The HF transceiver I couldn't finish | Słomkowski
Originally published on 2009-08-10, updated on 2026-06-26 , 24 min<br>read, ~5000 words.
I’m slowly thinking about getting back into amateur radio, so I started digging out old projects. I thought, why not publish my struggles with<br>radio-frequency design on my website? There is some value in showing failed projects, after all.
Introduction
At the beginning of my education in liceum, I was rapidly becoming interested in the shortwaves. A year earlier, I had discovered<br>the ham radio club SP2KPD in my hometown and met many helpful hams there. Since electronics and tinkering had always been<br>alluring to me, I never considered simply buying a factory-made transceiver. What I wanted instead was to build a general-coverage receiver, or even a<br>transceiver, myself — so that my creation would be truly mine and I would have complete control over how it worked. That goal proved to be<br>foolish: the scope of the project overwhelmed me.
I oscillated (pun intended) between a classic 0–30 MHz receiver with a traditional quartz filter, an SDR with a Tayloe mixer, and a full-fledged<br>heterodyne transceiver. None of these attempts were fully successful; nonetheless, I think I learned a lot in the process before abandoning them. Shortly<br>afterward, I moved to Poznań to begin my university studies.
It took me over sixteen years to finally return to radio. A lot of things have changed in the meantime. Access to literature and tutorials is now<br>superb. Tremendously useful tools like NanoVNA are available cheaply on AliExpress, and a wealth of RF<br>modules is available as well. LLMs can answer the majority of my design questions.
Before starting fresh again, I unshelved my old schematics and photos and pieced them together as a memorial to my teenage fascination. The majority of<br>the photos were taken before 2010 and were intended not for publication, but for my own documentation. Admittedly, they are of low quality. A lot of<br>information has been lost (such as firmware), because I did not keep proper notes at the time.
I also realized that no other project on this website, except Octoglow, has<br>a similar level of complexity to the transceiver project. Unfortunately, adult responsibilities have limited the scale of my endeavors.
First attempt at 9 MHz heterodyne receiver
At the beginning, I wasn’t thinking about transmitting or becoming a licensed operator, so I focused on the receiving side. Receiving also allowed for<br>smaller antennas; I could get away with a random wire.
PCBs for the first iteration of the heterodyne receiver.
After some encouragement and receiving some electronic components from members of the SP2KPD club, my plan settled on a classic heterodyne receiver with<br>a 9 MHz crystal filter, a double-balanced diode mixer in the front end, and an intermediate-frequency amplifier built with double-gate MOSFETs. This<br>resulted in several modules, each described below.
The VFO based on AD9852 DDS
I think I first started entertaining the idea in 2007. I realized that, in order to play with radio frequencies, I needed some measuring equipment.<br>Fortunately, I already owned a 100 MHz four-channel analog oscilloscope at the time, but apart from that I didn’t have much electronics equipment (except<br>mountains of scrap electronics to extract components from, of course). To fill this pressing need, I decided to build a signal generator that would serve<br>not only as a VFO in the finished transceiver, but also as a device for testing other receiver<br>modules.
A classic capacitor-tuned generator wouldn’t provide a precise frequency readout. Moreover, building a proper LC VFO is an art in itself — I knew<br>beforehand that any attempt I made would result in a wobbly, unstable oscillator.
Instead, I opted for a more modern approach and jumped straight into direct digital synthesis after stumbling upon a fairly popular design by SP3SWJ (formerly SP2SWJ ), leveraging the AD9851 chip from Analog Devices. I was somewhat concerned about phase<br>noise, but I realized that many hams had already been successfully using DDS-based synthesizers in their transceivers, so it was unlikely to be a<br>problem for a beginner like me.
2026 photo of the AD9852 DDS board, found in my "failed projects" box.
In retrospect, I should have simply replicated SP3SWJ’s synthesizer boards and assembled them. That would have solved the VFO problem before I moved on<br>to anything else. The LCD and front panel would have been a bonus. But, being myself, I designed my own board instead.
This was my second or so home-made PCB. Previously, I had built all my circuits exclusively on universal PCBs. This was a major step forward in my<br>electronics hobby. I learned CadSoft Eagle in the process, a commercial software package that was popular at the time. Nowadays, I use<br>KiCad for the same purpose.
I was pleased that the thin traces turned out well.
The backside of the DDS board.
My synthesizer implementation is based on the AD9852...