My 10 GHz Narrow-Band Transverter System
Taking a cheaper and longer road...to 3 cm microwaving!
   

By: Bertrand Zauhar, VE2ZAZ / VA2IW


This was a winter of 2024 project, and a commitment to myself: Put together a 10 GHz transverter and dish system, and spend as little as possible in the process! This meant re-use what I had on hand, design and/or assemble as much as possible. In the end I got a good performance system, and had a lot of fun putting it together.

Updated - 2024/08/06: On the Sequencer board, lowered the value of the current setting input resistors on the PhotoMOS relays, from 4.7K to 2K. This guarantees proper relay operation at hotter ambient temperature (in the sun!).

UPFRONT CONSIDERATIONS

When reading through this web page, please put the following into perspective:
  • I already owned the required microwave test instruments to properly test the project,
  • I had Surface-Mounted Device soldering skills and tools,
  • I already had many of the required components and hardware,
  • RF design is in my skill set, and had no fear of tackling this project.
Simply put, this project is not for the standalone beginner. A lot of it is custom-designed for this particular system, and may not work on other setups. However, I share the following in hope that others can re-use some of these concepts and approaches on their own setup.

YouTube VIDEOS

This web page provides some details of the final system. However, if you are interested in knowing much more on the project, follow the design steps I took and the measurements I made, you MUST watch my series of 25 (yes!) short progress report videos. The YouTube playlist is accessible by clicking on the thumbnail to the right.  



OBJECTIVES

These were the objectives I set early on in the project:
  • Use the 10GHz dish I already owned and fit everything in the rear compartment,
  • Design and/or assemble as much as possible,
  • Shoot for at least 1Watt of transmit power (2-3 W would be great),
  • Try to 3D-print the various enclosures. EM shielding practical?
  • Use my existing FT-817ND radio as the 144 MHZ IF radio (note that I eventually switched to an IC-705 for the IF),
  • Spend AS LITTLE AS POSSIBLE! Will I use the system often?

SYSTEM DESCRIPTION

See video #5 and #18 for more detail on system overview:
https://youtu.be/8DFl8cSzDZE
https://youtu.be/dRT_LpgJlLo

High Level Block Diagram

Detailed Block Diagram

THE HARDWARE

The Parabolic Dish

I bought it used at a hamfest, paid $20 USD including the feed. It is a 22-inch (57 cm) diameter prime focus dish with the following characteristics:
  •  f/D ratio: 0.27 (a deep dish)
  • Calculated gain: ~34dB @ 10.368 GHz
  • Had a side shield (ended up cutting it off later),
  • Included a 10 GHz dual-dipole feed, offering linear polarization,
    • Good for deep dish,
    • Built on a WR-90 waveguide,
    • Has a standard UG-39/U Flange,
    • Cut to focal length f = 6 inch distance from the bottom of dish,
    • Optimal for my dish?
  • There is a rear compartment with cover, very convenient!
  • Has two welded on mounting plates, good for tripod mount and IF radio side shelf addition,
  • Is all aluminum, however still somewhat heavy for hand carried expeditions.
See video #1 for more detail: https://youtu.be/cHVOxPAIHcA


Mail Order From W1GHZ

The following components were ordered from Paul Wade - W1GHZ's website. Prices are very reasonable
, and it saved me from having to re-design these blocks and having to separately order some of the key components I did not already had on hand.
  • Transverter PCB,
  • Tripler PCB,
  • 10x NLB-310 10 GHz MMIC Amps, used on the above PCBs,
  • 2x MCA1-12G Mixers (one spare…), used on the transverter PCB.
The transverter and LO Tripler designs are well documented on Paul's website. See http://www.w1ghz.org/xvtr/transverter.htm, and also browse other pages of his website, as all the 10 GHz stuff is not regrouped on the same page.

See video #8 for more detail: https://youtu.be/Fq8jSmYnjW8


Local Oscillator (LO) Design

It is based on a Chinese-made ADF4351 Board, which sells for around $25 USD on AliExpress,

  • Covers 34 – 4400 MHz, with Max. output of about +3 dBm,
  • Can accept a 10 MHz input reference,
  • Needs programming at power up. I custom-designed a PIC12F683 micro-controller firmware that:
    • Sets output to 3408 MHz,
    • Sets Maximum output power.
    • Sets the PLL loop current to minimize output spur generation.
  • ADF4351 board also requires a few physical mods to improve phase noise performance
    • Disable on-board oscillator,
    • Add bulk capacitive filtering on +5V rail.
    • Add pull-up resistor to force output to be always on.
The firmware source file was written in XC8 C language, for a PIC12F684 micro-controller target. It can be compiled in the MPLABX IDE.


See videos #6 and #12 for more detail:
https://youtu.be/DzZl1HtuDBA
https://youtu.be/3q-5-Ht0lL4

LO Tripler
It is based on the W1GHZ Tripler board.
  • Has three MMIC amplifier stages, the first one being saturated to produce harmonics,
  • The two 10GHz Pipe Cap Filters are tuned to 10224 MHz LO frequency,
  • Produces a +6 dBm (4 mW) output power, enough to drive the Double-Balanced Mixer on the transverter board.


