Silicon ChipLow-cost, Wideband Digital RF Power Meter - August 2021 SILICON CHIP
  1. Outer Front Cover
  2. Contents
  3. Subscriptions: PE Subscription
  4. Subscriptions: PicoLog Cloud
  5. Back Issues: PICOLOG
  6. Publisher's Letter
  7. Feature: The Fox Report by Barry Fox
  8. Feature: Techno Talk by Mark Nelson
  9. Feature: Net Work by Alan Winstanley
  10. Project: Ol’ Timer II by Tim Blythman
  11. Project: Low-cost, Wideband Digital RF Power Meter by Jim Rowe
  12. Project: Switchmode Replacement for 78xx regulators by Tim Blythman
  13. Feature: KickStart by Mike Tooley
  14. Feature: Microphone Preamplifier (for Vocoder) – Part 4 by Jake Rothman
  15. Feature: Make it with Micromite by Phil Boyce
  16. Feature: Circuit Surgery by Jake Rothman
  17. Feature: Max’s Cool Beans by Max the Magnificent
  18. PCB Order Form
  19. Advertising Index

This is only a preview of the August 2021 issue of Practical Electronics.

You can view 0 of the 72 pages in the full issue.

Articles in this series:
  • (November 2020)
  • Techno Talk (December 2020)
  • Techno Talk (January 2021)
  • Techno Talk (February 2021)
  • Techno Talk (March 2021)
  • Techno Talk (April 2021)
  • Techno Talk (May 2021)
  • Techno Talk (June 2021)
  • Techno Talk (July 2021)
  • Techno Talk (August 2021)
  • Techno Talk (September 2021)
  • Techno Talk (October 2021)
  • Techno Talk (November 2021)
  • Techno Talk (December 2021)
  • Communing with nature (January 2022)
  • Should we be worried? (February 2022)
  • How resilient is your lifeline? (March 2022)
  • Go eco, get ethical! (April 2022)
  • From nano to bio (May 2022)
  • Positivity follows the gloom (June 2022)
  • Mixed menu (July 2022)
  • Time for a total rethink? (August 2022)
  • What’s in a name? (September 2022)
  • Forget leaves on the line! (October 2022)
  • Giant Boost for Batteries (December 2022)
  • Raudive Voices Revisited (January 2023)
  • A thousand words (February 2023)
  • It’s handover time (March 2023)
  • AI, Robots, Horticulture and Agriculture (April 2023)
  • Prophecy can be perplexing (May 2023)
  • Technology comes in different shapes and sizes (June 2023)
  • AI and robots – what could possibly go wrong? (July 2023)
  • How long until we’re all out of work? (August 2023)
  • We both have truths, are mine the same as yours? (September 2023)
  • Holy Spheres, Batman! (October 2023)
  • Where’s my pneumatic car? (November 2023)
  • Good grief! (December 2023)
  • Cheeky chiplets (January 2024)
  • Cheeky chiplets (February 2024)
  • The Wibbly-Wobbly World of Quantum (March 2024)
  • Techno Talk - Wait! What? Really? (April 2024)
  • Techno Talk - One step closer to a dystopian abyss? (May 2024)
  • Techno Talk - Program that! (June 2024)
  • Techno Talk (July 2024)
  • Techno Talk - That makes so much sense! (August 2024)
  • Techno Talk - I don’t want to be a Norbert... (September 2024)
  • Techno Talk - Sticking the landing (October 2024)
  • Techno Talk (November 2024)
  • Techno Talk (December 2024)
  • Techno Talk (January 2025)
  • Techno Talk (February 2025)
  • Techno Talk (March 2025)
  • Techno Talk (April 2025)
  • Techno Talk (May 2025)
  • Techno Talk (June 2025)
by Jim Rowe Low-cost, Wideband Digital RF Power Meter Simple to build and low in cost, this Digital RF Power Meter will be very useful for anyone who needs to measure radio frequency signals from 1MHz to 6GHz. By itself, it can only handle power levels up to about 3mW (5dBm), but its range can easily be extended using fixed attenuators. W hile examining a Banggood RF Power Meter, it occurred to me that we could design a similar device that wouldn’t cost much more to build, but would handle much higher frequency signals. I also realised that its construction could be made easy by using other low-cost prebuilt modules that I had reviewed. The concept quickly solidified around using an Arduino Nano module as the ‘brains’, together with the Banggood RF Detector module I reviewed in the March 2019 issue. It offers useful capability, such as a frequency range up to above 6GHz, the ability to send the results of each measurement to your PC for data logging, and an allowance