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.

IWR1642: Independent contractor for person detection at 100 meters

Part Number: IWR1642
Other Parts Discussed in Thread: IWR1443, IWR1443BOOST,

Our customer has an immediate need to a product using either the 1624 or the 1443. The application has the following general specifications

1st, which would TI recommend, the 1642 or 1443?

2nd, standard antenna pattern shown in the dev docs, or one modified for range and narrower FOV?

3rd, any suggestions for a consultant to speed things along?

General application description

General Application Goal

 

Detect an adult human moving at a walking pace at 100 meters within 5 seconds

 

 

 

General Application Conditions

  • We have no power considerations

  • We have no size constraints

  • We are a stationary application

  • We have a stable operating environment

  • We are a 24/7 application, constant operation

  • We are a 2D (two dimensional X-Y) application

  • We are a 360-degree horizontal and +/- 15 degree vertical application

  • We believe will use 4 chips per system, in other words we intend to use a 90 degree FOV sensor

  • Each chips target sampling time (chirp profile) can be up to 4 seconds long

  • We do not need high accuracy target edge resolution

  • We do not need high accuracy direction of travel resolution

  • We do not need high accuracy target velocity resolution

  • We do not need target velocities over 40mph

     

  • Hi John,

    You can find a People Counting Demo that detects humans on TIREX.  We have a chirp that supports detection at 50m.  The customer will have to use the IWR1642 to support detection of people on device. To extend the detection range further, the customer will want to look into designing an antenna with a more focused radiation pattern.

    Regards,

    Justin

  • Hi Justin - thank you. We've seen that demo. We've used the IWR1642 demo kit to do the short range people detection.

    Can use suggest a resourse we can employ to help with the more focused antenna design? We are hoping that an antenna design for a 90 degree device FOV can get us the extra range. 70-80 meters is initially acceptable, but the 90 degree FOV will be pretty much fixed and needed . We'll use 4 IWR1642 to get our required 360 degree coverage.

    1- antenna design resource?

    2- Does the 1442 give us other better options for people detection at 100 meters?

    I look forward to your answers.

    Thanks

  • Hi John,

    We do not have an antenna design guide. However, we do have an antenna design database that can be found here.  For long range, consider a horn antenna. A lens could also be used to focus the beam.

    The IWR1443 has 3 transmitters, which will increase the number of virtual antennas by 4.  In the EVM antenna design, the extra virtual antennas are used to determine elevation of the detected objects.  However, with a different configuration, the angular resolution of the sensor could be increased.  However, the output power of the IWR1443 and IWR1642 is the same.  From an RF standpoint, the third transmitter won't increase the detection range. The extra data provided by having a third transmitter could be used to improve the human detection algorithm - elevation data or increase angular resolution would make it easier to separate objects that have multiple detection points. Finally, the IWR1642 has a DSP, which the IWR1443 does not.  This allows much more computation to be done on the device itself. In the People Counting example, all computation is done on the device. 

    Regards,

    Justin

  • Thank you Jason - this is the piece of data I / we needed. Given our circumstances we'll begin with the 1642 with boost antenna. We were thinking of the Rogers RT/Duroid 5880 or 5870. A bit more expensive but we thought it would help. Thoughts?
  • Hi John,

    Sorry for the delayed reply. This is a complex topic - so I am speaking very generally. The performance of the antenna should improve. (Better SNR, etc). However, this material has a different dielectric constant than the materials TI uses in the EVMs. This means you would have to redesign the antenna to account for this.

    Regards,
    Justin
  • Hi Justin - thanks for the good advice. I know just enough about RF to be dangerous, and to understand this is an area for experienced wizards….

    FM= Friggin Magic

    AM= Another miracle   :)

    At a minimum we'll do our beta system with the PCB material suggested in the notes.

    So... I believe you said you didn't know of any but.. how do I find an antenna consultant with experience with this product, or type of RF system? Thoughts?

    Thanks,

  • John,

    Sure, we do have Ecosystem partners who has very good experience with our devices and antenna design. I am pointing two of such partners here:

    D3 Engineering - www.ti.com/.../thirdpartydevtoolfolder.tsp
    SmartRadarSystems - dev.ti.com/.../ (3rd row in the table)

    They both have antenna design and RF system expertise with our devices.

    Regards,
    Bhavin
  • John,

    We would highly recommend to use the antenna materials RO3003 and RO4835 LoPro as recommended in the PCB Design and Manufacturing App Note pointed below:

    www.ti.com/.../spracg5

    Also the antenna database pointed below provides you full details to reuse the antenna which we designed for IWR1443BOOST and IWR1642BOOST EVMs.

    dev.ti.com/.../

    Please let me know if there is more information needed on this.

    Regards,
    Bhavin