ADX rp2040 Digital Transceiver (ADX_rp2040)

Overview to Version 2.0

A brief story of the project start with the excelent ADX Transceiver from Barb (WB2CBA) which can be found at

• Github site link.
• ADX transceiver blog link

The ADX transceiver is powered by an Arduino Nano (ADX) or Arduino Uno (ADX_UNO) boards using both the ATMEL ATMEGA382p processor.

In order to leverage the capabilities of the transceiver with a powerful processor such as the Raspberry Pi Pico which uses the rp2040 architecture this project was started.

Then a map between the Arduino board I/O and the rp2040 I/O was made showing some differences needs to be addressed which requires additional circuitry.

Once the hardware platform was defined the firmware was ported using the ADX_UnO_V1.3 firmware as a baseline, the porting didn't introduce any new feature or function, just the minimum number of changes to the code to accomodate the architecture differences between both platforms.

*New in release 1.0 build(23) and higher*

* Autocalibration mode has been added check the appropriate section on how to enable and operate

*New in release 1.0 build(40) and higher*

* CAT support, TS2000 protocol

Support and issues

This is an experimental, work-in-progress, non-profit, project performed as closest to the ham spirit as possible. Only spare, hobby, time is available to move the project forward or to provide support on usage or issues.

If anybody has questions or issues please:

• Be sure you read the documentation first.
• Check on the issues list of the GitHub site, the issue might have been described there or even a workaround might exists for it.
• State it as an English request. English isn’t even my fourth language, still I do my best to adhere to it.
• For casual question you can use the groups.io uSDX forum and for a longer ones please open and issue at the GitHub portal of the project.
• Express very clearly which version and level the firmware has. In most cases using the latest would solve the issue.
• Ensure the issue happens with a freshly downloaded last version of the firmware, don’t expect me to debug any modification you did.
• Describe in your own words the problem and what you did to expose it and what workarounds you attempted.
• Present a photo of the issue if it can be seen in the TFT display.
• Add the content of the monitoring terminal session with DEBUG enabled to help me understand what is going on.
• Attach any other documentation you think might help debugging the issue.

Please report ONE (1) issue per entry, and proceed as clean as possible with the debug instructions given to you to further understand or to fix the problem. I’ll address your issue as soon as my available time allows, not necessarily in a FIFO way.

Hardware

I started the porting of the firmware assuming an ADX transceiver board, no more but no less features, but being powered by a Raspberry Pico board; the porting of the software was a great deal of a learning curve not only for the rp2040 architecture being different from the ATMEGA382p and being more powerful, but also a substantially different build chain which in some cases is implementing partially some features. For migration purposed the Raspberry Pi Pico board was used.

Although the first version of the new transceiver board was made after a wired (but functional) version of it Barb helped very substantially by creating the PCB design for the ADX2PDX board (included in this site). The following illustration shows the wired prototype used for the firmware development and other evaluation purposes for future projects.

The ADX2PDX daughter board connect with an ADX board with minimal modifications using the Arduino Nano socket and provides a functionality similar to the main board.

Firmware

Build environment

The development environment used is the Arduino IDE, even with some limitations it's far more easy to setup and operate than the Eclipse IDE alternative and present a much smoother transition from development for the Arduino environment into the rp2040 environment.

The usage of the Arduino IDE is based on the arduino pico core libraries developed by Earle F. Philhower, III.

In order to install it a tutorial can be found here or here.

Pre-requisites and libraries

In order to build the firmware some libraries are needed, the dependencies are shown as follows

Basic support

• Earl E Philhower rp2040 core porting to Arduino IDE (https://github.com/earlephilhower/arduino-pico)
• SI5351 Library by Jason Mildrum (NT7S) - https://github.com/etherkit/Si5351Arduino
• Arduino "Wire.h" I2C library(built-into arduino ide)
• Arduino "EEPROM.h" EEPROM Library(built-into arduino ide)

Code excerpts gathered from manyfold sources to recognize here, large pieces of code were extracted from former projects

• PixiePi.
• Pixino.
• OrangeThunder.

