• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 32 subscribers
  • Views 1816 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

ldc1614 frequency counter principle

Tim Gerbert
Intellectual 330 points
Other Parts Discussed in Thread: LDC1614, LDC1612

Hello,

I got a new LDC1614 EVM with MSP430 from TI.

I use LDC1613/1614 GUI EVM.

I followed page 40 in datasheet (design procedure) to get frequency count. I use example coil with L=33 µH and C=330 pF. The sensors frequency is about 1,6 MHz. and I set the registers and sample time to measure 1 ms.

With 1,6 MHz frequency I should get 1600 counts every 1 ms. But I got higher counts about 20380840 counts. I assume that the frequency counter will count more than 1 time of one complete amplitude. Where is this amount of cycles defined? How can I adjust sampling rate within one cycle? How does the frequency counter use external clock to count fsensor?

Thanks in advance.

Best Regards,

Tim Gerbert

  • 0
    TI__Genius 14270 points
    Hello Tim.

    To convert the LDC1614 or LDC1612 output code to sensor frequency, use equation 8 in the datasheet (in section 7.3.2).

    The LDC161x devices output code is a ratio of sensor frequency/reference frequency, similar to an ADC, in which the output code is Vin/Vref.

    The LDC161x devices are different than the LDC1000; the LDC1000 outputs a direct count of reference cycles for a specified number of sensor cycles.

    The LDC161x devices' sample rate is based on the reference frequency, clock divider, and RCOUNT setting for the device. Refer to equation 6 and figure 11 in the datasheet.

    Regards,

    ChrisO
  • 0 in reply to Chris Oberhauser66
    Intellectual 330 points

    Hi Chris, thanks a lot for your answer.

    Unfortunately, I think I have a specific question in detail:

    For example I use 20 MHz as fref and sensor frequency of 1,6 MHz. In EVM GUI I set sample time to 1 ms (register 0x08).

    In sample time of 1 ms the LDC 1614 should count: fref=20000 and fsens=1600 counts.

    In relation to your example with an ADC converter I get:

    count fsens/ count fref = 1600/20000 = 0,08

    This ratio is mapped by multiplication of 2^28 to value of DATAXCH0=21474836,48

    But if fsensor is changing only full periods are counted: so instead of 1600 counts now 1601. I assume that fref count is stable so we get

    a new ratio: 1601/20000=0,08005 mapped to DATACH0= 21488258,2528

    This seems to be a huge step between two different counting data.

    Does the LDC1614 contain any part of PLL to multiply reference frequency?

    In my measurements I got much better resolution with gap of only 10 in DATACH0(see attached figure)

    This means, that either in LDC1614 core a higher frequency is used to get this

    resolution or another principle of frequency count of fsensor and fref is used.

    Until now I thought that only every amplitude level of i.e. 0 V is sampled and count.

    Can you give me more detailed explanation of this procedure? It isn't written

    in datasheet. There I only found the formula. I would like to know what is inside

    the core and how does it count fsens and fref.

    Best Regards,

    Tim Gerbert

  • 0 in reply to Tim Gerbert
    TI__Genius 14270 points

    Hello Tim.


    Please note - the sample time of 1ms in GUI does not change the RCOUNT value. In fact, at the maximum RCOUNT value with the 40MH fREF(and with the reference divider=1), it will take 40ms to perform a conversion.

    I'm am sorry that I cannot go into detail on the implementation of the LDC core, but it does not use a PLL.

    Regards,

    ChrisO

  • 0 in reply to Chris Oberhauser66
    Intellectual 330 points

    Hello Chris,

    I set sample time (conversion time) in LDC1614 to 1 ms when I changed register 0x08 (reference count) to value 0x4D6. So the LDC1614 should stop counting after 1ms.

    Moreover I stream data to GUI every 1 ms without logging and 10 ms with logging. I set clockdivider for reference frequency to 20 MHz. So why it will perform a conversion of 40 ms?

    Furthermore I have a question about the quantization to a digital signal.

    I measured with different sample coils a noticed that the resolution (gap between to counting values) is different to sensor tanks and also frequency (see attached figures).

    Best Regards,

    Tim Gerbert

    With different coils, but same reference frequency I should only see difference in counts for fsens. So the quantization should be equal all time. Unfortunately I can´t see this in my attached figures. Why?

  • 0 in reply to Tim Gerbert
    Prodigy 30 points
    hi my name is Matteo,
    I'm using a LDC1614 board with the LC sensor already attached on the board to measure oscillations of a mechanical structure.
    I ve already calibrate the sensor with a specific target that i fixed on the structure.
    The data are acceptable, but i ve problem regarding the sample rate.
    Even setting a sample rate of 1 ms, 1000 samples per second i get around 1/3 of samples is should have (1000).
    I set the Rcount and settlecount according to the formulas provided in the datasheet, i'm using the internal oscillator, divider code 1, Fref 43,4 MHz.

    Another question is about the sampling time, if i sum the sampleDeltaMs i should get the total time of acquisition? And if sampling rate is constant why is the DeltaMs different for every step of data recorded?




    x,evmSampleDeltaMs,logDeltaMs,DATA0_uH,DATA1_uH,DATA2_uH,DATA3_uH,DATA0,DATA1,DATA2,DATA3,DIFF0,DIFF1,DIFF2,DIFF3,DATA0..........................................

    62662,0.014928001910448074,0.014928001910448074,5.983889572633201,8.41723469887236,Infinity,57.39921480561446,24035395,842491.......................................

    62663,0.00466499850153923,0.00466499850153923,5.983889572633201,8.41723469887236,Infinity,57.384281335435425,24035395,2026554..........................................

    62664,0.0027990005910396576,0.0027990005910396576,5.983889572633201,8.41723469887236,Infinity,57.40820025804595,24035395,2026..........................................

    62665,0.003732999786734581,0.003732999786734581,5.983889572633201,8.41723469887236,Infinity,57.49360376416261,24035395,202655..........................................

    62666,0.0027990005910396576,0.0027990005910396576,5.983889572633201,8.41723469887236,Infinity,57.58071427523693,24035395,2026..........................................

    62667,0.003265000879764557,0.003265000879764557,5.983889572633201,8.41723469887236,Infinity,57.74027255545999,24035395,202655..........................................

    62668,20.70799699984491,20.70799699984491,5.983889572633201,8.41723469887236,Infinity,57.90070931016015,24035395,20265547,0,81..........................................