• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 38 subscribers
  • Views 575 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.

TIDA-00176: question of TIDA-00176

Jack Li-SH
Expert 4575 points
Part Number: TIDA-00176
Other Parts Discussed in Thread: TLV3201

Dear sir,

would you please help on the two questions from customer?

Q1: In chapter 1.3.1.6, the calculation of angle resolution cannot match the figure value. For example, if line count is 512, angle resolution may be 10 rather than 11 calculated by formula log2(2xN). So the question is why the resolution can be increased by 2-bit and how to reach it?

Q2: 2. This document mentioned Quadrature encoder pulse counter, how does it work to increase resolution? Could you help show the method about quadrature encoder theoretical approach?

Regards,

Jack

  • +1
    TI__Expert 3580 points

    Hi Jack,

    Thank you for your interest in TIDA-00176. Please find below your answer to your question. I also refer to the Design guide — TIDA-00176

    Question 1:  Section 1.3.1.6

    Please note that the formula (6) in section 1.3.1.6 includes the ADC resolution. It could written differently to make it better understand:

    Resolution = log2(4xN) + ADC resolution[bit] – 1-bit

    If we only use the incremental counter we need to consider the two comparators TLV3201 (which convert the analog Sin/Cos into 0 or 1), so comparators are  
    equivalent to a 1-bit ADC.

    So the resolution without interpolation is:

    Resolution = log2(4xN) = log2(2xN) + 1-bit.

    For N= 512 line count and no interpolation just using the incremental up-down counter “coarse angle”, we get 11-bit of resolution.

    Question 2:

    Basically the Sin/Cos interpolation with high-resolution dual ADCs using the tan-1(Cos/Sin) interpolates the angle within one line count. The resolution increase versus
    the incremental line count using an up-down counter by ADC[bit] -1[bit].  

    Please see the formula (5) and the description on interpolation in section 1.3.1.2 to 1.3.1.4.

    A quadrature encoder counter is a directional up-down counter, which counts the digitized signals from Sin/Cos encoder. It counts up to 4xN-1 and then roles over to zero.

    You can find this described for example on the C2000 eQEP module, e.g. TMS320F280013x Real-Time Microcontrollers Technical Reference Manual (Rev. A)
    Chapter 15 Enhanced Quadrature Encoder Pulse (eQEP)

    Does that answer your questions?

    Thanks.

    Regards,
    Martin

  • 0 in reply to Martin Staebler
    TI__Expert 4575 points

    Hi Martin,

    Why do we need to subtract 1-bit when using ADC to do interpolation?

    Regards,

    Jack

  • 0 in reply to Jack Li-SH
    TI__Expert 3580 points

    Hi Jack,

    this depends on how we write the formula. 

    • Resolution = log2(4xN) + ADC resolution[bit] – 1-bit         (EQ 1)
    • Resolution = log(N) + log2(4) + ADC resolution[bit] – 1-bit  = log(N) + 2-bit + ADC resolution[bit] – 1-bit (EQ 2)
    • Resolution = log2(N) + ADC resolution[bit] + 1-bit   (EQ 3)

    I will explain with equation (EQ 3): 

    We get N-bit resolution due to the encoder line count N, which is number of sinusoidal periods over one revolution

    Within one period (one line count) we interpolate with two ADCs using the tan-1(Cos/Sin). Since we have two ADCs we get a an
    additional log2(2) = 1-bit resolution.  So the resolution due to interpolation within one period is ADC resolution[bit] + 1-bit
    Now adding the line count N gives us equation (EQ 3). 

    As mentioned we can write the above equations in different forms but same result.

    Regards,
    Martin