• State
  • Replies 9 replies
  • Subscribers 41 subscribers
  • Views 180 views
  • Users 0 members are here
Support feedback
Options
Options
Related

CC3351MOD: Wi-Fi TX Power issue by using Calibrator Tool

Parv Patel
Prodigy 230 points

Part Number: CC3351MOD
Other Parts Discussed in Thread: CC3301

Hi TI Team, 

We have created this new Thread for Wi-Fi (Old Thread).

We are currently performing Wi-Fi pre-compliance testing by using calibrator tool with version 1.0.0.93 on CC3351MOD TI-based Wi-Fi module, and we are facing an issue where the TX power configured using the calibrator tool. 

We attempted multiple power configurations through the calibrator utility, but the measured output power during testing does not match the configured values consistently. 

Test 1:
Wifi on 0dbm & 5dBm TX Power with 2.4GHz commands -->
calibrator wlan0 cc33xx_plt stop_tx
calibrator wlan0 cc33xx_plt tune_channel 1 0 0
calibrator wlan0 cc33xx_plt set_manual_calib -tx 1 -rx 1
calibrator wlan0 cc33xx_plt set_tx -default 0
calibrator wlan0 cc33xx_plt set_tx -preamble_type 1 -phy_rate 1 -tx_power 15 -gi_ltf_type 0 -pkt_mode 0 -length const packet 1500 -delay 50 -data_mode 0 -data_const_val 100 -cca 0
calibrator wlan0 cc33xx_plt start_tx

Output on Spacture Analyzer:
   

Test 2: 
Wifi on 0dbm & 5dBm TX Power with 5GHz commands -->
calibrator wlan0 cc33xx_plt stop_tx
calibrator wlan0 cc33xx_plt tune_channel 36 1 0
calibrator wlan0 cc33xx_plt set_manual_calib -tx 1 -rx 1
calibrator wlan0 cc33xx_plt set_tx -default 0
calibrator wlan0 cc33xx_plt set_tx -preamble_type 2 -phy_rate 5 -tx_power 15 -gi_ltf_type 0 -pkt_mode 0 -length const packet 1500 -delay 50 -data_mode 0 -data_const_val 100 -cca 0
calibrator wlan0 cc33xx_plt start_tx 

Output on Spacture Analyzer:
   
 
Question: 
1. We are getting different wave-form on 0dBm & 5dBm power on 2.4 GHz & 5 GHz. While increasing the Tx power signal is getting distorted. Additionlly TX signal transmission is not continues, kinldy suggest the configuration for contineous transmission.

2. We are configuring this(-preamble_type 2 -phy_rate 5 -tx_power 15 -gi_ltf_type 0 -pkt_mode 0 -length const packet 1500 -delay 50 -data_mode 0 -data_const_val 100 -cca 0) parameter for Wi-Fi certification testing, is it sufficient or do we need to modify any commands, kinldy confirm or suggest the required configuration parameter for 2.4GHz and 5GHz Wi-Fi.

3. Is we need to configure CountryCode & other parameters in INI file for pre-complience testing by using CC33CONF tool?

Please let us know if any additional calibration data are required to do. 

Regards,
Parv

  • 0
    TI__Expert 3445 points

    Hi Parv,

    Continuing to look into this, see my comments:

    Parv Patel said:
    We are getting different wave-form on 0dBm & 5dBm power on 2.4 GHz & 5 GHz. While increasing the Tx power signal is getting distorted. Additionlly TX signal transmission is not continues, kinldy suggest the configuration for contineous transmission.

    Have you also measured board trace loss from module pin to RF port? We want to make sure we are accounting for the trace loss on your board and that it is reflected in the INI. You already have the device transmission configured to continuous with "-pkt_mode 0". Also please ensure your spectrum analyzer is on peak detector and max hold to view a steady plot.

    Parv Patel said:
    2. We are configuring this(-preamble_type 2 -phy_rate 5 -tx_power 15 -gi_ltf_type 0 -pkt_mode 0 -length const packet 1500 -delay 50 -data_mode 0 -data_const_val 100 -cca 0) parameter for Wi-Fi certification testing, is it sufficient or do we need to modify any commands, kinldy confirm or suggest the required configuration parameter for 2.4GHz and 5GHz Wi-Fi.

    The settings you are modifying are fine, but honestly some are not needed. For example, settings like "-gi_ltf_type" should only be used when transmitting 11n or 11ax packets. Our "...set_tx -default 0" is the baseline you should use to configure your device for testing. You then can call additional parameters to change based on what you desire. We run our testing by only changing preamble type, phy rate, tx power, length const packet, and delay. Everything else stays at its default value.

