PGA460 ultrasonic sensing: EVM GUI data monitor
This video provides a demonstration of ultrasonic echo data dump data and ultrasonic measurement results for the PGA460 Evaluation Module (EVM) using the GUI software. The ultrasonic activity can be depicted as an echo data dump plot and/or ultrasonic measurement results, which includes the distance, peak amplitude, and width of the returning echo. To maximize the scaling of the signal-to-noise ratio, optimizing the gain is necessary to set the threshold for reliable and repeatable results.
Resources
Now that we have configured the critical device settings, let us go to the Data Monitor page to view the performance of our device with the current settings. Since I've configured preset 1 using the block diagram, I must ensure that the preset 1 radio button is selected in my run options. I want to run a burst and listen command to transmit and record a round trip time of flight based echo.
To confirm my GUI will visually print or plot results, I must confirm that either the plot data dump or list measurement results are checked. I will hit Start to initiate the burst and listen command sequence.
The data monitor has been updated to plot the echo data dump results. Here we can see that across time, the echo data dump profile changes. Initially, we are bursting, which is why our signal appears saturated, followed by the ringing decay period of the transducer to reach the noise floor, at which point we are then able to see our returning echo. The returning echo is at an equivalent distance of approximately 1.99 meters, rounded up, 2 meters.
At this time, the peak of the echo is at a value of about 53 out of 255, based on the chart cursors. This indicates that we have successfully captured the reflection from the ceiling at about 2 meters. However, we are able to make improvements to this profile by modifying the device settings. We will mainly be tuning the analog gain profile, as well as the digital gain profile.
Let us quickly modify our digital gain multiplier to see what type of immediate impact that setting will have. Note that the digital gain short-range versus long-range application time is designated by the green dotted vertical line. Recall that our digital gain settings requested the long-range multiplier to be applied at the point equivalent to threshold 9. If we were to count our threshold values, we would see value 1, 2, 3, 4, 5, 6, 7, 8, 9. The echo response is already in the long-range multiplier window. Therefore, let us increase our long-range multiplier on the general page.
While still using preset 1, I will increase the long-range gain multiplier to times 4. Going back to the data monitor page and hitting Start, we will see that we have slight improvements to our echo data dump plotting of the actual returning echo.
Let us further increase the long-range gain to a multiplier of times 8. We have scaled the echo data to plot even more so. Now our signal surpasses and crosses the threshold. This will allow us to use the list measurement results.
By checking the list measurement results, we will now be able to pull the ultrasonic measurement results and print the distance width and peak amplitude of the reflected object. Keep in mind that the echo data dump plotted will not be identical to the ultrasonic measurement results, given the ultrasonic measurement results and the echo data dump cannot be plotted and computated for the same burst and listen cycle. The PGA460 is only able to either plot the echo data dump or calculate the measurement results. Therefore, by checking both of the display format options, the burst and listen cycle is run twice, first, for the echo data dump, and then secondly, for the ultrasonic measurement results.
Here we can see that our distance read out for the object crossing the threshold is at 2.024 meters with a width of a 172 microseconds and a peak amplitude of 212. This equates very closely to the values that we see on our echo data dump.
If I were to increase my digital gain by a larger multiplier of times 16, we are now experiencing a saturated echo point. We are still able to utilize the threshold to compute the distance, but notice that our peak amplitude readout is now 255, which is the maximum read out the PGA460 is capable of detecting. As a result of this multiplier, we are artificially decreasing our SNR, such that only our noise floor will continue to increase by applying any additional gain, regardless if it be digital or analog. Therefore, a multiplier of times 8 is most ideal for the current analog front end gain conditions.
Now that we have configured our AFE and digital gain for the best case SNR according to our echo data dump, let us fine-tune the threshold to ensure we are able to read out the most stable ultrasonic measurement results, while ensuring we leave enough margin in case the noise floor fluctuates significantly. Using the preset 1 timing and level options, let us drop our level down to a value of 24. Then let us set all the other values to 24 by clicking the All Level 1 and Time 1 button.
Here we can see that the threshold gives significant margin to the noise floor, which is at a value of practically 0. But as the analog front-end gain has increased, so has our noise floor. At some point, our threshold may be at the same level or below the noise floor. Therefore, let us increase the threshold as the analog front-end gain increases in the long distance.
First, let's add additional space in between each threshold level. My threshold now scales across the entire profile. Starting at threshold level 6, I will increase my levels such that they allow for sufficient margin from the noise floor fluctuations.
The offset is not necessary for this particular use case. By rerunning our burst and listen command, we will notice that our return echo decreased. That is because the long-range digital gain multiplier has been moved to initiate at a distance equivalent of about 4.4 meters. Since threshold 9 is the earliest threshold point at which the long-range digital gain multiplier can be activated, we must apply a digital gain multiplier of times 8 for the short-range.
Now we have recreated the scaled SNR signature, as previously seen. You will notice that our distance is repeatable within 1 millimeter precision. By setting my number of loops to 0, I can infinitely run my burst and listen command to monitor the distance width and peak amplitude repetition. I am also able to log my ultrasonic measurement results and view these results in the utilities data log window. By logging these results, I can post-process and analyze the stability, tolerance, and effectiveness of my threshold.
You will notice that two returns are actually present, one at 2 meters, and the other at about 4.2 meters. This second reflection at 4.2 meters is my ultrasonic echo traveling to and from the ultrasonic transducer twice. Since the second round trip echo is registered and appears as a valid signature, we could increase our number of objects to detect to measure the true distance width and peak amplitude of this second reflection.
You'll see our ultrasonic measurement results table has now been updated with the results of two objects, our initial expected object and our second residual object. The ultrasonic measurement results is able to print the distance, width, and peak amplitude for up to eight objects. If no object is detected, the maximum value will be displayed instead, indicating that the results calculated was not triggered for these objects.
This video is part of a series
-
Ultrasonic sensing with the PGA460-Q1
video-playlist (6 videos)