Code structure

The main functionality is quite similar to the baseline ADX firmware used, there are three changes needed to adapt the firmware to the rp2040 architecture.

• I/O. A mapping between the Arduino I/O pin and the rp2040 GPIO ports has been made, symbolic names were adjusted and coding macros used to replace the primitives to operate it. Proper initialization was initroduces for all ports used.

• EEPROM. The rp2040 based Raspberry Pi Pico board used to host the porting doesn't have EEPROM available, however the arduino core is able to emulate it on flash memory, however additional definitions to initialize and commit values is needed and thus included in the ported code.

• Tone frequency counting. The rp2040 processor lacks the zero cross comparator interrupt used by the ADX transceiver and thus it has been replaced using a firmware definition on one of the PIO of the processor (RISC processor).

The code port was made starting with the ADX_UnO_V1.3 as available at the GitHub site by Nov,15th 2022, no automatic synchronization mechanism has been established with it. The overall logic cycle of the firmware can be seen in the following figure.

The main functionality is contained in the file ADX_rp2040.ino which is compiled by using the Arduino IDE supplied with the stated libraries, different subsystems are made dependent on configuration directives (#define on the microcode, typically to signal the introduction of a porting segment by means of the #define RP2040 directive) which mades the relevant code segments to be included or excluded from the build process.

mandatory files

ADX_rp2040.ino

PIO programming (counting method)

freqPIO.cpp
freqPIO.pio
freqPIO.pio.h

Transmission Algorithms

The ADX transceiver by Barb (WB2CBA) owes in part it's popularity to it's simplicity, and no small part of it derives from the very simple, yet effective, way to process incoming audio signals to derive what is the tone being sent by the host program and direct the transmitter to operate at a base frequency plus that tone frequency in order to achieve direct synthesis USB transmission with a very low overhead mechanism which involves little or no DSP programming which can be very taxing for the original Arduino board the transceiver has been designed with. The algorithm comes actually from Burkhard Kainka (DK7JD) link cleverly uses the COMPA timer of the ATMEGA328p processor architecture into a very effective frequency meassurement with very little overhead, hardly can be considered a DSP processing, then the information of the frequencies is applied to the frequency output to achieve the digital modulation desired. This can be made because this is a digital transceiver and it's assumed it's used to transmit some form of a weak signal mode such as WSPR, JS8, FT4 or FT8, it can be used for CW and in general with some limitations for any mode where the amplitude carries no information. Unfortunately the rp2040 doesn't have such a good resource, so the frequency needs to be measured using other strategy, this in turns become a great source of experimentation.

A PIO processor (16 of them available with the rp2040 architecture) is configured to generate an interrupt with the rising flank of the signal, a high resolution timer (capable of +/- 1 uSec resolution) is used to compute how many ticks can be counted between two sucessive flanks. There is actually no "zero crossing" detection as the GPIO isn't capable to trigger interruptions with that condition, but it's triggered when the signal level achieves the ON condition of the input pin. It's assumed the trigger level is the same for all cycles at it is not a signal subject to fading, therefore the frequency can be counted by measuring the time elapsed between similar parts of sucessive cycles. In order for the algorithm to work a signal conditioning needs to be performed by means of external hardware, this can be made either thru a MOSFET or an IC comparator based circuit (see hardware for details). The actual measurement accuracy is in the order of +/- 2 Hz in the worst case, which is accurate enough for digital modes.

Sources of error

The counting has +/- 2 Hz error as the start and end of the counting window might include or exclude the starting and ending cycles. Also
the trigger point might suffers some variations making the actual timing between sucessive cycles inaccurate. Finally the SNR of the signal
might present noises which trigger false counts.

