DigitalVFO

Digital VFO

About

I'm an amateur radio operator, but since I'm not in the country in which I earned my licence I can't actually build or operate a transmitter, so I build receivers and test gear. I'm working on being able to operate from my country of residence, but that's another story.

One piece of test gear that would be useful is called a Signal Generator. This is a piece of kit that generates a radio signal of a known frequency. It can also be used as a VFO (variable frequency oscillator) in a receiver.

The modern way to accurately generate an RF signal is to use the AD9851 DDS chip which digitally generates a known frequency from a crystal source. There are quite a few people offering small boards using this chip but I like the DDS-60 daughterboard <http://midnightdesignsolutions.com/dds60/>_ from Midnight Design Solutions <http://midnightdesignsolutions.com/>_. All I need to do is use a microcontroller to control the DDS-60 with some sort of frequency display!

The obvious solution is to use an Arduino and the ubiquitous 16x2 display. Since the Arduino is a little large and I hoped to put the kit into a small case I decided to use a Teensy microcontroller <https://www.pjrc.com/store/teensy32.html>_ which is programmable through the Arduino IDE. It's about 35mm x 18mm! It's a very capable little microcontroller and is overkill for this project, but it's what I have at the moment. I'm working on a smaller "generic" design <https://github.com/rzzzwilson/KiCad_Projects/tree/master/Iota>_ that I can fit into any PCB projects that need smarts.

I got a couple of generic 16x2 displays from AliExpress <https://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20170504210259&SearchText=display+1602>. I got the rotary encoder from the same place <https://www.aliexpress.com/wholesale?catId=0&initiative_id=AS_20170504210300&SearchText=rotary+encoder+switch>.

Schematic

The latest code runs on this hardware arrangement:

.. image:: DigitalVFO_1.4_Schematic.png

PCB

The 1.4 PCB looks like this:

.. image:: DigitalVFO_1.4_PCB.png

Status

.. image:: DigitalVFO_case.jpg

Version 1.4 is running as shown above! There is provision for a battery plus monitoring of the voltage and calibration of various things. There will be another version called "HeadlessVFO" that will be controlled only by a PC via USB. This version won't need a display or rotary encoder, plus it'll always be connected to a PC and get USB power, so no battery. I probably don't need a Teensy 3.2 to drive HeadlessVFO, so I'm searching for cheap microcontrollers on AliExpress that can accept text commands from the USB connection.

13 Sept: Looking at the AtMega32U4. A similar chip is used in the Arduino UNO R3 instead of the FTDI FT232RL USB interface chip, but the AtMega is quite capable and can control the HeadlessVFO so I won't need any special USB interface chip. Anyway, that's the current plan.

Version 1.5

5 May 2019 Update

Noted corruption of the display and things like saved frequencies. The DDS-60 was being set to a negative frequency! Investigated, but suspect this problem was caused by interference by connecting USB power to the Teensy while the VFO was running from the battery. Debug by connecting USB power AT THE PC while the VFO is off. See if the problem reoccurs whie doing that.

Early 2019

Modified the PCB layout to include test points for battery voltage, RF out, ground, etc.

Version 1.4

13 September Update

The battery arrived but it won't fit into the case. Instead I'm using the two 18650 test cells with a small BMS PCB, all wrapped in a piece of heat-shrinkable tubing. I've bypassed the polarity protection diode as it eats a volt from the battery. Working well, but I've found that the power switch I used is a little unreliable. That needs to be replaced - the replacement is on a slow boat from China.

03 July Update

Still waiting for a 7.4v battery from China. When it arrives I'll experiment with the monitoring software. I may remove the 'polarity' diode once I have cables made up. It removes about a volt from the battery and the DDS-60 was cutting out long before the battery was empty.

21 June Update

The case is finished. When/if a permanent battery is installed I'll drill a few holes for a switch and a 9v charging port. The software for commands from the PC is basically finished. I removed a lot of stuff not necessary for using the DigitalVFO in an instrument cluster.

19 May Update

The 1.4 board is back from Seeed, built and operating. This version has provision for the micro to sense the battery voltage so it can display a battery symbol with full/empty options. Software development time!

Version 1.3

A redesign to use another BNC connecter as the first was junk. Not sent to Seeed as I also decided to add battery monitoring.