for fixed attenuators at the Meter’s input to extend its power range. I freely admit this last idea was copied from the Banggood RF Power Meter... gain of 8.7dB. The outputs of each amplifier stage are connected to a full-wave detector cell, and the output currents of the detector cells are summed and fed to a current-to-voltage converter which produces output voltage VOUT. The voltage-to-current converter at upper right allows adjustment of the slope of VOUT. For example, when the VSET and VOUT pins are tied together, this sets the output slope to a nominal −25mV/dB. The AD8318 also includes an internal temperature sensor and bias stabilisation on the cascaded gain stages, so that changes in ambient temperature do not unduly affect accuracy. And all of this impressive technology is squeezed into a tiny 4 × 4mm 16-lead LFCSP surface-mount package. Fig.2 shows the measured transfer characteristic of an AD8318 at four different frequencies: 100MHz, 1GHz, 2GHz and 4GHz. It’s very close to linear at −25mV/dB at all four frequencies, between 0dBm and −60dBm. Fig.3 is the full circuit of the Banggood log detector module we are using. There’s very little in it apart from the AD8318 and a 78L05 regulator, which provides the AD8318 with a regulated +5V supply. (We are actually bypassing the 78L05 in this project, as you’ll learn shortly.) The Meter’s heart The Analog Devices AD8313 demodulating logarithmic amplifier IC in the RF detector module forms the heart of the Meter. It accurately converts an RF signal into a decibelscaled DC output voltage. It maintains accurate log conformance for signals from 1MHz to 6GHz and provides useful operation to 8GHz. The input range is typically 60dB (referenced to 50), with errors less than ±1dB up to 5.8GHz. Fig.1 shows how the AD8318 works. Fig.1: internal block diagram for the AD8318 log detector IC. The differential It has nine cascaded input signal passes through a string of nine amplifiers/limits and the outputs of amplifier/limiter each one go to full-wave detectors. The direct currents from each detector are stages, each with a summed and converted to a voltage which appears at VOUT. 24 The full circuit The full circuit for our new Digital RF Power Meter is shown in Fig.4. The Banggood AD8318based log detector module is at upper left, connected to the rest of the circuit via Practical Electronics | August | 2021 CON2. The Arduino Nano MCU ‘brain’ is on the right. IC1 in the centre, an LTC2400CS8 high-resolution (24-bit) ADC (analogue-to-digital converter), is used to digitise the output voltage from the log detector module. This ADC requires a reference voltage to set its input scaling, and this is provided by accurate 2.500V reference SC REF1, an LT1019ACS8. Ó IC1 digitises its input voltage under the control of the Arduino MCU via an SPI interface using Nano pins 1 (SCK), 30 (MISO) and 28 (SS). After the MCU processes the digitised sample data, it displays the calculated RF power and voltage levels on the 16×2 LCD module via CON1. This is via an I2C interface using MCU pins 8 (SDA) and 9 (SCL) – the LCD module is an I2C serial type. Three pushbutton switches (S1-S3) are connected to MCU pins 25, 23 and 21. These are used to tell the unit when you have connected one or more external RF attenuators ahead of the Meter’s RF input, to increase its measurement range. It then adjusts its display to give correct readings. Since the Meter is designed to operate from a 5V DC supply derived via the USB cable connected to the Arduino Nano, the supply for the rest of the Meter circuitry is taken from MCU pin 12. This goes directly to the LCD module (again via CON1). For the rest of the circuitry, it is filtered by inductor RFC1 and several bypass capacitors. We are making a few minor modifications to the Banggood Log Detector module to simplify using it in the RF Meter Fig.2: plot of VOUT vs input signal level for the AD8318 at four different frequencies (default slope setting of −25mV/dB). As project. The 78L05 regulator on the module needs an input you can see, the linearity is excellent, and the frequency has of at least 7V for proper regulation, but we don’t have that. minimal