CAT Support

When enabled by removing comments from the "#define CAT 1" statement the UART (RX/TX) pins will be used to operate a CAT (Computer Aided Tuning).

The parameters of the support would be:

• Protocol: Kenwood TS-2000.
• Speed: 115200
• Parity: 8N2
• Control lines: DTR/RTS lines On. Handshake: none.

In order to connect the transceiver board with the computer the "Test CAT" button needs to be pressed, depending on the sync status more than once might need to be pressed. The board will detect the CAT commands being sent by the computer and turn itself into "CAT" mode, this is signaled by lighting the JS8 and FT4 leds while turning off the FT8 and WSPR leds. Once the CAT mode is activated the board no longer reacts to the board buttons. The CAT settings will be lost when the board is turned off. When a frequency change is made the FT8 & WSPR leds will blink for 10 seconds to remaind the operator that the output filters might require a review prior to the operation.

Automatic calibration (autocalibration)

Starting on version 1.0 build(23) and higher a new capability to perform an automatic calibration of the Si5351 VFO has been added.

Enabling

The firmware defaults to the standard manual calibration procedure, in order to activate the automatic calibration the following sentence needs to be included or uncommented

#define AUTOCAL            1

The firmware needs to be recompiled and flashed into the Raspberry Pico board after the modification.

Operation

When started the firmware will look during the setup stage if the DOWN pushbutton is pressed, if so all the on-board LEDs will be lit with the exception of the TX LED indicating a waiting pattern, the autocalibration procedure will start as soon as the push button is released.

If the board is powered off before the push button is released the previous calibration stored in EEPROM (flash memory) will be reset to zero.

The calibration can be monitored either by the LED pattern exhibited or thru the USB serial port (Arduino IDE Serial Monitor), once the calibration is completed the results will be written in EEPROM (flash memory) as in the manual calibration in order to be used on sucessive starting cycles. While the calibration is being performed the TX LED will blink once per second, the rest of the board LEDs will mark how large is currently the difference in the calibration mode:

         WSPR,JS8,FT4,FT8 lit       error > 75 Hz
         WSPR,JS8,FT4     lit       error > 50 Hz
         WSPR,JS8         lit       error > 25 Hz
         WSPR             lit       error > 10 Hz
         All LED off                error < 10 Hz  (final convergence might take few seconds more)

When monitoring the calibration thru the USB Serial monitor the messages will look like:

Autocalibration procedure started
Current cal_factor=0
Current cal_factor=0, reset
Si5351 clock setup f 1000000 MHz
n(12) cal(1000000) Hz dds(1000071) Hz err (71) Hz factor(0)
n(12) cal(1000000) Hz dds(1000074) Hz err (74) Hz factor(500)
.............[many messages here]................
n(11) cal(1000000) Hz dds(1000001) Hz err (1) Hz factor(71500)
n(10) cal(1000000) Hz dds(1000001) Hz err (1) Hz factor(71500)
n(9) cal(1000000) Hz dds(1000001) Hz err (1) Hz factor(71500)
n(8) cal(1000000) Hz dds(1000002) Hz err (2) Hz factor(71500)
n(8) cal(1000000) Hz dds(1000000) Hz err (0) Hz factor(72000)
n(7) cal(1000000) Hz dds(1000001) Hz err (1) Hz factor(72000)
n(6) cal(1000000) Hz dds(1000001) Hz err (1) Hz factor(72000)
n(5) cal(1000000) Hz dds(1000001) Hz err (1) Hz factor(72000)
n(4) cal(1000000) Hz dds(1000001) Hz err (1) Hz factor(72000)
n(3) cal(1000000) Hz dds(1000001) Hz err (1) Hz factor(72000)
n(2) cal(1000000) Hz dds(1000001) Hz err (1) Hz factor(72000)
n(1) cal(1000000) Hz dds(1000000) Hz err (0) Hz factor(72000)
Calibration procedure completed cal_factor=72000
Turn power-off the ADX board to start

Upon finalization a message will be sent thru the serial monitor and the TX led will stop to blink, the board power needs to be cycled to restart the operation.