    I see you are using the CC3351MOD. What certification testing are you planning to conduct on the module? The CC3351MOD device is already FCC, ISED, CE, and Japan certified so there should not be any RF related testing needed for certification. Only EMC and system level testing. Can you further detail your plans for me?

    Parv Patel said:
    3. Is we need to configure CountryCode & other parameters in INI file for pre-complience testing by using CC33CONF tool?

    You need to ensure your insertion loss is correctly measured and inputted into your INI as I mentioned in response 1. Additionally since you are using the module, you should use the dedicated region conf files for the region you are testing for (e.g cc33xx-conf-fcc for FCC or cc33xx-conf-etsi for EU and Japan)

    Looking forward to your response.

    Josh 

  • 0 in reply to Josh Smith
    Prodigy 230 points

    Hi Josh, 

    Josh Smith said:
    Have you also measured board trace loss from module pin to RF port?

    We have measured the return loss, which is less then -10dBm. 

    Josh Smith said:
    You already have the device transmission configured to continuous with "-pkt_mode 0".

    We are still observing signal getting off in between continuous transmission, while setting -pkt_mode 0 with set_tx command. Is this expectable for FCC pre-complience testing?

    Josh Smith said:
    The CC3351MOD device is already FCC, ISED, CE, and Japan certified so there should not be any RF related testing needed for certification. Only EMC and system level testing. Can you further detail your plans for me?

    We are planning for full certification as we have multiple radio in our device. 

    Josh Smith said:
    dedicated region conf files for the region you are testing for (e.g cc33xx-conf-fcc for FCC or cc33xx-conf-etsi for EU and Japan)

    We are using below path cc33xx-conf-fcc.bin for testing FCC: 
    /home/root/ti/simplelink_wifi_toolbox_lin_4_1_16/radio-tool/binaries/CC3351MOD/cc33xx-conf-fcc.bin

    We are still facing the same issue. Even when we set higher power, the signal transmission is not as expected. (Refer the image as attached above)

    Can you suggest the possible solution?
    Is it possible to connect on virtual call?

    Your quick support is highly appreciated.

    Regards, 
    Parv

  • 0 in reply to Parv Patel
    TI__Expert 3445 points

    Hi Parv,

    Parv Patel said:
    We have measured the return loss, which is less then -10dBm. 

    Thanks for measuring the return loss. However, have you measured the insertion loss? This can be done by conducting a S21 measurement on a VNA. The value of your S21 measurement should be filled out in the INI parameters:

    - InsertionLoss_2_4GHz

    - InsertionLoss_5GHz
    Parv Patel said:
    We are still observing signal getting off in between continuous transmission, while setting -pkt_mode 0 with set_tx command. Is this expectable for FCC pre-complience testing?

    Yes this should be expected based on your Spectrum Analyzer Settings. If you do not have Max Hold configured, you will see the transmission operate on it's duty cycle based on your packet length and delay time. Just FYI, the most optimal duty cycle is running at 30% D.C using these parameters Feel free to imitate this setup in your certification:

    You should not see the D.C operating with Max Hold enabled on S.A.

    Parv Patel said:
    Even when we set higher power, the signal transmission is not as expected.

    What is your expectation for the power level? If you are expecting to see 0 or 5dBm at the antenna, that is not the correct expectation. You expectation at the spectrum analyzer should be <Power_Level set in calibrator> - <Insertion Loss of board> - <cable loss to SA>. Your results in the images are difficult to see and I would request you submit them in .pdf or .ppt format so I can zoom with clarity of image. Please keep in mind as you adjust to "higher power", that CH.1 802.11b rates will not exceed 18dBm and CH.1 802.11g/n rates will not exceed 13.5dBm as defined in the FCC INI file.

    Additionally, On our side we usually measure power with a S.A  by configuring the SA to zero span to ensure we are measuring power in reference to time. I then would make sure that my Trace/Detector is set to “Max Hold” and change the Detector settings from ”Peak” to “Average”. Maybe your tool has similar settings to mimic my configuration, but this is how you can ensure you are measuring the correct output power from our module using a SA. 

    Hopefully this helps and provides some guidance.

    Josh

  • 0 in reply to Parv Patel
    Prodigy 230 points

    Hi Josh, 

    During testing, we observed the following:
    2.4 GHz Wi-Fi : 
    1. While increasing the transmit power using the calibration tool, the transmitted signal bandwidth is also increasing.
    2. The bandwidth expansion becomes more noticeable at higher TX power levels.


