EVM430-FR6047: Have great ADC Capture but still getting error 135 and non-stable flow waveforms

Hi Anthony,

Agreed with many of your points throughout your post. The ADC capture certainly looks good. Agreed with your bulleted list at the end as well I don't suspect these adjustments will fix things.

As a brief comment on the last point - I would still recommend performing a frequency sweep with your transducers to make sure you are using the optimal frequency. Your transducers may be rated for 1MHz but (I am not sure what tolerance applies to these) the best frequency could be 990kHz, 1.02Mhz, etc. Just helps to make sure you're getting the best signal.

This may seem obvious but - you mean that you have one upstream and one separate downstream transducer? This would be the typical arrangement expected by the USS Demo firmware and the template example. But if you are expecting the transducers to transmit in a specific order that isn't one upstream set of pulses and one downstream set of pulses, you'll need to reconfigure the sequence selection. That is all just to be clear, it sounds like you are using a standard configuration. Just something to be aware of.

What is the expected absolute time of flight? Are either the upstream or downstream showing the expected one or are they both wrong? 

If you view the channels on an oscilloscope do the signals actually have a much different time of flight? Or do they look the same - indicating that the error is in the inputs to the algorithm rather than the hardware?

Hello Dylan. Thanks for your prompt reply.

I am using the TIDA-01486 amplifier board with the EVM430-FR6047. The TIDA-01486 can use 2 pairs of transducers. In my original post, I was clearly describing my setup. The implication is that I have made modifications to the TIDA-01486 so that it doesn't do the switching between the 2 pairs of transducers since the EVM430-FR6047 itself can only handle 1 pair of transducers.

So yes, I have one upstream transducer and one downstream transducer which makes up my single transducer pair.

I have some new information regarding my setup. Since one of the suggestions was to increase the value USS_ALG_MAX_SAMPLE_SHIFT, I was looking through the USS gui to find a setting I could change that might have something to do with capturing the received signal and possibly improve the system performance.

So I reduced the Capture Duration from 60 us to 40 us and the 135 error stopped. The ADC Capture waveform was unchanged, but the flow measurement waveforms were suddenly stable and I was able to measure the correct absolute and delta TOFs for the piece of 1-inch PVC pipe. I shortened the Capture duration more and the system kept working properly. Now here is the interesting part - I changed the Capture Duration back to 60 us along with any other setting I had tweaked, and the system kept working. It's as if the measurment algotithm needed to be primed somehow.

Is this possible? Are there calculated values in the algorithm that are hidden to the USS gui that need correct "starting" values for the algorithm to work properly?

My system is working again and I can now clamp my transducers to my stainless steel pipe. But it makes me nervous that I had to "kick" the system like this. Can you comment on this?

Hi Anthony,

Thanks for the added clarity about your set up and testing. I am glad to hear that you were able to find a configuration that improves your results.

I am sure you will be disappointed to hear - unfortunately I am unable to comment specifically on how the algorithm works and finds time of flight. I would recommend re-flashing the device, performing a POR, then requesting an update with your given parameters again. My expectation is that your original error would return at this point until you reduce the capture duration again.