SI4732A MINI-RX MODIFICATIONS

When hunting for documentation to pair a DSP-IC SI4732A with an ESP32 to build a small shortwave receiver, I stumbled upon a Chinese developer’s project that had already cracked the code. His work became the foundation for many hardware and software evolvements, and eventually, Chinese manufacturers began to offer ultra-affordable hardware powered by increasingly impressive community-driven firmware for the utilized ESP32S.

At ~€30, this pocket-sized companion is not only perfect for field shortwave listening, but now it even could outperform my bulky stationary Kennwood R600 in some respects … if there were not a few issues :

I have modified my V3 MiniRx hardware to address five issues:
 1. To circumvent the battery drain issue,
 2. To protect the input circuit against high voltage,
 3. To implement a solution for the VCO wobbling effect,
 4. To replace the old tuning knob, and
 5. To exchange the original by  a custom firmware

Many users faced this issues too, as I could read on the forums. It was quite interesting to investigate, test, discuss at Githup, come in contact with enthusiastic developers and tinkerers again. Her a brief version of my results -more in depth at my blog.

1. The battery drain issue
Some experiences with the first Hardware lead to a now sold V3 model . Presently being not very good documended, it especially features now an added High-Z impedance converter at the HF input. 

But attention! It’s wired straight to the battery, draining 5mA continuously – and so I received it with an empty Bat. 

Fix: Cut a PCB trace and reroute it to a switched 3.7V–4.2Vbat line, as shown here:

2. High voltage protection
My enthusiasm was dampened until I was able to  replace the SI4732A that  I had fried by only one antenna experiment. So this protection circuit was implemented at the SMA-socket before the next experiments:

3. VCO wobbling effect
Very long story short: The Si4732A uses a 3.4 GHz VCO in a PLL, and this signal leaks through the FM antenna pin and the diplexer to the antenna—and then even back again. So, a changing load at the antenna socket affects the VCO frequency, which is then periodically corrected by the PLL . And this  sounds like a Theremin when SSB mode is active and you touch the first centimeter of the antenna.
A simple air coil inductor—mechanically not very easy to implement—was the solution: The leakage got reduced by 13dBm! The Theremin is gone!

4. Tuning knop
It proved difficult for me to surf the waves continuously with the original tuning knob. I replaced it with a 4cm long piece of a wooden toothpick, which was pressed and glued into the axle gap. The wooden arm got protected with a piece of shrink tube. Now I can turn the axle very sensitively and continuously with the tip of your finger.

5. The apropriate firmware.
This project was fun, as I came in touch with some very enthusiastic and technically sopisticated guys. I’d like to mention Dave, G8PTN who wrote a fantastic firmware on Github as a foundation for several other specialized firmwares. And he even seems to know some of the darkest secrets of the DSP-chip ;-)

One of the forked projects was developed by H.J.Berndt who pushed it to very interesing limits. Or did you expect the firmware of this pocket sized companion to decode Morse, RTTY, RDS, …? 

So, here are some of my links:

More details at https://peterneufeld.wordpress.com/2025/06/13/si4732a-minirx-modifications/