MSP430FR6043: The values are not accurate

Hi Cherry,

 -129464 in iq44DeltaTOF: -129464/2^44 = -7.35ns. 77026092 in iq40TotalTOF_UPS :77026092/2^40 = 70.05us, 77027416 in iq40TotalTOF_DNS: 77027416/2^40 =  70.05us, 

47902 of iq16Temperature: 47902/2^16 = 0.73 degC.

You can call the USS_getResultsInFloat() function to get these values in float. 

Best regards,

Cash Hao

Hello Cash,

Thanks for your support.

What if the calculated speed does not match the speed at which the ultrasonic wave travels in water? Does the calibration also fail to change the measured propagation time?

How to understand 0.73 degrees Celsius? The customer's water is placed in a room at 20 degrees Celsius, and it is not possible to approach 0 degrees Celsius. And sometimes the temperature measurement here is 0.0, why?  What do they need to do so that it can measure the temperature? 

Thanks and regards,

Cherry

Hi Cherry,

Cherry Zhou said:

What if the calculated speed does not match the speed at which the ultrasonic wave travels in water? Does the calibration also fail to change the measured propagation time?

The absTOF might not be the time of flight of 1st lobe receive time. It could be the time of flight for the 2nd lobe, 3rd lobe and other lobes depends on the configuration. You can refer to the algorithm training video on the b23.tv/ZBMXaOo

Cherry Zhou said:

How to understand 0.73 degrees Celsius? The customer's water is placed in a room at 20 degrees Celsius, and it is not possible to approach 0 degrees Celsius. And sometimes the temperature measurement here is 0.0, why?  What do they need to do so that it can measure the temperature? 

Has the customer change the value of USS_TEMPERATURE_LUT_TRANSDUCER_CONST in the USS_temperatureLUT.h. And enable the USS_ALG_ENABLE_ESTIMATE_TEMPERATURE in the USS_userConfig.h

Best regards,

Cash Hao

Hello Cash,

Cash Hao11 said:

The absTOF might not be the time of flight of 1st lobe receive time. It could be the time of flight for the 2nd lobe, 3rd lobe and other lobes depends on the configuration. You can refer to the algorithm training video on the b23.tv/ZBMXaOo

The customer's threshold should be small enough to measure the first waveform.

As shown in the figure below, the maximum value of this measurement is 847, taking 11% as 847 x 11% = 93.17, which is closest to the first wave (88). The absolute time of flight is 57.21 us, with clear water, and the speed is the distance between the transducers divided by the absolute time of flight, v = s / t = 0.062 / 0.00005721 = 1084 m/s. However, the theoretical value of 1450 m/s is still not met, and the ambient temperature is 20 degrees Celsius. 

Cash Hao11 said:

Has the customer change the value of USS_TEMPERATURE_LUT_TRANSDUCER_CONST in the USS_temperatureLUT.h. And enable the USS_ALG_ENABLE_ESTIMATE_TEMPERATURE in the USS_userConfig.h

The temperature read when enabled is 215 degrees Celsius and USS_temperature_LUT_transfer_const uses the default value of 0.04716. What needs to be done for changing it?

Thanks and regards,

Cherry

Hi Cherry,

How does the customer measure the distance between the transducers. In the first forum, they use 72mm as the distance between transducer for calculation. However, in the latest one, they use 62mm for calculation. 

Also, check this parameter USS_ALG_ADC_ADDITIONAL_CAP_DLY in the USS_userConfig.h. If it is not 0, change it to 0.

The USS_TEMPERATURE_LUT_TRANSDUCER_CONST should be half of the distance between the transducers.

Best regards,

Cash Hao

Hi Cash,

The distance of the transducer is actually 62 mm.

Customer changed USS_ALG_ADC_ADDITIONAL_cap_DLY to 0 and USS_temperature_LUT_transit_const to half the distance, resulting temperature of -72. Is that also incorrect? How to read the correct values? 

Thanks and regards,

Cherry

Hi Cherry,

What is the absTOF result now with -72 temperature reading?

Best regards,

Cash Hao

Hi Cash,

Yes, I don't understand why the value of temperature is -72, but the room temperature is about 20℃, I have changed USS_ALG_ADC_ADDITIONAL_CAP_DLY to 0 and USS_TEMPERATURE_LUT_TRANSDUCER_CONST to 0.31. Could you tell me what else I need to do to get the temperature properly?

Thanks and regards,

Cherry

Hi Cherry,

I am not sure why they get the temperature value at -72. If the absTOF results are good they should get the correct velocity data. Then it will search the lookup table to find the temperature data. That's why I am asking what is the absTOF result at that point. 

They can also check why they get the wrong temperature by their self. The temperature calculation code is open to the customer. They can find it in ussSWLib->source->ussSwLibTmeperature.c

I would suggest you add a breakpoint at line 80. They can debug the inputs and outputs of the USS_calculateVelocity function. And will know where the error is coming from. 

Best regards,

Cash Hao

Hi Cash,

The absTOF results are 57.10us when the temperature is -72 as shown in picture2. I don't know if the data of 57.10us is correct because envelope crossing threshold is 11%.

In theory, the first lobe is taken. Since clean water is used, the propagation speed calculated based on 57.10 is 0.062 / 0.00005710 = 1085 m/s, not 1450m/s.

I added a breakpoint at line 80, and the result is shown in picture3, I don't know if this is correct, I don't know what criteria to check it.

picture 1

picture 2

picture 3

Thanks and regards,

Cherry

Hi,

Let the customer send their USS_user_config.h file here. 

Cherry Zhou said:

I added a breakpoint at line 80, and the result is shown in picture3, I don't know if this is correct, I don't know what criteria to check it.