Version 1.2

After building the v1.1 PCB and paddling around on aliexpress.com to find different components, I'm starting on the v1.2 PCB. Should be smaller, have the LCD located in a more natural place and place controls in more natural positions. May also rethink the power supply - perhaps run both board and DDS-60 from a 78L05 instead of running 9v to DDS-60 and using the onboard regulator to generate 5v for the main board. Use the same case as before.

29 March Update

Version 1.2 of the PCB submitted to Seeed today. Changes were:

• moved components around to minimize vias
• on-board BNC socket, saves on internal cables
• moves Teensy to edge so USB goes through case, saves on internal cables
• added provision for internal 2s Li-Ion battery and manager PCB feeding the DDS-60
• added provision for LED power in battery power switch

Version 1.1

Need to:

• new case allows in-case programming
• new case has ventilation (passive at first)
• better packaging (printed circuit board)
• better, more linear, contrast control

And, of course:

• allow control over USB
• maybe add a "sweep" function to the in-case code (from, to, step, pause)
• look for bugs in the software

With PC control we can measure things with a "sweep" function. Allow the PC to control the VFO frequency and the PC can sweep the frequency to whatever limits required. At the same time, measure voltage/current/whatever from other Teensy instruments. This means we could build:

• an antenna analyzer
• measure filter response curves
• etc

This version will use a printed circuit board to address the general unreliability of the 1.0 version. The LCD will plug into the back of the main PCB and the DDS-60 will plug into the front. The rotary encoder will have a cable with header that will plug into the PCB.

The file DigitalVFO_Schematic_1.1.pdf is the schematic of the circuit. The directory kicad/DigitalVFO/Gerbers holds the output files produced by KiCad. The Gerber files were sent to a Chinese PCB fab factory.

The PCB has now returned from China, and it does look good. Better yet, assembling the board and testing it shows the thing WORKS! I had to change the code to match changes I made while setting out the PCB, BUT IT WORKS!!

Version 1.0

The VFO hardware is complete. Everything is inside a small metal case. Perhaps too small a metal case as running at 12v the thing overheats and becomes unreliable. It seems OK at 9v. I also didn't think about accessing the Teensy USB port when the case is closed, so I can't tweak the software without cracking the case, which I can't do often as the wiring inside is fragile.

Completed:

• Interface display to teensy, test writing, etc
• Interface rotary encoder, get rotate and button press events
• Basic display of 8-digit frequency
• Implement a simple 'event' system to produce system events
• Get frequency column select and increment/decrement working
• Save state in the EEPROM, restore state on start up
• implement a simple menu system
• extend the menu system to allow extra functionality
• added vfo_DClick double-click event
• add code to toggle standby/ON mode (from vfo_DClick)
• add in the DDS-60 control code
• add a "Calibrate" setting to adjust the divider value in the DDS-60 code
• test the thing fully!
• make up a metal case and rebuild into it

Perhaps:

• make another case with USB access
• think about making a circuit board for it (Fritzing, cheap Chinese PCB maker?)

Ongoing:

• look for bugs in the software
• maybe add a "sweep" function to the code (from, to, step, pause)
• split out the Arduino code samples showing RE handling and menu system
• think about hardware assembly methods to make DigitalVFO version 2!

Interface

I don't need a lot of functionality in this VFO, I just want to:

• set a frequency from 1.000000Mz to 30.000000MHz with steps down to 1Hz
• save and restore a number of frequencies

It should be possible to do all this with an interface made up of:

• A 16x2 display, and
• A rotary encoder, with switch.

Using the above the user can move a 'column selection' indication in the frequency display by pressing down on the encoder knob and then rotating it. With the knob up, rotation just increments or decrements the selected digit with over- and under-flow occurring to the left of the selected digit. This is pretty much the way everybody does it.

It would also be nice if the VFO remembered the frequency and selected digit if the power is lost and then restored. Use the on-board EEPROM for this, as well as remembered frequencies.

We may have to come up with some way of having a simple menu-driven method of:

• Saving/restoring frequencies
• Adjusting various internal parameters such as clock scaling, etc.

We could possibly drop into the menu system if the encoder knob is held down for some length of time.