DIY RF Condenser Microphone


• Just click on the image below for a PDF of the schematic and PCB parts list •

The schematic shows the current version of the circuitry employed in the creation of this simple amplitude modulated RF oscillator based microphone. This latest version (v.5.0) includes a number of revisions which have been made to further improve the overall perfomance of the microphone.
The PCB layout remains the same.

• You can find inductor construction details - including alternative source suggestions: HERE

• You can find some notes on the reasons behind the latest versions: HERE

• You can find a copy of Version 4.3 of the schematic and associated parts list: HERE

• You can find a copy of the original version of the schematic and associated parts list: HERE

• Some additional project concept notes can be found: HERE

The circuit concept is based on some of the techniques described by audio engineering legend Peter Baxandall in THIS PAPER from 1963.
This circuit is designed to be connected to a low impedance, balanced XLR microphone pre-amplifier, which must have the option of providing 48V phantom power. It consists of four functional sections:

• RF Oscillator   • Modulator/ Detector   • DC Power   • Audio Output
(The last two functions utilise the elegant and much copied 'Schoeps' style interface circuit).

• RF (Radio Frequency) Oscillator:
The circuit components connected to Q1 form a crystal controlled Colpitts type oscillator. Around 40% of the c.6V(p-p) sine wave present at Q1 emitter is connected to the primary winding of T1 via R3. With the values shown, the oscillator draws around 2mA from the 48V phantom power supplying the circuitry.

• Modulator/Detector:
TI and T2 are 'off the shelf' 5.3uHH tunable IF cans, available from Spectrum Communications here in the UK.
One end of the centre tapped secondary winding of T1 is connected to one terminal of C4 - a capacitor selected to have a value close to the measured capacitance of the capsule (see schematic for more details). The other C4 terminal is connected to one termination of the microphone capsule. The second termination of the capsule is connected to the remaining end of T1 secondary.
This arrangement provides a slightly unbalanced tuned 'bridge' circuit, where the 'balance' is further modified by the changing capacitive value of the capsule, as it responds to audio stimulation.
This change of the bridge balance state will enable amplitude modulation of the RF oscillator in proportion to the applied audio stimulus.
These changes in capacitive value tend to be extremely small - in the order of 0.001pF for an alternating pressure of around 1 dyne/cm² (i.e. normal speech level signals at around 30cm.) - according to Baxandall.
T2 is an identical tunable IF can to T1, and one end of the primary winding is connected to the centre of the capacitor 'bridge' across T1 secondary. The other end is connected to the centre tap of T1 secondary (which is also referenced to ground).
T2 secondary winding is loaded with C8 - a 47pF capacitor - which allows the inductor core to be adjusted to resonate at the crystal frequency. The high 'Q' of T2 when tuned will allow the tiny changing bridge imbalance signal to be 'stepped up' by T2 to a higher AC voltage on its secondary winding - without introducing any further noise. - This voltage is presented to the gate of the JFET Q4.
Note that the very high impedance of the JFET gate essentially presents no additional load to T2.
Q4 acts as both an infinite impedance detector and an audio phase splitter. The JFET is 'self biased' so that it is always biased around Vp, the JFET 'pinch off' point.
The infinite impedance detector function of the JFET takes the RF carrier present on the gate - which is amplitude modulated by the varying capsule capacitance - and uses it to charge both C9 and C13 on every negative going half cycle of the carrier. These two capacitors are then partially discharged - through R10 and R4 respectively - during the postive half cycle of each carrier wave, where the JFET is turned off. The capacitors are then recharged on the next negative half cycle, to a level set by any change in RF carrier level (amplitude modulation) caused by any new variation in the capsule capacitance.
The end result of this process is a demodulation of the RF waveform, so that only the audio portion is retained. The values of R4/C9 and R10/C13 are selected to allow an appropriate maximum high frequency audio limit to be applied to the signal.
Because the inherent self bias of the JFET ensures that Q4 is always biased around cutoff, the rectification of the signal does not require the bias of a conventional recifier diode.
In addition to the infinite impedance detector function, Q4 also serves as an audio phase splitter, presenting opposite phase audio signals at the source and drain terminals.

• DC Power:
Power for the unit is derived from the 48V phantom power presented to the circuit via the 3 pin XLR male connector XLR1. Q2 and Q3 are configured as emitter followers, with commoned collectors connected to the positive terminal of C7. This arrangement is generally known as a 'Schoeps' style output circuit, and is widely used for this type of device.
R11 and C15 - together with the 100nF capacitor now fitted into the 'D1' location - form additional low pass filtering which will help to decouple and remove any noise that may still be present from less than perfect 48V phantom power supplies.
This smoothed DC voltage of around 20V is connected to the junction of R2 and R10, to provide DC power to both the RF oscillator based around Q1, and the infinite impedance detector/ phase splitter based around Q4. The oscillator circuitry is further decoupled by C6.

• Audio Output:
The audio outputs present at the drain and source of Q4 are connected - via C5 and C10 - to the bases of Q2 and Q3. These are configured as emitter followers, and present a low impedance audio path via R7 and R8 to pins 3 and 2 of XLR1 respectively.
L1 and L2, C11 and C12 are included as lowpass filters to remove any residual RF from the audio output.