     

See video #9 for more detail: https://youtu.be/E8fPbDX_8-A



10GHz Transverter

It is based on the W1GHZ Transverter board.
  • Has three amplifier stages on both Rx and Tx paths. Amplification is kept linear to reduce harmonics generation,
  • Uses:
    • An on-board resistive splitter/combiner,
    • A Mini-Circuits MCA1-12G Mixer,
    • Four 10GHz Pipe Cap Filters tuned to LO frequency,
    • Independent +8V supply rails for Rx and Tx strips,
  • Produces a 0 dBm (1 mW) output power, somewhat lower than ideal, however is enough to drive the Alcatel Power Amplifier.
See video #10 for more detail: https://youtu.be/oA-fHLZmvXY



10GHz Low Noise Amplifier (Rx Preamp)

It is a recovered Bell-ExpressVu (Dish-Network brand) DSS offset dish LNB (see picture to the right),
  • Natively covers the Ku band, 12-13 GHz. Offers two usable broadband preamplifier strips (cross polarization satellite coverage), which work great at 10 GHz.
    • Uses a low noise NEC Hetero-Junction FET transistor front-end,
    • Offers a 20+ dB Gain and a 0.35 dB NF!,
  • Mods performed:
    • Disconnected the downstream band-pass filter and mixer.
    • Disabled the on-board Local Oscillator.
    • Kept the existing supply and FET bias circuits. Fed +8 VDC into the +5 V linear regulator,
    • Cut off the existing input probes,
    • Interfacing the input and output coaxial cables was the challenge. Made perpendicular-to-board-surface connections with conformable coaxial cable, from the PCB bottom side, and through a tiny hole aligned with input and output copper tracks. Conformable cable shield is soldered to bottom side copper plane.
  • Careful! Thin, brittle PCB material and copper foil is used on the LNB PCB. Avoid excessive mechanical stress, and do not overheat. 
See video #13 for more detail: https://youtu.be/EfDosDKckak



144MHz IF Transfer relay and Attenuator
  • Is hand-made on universal PCB,
  • Offers a 30 dB Pi-shape Attenuator in Tx and a straight through path in Rx,
  • Defaults to attenuated path (Tx) to protect the transverter mixer,
  • Uses a 50-Ohm power resistor, purchased on AliExpress,
  • Uses a standard 3-Amp DPDT power relay,
  • The measured VSWR is 1.4:1 Good enough for ZAZ!

  

See video #16 for more detail: https://youtu.be/xlSiw_hP1tg


10 GHz Power Amplifier

The PA was recovered from an Alcatel MDR-6000 point-to-point microwave system.
  • RF Power: -5 dBm in, 3+ Watts out,
  • High input sensitivity saves from adding an extra boosting stage between the Transverter and the PA,
  • It is a Class A amplifier, so is DC power hungry
    • Amplifier is 13% efficient (best case!),
    • +10.5V @ 2.2 Amps,
      • I used a Buck-Boost converter for the +10.5 V supply, with added input and output low pass LC filtering,
    • -5V @ 10 mA,
      • I used a +5 V to -5 V converter, with added input and output low pass LC filtering.
  • Has a thick bottom cooling plate, so no heat sinking is required for our intermittent ham operation,
  • Measures 8 inches long, and is rather heavy at 3.5 lbs,

 

See video #15 for more detail: https://youtu.be/viyW5b-eYng


3D-Printed Enclosures

Various parts of the system were contained in custom-designed enclosures designed in FreeCAD.
  • Printed in PLA plastic. Took 2 to 3 hours for each piece of an enclosure,
  • Inside walls were lined with self-sticking aluminum duct tape. The tape extended to the outside of the bottom surface,
  • The enclosures have an inside lip that squeezes the board between the top and bottom parts,
  • All holes were drilled afterwards,
  • Feed-through capacitors were added for precaution on the DC supplies,
  • Power on LEDs were added afterwards,
  • 3mm Hex head screws hold the end plates.



See video #14 for more detail: https://youtu.be/N8kBVO035X0


Coaxial SMA Relay
  • Used a standard 1:2 microwave SMA coaxial relay,
  • Good for several Watts of RF when not hot-switched,
  • The 24 VDC version of these relays is much cheaper than the equivalent 12V version,
  • However a 24V source is required,
    • Used a small Adjustable DC-DC Boost converter,
    • Costs less than 1$,
    • Is a switching supply: Noisy, however not too critical since only activated in Tx. I added some extra capacitive filtering on input,
    • 8-14V in, adjust to 20-26V out, depending on relay requirement,
    • A 20V output is fine for 24V relays. Reduces coil current,
    • Also added a back-EMF protection diode on converter output,
    • Mounted it to the relay using a small screw and some double-sided sticky tape.


