"Knobless Wonder Design" 7 MHz SSB transceiver

This is my construct of the KNOBLESS WONDER type design:
single-fixed frequency, low power (2~3 Watt), 7 MHz Amateur Band, LSB transceiver.
(6 FET/Bipolar transistors & 1 LM386 I.C.)
It has no external adjustable user controls whatsoever, hence the name "knobless"
Based mainly on a design by Peter Parker (VK3YE) June 2013
but also borrowing from Drew Diamond's (VK3XU) excellent 1994/5 design of a TCF
(twin crystal filter) 7MHz transceiver. (see schematic below)
My alterations (purely for practical reasons), to the original VK3YE circuit design, are as follows:-
An additional relay (DPDT: "2 sets c/over contacts") has been added to switch the transmit or receive signal thru the crystal lattice filter, and thus eliminate 2 redundant transistor stages; here-to-fore used for amplification and "steering" (but now via relay contacts ).
In the original Bit-X design (from which some of the knobless wonder's circuitry derives) a transistor amplifier's input and anothers output are capacitivley "bussed" together at both ends of the xtal lattice filter.
A series diode in the (de-energized) collector load is added to prevent signal-loss back through its 220 ohm collector resistor and the supply rail.
Improved signal isolation and less stray path loss, will be evident by the use of mechanical relay contacts.
A single un-buffered JFET carrier oscillator is used, requiring 1 less transistor than previous.
The RF operating frequency is 7,200 KHz (on account of an Ebay Chinese listing of 50 xtals/$5 freepost - 10c each!)
A 39pF series capacitor with the xtal 'pulls' resonance up to approx 7201 KHz.
Then audio is mixed with this carrier frequency, in the balanced modulator.
Only the lower side band will pass thru the 7200 KHz xtal lattice filter, the carrier and opposite sideband being rejected.
Apparently some variation in passband is obtained by changing the capacitors in the xtal-lattice filter.
Higher capacitance = less bandwidth and vice-versa.
i.e. slight reduction in caps would result in a less "nasaly" sounding transmission.
I have used a manufactured balanced mixer/product detector (either SBL-1 or ADE-1 types suitable) as I had some on-hand, in lieu of the (otherwise satisfactory) half-bridge discrete component type originally used.
On receive the direct conversion principle is used, where the product detector mixes the incoming signal with the L.O. to produce audio output.
A power FET (mounting hardware 'heatsinked' to the metal casing) is used as the P.A. (VK3XU cct) instead of a bipolar transistor.
A trim-pot is used to adjust the correct gate bias voltage.
I have used a chinese manufactured s.m.c. LM386 audio output stage. Cheap & compact.
A handheld speaker/mic with PTT (push to talk) switch is used.
I have deleted the muting transistor from the LM386 audio circuit; since it's de-energized on 'transmit' so why the need for muting ?
Circuitry is constructed on double sided vero board. (see design worksheet)
Housed in re-cycled PC power supply; metal case.
On-air testing is yet to come, alas QRM from AM broadcast stations, might be encountered at 7.2MHz !

Original hand drawn circuit
Re-draft of original
Modified cct uses switching relay
VK3XU TCF Xcvr cct 1995 s
Final combined design uses Fets/relays
Vero design worksheet


Diode type Balanced Mixers - Technical
ELECTRET microphones - an explanation