effect on the measured RF power level. Instead, we have a well-filtered 4.75V rail after the 4.7 series resistor. So we are bypassing the 78L05 in the module The only other modification needed is to fit a 1nF SMD by connecting the supply wire from CON2 directly to its ceramic capacitor (2012/0805-size) across the two pads just output pin 1. to the left of the 2-pin output connector on the log detector To make sure that the 78L05 isn’t damaged by reverse cur- PCB. This provides additional filtering for the AD8318’s rent, it’s a good idea to remove the 10k resistor in series with internal feedback loop – it’s shown as COBP on Fig.4. the LED at the input of the 78L05. It’s pretty unlikely that All of these modifications should be clear from both the such a small current notes on the circuit (Fig.3) and the close-up photo of the log would damage the detector module PCB. regulator, but the Pin 8 of IC1 (the LTC2400 ADC) is taken to the centre pin of LED won’t be visible JP1, a three-pin header. This allows the sampling frequency once the case is on of IC1 to be set for optimum rejection of any power line anyway, so it just frequency components in its input signal. Function: wastes power if left When the jumper shunt fitted to JP1 is in the lower posiA compact, low-cost RF power in-circuit. tion, the sampling frequency is set to reject 60Hz components and level meter with LCD screen and USB interface Frequency range: from 1MHz to over 6.0GHz Input impedance:  50 nominal Maximum input power level:  +5dBm (3.2mW/398mV RMS) Minimum input power level: −60dBm (1nW/224µV RMS) Measurement range: −60dBm (224µV RMS) to +33dBm (10V RMS) with recommended  attenuators Measurement linearity:  about ±1dBm, 10MHz to 1GHz, +6dBm/−4dBm, 1MHz to 4.0GHz (see measurement plots) SC 1MHZ – 8GHZ LOGARITHMIC DETECTOR MODULE  Measurement resolution: approximately ±0.1% Fig.3: the circuit of the pre-assembled log detector module is very simple. The RF Power supply: signal is terminated with a 51Ω resistor (52.3Ω might be better) and coupled to the inputs of IC1 via a pair of 1nF capacitors. The output from IC1 is fed to a pin 5V DC at less than 120mA via header, while power is supplied via a 2-way terminal block. We’re bypassing 5V USB micro-B socket Features and Specifications 1Mhz – 8GHz Logarithmic Detector Module regulator REG1 in this project. Practical Electronics | August | 2021 25      Reproduced by arrangement with SILICON CHIP magazine 2021. www.siliconchip.com.au   16 x 2 LCD Wideband Digital RF Power Meter SC WIDEBAND DIGITAL RF POWER METER  Fig.4: thanks to the use of three prebuilt modules, the circuit of the RF Power Meter is not too complicated. The Arduino Nano uses 24-bit analogue-to-digital converter IC1 to read the output of the log detector with high precision. 2.5V reference REF1 ensures that IC1 measures that signal with reference to a very stable voltage. The whole circuit is powered from the 5V pin of the Nano, which gets its power from a USB charger or computer. (as you’d need in the US), but if the jumper shunt is fitted in the upper position, the sampling frequency is set to reject 50Hz components. So the latter position is the best one for use in Australia, New Zealand or the UK. What the firmware does The firmware sketch for the Digital RF Power Meter is called RF_Power_Meter_sketch.ino and is available for free downloading from the August 2021 page of the PE website. When uploaded to the Arduino Nano’s ATmega328P micro, it does several things. Its main task is to direct IC1, the ADC, to take a sequence of 10 measurements of the output voltage VOUT from the log detector module. It then averages each group of measurements and calculates from that the corresponding RF power level in dBm and the equivalent voltage level in millivolts or microvolts. These figures are then sent to the LCD module for display, and are also sent out via the Meter’s USB data line for display and possible logging on a computer. The firmware’s other main task is to check between measurement cycles for any presses of the Select Attenuation pushbutton switch, S1. If S1 has been pressed, it then swings into ‘change attenuation’ setting mode and it monitors any presses of switches S3 (‘Increase’) or S2 (‘Decrease’) and adjusts its setting for the external attenuation in steps of 1dB. Then, when S1 is pressed again, it saves the new external attenuation figures and returns to its normal measurement mode. The attenuation value is set to zero each time the unit is powered up. 26 Construction The complete Digital RF Power Meter is housed in a diecast aluminium box measuring 119 × 93.5 × 56.5mm. Pushbutton switches S1-S3 and the LCD module all mount on or behind the box lid/front panel. All of the other modules and components are mounted on a double-sided PCB measuring 109 × 83mm, coded 04106201 and avaiable from the PE PCB Service. This also mounts behind the box lid/front panel, via four pairs of spacers. CONNECT +5V WIRE TO THESE PADS REMOVE THIS RESISTOR ADD 1nF CAPACITOR ACROSS THESE PADS A few minor modifications need to be made to the Banggood module before fitting it to the PCB. Practical Electronics | August | 2021     SILICON CHIP Fig.5: this PCB overlay diagram and the photo below show which parts go where. The only polarised parts are IC1, REF1 and the Arduino Nano module. Pushbutton switches S1-S3 are mounted on the lid and wired back to the board using flying leads, while the header on the LCD screen (also mounted on the lid) is soldered directly to the pins of CON1 as the last step in the assembly. Case preparation There are only two holes to be cut in the box proper: an 11mm-diameter round hole in the front, and a 9 × 11mm rectangular hole in the rear. The location of each of these holes is shown in Fig.6. There are 12 holes to be cut in the box lid, which becomes the Meter’s front panel. The locations and sizes of these holes are shown in Fig.7. There are three 12.5mm holes for the three pushbutton switches and a 65 × 15mm rectangular hole for the LCD ‘window’. The remaining small holes are for mounting the LCD module and the main PCB. After you have made and deburred all the holes in the lid/front panel, it’s a good idea to attach a dress front panel to the front for a professional appearance. We have prepared full-size artwork for this, which can be downloaded from the August 2021 page of the PE website as a PDF file. You can print this out in colour and then hot-laminate it in an A5 laminating pouch. After this you can cut it to size, punch four 3mm holes (one in each corner) and then attach it to the front of the lid using either thin double-sided cellulose tape or contact adhesive. Once it is securely attached, cut out the remaining holes using a sharp hobby knife. The next step is to attach an 80 × 25mm rectangle of thin clear plastic (say, 0.4mm thick) behind the LCD window Practical Electronics | August | 2021    Begin construction by first fitting the passive SMD components to the main PCB, using the overlay diagram of Fig.5 and the matching photo as a guide. Then fit RFC1, which is larger and will probably need a hotter iron. It’s best to smear a thin layer of flux paste on its pads before soldering it in place. After this, install IC1 and REF1, which are both in SOIC-8 SMD packages. Next, mount 4-pin SIL headers CON1 and CON2, along with the 3-pin header for JP1. Then you can fit the four PCB terminal pins, which all push through their matching holes in the main PCB and are soldered to the pads underneath. Two are to the left of RFC1 (TPGND and TP5V), while a third pin (TP2.5V) is to the right of REF1 and the fourth (TP VOUT) is to the right of CON2. You should then be able to fit the Arduino Nano module to the PCB, with its 30 pins passing down through the matching holes and soldered to the pads underneath. The final step in assembling the main PCB is to fit the AD8318 log detector module. It mounts on the top of the main PCB using four 10mm-long M3 machine screws, with an M3 nut used on each screw as a spacer, and then further M3 nuts underneath to complete the job. Once it has been secured, plug a 4-pin SIL socket into header CON2 and solder four short lengths of light-duty hookup wire to its pins, then to the matching points on the module using Fig.5 as a guide. By the way, although the log detector