Check the iq16Velocity value at the breakpoint. You should get 1085m/s based on the customer's configuration.

Best regards,

Cash Hao

Hi Cash,

Cash Hao11 said:

Let the customer send their USS_user_config.h file here. 

https://e2echina.ti.com/cfs-file/__key/communityserver-discussions-components-files/55/USS_5F00_userConfig.h

Cash Hao11 said:

You should get 1085m/s based on the customer's configuration.

The customer indeed can't understand why the calculated 1085m/s is so different from the theoretical value 1450m/s.

Thanks and regards,

Cherry

Hi Cherry,

I check their config file. It looks good to me. The absTOF result should be correct. So based on the formula V = distance/absTOF. I would double check the distance data between transducers. Is their anything can be shared by the customer with us about how they calculate the 62mm distance?

Best regards,

Cash Hao

Hi Cash,

They read 62mm on the website: https://item.taobao.com/item.htm?spm=a1z10.3-c-s.w4002-22815803607.11.33fd3856ibr0f3&id=673508863286, and  measured the distance with a ruler to be 62mm.

Thanks and regards,

Cherry

Hi Cherry,

The 62mm is the distance the reflectors. What is the distance between the transducer and reflector?

Best regards,

Cash Hao

Hi Cash,

Thanks for your diagram.

They actually don't know this distance. The document TIDM-02005DesignGuide states that the distance is small enough to be negligible.

Thanks and regards,

Cherry

Hi Cherry,

In order to facilitate the derivation of the formula, we neglect that distance in the document. However, it is not negligible in an actually design.

Best regards,

Cash Hao

Hi Cash,

if it's 10mm, the calculated velocity is about 1450m/s. But how to get the right temperature? Is it related to the USS_computeTemperature function? I see there is a function USS_calculateVelocity to measure the velocity in it. Is the temperature related to the velocity? I don't understand why velocity was used as the 4th parameter to look it up in the table.

Thanks and regards,

Cherry

Hi Cherry, 

Yes. the temperature is related to the velocity. 

Please refer to this figure in https://www.ti.com/lit/an/slaa959/slaa959.pdf.

Best regards,

Cash Hao

Hi Cash,

From pic1, C can be calculated from formula 3: C = L/2 x (1/T12 + 1/T21) = 0.62 / 2 x (1 / (5.71 x 10^-5) + 1 / (5.71 x 10^-5)) = 10858 m/s. In picture 2, there is no corresponding value of 10858, so it shows that the measured value is -72. If this is the reason, how do I get it to measure normal temperature?

picture 1

picture 2

Thanks and regards,

Cherry

Hi,

Please use the following formula for calculation. 

At zero flow, v = 0. So, C = (2L0+L1)/T1 or C = (2L0+L1)/T2. To do an average method, C = 0.5*((2L0+L1)/T1+(2L0+L1)/T2).

The L0 is not negligible in an actually design.

Best regards,

Cash Hao

Hi Cash,

when L0 = 10mm, C = 0.5*((2L0+L1)/T1+(2L0+L1)/T2) = 0.5 x ( (2x0.01+0.062)/(5.71 x 10^-5) + (2x0.01+0.062)/(5.71 x 10^-5) ) = 1436 m/s

Figure 1 shows that 1436m/s corresponds to a temperature of about 8 ℃,  but why is the temperature in the structure displayed as -72?

1

2

Thanks and regards,

Cherry

Hi,

Could you check if the customer call the USS_computeTemperature function after calling USS_runAlgorithmsFixedPoint?

Best regards,

Cash Hao

Hi Cash,

Yes, USS_computeTemperature was called  after calling USS_runAlgorithmsFixedPoint:

Thanks and regards,

Cherry

Hi,

May i know is there any feedback regarding the latest response?

Thanks and regards,

Cherry

Hi Cherry,

Sorry, I forgot to reply on this thread. 

So, where do they add the breakpoint to observe the temperature data? Could they add the breakpoint right after they call the USS_computeTemperature function.

Best regards,

Cash Hao

Hi Cash,

Thank you.

Cash Hao11 said:

Could they add the breakpoint right after they call the USS_computeTemperature function.

Yes, they add the breakpoint after USS_computeTemperature:

Thanks and regards,

Cherry

Hi Cherry,

Could you let the customer send the USS_temperatureLUT.h here?

Best regards,

Cash Hao

Hi Cash,

Please see here:

https://e2echina.ti.com/cfs-file/__key/communityserver-discussions-components-files/55/USS_5F00_temperatureLUT.h

Thanks and regards,

Cherry

Hi Cherry,

Customer should change the USS_TEMPERATURE_LUT_TRANSDUCER_CONST to 0.041 as discussed in early thread. (2*L0 + L)/2 = 0.041

Best regards,

Cash Hao

Hi Cash,

Customer placed breakpoints after USS_computeTemperature, USS_temperature_LUT_transfer_const changed to 0.41 with temperature showing -4. But the air conditioning on the customer's side is 26 degrees Celsius. 

They think the temperature and distance are not related, because it was 0.04716 before, and the temperature was still not right.

Thanks and regards,

Cherry

Hi Cherry,

The temperature is related to this distance. If they are not related, changing the distance data would not affect the temperature result. 

And based on the screenshot provided from the customer, the distance setting is 0.041. And the average of the absTOF_UPS and absTOF_DNS is around 59.29us. 

So the velocity should be distance*2/average absTOF = 2*0.041/59.29us = 1383m/s. Based on the temperature and velocity figure, the temperature -4.75 degC makes sense to me.

Customer may need to adjust the distance data to get the correct temperature result. The 0.041 is not that accurate.

Best regards,

Cash Hao