EVM430-FR6047: USS parameter tuning and errors

Hi Cash,

Thanks for getting back to me. I will investigate this parameter next time I have access to the flow meter test rig.

The most frustrating thing we found was that we would have a working configuration and then change a setting that would cause an error. We would then revert to the original settings and the measurement would remain broken. Can you comment on that? Is there any sequence we should perform to ensure the selected parameters are written to the device? Does it need to be stopped before writing parameter changes?

Thanks for your help.

Hi Cash,

Yes, we are using the USS GUI to update the parameters. 

When we click 'Request update' is that not the same as reprogramming the device?

To clarify:

We would have a working configuration, we would then change a parameter, click 'request update' and it would break. We would then change it back to the value that worked previously, again, clicking 'request update' and it would remain broken.

The only way we were able to fix it was by clicking 'reset values' and re-entering the meter constant.

Have you experienced this issue before? We would really like to figure out why changing the parameter back to the original value does not fix the issue.

Is there any sequence we should perform to ensure the selected parameters are written to the device? Does it need to be stopped before writing parameter changes?

Thanks

Hi Cash,

Can you please expand on how to reprogram the device?

Thanks

Hi Cash,

OK thanks I can try this next time I have access to a test rig.

Is it only the envelope crossing threshold parameter that can affect cycle slip? Assuming this is what we are experiencing.

When we tune parameters should we be able to do it with water flowing through the system? Or should the flow rate be zero? Or should it not matter?

I have had success with the standard deviation test here Ultrasonic Water Meter - Design and Testing where we fill the flow meter with water and take measurements at 0 flow rate. I did not experience any issues at no flow rate and I was able to tune parameters freely.

The ADC capture at no flow rate looked like below. 

Still very interested in trying to figure out the relationship between the problem we have witnessed and tuning the parameters. Is there any data we can log during tests that could help you debug this for us? The problem we have is that we only have infrequent access to the test rig and its very expensive so we need to be prepared with solutions in case we encounter this issue again. 

Lastly, I was wondering if you could help me understand how important temperature is when measuring flow rate using this system. Do you recommend the calibration type to be 'Flow' or 'Flow + Temperature' ? I read that temperature can improve flow rate accuracies at <100 l/h. On our first set of tests we calibrated solely using 'Flow'.

Thanks

Hi Cash,

OK we can try and reconfigure at zero flow and see if it makes any difference.

So for the standard deviation test at zero flow i used the parameters in the below screenshots.

I increased the number of pulses to as many as possible as i read this was the best way to improve SD, i also increased capture duration to 80. I reduced UPS0 to UPS1 gap to 50 so we would take more frequent measurements.  We do not care about power consumption in our application.

I increased the GAP between pulse start and ADC capture to 80 in order to make the capture more centered on the plot. Then i adjust the Gain so that the ADC counts are about 900. 

Can you tell me what looks abnormal about it? I would like to try and understand as best as possible how to make the capture as good as possible.

We might be able to set up something very basic in the office. What sort of flow rates would cycle slip occur at?

Thanks

Jonathan