module shown in the photos and diagrams is fitted with a small two-way terminal block for power and a two-pin header for VOUT, the module as supplied may not have these. Neither connector is required in this application, as you can simply solder the wires to the pads on the PCB. cutout, to protect the screen from dirt and damage. This can be attached using standard cellulose tape, taking care not to cover the LCD module mounting holes. The lid assembly can now be finished. Mount the LCD module behind the window using four 16mm-long M2.5 countersunk screws, four 9mm-long untapped spacers, three or four nylon washers and then four M2.5 nuts, as shown in Fig.8 and the photos. Then you can mount the three pushbutton switches using the supplied plastic nuts, and finally attach a 25mm-long M3 tapped spacer near each corner using a 6mm-long M3 machine screw. The rear of your lid/front panel should now look like the photo. Next, cut six 25mm lengths of single-core hookup wire (three red and three black) and strip off about 4-5mm of the insulation at both ends of each. Then solder one end of each 27 Photos of the front (above) and rear (below) of the assembled project showing the holes required, These photos match Fig.6, left. Fig.6: only two holes need to be made in the main part of the case, with the locations and sizes shown here. The round hole is for the SMA RF input connector, while the rectangular cutout allows a USB micro-B plug to be inserted into the socket on the Nano board After plugging a four-pin SIL socket into CON1, attach the main PCB using four 12mm M3 screws through each corner of the PCB, with a 6mm-long untapped spacer between the PCB and each 25mm-long tapped spacer – see Fig.8. The only trick is making sure that the wires from each pushbutton pass through their matching holes in the main PCB, although you can adjust them later if necessary. Once all the switch wires are through their corresponding PCB pads, upend the assembly and solder the wires to those pads. The final step is to solder the four pins of the SIL header on the LCD module to the corresponding pins at the top of the SIL socket you fitted to CON1. You may need to slightly bend the LCD header pins using a pair of needlenose pliers, so that they are close to the pins of the SIL socket, allowing them to be soldered together. If this proves a little tricky, it can help to temporarily remove the nearby tapped spacers, which can be replaced easily once the connections have been made. Don’t fit this assembly into the box just yet, since it’s a good idea to check a few key voltages at this stage. It may also be necessary to adjust the contrast of the LCD to get the clearest display once the Meter firmware has been uploaded. Testing and setup First, connect the Meter up to a USB 5V power supply using a mini-B cable. As soon as power is applied, the LCD’s backlight should illuminate. Get out your DMM and check a few voltages relative to the TPGND pin at the left rear of the main PCB. You should measure close to 5V on the adjacent TP5V pin, around 4.75V on the VCC pin of the socket plugged into CON2, and very close to 2.5V at TP2.5V. Fig.7: most of the holes that need to be made are actually in the case lid, including a large rectangular cutout for the LCD screen. This is best made by drilling a series of small (say 2mm) holes around the inside of the perimeter, knocking the inside part out, then filing the edges to shape. You can use a similar technique for the USB socket hole in the base. red and black pair of wires to the connection lugs at the rear of each pushbutton switch. These are to connect the switches to their matching pads on the main PCB. 28 Practical Electronics | August | 2021 If you get all of these readings, remove the power and download the Meter’s Arduino sketch from the August 2021 page of the PE website. You will need the Arduino IDE (Integrated Development Environment) to compile and upload the sketch. If you don’t have it already installed, it’s a free download from www. arduino.cc/en/Main/Software Our sketch, RF_Power_Meter_sketch.ino, uses libraries: SPI.h, Wire.h and LiquidCrystal_I2C.h. The first two come as