MSP430FR6043: Water Meter USS Configurations

Hi Dylan,

Apologies for the late response.

There are no bubbles in the line and the envelope crossing threshold is set at 35(%), But the geometry of the curved reflective surface of the pipe and the sensor arrangement was a little off, which was giving me Peak-to-Peak amplitude of the Rx signal of approx. 200mV as observed on the PicoScope which is way below the threshold specified in the Audiowell sensor datasheet of 350mV.

On correction of this, the observed peak-to-peak amplitude is approx. 750mV. I am attaching the PicoScope capture and the updated TI USS Design Center Software Parameter Configurations of the same. I am using this for flow testing and validation as well.

The problem I am now facing is that at low flowrates of less than 25 or 30 LPH, the flowrate is computed correctly but at anything beyond that it is either zero or a negative value. I think I should adjust the Gap between Pulse Start and ADC Capture Parameter to ~90us. Do I have to vary any other parameter for correct delta TOF detection? (Based on the PicoScope Capture attached)

MSP430FR6043_RawWaveformCaptures_GeometryCorrection.pdf  

Hi Rishab,

No worries.

Are you using the EVM here or a custom PCB? If you are using the EVM we could take a look at the ADC capture window to see if there is some issue there. Looking at the picoscope window it is a bit harder to pinpoint the issue but the waveform in your pdf looks fine and the amplitude you mention is pretty typical. You could try increasing the programmable gain to see if this helps but I am not immediately concerned with the amplitude.

You may need to perform flow VFR calibration, as shown in the USS Design Center Users guide for the varying flow rate.

Hi Dylan,

The EVM is being used here. I had to vary the Gap between Pulse Start to ADC Capture parameter to 82us. Now, the ADC capture is stable even on water flow. 

I am attaching the waveforms observed. Are the Abs ToF and delta ToF acceptable here?

The Reference flowrate was approx. 600 LPH but the EVM detected only approx. 25 LPH. To perform the VFR calibration, I am calibrating with only one range to start with and the Range 1 Calibration Points are:

The Calibration Parameters are:

Even after I update/write this to the EVM, it still detects only approx.25 LPH. Any ideas on what I should change here such that it comes close to my reference meter range of approx. 600 LPH?

Thanks for the info here. 

Whether the delta and absolute ToF are acceptable is really up to you - the questions to ask are 

1) Is it resulting in the correct flow rate with acceptable error? If yes then great, if not then go to #2 and #3

2) Is the absolute time of flight accurate based on the distance between transducers, the flow rate, and the speed of sound in the medium you are testing?

3) Is the delta time of flight accurate based on the distance between transducers, flow rate, and speed of sound in the medium?

If 2 and 3 are are answered with yes, then you probably just need to change the meter constant for the given flow range. If either of them are answered with no, you need to go back to adjust some of your other parameters. The best way to determine if your parameters are acceptable with the EVM is to look at the ADC capture with your expected flow rate and with zero flow. You can see quite a few examples of a good ADC capture in various documents on the product page. 

I see a bunch of jumps in your aToF, dToF, and flow rate, but it sounds like this image includes you adjusting the flow rate (so they should be changing? please confirm).

If you can confirm the above, and the calculated flow rate is just off by a scaling factor, then you need to adjust the meter constant. The recommended way to find a meter constant is to use a reference meter so you know the flow rate, then tweak the meter constant until the flow rate is correct. The meter constant is just something that scales the VFR linearly to account for different pipe sizes.

Hey Dylan,

An update here, so the flow rates are linearly varying with different flowrates set on the reference meter but not at a correctly scaled value. I will try your suggestion on varying the meter constant during calibration and I will get back to you on the results. Is the meter constant directly proportional to the flowrate? In the sense when I increase the meter constant, does this increase my flow rate or does it decrease it? The existing meter constant that is configured is 19296724.00 L/h.

On points #2 and #3, I have set them based on whatever value is suitable for my use-case but I will re-check and confirm on the same.

Yes, I am varying the flow rates which is why there are a bunch of jumps you see on the aToF, dToF and flowrate parameters.

Hey Dylan,

These are the recorded values for flowrates varying between approx. 175 LPH to 1300 LPH. The repeatability of each of these flows are also verified by flowing each flowrate 3 times.

Since the values of Actual VFR are increasing linearly with little noise, should they be linearly fitted (y=mx+c) before inputted on the GUI for flow calibration?

Here they (Actual VFR) seem off by a scaling factor. Then I input these values on the GUI for flow VFR calibration, it shows me errors (+/-25%) on all the flows except one or two flows where it is <1%. The acceptable error threshold for my use case is +/- 1%. I am using only one range (Range-01) for this calibration.

Even after calibration, it is still detecting between 0 to 50 LPH when the expected flowrates are supposed be between 175LPH and 1300LPH.

What parameter do I vary to fix this on calibration?

Hi Team, any updates here?

Hey Dylan,

No worries.

I will try varying the meter constant, although I wanted to get your thoughts on the data I've sent on the previous message (Expected VFR vs Actual VFR).

Do you think this will be resolved just by varying only the meter constant parameter?

When I take a look at your data in excel, I can see that the ratio of the actual flow vs measured flow is in a range of about 25-40, with some noise.

Without knowing your exact use case, I would assume that using 1 flow rate range, with 1 meter constant, would give too much error, as some ranges seem to need a constant of 40, some 25, average is maybe 30. You could retry with more ranges, but of course you dont want to characterize 10+ ranges either. 

Are you sure that your ADC capture looks correct at all of these flow rates? You may want to check that at each flow rate to get an idea of whether the flow rate is correctly calculated before the meter constant scales the value.

Yes, the ADC capture is stable throughout all ranges without sudden noisy signals.

I also tried varying the Meter Constant throughout its min and max limits. It does scale the flowrate but not to 25 or 30 times the measured flowrate like I need it to. It only scales it by +/-10%. 

Attached is an example of the flow waveforms captured (Volume Flow Rate, AbsToF, dToF):

Also once I capture the data and update the VFR Flow calibration to the EVM-MSP430 board, it says "Calibration Update Successful" but I don't see any variation of the flowrate whatsoever regardless of what calibration value is written to the board. I am attaching an image of the calibration parameters for reference.

Is there anything else I have to do before writing the VFR calibration to the board?

Team,

Any update on this query?