ADAR1000-EVAL Measurement result

Could anyone clarify and give me any advice?

Part : ADAR1000-EVALZ

We tested the method provided by the user guide.
Test was conducted using ADAR1000 Max Gain Settings provided by ADI web.
Based of Freq. = 9.5GHz @Tx1, the gain was 15.9dB. The board loss was 3.2dB.
The datasheet says 21dB.
can you advise me

Please kindly help.
Please. Can someone tell me?

Parents
  • 0
    •  Analog Employees 
    on Jul 27, 2021 4:20 PM

    Have you had a chance to remeasure the Tx gain with the bias values outlined in my last response?

  • I tested according to your advice.
    However, the same result was obtained.

    The file you advised and the file provided by ADI Web was the same.

    If we change the Bias Current, MaxGain is similarly measured.
    However, with the Bias Current change, the current increased significantly.
    Transmit Mode: 410mA
    Receive Mode: 330mA

    Is it okay to use it this way?

    And, how can I check the TEMPERATURE SENSOR?

  • +1
    •  Analog Employees 
    on Jul 30, 2021 4:17 PM in reply to Lee Jaegook

    The lower gain could be explained by part-to-part variation and/or board-to-board variation. 

    It is OK to use the part with higher bias settings, but know that the RF metrics will change as well, which we didn't characterize.  We only tested at nominal and low bias settings. 

    Here is a simple Sequence to read the Temperature Sensor value:

    • Write 0x60, to Register 0x32 
      • Enables ADC, 1MHz clock, temp sensor input
    • Write 0x70, to Register 0x32   
      • Initiates conversion cycle 
    • Read ADC_EOC bit from Register 0x32
      • Low during conversion cycle. High when complete.
    • Read Reg0x033 
      • This is the ADC output
  • We are testing with two EVAL boards.
    Both board of our symptoms are the same.

    There is a table for Power Modes in Datasheet page26.
    Even using this method, different values from the datasheet were measured.
    Is it the correct way to use the Bias Current value to match the same measurement value as the datasheet?
    Please tell us in detail what it means to change the Bias Current Value.

    And,
    When measuring NF (TX1 measurement standard), if only Tx1 is ON, it is similar to the datasheet, but if TX2 is ON and TX1 is measured, a 2dB difference occurs. (TX2 Term state)
    In this way, when TX3 and TX4 are turned on, NF gradually increases.
    This is quite strange. Is this correct?

Reply
  • We are testing with two EVAL boards.
    Both board of our symptoms are the same.

    There is a table for Power Modes in Datasheet page26.
    Even using this method, different values from the datasheet were measured.
    Is it the correct way to use the Bias Current value to match the same measurement value as the datasheet?
    Please tell us in detail what it means to change the Bias Current Value.

    And,
    When measuring NF (TX1 measurement standard), if only Tx1 is ON, it is similar to the datasheet, but if TX2 is ON and TX1 is measured, a 2dB difference occurs. (TX2 Term state)
    In this way, when TX3 and TX4 are turned on, NF gradually increases.
    This is quite strange. Is this correct?

Children
  • 0
    •  Analog Employees 
    on Aug 3, 2021 4:29 PM in reply to Lee Jaegook

    The low gain cause must be something common to how you are setting up the part. 

    The bias levels that I already gave should provide gain values that are close to the datasheet (minus the insertion loss of the lines).

    The various bias current values controls the current in the several subcircuits (vector modulator, VGA, LNA, Driver).

    The NF is measured with just the measured channel enabled.  So if you are measuring Tx1, only enable Tx1 to get a NF close to datasheet specification. 

  • Hi, jdobler

    I checked the data of the S2P file provided by ADI Web.
    This data is worse than the values in the Datasheet.

    Is this correct?

    And,

    How is the switching time measured?

  • +1
    •  Analog Employees 
    on Aug 9, 2021 3:40 PM in reply to Lee Jaegook

    The Tx S2P data on the website should match reasonably well the data in the datasheet as the RF lines' insertion loss is deembeded.  

    However, I just checked Rx data and I believe this data set was taken with old subcircuit bias values and thus is is slightly low (about 1 dB).

    I have attached updated S-params below that used the Nominal bias values shown in the datasheet.

    ADAR1000_S-PAR.zip

  • +1
    •  Analog Employees 
    on Aug 9, 2021 3:49 PM in reply to Lee Jaegook

    Transmit and Receive Switching time is defined as the time from the TR pin at 50% of its final value to the RF signal at 90% of its final value.  The RF signal was detected using the fast envelope detector ADL6012. 

    The Phase and Gain Switching time was also measured using the ADL6012, but it was referenced at 50% of the TX_LOAD or RX_LOAD pins. 

    Phase switching time can be measured by detecting the resultant magnitude of 2 combined Rx channels.  If the channels are in-phase, the resultant output will be twice as big relative to a single driven channel.  If the channels are out-of-phase by 180 degrees, the signals will cancel each other out and the resultant magnitude will just be noise. 

  • Hi, jdobler

    I have additional questions.

    1. SPI Control Speed is 25MHz, is there any way to improve the speed? (I want it to run faster.)
    2. I want to check PA Bias in ADAR1000-EVAL. Is there any way?
    3. We are testing three ADAR1000-EVALs. What is the current applied to -5V? It shows 0mA from the power supply. It seems that this value is too small to be checked. Am I interpreting it in the right way?