standard with the Arduino IDE, but you’ll probably have to install the last one via the Library Manager or download it from github: http://bit.ly/pe-aug21-rf Once ready, plug the Meter’s USB cable into a free port of your PC. If you are running Windows 10, go into Settings -> Bluetooth & Other Devices and then go down to Other devices. You should find an entry like USB-SERIAL CH340 (COMxx), where the digits after ‘COM’ indicate the virtual COM port that Windows has assigned the Meter’s Nano – or strictly, its CH340 USB-serial interface chip. Next, start up the Arduino IDE, and go into the Tools menu. Then click on Board, which will produce a list of possible Arduino modules; select Arduino Nano from that list. Then click on Processor and select ‘ATmega328P (old Bootloader)’, since this is the appropriate one to communicate with the Meter’s Nano MCU via its CH340 serial interface. After this, click on Port, which should give a listing of any virtual COM ports that the IDE has found available. Select the COM port address that corresponds to the Meter. If you haven’t already loaded the LiquidCrystal_I2C library via the Library Manager, do so now. If you downloaded the ZIP file instead, add it via the ‘Add .ZIP Library’ option near the top of the Sketch -> Included Library list. Now open the downloaded sketch file and click Sketch -> Verify/Compile, After 20 or 30 seconds, you should get the message ‘Done compiling’ in the box near the bottom of the IDE window, plus some statistics regarding the compilation. If all has gone well, the final step is to go into the Sketch menu again and click on Upload. When this is completed, the Meter should spring into life. The LCD should first display the initial greeting: Silicon Chip RF Power Meter After a few seconds, it should begin displaying the results of its RF input sampling and calculations. With nothing connected to the Meter’s RF input, you should get a display like this: RF Pwr= −68.5dBm =83.2uV At=00dB If the display on the LCD is not clear and well defined – perhaps just two lines of blocks – that indicates that the contrast trimpot on the back of the LCD module needs to be adjusted. Rotate the trimpot in one direction or the other using a small screwdriver. The trimpot is just above RFC1 and the TP5V and TPGND terminal pins. The last thing to test before fitting the Meter assembly into its box is to make sure it is sending the test readings back to the PC. To do this, go to the Arduino IDE and open the Tools menu. Click on Serial Monitor and it will open up another window. This should show the Meter’s virtual COM port address at the top, and at the top of the centre area you should see: Silicon Chip Digital RF Power Meter Then, after a few seconds, you should see the results of the first reading on a single line: Practical Electronics | August | 2021 RF Pwr= -68.6dBm = 82.6uV At=00dB Further readings will appear every few seconds. If you don’t see this display in the Serial Monitor window, or if all you see is a string of weird graphic symbols, check at the bottom right of the window to make sure that the serial data rate is set to 115,200 baud (bits per second). This is the data rate at which the Meter’s Arduino Nano sends the reading data. If you click on the ‘Show timestamp’ checkbox at bottom left of the same window, a timestamp will be added to the start of each line of readings to allow data logging. If you have access to the equipment necessary to finetune the Meter’s calibration, as described at the start of the section below, you may wish to do that now. Otherwise, you can accept the default calibration we have built into the firmware. In that case, unplug the USB cable and lower the Meter assembly into the box, securing it with the four supplied mounting screws. Your Digital RF Power Meter is now ready for use. Parts list Wideband Digital RF Power Meter 1 diecast aluminium box, 119 x 93.5 x 56.5mm [Jaycar HB5064 or similar] 1 double-sided PCB coded 04106201, 109 x 83.5mm, available from the PE PCB Service 1 Arduino Nano or compatible module 1 1-8000MHz AD8318-based RF Logarithmic Detector module [eBay, AliExpress, Banggood] 1 16x2 LCD module with blue