                                     *** Warning ***

Calibration time might vary depending on the unique factory characteristics of the Si5351 chipset being used.

Algorithm

The Si5351 CLK2 output is connected internally to the GPIO09 I/O port of the rp2040 board. When in autocalibration mode the firmware is configured to measure the frequency on that pin.

The calibration starts by setting the calibration factor of the Si5351 chip as zero.

To start the Si5351 clock is set to a frequency of 1 MHz and the result is meassured, differences between the value obtained and 1 MHz are assumed to be because of calibration differences, the calibration constant is then modified and the cycle repeated.

The calibration continues with small variations and finishes when the resulting value converges into a 1 MHz measurement.

The resulting calibration factor to achieve that is then stored in EEPROM (flash memory) to be used in the next re-start.

Hardware

The hardware required by this transceiver derives directly from the ADX Transceiver (WB2CBA), the implementation can take basically two forms:

• Build a hand wired version of the circuit.
• Build an ADX transceiver, any using the Arduino Nano board, and replace it with the ADX2PDX daughter board created by Barb (WB2CBA), see below.

ADX_rp2040 circuit

The circuit used is esentially the ADX transceiver with the minimum set of modifications to accomodate a Raspberry pico (rp2040 processor) instead of an Arduino Nano (ATMEGA328p processor). The following diagram has been originally conceived by Dhiru (VU3CER) and put together by Barb (WB2CBA):

PDX_V1.0_Schematic.jpg

The receiver, Si5351 clock, RF driver and final stages are identical to the standard ADX Transceiver, whilst changes are made around the rp2040 processor to accomodate the different signaling and voltages used.

rp2040 pinout assignment

For circuit design and future expansion several assignmentes has been made on the rp2040 pinout for the following assignment assignment

CAT Support

In order for the CAT to work an USB-TTL dongle is recommended. The connection wiring has to be

• GPIO12 (pin16) UART0 RX to board UART RX.
• GPIO13 (pin17) UART0 TX to board UART TX.
• GND (Any GND) to board GND.

The board +5Vcc and +3.3Vcc are left unwired.

Power supply

The Raspberry Pi Pico operates with +3.3V logic as opposed to the Arduino Nano used by the ADX transceiver, still it has an internal +5V/+3.3V regulator which is used by the design. The standard +12Vcc voltage is used to feed only the final amplifier (3xBS170), a +12Vcc/+5Vcc regulator is used to obtain the voltage to feed the Raspberry Pico (VSYS) pin, the +3.3Vcc obtained from the board is then used to feed the receiver logic, the Si5351 clock, the 74ACT244 RF driver and the CD2003GP based receiver as well as the miscellaneous circuitry such as LED indicators, switches and signal comparator.

Switches

Like the standard ADX transceiver the design carries three (3) push (normally open) switches to signal:

• UP change mode, up band in band setting mode and up frequency in CW mode. Also used to signal the start of the calibration mode on start-up (see below).
• DOWN change mode, down band in band setting mode and down frequency in CW mode. Also used to signal start of serial configuration terminal on start-up (see below).
• TX manual transmit, manually set the transceiver in transmission mode, also keyer in CW mode.

The UP/DOWN buttons can be used, if pressed simultaneously to place the transceiver in Band change mode, when pressed at the start they can be used to place the transceiver in calibration mode, this behavior is identical to the ADX transceiver with it's origiinal firmware.

¡WARNING!
The three resistors pulling up the switch voltage from +5Vcc in the ADX transceiver **needs to be omitted (not populated)**
on the PDX transceiver as they will feed the corresponding GPIO pins with +5Vcc instead of +3.3Vcc and might result in the
**damage** of the processor. The firmware uses internal pull up resistors to replace them.

