The EVAL-ADL5920-ARDZ shield is a dual RMS power detector with integrated directional bridge featuring the ADL5920 IC. The voltage outputs of the ADL5920 are routed to the ANALOG IN connector of the Arduino base board. This allows the RF power detector’s output voltage to be easily digitized and processed by the Arduino base board’s integrated six-channel ADC.
The ADL5920 with integrated bridge simultaneously measures forward and reverse RMS power up to 7GHz, and provides return loss results. The detector has 50dB of dynamic range at 1GHz. The demo board requires an external power supply that connects to the Linduino board.
The EVAL-ADL5920-ARDZ is compatible with EVAL-ADICUP3029 and DC2026C (also called Linduino One). For both platforms, PC software GUI applications (EVAL-ADICUP3029, Linduino) are available, with which the user can make RF power measurements and also calibrate the device to decrease measurement error. Device drivers for EVAL-ADICUP3029 and for Linduino Uno are also available, which the user may use to develop their own code for RF measurement, device calibration, and more.
1. Remove the DC2847A from its protective packaging in an ESD-safe working area (see Figure 1).
2. Connect external wall wart power supply to J2 on Linduino board DC2026C. Set JP1.
3. Connect USB cable to PC and J5 on Linduino board.
4. Connect signal generator to RF_IN(J1) on the ADL5920 shield.
5. Connect RF load to RF_OUT.
6. Turn on signal generator, set frequency between 9KHz and 6GHz. Set RF power below 30dBm.
7. Go to www.analog.com and download and install quikeval.
8. Open quikeval, set frequency, click “READ” to measure forward and reverse RMS power using default calibration. See Figure 2.
9. User calibration can be performed to improve accuracy. Click “Calibrate” to calibrate the device across frequency. Linear interpolation is used to calculate the slope and intercept for frequencies between the calibration points. The calibration coefficients are stored in the GUI that can be re-used later. See figure 3.
Figure 1. EVAL-ADL5920-ARDZ setup
Figure 2. User Interface
Figure 3. Calibration
The EVAL-ADL5920-ARDZ shield converts the measured ADC code to RF input power in dBm using stored calibration coefficients. A 3-point calibration methodology is used. The software program includes default calibration coefficients that correspond to the default response of the ADL5920 across RF power level and frequency. datasheet specifications of ADL5920. Because of part-to-part device variation, observed accuracy using the default calibration coefficients will be sub-optimal. By availing of the software program's 3-point calibration function, measurement accuracy can be increased.
Related topic: Calibration of EVAL-ADL5920-ARDZ
Click “Connection” or “Calibration” to switch to respective window.
If you plan to operate at a frequency not on the list, make sure to calibrate at least on the adjacent upper and lower calibration frequencies (the software program will interpolate these data to ensure accuracy at the operating frequency). If the operating frequency is higher or lower than the available calibration frequencies, calibrate only on the highest or lowest calibration frequencies.
Two-point calibration is the simplest calibration technique. This models the transfer function of the ADL5920 and ADC as a single straight line
PIN = (CODE/SLOPE)+INTERCEPT
PIN is the RF input power being measured
CODE is the ADC code
SLOPE is the slope of the ADL5920 transfer function's linear model (unit is LSBs/dB)
INTERCEPT is the (extrapolated) input RF power level which would yield and ADC code of 0 (this is a theoretical value with a unit of dBm)
SLOPE and INTERCEPT are calculated and stored during the calibration process by applying two known RF power levels, PIN1 and PIN2 (these RF power levels should be within the linear input range of the ADL5920) and measuring the corresponding ADC codes, CODE1 and CODE2. The equations for calculating SLOPE and INTERCEPT are as follows:
SLOPE = (CODE1–CODE2)/(PIN1−PIN2)
INTERCEPT = PIN1-(CODE1/SLOPE)
If there is some non-linearity in the transfer function of the RF detector, the number of calibration points can be increased to improve measurement accuracy. To implement three-point calibration, three known power levels are applied PIN1, PIN2 and PIN3 (PIN1 should be greater than PIN2 which should be greater than PIN3) and the corresponding ADC codes are noted (CODE1, CODE2, CODE3)
This results in two SLOPE values and two INTERCEPT values which are calculated using the equations
SLOPE1 = (CODE1–CODE2)/(PIN1−PIN2)
SLOPE2 = (CODE2–CODE3)/(PIN2−PIN3)
INTERCEPT1 = PIN1-(CODE1/SLOPE1)
INTERCEPT2 = PIN2-(CODE2/SLOPE2)
After calibration when measuring RF input power, the power is calculated using the appropriate equation
PIN = (CODE/SLOPE1)+INTERCEPT1 (if CODE > CODE2) or PIN = (CODE/SLOPE2)+INTERCEPT2 (if CODE < CODE2)
To decide which equation and calibration coefficients to use, the CODE from the ADC should be compared to CODE2 (CODE2 is the demarcation point between the two calibration regions). This will indicate which region of the ADL5920's transfer function the RF input power is located. For example, if the ADC CODE is greater than CODE2, this will indicate that the input power is greater than PIN2. So SLOPE1 and INTERCEPT1 should be used to calculate the input power. Because of the need to identify the region in which the measured RF input power is located, the CODE2 value should also be stored after calibration along with the SLOPE1, SLOPE2, INTERCEPT1 AND INTERCEPT2.
This technique can be extended to four or more calibration points. This may improve measurement accuracy at the cost of more complex calibration.
Development drivers are available for C and Python. Other development environments may be used but this development guided is focused on software development on CrossCore Embedded Studio (for C) and on Pycharm(for Python).
Assumes a fresh installation of all required software
For any queries regarding the hardware and evaluation software, contact us at EngineerZone.