Hi Andrew,

    I'm talking to our USS expert about this and will get back to you on the recommendation for setting changes.

    Regards,

    Luke

  • Hi Andrew,

    Reduce the number of pulses down to 12.

    The total time for the excitation pulses overlaps with the ADC start time, and if you increase the time to start ADC sampling you will miss results. So by reducing the number of pulses, the total excitation time will not interfere with the ADC start.

    Regards,

    Luke 

  • Hi,

    That's what I suspected as well, but no combination of "Number of Pulses" and "Gap between pulse start and ADC capture" seems to be enough.

    For example, here is the capture for 12 pulses:

    And here is the capture for 4 pulses:

    As the issue appears to be an incorrect capture of the pulses being sent, I would think the advanced parameters would might aid in this issue as well. A multitude of different "Start PPG Count" and "Turn on ADC Count" settings does not seem to clearly capture the signal properly either though.

  • Hi Andrew,

    Let's adjust some settings here to better match our tested states.

    • Set the transmit frequency to 170 and 220 (if you're using 200kHz transducers)
    • Set the number of pulses to 8
    • Gap between pulse start and ADC capture to 82 (if you're still using Hydrogen)
      • Derived from the equation you already stated, ADC Start Capture Time = (Ultrasonic Path Length/Speed of Sound) – 40us. Where path length is in meters and speed of sound is meters per second.
    • UPS and DNS gap to 4,000 (this is 1/2 the default value, your setting differs from the Academy as you have ~twice the distance but with hydrogen it is 4x as fast)
    • Gain to 1 dB (if the ADC code swing is large still, lower this. Our target code range is 1000. The minimum gain is -6.5 dB)
    • User Param #8 to 128.
    • Search range to 8.

    I am setting these to the default for most parameters, lowering the pulses, and decreasing the gap to ADC capture. The goal is to center your waveform and get the image to look similar to the tested waveform.

    Regards,

    Luke

  • Thanks for quick response! I am using 200 kHz transducers.

    For Hydrogen, the above settings now yields the waveform shown below.

    These changes seemed to have begun to clean up some of this signal (at least in the UPS case), but I still have the large overshoot at the beginning of the capture.

    Just to see what is happening at the beginning of this ADC capture, I lowered the gain and gap even further, and the waveform then looks like:

    I have little intuition on how to correct for this, as changing pulse number and general timing settings. It also does not appear to be a transducer issue as I am still able to get a proper signal for Argon.

  • As an aside, the title/part-number/and tag of this thread should be EVM430-FR6043 instead of EVM430-FR6047. I selected the wrong part number when I made this post and am unable to edit it. Sorry for any initial confusion because of that!

  • Hi Andrew,

    Can you put an oscilloscope on the AFE input channel0, see below diagram.

    This will let us see this signal that the ADC is receiving.

    Regards,

    Luke

  • Hi,

    I was able to borrow an oscilloscope and capture the signal for both Air and Hydrogen.

    Below are the settings and ADC capture for Air:

    This was the corresponding oscilloscope signal for air. (The x and y axis information is on the top line. i.e., the window shown is 200 us in total)

    When looking at Air, I was curious to what the ADC capture would look like for a similar gap to that of hydrogen. This is show below:

    Obviously I'm not expecting to capture the signal with this gap, but is seeing this much "overshoot" typical? From the discussion above, this seems to be the excitation pulses overlapping with the ADC start time. Is this overlap larger than expected and inhibiting my ability to capture high sound speeds or short path lengths? Would this be a transducer specific issue?

    Next, I repeated the above for Hydrogen instead of Air. Below are the settings and ADC capture for Hydrogen:

    Here is the corresponding oscilloscope capture for Hydrogen. The three images below are the same signal, but progressively zoomed out in time (1 us, 2 us, and 50 us):

    When looking at the oscilloscope capture, is there anything specific I should be looking for to help troubleshoot this issue?

  • Hi,

    The first oscilloscope capture in air looks good to me. And I was expected that you will use another channel to capture the CHx_OUT signal at the same time. If so, you will get a capture similar to the following figure. 

    You can see there is a bias voltage is set up in the yellow signal. And that means the measurement sequence starts at that point. The excitation signal will be generated after that bias voltage set up. Then there is the receive signal.

    The rest of the oscilloscope captures are not good captures. And I am thinking that you are capturing noise signals in those captures. 

    From the discussion above, this seems to be the excitation pulses overlapping with the ADC start time. Is this overlap larger than expected and inhibiting my ability to capture high sound speeds or short path lengths? Would this be a transducer specific issue?

    Yes. This could limit to capture high sound speed or short path length. Lower the excitation pulses could help. 

    Best regards,

    Cash Hao 

  • Hi,

    I increased the path length (now 250mm) in hopes of being able to resolve some sort of useable signal, but seem to be running into the same issues. Below are the settings I used that resulted in the corresponding oscilloscope capture. Does this oscilloscope capture look incorrect in any way? Does it indicate what the underlying issue may be?

    Thanks for all the continued help!

  • Hi,

    I changed to the same configuration as yours. And here is my test result. 

    It seems I can only put three figures in one post. I will start another one to show the oscilloscope capture.

    Best regards,

    Cash Hao

  • As continue,

    You can still observe the bias voltage is set up before the excitation signal. And the time between the excitation signal and the receive signal matches what we see in the USS GUI. 

    However, in your figure, I do not see a valid receive signal there. And I want to correct what I said before. You need to capture the receive in the orange point instead of the red point to observe the voltage bias. Otherwise the voltage bias will be filtered out. 

    Best regards,

    Cash Hao

  • Hi,

    To confirm, are these correct points to capture the signal with the oscilloscope?
     

    Using these connections, here are the settings and corresponding capture for Argon:

    Using these connections, here are the settings and corresponding capture for Hydrogen:

    Comparing the Argon and Hydrogen captures, it appears that I'm able to get a small received signal for Argon but do not see anything for Hydrogen. The Argon received signal seems low in amplitude (which could be from signal attenuation from the increased pathlength). 

  • HI,

    The connections are good. And the receive signal is shown below with Argon.

     

    And for in Hydrogen, ether the receive signal is too small, or the sound speed is too fast and barrow the receive signal in the excitation noise.

    Best regards,

    Cash Hao

**Attention** This is a public forum