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.

MSP430FR6047: Zero Flow sensor readings with FR6047 not stable

Part Number: MSP430FR6047

Hi Team,

Posting on behalf of our customer.

I am using a MSP430FR6047 with my custom PCB design which as done according to the evaluation board design of the FR6047. Since I am using my cutom board it does not have the HID interface parts therefore I use a ez-FET debugger and use the FR6047_USSSWLib_template_example I am using a pair of these sensors with a clamp on configuration sensors facing each other at 45 degree angle P-T0710017流量(液位)换能器英文规格书.pdf I performed a frequency sweep using uart output to find which frequecny gave the highes and most stable peak which then I found out as 874000Hz. The below image shows four captured wave forms using my serial plotter I have adjusted the gain so that the amplitude is between +-900.  

ADC capture.png

Eventhough the waveforms is mostly overlap the main issue is zero flow drift I get negative values for DTof while the water pipe is still without any air bubbles so as a result the flowarate also jumps to negative and positive values. Same thing happen under flow even though my meter constant is not calibrated when I pass a 6L/min (300L/hr) constant flow the meter should have some sort of stable value for the flow right? But below image shows how the VFR keeps jumping around under the constant flow. Similar to this under zero flow also the VFR keeps jumping.

Flow data variation over time at flow.png

The VFR even goes to negative values even under positive flow. Now I tried changing channels of the sensors then all values become negative so that confirms the way i used was correct.

Please give me a solution on how to fix this drift the sensor Tof readings are  having only about less than 15ns variation in each sensor's reading over a fixed number of samples which then results in a DTof difference of about +-500ps which is also changing eveytime.

Regards,

Danilo

  • Hi Danilo,

    Could you please also upload your "USS_App_userConfig.h" and "USS_userConfig.h" files here? I want to check some of your parameter settings to see if there is any issues.

    BTW, since you are using the clamp on type meter, have you taken a look at our document "Clamp On Water Meter for PVC Pipes"?

    Best Regards,
    Peter

  • Hi Peter,

    Yes I have read the clamp meter guide in setting up the parameters. Since I am using a custom PCB I am using the USSWLib_Template_Example. Below is the uss_userconfig.h. The most concerning matter is since I have used a single ground plane in my custom pcb instead of the split ground with controlled connections as in the EVM is it a noise causing issue.

    For some reason, I can't upload the uss_userconfig.h here so I will share it offline.

    Regards,

    Danilo

  • Hi Danilo,

    I would say if not split the ground will definitely cause interference between the analog and digital part in your board, which will result in some noise with your waveform captured, so there is a need to make improvement to your hardware design.

    Best Regards,
    Peter

  • Hi Peter,

    Thank you for your feedback. Please see our customer's follow up inquiry below.

    I want to know one thing for sure. Now when water flow readings are taken per second do we need to take few readings and average to get the final result? Because now when taking one sample it usually become noisy. Can you tell me whether I need to average about VFR readings to take the final output.

    Regards,

    Danilo

  • Hi Danilo,

    You can apply average filtering to the final result to make the result looks like more stable, but in general, even with noise, the result should remain within a reasonable range and without causing significant deviations. So I would suggest if in the current evaluation and testing stage, we can just capture the raw data to analyze if there is any potential issue, and if everything goes well and in the final stage, we can perform some filtering to the result to make better performance which depends on your requirements. 

    Best Regards,
    Peter

**Attention** This is a public forum