LED

Like the standard ADX transceiver the design carries four (4) LED to signal the transceiver state:

• WSPR LED, to signal WSPR mode, band1 in band setting operation, calibration mode and terminal mode. Also frequency indicator in CW mode.
• JS8 LED, to signal JS8 mode, band2 in band setting operation, end of calibration mode and terminal mode. Also frequency indicator in CW mode.
• FT4 LED, to signal FT4 mode, band3 in band setting operation, end of calibration mode and terminal mode. Also frequency indicator in CW mode.
• FT8 LED, to signal FT8 mode, band4 in band setting operation, end of calibration mode and terminal mode. Also frequency indicator in CW mode.
• TX LED, to signal transmission, band setting operation, end of calibration mode and terminal mode, watchdog activated, also keying in CW mode.

  ¡WARNING!
  Due to the smaller drawing capability and lower output voltage of the GPIO pins 3mm red LED are
  recommended for all five positions

Si5351

Operation of the Si5351 clock generator is identical as in the ADX transceiver with minor differences, it's being used as:

• CLK0, transmitter oscilator and FSK generator.
• CLK1, receiver oscillator.
• CLK2, calibration reference signal.

¡WARNING!
The calibration process is manual as per the ADX transceiver (see the original method as described on Barb's web page and blog), the hardware
support a future implementation of an automatic calibration procedure which isn't been implemented in the firmware.

Receiver

The receiver sub-system is identical than the ADX Transceiver.

RF Power

The RF power (driver and finals) is identical than the ADX Transceiver.

SWR protection

A Zener Diode (D10,1N4756) located where the board TP3 is defined would prevent a situation of high SWR to damage the finals.

Low Pass Filter

The Low Pass Filter (actually more than that) is needed to suppress unwanted spurious responses and also to achieve high efficiency class E operation. The design is identical than the ADX Transceiver.

ADX2PDX daughter board

Barb (WB2CBA) created a small daughterboard, dubbed as ADX2PDX, which can be used to transform a standard ADX transceiver into a PDX transceiver with minimal modifications.

The board replaces the Arduino Nano on a standard ADX board, the signals in that socket are rerouted according with the rp2040 pin assignment. Additional support circuits are added as well:

• +5Vcc regulator, the ADX depends on the +5Vcc regulator present on the Arduino Nano.
• Comparators (MOSFET and LM393) needed to process the signal for the rp2040 firmware counting methods to process.
• Miscellanea hardware needed for the rp2040 to operate.

A picture of the prototype daughter board is shown

The daughterboard schematic is as follows:

¡WARNING!
**ADX2PDX daughter board is at the prototype stage**

¡WARNING!
In order to accomodate the daughterboard on top of the standard ADX board some construction modifications are recommended:
* Bend L2,C16,C17 and C19 as they are large components with might interfere with the daughterboard.
* Use a 2x pin strip (female) instead of a socket for the Arduino nano as it creates better grip of the daughter board.
* Remove R12,R13 and R14 to avoid +5Vcc to reach the GPIO pins of the rp2040 rated for operation at +3.3V and prevent potential damage. 

This approach allows an existing ADX board to be upgraded with the new processor but to develop either a wired prototype or a custom board using the rp2040 processor instead of the ATMEGA328p are also options.

Construction note
For building flexibility both the MOSFET and CI based comparators are provided in the daughter board but only one
must be connected, even if both can physically be present the selections of the JP1 and JP2 jumpers will define
which one is actually used.
Some modifications on the ADX2PDX daughter board and the ADX board needs to be applied depending on the 
comparator used, please see text below for further details.

ADX2PDX daughter board prototype fixes

The following fixes needs to be applied to the prototype daughterboard

Connect Raspberry Pico +3.3V (pin 36) to the +3.3V pin on the Arduino socket.
Cut the trace between Raspberry Pico GP15 (pin 20) to the pin header labeled as ATU
Run a wire from the Raspberry Pico GP27 (pin 32) to the pin header labeled as ATU
Solder pin 5 & 6 of the LM393 chip together and wire them to ground.
Solder with a small blob of tin the proper connection of the JP1 and JP2 according with the comparator to be used.