    I also checked the same on the CC3301 EVM module and observed the same result. I have attached the CC3301 signal observations as well.

    WLAN_2.4GHz_Signal_IssueAM62L.pdf
    WLAN_2.4GHz_Signal_Issue_CC3301.pdf

    Please let us know if any additional information need to do. 

    Regards, 
    Parv

  • 0 in reply to Parv Patel
    TI__Expert 3445 points

    Hi Parv,

    Thank you for providing this. As I have suggestions that applies to both BLE and Wi-Fi conditions, I will post this message on both threads, but contents will be the same.

    I do see that as you increase power, the bandwidth increases drastically. This has not been replicated or found on our side through my debug or this device's lifetime so my goal now is to help clear up what could be causing the incorrect signal measurements.

    The first thing I would like to do is ensure the PA is getting enough supply on your CC3301EVM. Can you please connect the additional USB supply to the EVM and run a 5, 10, and 20dBm test again?

    The second thing I would like to try here is to attenuate the RF power as it reaches your SA. Do you have an external attenuator? Can you please transmit 20 dBm from your custom board, add a 20dB external attenuator, and have the output of the attenuator connected to your SA? My goal is to try to match the power level of the 0dBm attempt(which looks the most correct), and see how the signal looks.

    Please try those two things and let me know how that works.

    Best,

    Josh

  • 0 in reply to Josh Smith
    Prodigy 230 points

    Hi Josh,

    Thank you for your suggestions.

    We changed the Reference Level to 30 dBm and the Internal Attenuation to 45 dB in the Spectrum Analyzer, and now we are getting proper signals.
    We have attached the waveform for:

    • B mode
    • Channel 1
    • 17 dBm transmit power




    Kindly check the attached waveform and let us know your observations.
    Also, could you please confirm what Reference Level setting you are using in your Spectrum Analyzer setup?

    Regards, 
    Parv

  • 0 in reply to Parv Patel
    Prodigy 230 points

    Hi Josh,

    We have checked the 2.4GHz and 5GHz waveforms using the digital spectrum analyzer. We observed that in different modes, the measured TX power is lower than the configured TX power.

    We are using cc33xx-conf-fcc.bin from: /home/root/ti/simplelink_wifi_toolbox_lin_4_1_16/radio-tool/binaries/CC3351MOD/

    Below are some observations:
    802.11ax mode - 2.4GHz, CH1
    Set TX power: 17 dBm → Measured TX power: 11.39 dBm
    Set TX power: 10 dBm → Measured TX power: 8.96 dBm

    802.11n mode - 2.4GHz, CH1
    Set TX power: 17 dBm → Measured TX power: 12.24 dBm

    802.11ax mode - 5GHz, CH36
    Set TX power: 15 dBm → Measured TX power: 13.54 dBm

    BLE 1M, CH37
    Set TX power: 10 dBm → Measured TX power: 3.4 dBm

    BLE 125 kbps, CH37
    No signal observed.

    Based on these observations, in some mode difference between configured TX power and measured power is high, which is unexpected.

    WLAN_2.4GHz_BLE_Signal_AM62L_DSA.pdf

    Could you please provide suggestions on it?

    Regards,
    Parv

  • 0 in reply to Parv Patel
    TI__Expert 3445 points

    Hi Parv,

    Thanks for reaching out here.

    The power levels you are seeing are expected. Please remember that the power level you see at the litepoint should be 

    <Power_Level from module> minus <Insertion Loss of board> minus <cable loss to SA>

    Also please note that the CC33XX-CONF-FCC.BIN contains a pre-defined power limit array. This means that the chip will not transmit higher than what is set in the conf file. For example, 802.11ax SU will not transmit higher than 12.5dBm on CHANNEL 1 even if you set 17dBm in calibrator.

    12.5dBm - <trace and cable loss> = ~ 11.39dBm

    Another example, 802.11n will not transmit higher than 13.5dBm on CHANNEL 1:

    13.5dBm - <trace and cable loss> = ~ 12.24dBm

    Please take a moment to review the power limit array set in the conf file. You can use a program like VS Code to see the PLA in easier manner. Please also review our INI composer guide where you can find the INI Parameters and descriptions on how to calculate the dBm value of the PLA.

    Let me know if this helps.

    Best,

    Josh

  • 0 in reply to Parv Patel
    TI__Expert 3445 points

    Hi Parv,

    This is channel 1? I see the center frequency set to channel 11 (2462MHz) instead?

    We set our reference level to 23dBm which is the max of our Agilent Technologies N9010A SA

    Best,

    Josh