Waveguide-to-Coaxial SMA Transition

I made a WR-90 waveguide to SMA transition using a surplus waveguide piece with existing UG-39/U flange at one end.
  • I followed W1GHZ’s recipe (QST magazine article, Nov/Dec 2006,
  • I used a SMA connector/probe I already had, cut it to proper length, drilled the wall and and positioned the probe as per the article. Note that a lot of heat is required to solder on a silver-plated brass waveguide! So I pre-heated the unit on a heating plate prior to soldering the connector.
  • Using a signal generator and directional coupler, I tuned the transition by pressing down on the copper foil rear wall (see photo of VSWR once installed on dish).
See videos #3 and #4 for more detail:
https://youtu.be/rtLG-l7ExIQ
https://youtu.be/LJnOWWF39iE



Everything else

Over the 40+ years of doing ham radio, I have accumulated a lot of electronic components, hardware, cables, wires, etc. Obviously, for anyone building such system, the cost of these elements would have to be taken into account, something I did not do.


POWER

Power Distribution Diagram

See video #19 for more detail: https://youtu.be/VmMuDBpOcvw

T-R SEQUENCER
It is inspired from “FT-817 Transverter Sequencer” paper by VK4CP / VK4GHZ, and re-designed by me for this application,
  • Uses a PIC16F1824 micro-controller, firmware written in Microchip C language.
  • Fixed 40 ms delay between each channel, modifiable by recompile.
  • Uses versatile AQV212 PhotoMOS Solid State Relays for the outputs.
  • Detects the presence of the FT-817 (or IC-705), senses PTT, inhibits IF Tx power until Tx sequence is completed.
  • Provides +8V_Tx, +8V_Rx, +12V_Tx, +12V_Rx, +5V.
  • Controls the +10.5V DC-DC converter (for the Alcatel Power Amp.)
I produced two versions of the design for different IF radios, the Yaesu FT-817/818 and the Icom IC-705. The main difference resides in how the Tx RF power is inhibited until everything is switched.  On the FT-817 version, the TX Inhibit input pin is used. On the IC-705 version, the radio's ALC input voltage is controlled instead.

Circuit Schematic for FT-817


Circuit Schematic for IC-705


Sequencer PCB
  • Made 100x100mm in size, outline and features meet usual PCB manufacturer low cost deals,
  • Includes a prototyping area, allowing for expansion or feature addition,
  • The +8V regulator must have proper heat sinking, as that voltage rail could provide up to 800mA.
  • Designed in KiCAD.


Firmware Flow Diagram


Firmware Source File

Here is the firmware source code, written in Microchip XC8 C language, for a PIC16F1824 micro-controller target. It can be compiled in the MPLABX IDE.
Microwave T-R Sequencer
IF Radio: FT-817/818

Microwave T-R Sequencer
IF Radio: IC-705


See video # 20 for more detail:
https://youtu.be/D3IxH2UfIGM

SYSTEM INTEGRATION
Floor space planning is a compromise exercise.
  • Optimized for the shortest RF links,
  • Marked, drilled and threaded all module mounting holes,
  • Segregated the switching converters,
    • Available RF-tight space between dish and rear compartment. Used it for the PA switching converters,
    • Installed Feed-through capacitors between compartments.
    • Added low-pass LC filtering on inputs and outputs to reduce switching noise.
  • Installed all modules, sequencer board and power terminal (under sequencer),
  • Wired up the various power lines: Always on, Tx and Rx,
  • Installed the proper coaxial lines:
    • Regular RG-type coax for IF and 10 MHz, BNC connectors,
    • “Conformable” PTFE coaxial or Semi-Rigid UT-141 PTFE coaxial for 3.4 GHz and 10 GHz, all with pre-installed SMA connectors.
See videos #21 and #22 for more detail:
https://youtu.be/dmCWcGQ1GIM
https://youtu.be/Hf6M1rQftpg




OPERATION

The first trials were receive tests of the VE2TWO beacon, located in grid square FN25uk, on the top of Mont Rigaud, some 120 Km away from my front porch! And it worked repeatedly! Hear the result here:
https://youtu.be/j7PxdoeTBKw


My first 10 GHz two-way QSOs took place on April 6th, 2024, from grid square FN15xi. Grid square FN03 was reached three times, with a maximum QSO distance of 376 Km between VE3KH and myself. Hear and see the result here:
https://youtu.be/Mf-WBhdVY7A

Another QSO with VE3KH took place in June 2024, this time using rain-scatter propagation.
Hear the QSO here:
https://youtu.be/uAOAHR4Jhyc





Other activities will follow. See the following video playlist for reports on my 10 GHz activities:
https://www.youtube.com/playlist?list=PLd6VW5EHf22c2C9gcL6SY1kPRs3QNjIl3

CLOSING COMMENTS

All in all, I am very pleased with system performance. I spent around $400 on that system, which is much less compared to buying off-the-shelf modules. Of course, there are improvement that could be made. However, I will make several more QSOs before changing anything. This will allow me to compare the results with other
hams' co-located 10 GHz systems, and assess areas of improvements.