Thursday, February 9, 2012

DL3PB's All-Tunnel Diode Parametric Triumph

My friend, DL3PB, has done something amazing with a pair of Russian-made tunnel diodes! Peter's imagination and patience are truly an inspiration. Congratulations OM! In his own words...

"I'd like to share with you a long-cherished dream, that recently came true, fourty years after I came to read about hams using tunnel diodes to make QSOs when I was aged twelve or so:

Finally I managed a first skywave QSO with my PARASAKI-transceiver, an 'all diode' rig: Christophe/F8DZY replied to my very first call on 20m band in REF-contest last weekend. I was running 2mW into a temporary vertical dipole on my balcony.
Distance between us is  918km - obviously OM Christophe has excellent ears.

Those interested in the cruel details of my circuit, please find attached a schematic and a photo of the pretty ugly setup. The circuit is designed straight-forward with exception of the parametric VXO, derived from Mike/AA1TJ's famous Paraceiver design. (see )

The low impedance of the high peak-current tunnel diodes make it very difficult to build a crystal controlled oscillator rather than an LC oscillator, synchronized by the crystal more or less, at least on the higher SW bands. The Parametric VXO provides a crystal-stable, chirp-free signal on expense of an output power of two milliwatts only instead of ten, but with an amazing spectral purity, no need for a low pass filter or such.

Of course it sounds pretty cool making a QSO with a 'bunch of diodes' and a parametrically excited crystal, but believe me or not, I'd preferred to bring that full ten milliwatt into the air - on the other hand that approach allowed to tune the rig a bit ( ~ 5kHz/per xtal ), which turned out to be much more valuable than a few milliwatts more while being 'rock-bound'.

The receiver in its 'gain-less' version works fine for strong signals - while listening to QRP(p) stations, the moderate gain of the audio amplifier helps a lot. A comfortable frequency shift between receive and transmit is realized by the 5┬ÁH inductor at the LO-port of the mixer, with little effect on sensitivity."


  1. Hi there,

    I was going to ask you has anyone ever design a complete transceiver using diodes and SCR's only?

    Diode ring mixer Tunnel diode VFO, SCR audio amp (I saw your 20db amp, by the way) but how would you amplify RF? Another SCR?


    Stephen Walters
    mister35mm at yahoo dot co dot uk

  2. Hi Stephen,

    SCR's have long been used to build efficient, high-power, VLF transmitter PA stages. In his book, "Practical RF Power Design Techniques," Irving Gottlieb suggests the maxiumum useful operating frequency using SCRs is ~30kHz.

    An article appearing in the September 1967 issue of, "The Westinghouse Engineer" discusses efficient SCR-based PA stages at frequencies approaching 90kHz. This is accomplished by firing a succession of SCRs "gatling gun style" and summing the output pulses.

    General Electric's venerable, "SCR Manual" lists VLF radio transmitters among the SCR applications. Chapter 13 provides detailed information on the topic of SCR inverter design (which, in a nut shell, is what a radio transmitter amounts to!).

    Unfortunately, their slow intrinsic turn-off time (~20uS) precludes the use of SCRs at anything approaching HF.

    On the topic of all-diode transcievers...I admit it isn't much, but some years ago I did make a 160km "skywave" QSO with an xcvr built around a single tunnel diode.

    Mike, AA1TJ