• State TI Thinks Resolved
  • +1 person also asked this people also asked this
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 32 subscribers
  • Views 657 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.

OPT3101EVM: Short Range Accuracy

Depeng Man
Intellectual 2030 points
Part Number: OPT3101EVM
Other Parts Discussed in Thread: OPT3101,

Hello,

We are considering the use of the TI OPT3101 for short range (0 to 500mm) distance measurement. I purchased the OPT3101EVM for testing after using the OPT3101-System-Estimator tool. According to the estimation tool, the accuracy should theoretically be 1 - 2mm. from ~20 to 500mm. Attached are screenshots from the estimation tool:

However, real world testing has yielded significantly poorer results for us with errors of upto ~25mm for short distances (<150mm) and upto ~8mm for distances from 150mm to 350mm. I have already checked the crosstalk measurement by covering the photodiode with multiple layers of black tape, and found it to be a 1 to 3 codes. All data was collected using an avg frame count of 4096 as I found that to return the most stable results.

Calibration on the unit is as it came from factory. I have also tried multiple targets with different amounts of reflectance, different modes, different avg frame counts, but the errors in the results have remained. 

Is it possible that the EVM is not optimized for short distances? What can we do to achieve better accuracy? We are even open to trying to different diodes/photoresistors, etc. We are hoping to achieve mm-accuracy with the unit.

  • 0
    TI__Expert 6800 points

    Hi Depeng,

    Could you share some of the data from your test? I am curious how the amplitude predicted by the tool matches up to the test.

    Thank you,

    Brent Elliott

  • 0
    Prodigy 20 points

    Hi Brent, The question was actually mine. I was having some account difficulties and was unable to login to e2e forums, so Depeng from customer service posted this for me.

    Here is the data from our tests. All distances measured are from the top of the PCB as I read in another post that distances returned by the sensor are from the top of the PCB.

    White target:

    Actual Distance Measured Distance (SHDR) Amplitude Measured Difference to prev.  Error Mode
    60.7 82 9885 21.3 SHDR
    110.7 154 22240 72 43.3 SHDR
    160.7 195 11213 41 34.3 SHDR
    210.7 243 6550 48 32.3 SHDR
    260.7 280 4360 37 19.3 SHDR
    310.7 331 13279 51 20.3 SHDR
    360.7 383 10231 52 22.3 SHDR

    Clear polycarbonate panel infront of white target:

    Actual Distance Measured Distance (SHDR) Amplitude Difference to prev.  Error Mode
    60.7 82 9885 21.3 SHDR
    110.7 135 24998 53 24.3 SHDR
    160.7 186 12219 51 25.3 SHDR
    210.7 216 7482 30 5.3 SHDR
    260.7 254 19478 38 -6.7 SHDR
    310.7 303 14360 49 -7.7 SHDR
    360.7 353 9835 50 -7.7 SHDR
    390.7 395 9835 42 4.3 SHDR

    We also observed that different measurements were returned for the same distance depending on the if the target was moving away from the sensor or towards. I suspect this is due to the amplitude threshold at which the emitter current changes.

    Also, switching the emitter current even when the target is stationary makes a considerable difference (as much as 10-15mm in some cases) to the measured distances. I assume that the greater the amplitude the better the accuracy?

    Both the above sets of data show target moving away from sensor.

  • 0 in reply to Praveen Sundaram
    TI__Expert 6800 points

    Hi Praveen,

    The EVM goes through calibration for medium to long-range distances. The EVM calibration is not optimized for short-range. Looking at the data provided it looks like the error is due to this reason.

    Thank you,

    Brent Elliott

  • 0 in reply to Brent Elliott
    Prodigy 20 points

    Hi Brent,

    So is it possible to optimize the EVM for short distances by recalibrating? Or are the EVM's short range capabilities hardware limited?

    Cheers,

    Praveen

  • 0 in reply to Praveen Sundaram
    TI__Mastermind 25895 points

    Hi Praveen,

    Yes you could try calibrating the EVM with targets at the shorter ranges you are using. The OPT3101 will have a minimum resolution of a few mm so the <2mm shown by the tool will not be achievable with this device. In addition there is some hardware limitation in the EVM for short range, but I would expect you to get a performance improvement by recalibrating over your current performance. You can find instructions for calibration along with all our collateral here: https://e2e.ti.com/support/sensors-group/sensors/f/sensors-forum/815766/faq-opt3101-what-are-the-resources-ti-provides-for-developing-with-opt3101 in the calibration section. Specifically the SDK tool includes a quick start PDF guide which allows you to use an MSP430 launchpad to perform the recalibration. However, this process still has a number of steps and requires looking through our other guides linked in the calibration section to understand the process. It is specifically phase offset calibration that you would want to do to get performance at short range.

    Best,

    Alex