ADX2PDX PCB

Barb (WBA2CBA) created an initial sketch for the PCB layout which is shown as follows

The Gerber files for this board can be found in the following link.

MOSFET comparator

In order for the ADX2PDX daughter board to work properly using the MOSFET comparator the following connections needs to be made on it.

• JP1 2-3
• JP2 1-2

Also the following modifications (mods) are required on the ADX board

MODS

* Replace R4 by a wire (short circuit).
* Replace R1 by 1K (instead of 1M).
* Replace R2 by 10K (instead of 4K7).

The above changes ensures that the MOSFET in the daughter board (Q1 BS170) is polarized in a way that a 1V pk-pk audio signal will make it conduct as shown in the following picture where the signals at the gate and drain of Q1 are shown.

LM393 comparator

In order for the ADX2PDX board to work properly using the LM393 comparator the following connections needs to be made on it.

• JP1 1-2
• JP2 2-3

No modifications are needed on the ADX board.

The above changes ensures that the LM393 in the daughter board (U1A LM393) is polarized in a way that a 1V pk-pk audio signal will make it conduct as shown in the following picture where the signals at pin 2 (input) and pin 1 (output) are shown.

Testing

The testing of the project requires several steps before being released as an alpha version

Test Setup
   ADX-rp2040 running on PDX V1.0 wired prototype (LT7D), WSJT-X on a MacOS. 
        MOSFET comparator JP1 2-3 && JP2 1-2
                Oscilloscope UNI-T UTD2052CL 50 MHz

   ADX-QUAD-V1.5 running on a ADX transceiver, WSJT-X on a Raspberry Pi
   The above test were repeated with the following setup
   ADX-rp2040 running on a standard ADX board, WSJT-X on a MacOS, the set is
   repeated with both the MOSFET and LM393 comparator setups (with the mods 
   applied as documented in the hardware section).

• Unit test. The following tests were performed with DEBUG mode on (#define DEBUG 1):

  • Initialization sequence.
  • Mode change (UP/DOWN sequences).
  • Band change (UP/DOWN sequences).
  • Transmission mode.
  • Recover of changed values on next startup (EEPROM saving)

• Integration test.

  • Audio in, FSK in

Using WSJT-X as a 1800 Hz tone generator, Channel 1 Audio in (AF), Channel 2 Comparator out (FSK) Verified voltage levels and operation of the VOX algorithm

*   Si5351 operation, transmission (CLK0 signal, 7.074 MHz, high level)

Using WSJT-X as a 1800 Hz tone generator, Channel 1 Si5351 CLK0 (TX), pressing the TX button Verified voltage levels and frequency of the clock

*   Si5351 operation, reception (CLK1 signal, 7.074 MHz, low level).

Using WSJT-X as a 1800 Hz tone generator, Channel 1 Si5351 CLK1 (RX), not pressing the TX button Verified voltage levels and frequency of the clock

*   Si5351 operation, calibration (CLK2 signal, 1 MHz, low level).

*   Calibration (enter calibration mode, UP/DOWN to calibrate, save in EEPROM).

• Communication test.

  • Communication, send 1800 Hz tone. Using WSJT-X as a 1800 Hz tone generator, sent TEST TONE 1800 Hz

  • Communication, reception of 1200 Hz tone. Using WSJT-X as a 1800 Hz tone generator, received TEST TONE 1800 Hz (generated by another station)

  • On the air test (OTA)

  • Communication, Answer a CQ from other station.

*   Communication, Send a CQ and being answered by another station.

• Results with ADX2PDX daughter board

Results using the ADX2PDX daughter board were identical as per the wired PDX prototype.