• TI Thinks Resolved

TIDA-00378: TIDA-00378: The correctness data and interpretation - again

Prodigy 50 points

Replies: 6

Views: 161

Part Number: TIDA-00378

My last thread has been locked ( I did not have access to a computer for a long time ) and I'm sorry to create new thread but I need a help.

I built a sensor based on your PCB and I need to test it now. I received data and I am not sure that my sensor works good.

Enclosed you will find screens of graphs with raw ADC samples. I do not know if received data are valid, because I did not find in guide user sample screens.

I did not run the firmware included with the TI design, because I need to use Nucleo STM32F4 with your PCB and these graphs were made by a loop 'while' and were drawn by program TelemetryViewer.

Before I use your algorithm, I'd like to know whether my samples are good.

Please for information on it.

  • Tom, hello and thank you for your continued interest in Texas Instruments.  Please allow another business day before we respond.  The original design owner has been informed about your question and should answer after the holiday weekend. 

    Best Regards

    J. Fullilove

    Reference Design Operations

    Texas Instruments

  • Hi Tom,

    Tom Waks
    I built a sensor based on your PCB and I need to test it now. I received data and I am not sure that my sensor works good.

    Can you provide more information your hardware setup?  What chamber/enclosure are you using?  For this test, are you covering the opening to the enclosure such that no dust particles are entering the sensing area?  Or this test with actual particles flowing through the chamber?

    Tom Waks
    Enclosed you will find screens of graphs with raw ADC samples. I do not know if received data are valid, because I did not find in guide user sample screens.

    This doesn't look right to me.  If the chamber is sealed off from dust particles, I would expect some ADC codes in the lower range due to noise in the system (see Figure 18 in the TIDA-00378 design guide).  I'm not sure why you have a second peak of ADC codes in the 180 range.  Are you pulsing the LED or is it always on? 

    Have you tried inserting an object into the sensing area (where LED intersects with photo diode)?  This should force the photodiode to sature and you should see the ADC generate codes in the high range.  I'm assuming you are you using the internal 12-bit ADC on the uC for these measurements?

    Gus Martinez

  • In reply to Gus Martinez:

    Maybe I will describe more specific my experience with this PCB. So, my tests were done with my room where there are not high concentration of dust. (another sensor shows results about 20-30 ug/m3). I am using a dedicated chamber, LED and photodiode from GP2Y1010AU0F. Device's hole is not closed, just a sensor lie down on the table. I am using the internal 12-bit ADC on my uC (stm32f446)
    Below there are few screens which are made in various conditions.


    1. LED on or off and chamber is opened

    2. LED on or off and chamber is closed ( no dust particles)

    3. In the chamber is an object ( a pen)

    Do you need more information? Please tell me. I am afraid of my PCB works bad and I do not have idea how to fix it.

    Thanks in advance for all help.

  • In reply to Tom Waks:

    Tom Waks
    1. LED on or off and chamber is opened

    Tom Waks
    2. LED on or off and chamber is closed ( no dust particles)

    Thanks for sharing additional information and details.  It almost looks like you have a lot of noise in the system.  Have you considered that maybe the system is picking up 60Hz noise from the lights?  See if the ADC output is oscillating at 60 Hz calculating the frequency of the ADC output using the sampling rate.  The metal shield on the chamber should protect against the 60 Hz noise.  The shield must be connected to the GND plane on your PCB.

    Tom Waks
    3. In the chamber is an object ( a pen)

    When you insert a pen I would expect the ADC to saturate.  So you should see ADC codes > 4000.  This not the case from the graph you sent.  What does and analog signal chain (trans-impedance, gain, and filter) look like on your PCB?  Did you copy the analog signal chain on TIDA-00378 as-is or did you make modifications?

    Gus Martinez

  • In reply to Gus Martinez:

    Below you can see a photo my device:

    Maybe I explain why I did that way: I made one modification, I changed over from PCB 4 layers to PCB 2 layers and I had to enlarge the size of PCB.

    If you are right and the shield is required, that may be the solution of my problem. As soon as possible I will try to improve my device and I will let you know about my progress.

  • In reply to Tom Waks:

    I would suggest you try attaching your PCB directly on the optical chamber and then re-install the metal shield over the PCB.  You should be able to solder the LED and photodiode directly the PCB.

    Gus Martinez