LED backlight, HD44780 or compatible controller and I2C serial interface 3 panel-mounting SPST pushbutton switches (S1-S3) [Jaycar SP0700 or similar] 1 100µH RF choke, SMD 12 x 12 x 8mm [Jaycar LF1402 or similar] 4 25mm-long M3 tapped spacers 4 9mm-long untapped spacers 4 6mm-long untapped spacers 4 M3 x 12mm panhead machine screws 4 M3 x 10mm panhead machine screws 4 M3 x 6mm panhead machine screws 8 M3 hex nuts 4 M2.5 x 16mm countersunk machine screws 4 M2.5 hex nuts 4 nylon flat washers, 3mm inner diameter 2 4-pin SIL headers, 2.54mm pitch 1 3-pin SIL header, 2.54mm pitch 2 4-pin SIL header sockets, 2.54mm pitch 1 2-pin SIL header socket, 2.54mm pitch 1 jumper shunt/shorting block 2 100mm lengths of light-duty hookup wire (red and black) Semiconductors 1 LTC2400-CS8 24-bit ADC, SOIC-8 (IC1) [Digi-Key LTC2400CS8#PBF-ND] 1 LT1019ACS8-2.5 voltage reference (REF1) [Digi-Key LT1019ACS8-2.5#TRPBFCT-ND] Capacitors 2 100µF 10V X5R SMD ceramic, 3216/1206-size 2 10µF 16V X7R SMD ceramic, 3216/1206-size 7 100nF 50V X7R SMD ceramic, 3216/1206-size 2 1nF 50V C0G or NP0 SMD ceramic, 2012/0805-size Resistors (all SMD 1%, 3216/1206 size) 1 5.6Ω (code 5R6 or 5R60) 1 4.7Ω (code 4R7 or 4R70) 29 Fig.8: this side profile view shows how it all goes together and fits into the case. If you don’t have untapped 6mm spacers, you could use tapped 6.3mm spacers instead. Note how the log detector module is spaced off the main PCB using nuts. The last step before dropping the whole thing into the case is to bend the 4-pin header on the LCD over to make contact with CON4 on the main board, then solder the pins together. byte S1 = 0; byte S2 = 0; byte S3 = 0; Calibration To fine-tune the Digital RF Power Meter’s calibration, you’ll need a DMM able to measure DC voltages up to 2.5V with high accuracy, and a UHF signal generator which can be set to provide CW signals at 1GHz (1000MHz) with an accurate amplitude of between +5dBm and −65dBm. The first step is to remove the Meter assembly from its box (if you’ve already finished the assembly) and apply 5V power via the USB cable. After allowing a few minutes for it to stabilise, use the DMM to measure the reference voltage at TP2.5V, up near the right rear corner of the main PCB, relative to the TPGND pin. This should be very close to 2.5000V, but whatever the reading you get, record it carefully as VREF. Next, transfer the positive test lead of the DMM to monitor the voltage at the TP VOUT terminal pin, just to the right of CON2 at the rear of the log detector module. Then connect the input of the Power Meter to the output of the signal generator via a short length (say 150mm) of SMASMA cable. The short length is to minimise cable losses. Set the generator to provide a CW (continuous wave, ie, unmodulated) signal at 1.000GHz, with an initial level of +5dBm (1.78V RMS). The DMM should show the log detector’s VOUT voltage to be around 0.5V. Record the actual value of this reading, this time with the label ‘Vo5dBm’. Next, reduce the generator output level to 0dBm (224mV RMS), and again record the DMM reading (it should be around 0.56V) with the label ‘Vo0dBm’. Repeat this exercise with the generator set to −55dBm (398µV), which should give a reading of around 1.9V, and −65dBm (126µV), which should give a reading of around 2.1V. These figures should be recorded as ‘Von55dBm’ and Von65dBm’ respectively. Now remove the DMM test leads and go back to the Arduino IDE, which presumably still has the Digital RF Power Meter sketch open. Scroll down about 50 or so lines from the top, where you’ll find three lines reading: Typical response plot After calibrating the prototype Digital RF Power Meter shown in the photos, we measured its response over a range of signal levels and between 1MHz and 4.0GHz (the upper limit of the Gratten GA1484B signal generator). The results are shown in Fig.9. This shows that the Meter response at most signal levels is within ±2dB up to 1.0GHz, rising to a peak of around +6dB at 2.5GHz, before falling away again. The peak at 2.5GHz is presumably related to the components (and possibly the PCB tracks) at the input of the This photo is from the same direction as Fig.8 above . . . . . . while this shot is from the opposite direction. 