MSP430FR6043: None-zero dTOF

Hello,

Let me check with our USS expert on this.

Hi Barry,

Can you please post the ADC capture you are getting? Can you also confirm that you have your tube in a metal/shielded can as described by the procedures at this link?

https://dev.ti.com/tirex/explore/content/msp_academy_msp430_1_00_05_06/modules/msp430/msp430_USS_training/msp430_uss_training.html#-introduction

BR,

Leo

Hi Leo,

Thank you for reply in short time.

Here is the ADC capture

I started my test as described in the design reference , till now I only try two pairs of transducers and they have same performence.

What should I do next?

Barry

Hi Barry,

Can you post a frequency sweep from 180kHz to 260kHz for these transducers? It looks like there is a lot of energy in the signal since there are more than 8 pulses in the rising excitation. We might increase F2 based on your frequency sweep results.

BR,
Leo

Hi Leo,

The frequency sweep result.

It seems not good enough right?

Thanks,

Barry

Thanks Barry,

Can you set F1 to 190 and F2 to 250kHz, reduce your gain by ~3dB and post an ADC capture with a snapshot of the waveforms at zero flow?

BR,
Leo

Hi Leo,

I made a mistake, the base gain in my first configuration screen shot should be 7.7dB.

The configuration and waveform with F1 190KHz and F2 260KHz .

Now my prolem is my board can not keep high accuracy in variable flow rate testing, for example when the flow rate is 10000l/h the measurement error is less than 1% but when the flow rate increase/decrease the deviation becomes much bigger than 1%.

Is this related to none-zero dTOF?

Thanks

Barry

Hi Barry,

No, I wouldn't expect this. Can you post a picture of your pipe design.  Do you see the same issue with the TI EVM? Was the other problem resolved?

BR,
Leo

Hi Leo,

This is the diagram of my pipe.

After modify the receive circuits of my board and change some parameters in USS configuration the dTOF decreased but can not come close to zero.

I want to make it clear that why I can`t  get the same accuracy when flow rate changes.

Thanks,

Barry

Hi Barry,

Modifications to the receive circuit may have changed the impedance.  The impedance looking back from the transducers to the Tx and Rx circuits must be the same or you will see an offset in your results.  If you look at the original design, you'll see that the RC network on the TX side is the same as the RC network on the RX side.

BR,
Leo

Hi Leo,

Is this the root cause of different accuracy in variable flow?

If it is true I have to modify my board again.

Thanks,

Barry

Hi Barry,

Can you test with our EVM(which is impedance matched)? This should make it clear whether the issue you are seeing is related to hardware modifications or some other aspect of the design.

BR,

Leo

Hi Leo,

I can not do my test with MSP430FR6043 EVM for large pipe size.

Now I`m sure the impedance are the same when transmitting and receiving.

I changed the USS_VFR_LARGE_PIPE_ADDL_SF_ENABLE to true and set USS_VFR_LARGE_PIPE_ADDL_SF to 2.000f.

Then I adjusted meter constant for testing, when the USS_VOLUME_SCALE_FACTOR was a big value(1000 for example) I got the negtive flow rate in USS GUI.

It seems the .volumeFlowRate got from lib API was over flow.

Can I fix it by modifying the code?

Thanks,

Barry

Hi Barry,

You can modify the scale factor in the GUI and in the USS_userConfig.h header file. Is the problem you are seeing that the actual flow rate is too large?  There is another variable called USS_ALG_MAX_SAMPLE_SHIFT which can be increased to accommodate larger shifts between the upstream and downstream signal.

BR,
Leo