Build An Audio Speech Processor


Build An Audio Speech
Processor

By

Although most modern HF rigs have
built in speech processors, a lot of the low power SSB rigs and
kits do not have any real speech processor circuitry. When one
operates at low power on SSB, speech processing becomes almost
essential to get through the QRM. The circuit in figure 1 is a
low cost speech processor that will perform well with a minimum
of construction effort. The circuit even includes the microphone
element so that a complete speech processor microphone can be
assembled in an old defunct desk mike.

The circuit consists of an audio
preamplifier and a feedback amplifier for the necessary speech
processing. To use the speech processor, S1 must be closed first.
Leaving S1 open gives us a basic preamplified microphone with R5
as the gain control. When S1 is closed, voice peaks from the
feedback amplifier trigger Q2 to conduct causing the gate of Q1
to become negative. This action causes the resistance between the
drain and source of Q1 to substantially increase. Resistor R1
allows a small amount of audio to pass during very large voice
peaks. This circuit is similar to the audio muting switch found
in many QRP transceiver circuits. Capacitor C2 smoothes out the
voice limiting action while C6 provides a time delay giving us
the fast attack, slow release action.

The gain of the preamplifier is
set by R3 and R4 giving us a voltage gain of 100. The feedback
amplifier’s gain is set by R6 and R7 and is set at a voltage gain
of 10. Because of the large amount of gain in the circuit,
background noises (even music) can easily be picked up in a noisy
room. It is recommended that this circuit only be turned on in a
relatively quiet environment. If desired, the level of
compression can be made adjustable by replacing R7 with a 1K
resistor and a 10K or 47K potentiometer as in figure 2. If a
different microphone than a condenser element is used, remove R13
and R14 from the circuit. It may be necessary to increase or
decrease R7 for different microphones or to use the circuit in
figure 2.

While speech processors can make a
difference getting through the QRM, they can cause unnatural
sounding voice audio and should not be used when chatting
locally. S1 provides an easy on/off switch for the speech
processor during the local rag chew.

This circuit can be easily
constructed on any universal PC or perf board and should be
housed in some type of metal enclosure to keep RF out. All parts
are available at Radio Shack for those who hate mail order.

DE

Parts List

R1
 
22K �
Watt Resistor

R2
 
1Meg �
Watt Resistor

R3,R6,R11

 
1K �
Watt Resistor

R4
 
100K �
Watt Resistor

R5
 
10K
Potentiometer (RS271-1721)

R7
 
10K �
Watt Resistor

R8
 
10 Ohm �
Watt Resistor

R9,R10,R12

 
47K �
Watt Resistor

C1,C2,C5
 
.1
Microfarad Ceramic Disk Capacitor

C3,C4,C6
 
10
Microfarad Electrolytic Capacitor (observe polarity)

C7
 
100
Microfarad Electrolytic Capacitor (observe polarity)

D1,D2
 
1N4148/1N914
(RS276-1122) Switching Diode

IC1
 
1458
(RS276-038) Dual Op Amp IC. Socket is recommended.

Q1
 
2N3904
(RS276-2016) NPN Transistor

Q2
 
MPF102
(RS276-2062) or 2N3819 (RS276-2035) JFET Transistor.
(observe different pin out for 2N3819)

S1
 
Any SPST
Toggle Switch

Build An Audio Speech Processor

Build An Audio Speech Processor #2


Sorry, comments are closed!