30 Then you’ll see a blank line, followed by a line reading: const float Von65dBm = 2.0451; In place of that figure of 2.0451, type in the reading you recorded for Von65dBm. Similarly, replace the values on the next four lines with the other readings that you noted earlier. Make sure that, in replacing these figures, you don’t remove the semicolons after each one. Otherwise, the sketch won’t compile. Save the modified sketch file and recompile it by going to the Sketch menu and clicking on Verify/Compile. Then if it compiles correctly as before, click on Sketch→Upload to load the revised firmware to Flash memory on the Power Meter’s Nano. Your Power Meter should now be calibrated. Just to verify that this has been achieved, you can set the signal generator output to say −40dBm (2.24mV RMS), whereupon the Meter should give a reading very close to this figure; within ±1dBm. The calibration is then complete. You can remove the power from the Meter assembly and reinstall it in its box, so it’s ready for use. Practical Electronics | August | 2021 End-on photo (above) with a shot showing the display board and pushbuttons, obviously before they were wired in! Note how the standoffs are lengthened to make the required spacing between the main board and front panel. log detector module. We wondered whether the 51Ω input load resistor was responsible, as the AD8318 data sheet suggests 52.3Ω instead. But swapping that resistor out with some 52.3Ω samples we bought did not eliminate the peak. So it’s probably a PCB layout issue. Even with a longer cable between the generator and the Meter (allowing for the cable losses), there was still a peak at 2.5GHz. But if you know the frequency of the signal you are measuring (as you usually would), you can use Fig.9 to make allowances for this behaviour. Suitable attenuators To make the Meter truly useful, you should ideally also get a few inline attenuators. These can be used to extend its measurement range above +5dBm. Banggood has a range of very compact SMASMA fixed coaxial attenuators, for the reasonable price of +20 +10  +5 398mV 0 224mV –10 71mV  –20 22.4mV around £6 each or £12 for three. They are rated at 2W and 0-6GHz, and are available with attenuation figures of 3dB, 6dB, 10dB, 20dB and 30dB. The 10dB attenuator could be used to extend the range of the Digital RF Power Meter to +15dBm (1.26V RMS, or 32mW), while the 20dB unit would extend its range to +25dBm (3.98V RMS or 316mW). Similarly, the 30dB unit would extend its range to at least +33dBm (10.0V RMS or 2W into 50Ω). I ordered the 10dB, 20dB and 30dB units, and thanks to the COVID-19 pandemic they took about seven weeks to arrive. But they did turn up eventually, and they seem to be well made. They’re pictured in the photo below. As mentioned earlier, when you power up the Meter, the external attenuation figure is set to zero – displayed as ‘00dB’ at the righthand end of the second line of the LCD. When you change the attenuation figure to allow for any attenuator(s) you are using via buttons S1-S3, the Meter will display this new figure on the LCD in the same position. If at a later stage you remove the external attenuator(s) and wish to reset the Meter’s attenuation figure to zero, this can be done either by using the trio of pushbuttons again, or simply by removing power from the Meter for about 10 seconds and then reapplying it. RF INPUT LEVEL in dBm –30 7.1mV –40 2.24mV –50 710mV –60 224mV –70 71mV –80 1 2 5 10 20 50 100 200 500 1GHz 2 5 10 FREQUENCY Fig.9: the measured performance of the finished product for nine different input levels over a range of frequencies from 1MHz to 4GHz. The readings are generally within about ±1dB up to 1GHz, but a peak at around 2.5GHz makes readings from higher frequencies less accurate. You can use this diagram to compensate the readings, as long as you know the signal frequency. Practical Electronics | August | 2021 A selection of attenuators, in this case 10, 20 and 30dB, which will significantly increase the power handling of your RF meter. These were sourced from Banggood, for around £12 for the three. 31