• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 576 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.

OPT4048: OPT4048 register data is abnormal

Roger Zhou65
Prodigy 70 points
Part Number: OPT4048

OPT4048  is designed for LED  backlight white balance monitor, now in working mode, the register data is incorrect,  the detailed is  as following , pls prompt   advise   how to debug the question, ths!

  • 0
    TI__Expert 6800 points

    Hi Roger,

    Do you have data or more details to go along with this?

    Thank you,

    Brent Elliott

  • 0 in reply to Brent Elliott
    Prodigy 70 points

    The parameter values of the light source read by using the Minolta spectrometer CL-500 are:

    X=183.2
    Y=198.9
    Z=183.7
    Lux=198.9
    CIEx=0.3238
    CIEy=0.3515

    Read the data through the sensor 00h-07h as follows:

    REG[0] = 0x00A2
    REG[1] = 0x4809
    REG[2] = 0x0159
    REG[3] = 0x80FC
    REG[4] = 0x0055
    REG[5] = 0x54F5
    REG[6] = 0x0473
    REG[7] = 0xECFF

    Convert to decimal to
    CH[0] = 41544
    CH[1] = 88448
    CH[2] = 21844
    CH[3] = 291820

    Using this set of data CH[1] = 88448, it can be calculated that the brightness is 190.249lux, which is basically the same as the instrument.

    Now we convert to get the XYZ value first, and then calculate the value of CIExy. But we can't get the correct XYZ value through the matrix formula and matrix coefficient given in the specification.

    I would like to ask how to calculate the correct XYZ value with the original value I read?

  • 0 in reply to Roger Zhou65
    Prodigy 70 points

    Pls  refer to the front  data ,confirm  where the question is from ?

  • 0 in reply to Roger Zhou65
    TI__Expert 6800 points

    Hi Roger,

    The XYZ values that result from the matrix (I got 15.39, 16.44, 13.82) are not scaled and are typically just used as an intermediate step in CCT calculation. To get the proper XYZ values, scale XYZ up by the factor that makes Y equal to lux. Therefore you would scale 16.44 up by 11.57 in order to have Y = Lux. Then the XYZ values would be 178, 190, 160.

    By the way, CIEx = X/(X+Y+Z), so XYZ doesn't need to be scaled to get to CIEx, but the above will hopefully help you understand how to scale if you are interested in the proper XYZ values.

    From this, I got CIEx = 0.337 and CIEy = 0.360.

    Thank you,

    Brent Elliott

  • 0 in reply to Brent Elliott
    Prodigy 70 points

    Hello ,

    pls check our test  result  is OK ?  the result is  gotten according to your advice  and guideness.

    According to the latest algorithm, XYZ is X=178, Y=190, Z=160, CIEx= 0.337 and CIEy = 0.360 respectively.

    There is still a certain deviation from the value read by CL500. Is this deviation within the normal range? In particular, the value deviation of Z is more.

    The following are the values read by CL500:

    X=183.2, Y=198.9, Z=183.7

    Lux=198.9, CIEx=0.3238, CIEy=0.3515

    Thanks 

    Broadvideo Roger

  • 0 in reply to Roger Zhou65
    TI__Expert 6800 points

    Hi Roger,

    In terms of CCT (formula I used for CCT is below), there is a 9% difference between the 4048 and the color meter.

    n = (CIEx-0.3320)/(0.1858-CIEy);
    CCT = 437*n^3 + 3601*n^2 + 6861*n + 5517

    Whether or not this is in the normal range depends on the details of the source. What source are you using, and what is the field of view of the source in relation to the device (how far is the device from the source, what is the beam pattern of the source like)?

    Additionally, do you have the device behind any glass or is the device exposed? The matrix provided in the datasheet is optimized for bare device. Placing the device behind a cover glass will require further calibration.

    Thank you,

    Brent Elliott