ON4CRD — Club Radio Durnal

LM386 — audio reception amplifier

File LM386 · Revision 1

A construction file to amplify an already demodulated mono audio output. A single reference version links the schematic, the component list, wiring, and testing.

Actual project status

Document design to be carried out and verified on a prototype. Numerical results are calculated; no power, distortion, stability, or sound quality has been measured. File validation does not replace electrical testing.

Reference configuration

LM386N-1 in PDIP8, 6 V isolated power supply with current limitation, nominal 8 Ω speaker and nominal internal gain of 20. Pins 1 and 8 remain without external connection. The 10 kΩ potentiometer adjusts the volume, not the internal gain.

Source audio compatible

Use a demodulated mono output, with low impedance and an appropriate level, referenced to GND. Do not connect a RF output directly, a bridged speaker output, or two stereo channels together. A high-impedance detector requires a separately designed interface.

Diagram and measurement points

C1 is a film capacitor of 1 µF before the potentiometer. C2 is 220 µF, positive towards pin 5. R1 and C3 form a series branch of 10 Ω and 47 nF between VOUT and GND. C4 and C5 locally decouple the power supply. C6 is 10 µF, positive towards pin 7.

Pinout and ground

PDIP8 viewed from above, notch at the top: 1 to 4 go down to the left; 5 to 8 go up to the right. Pins 2 and 4 to GND, 3 to the wiper, 5 to VOUT, 6 to VS and 7 to the bypass. Also check the physical terminals of the chosen potentiometer.

Short placement

Place the decoupling near pins 6 and 4 and keep the output and stability network connections short. Separate the speaker current return from the sensitive input path up to the common point GND. This folder provides a wiring diagram, not a qualified printed circuit board.

Step-by-step assembly

Assemble with power off and check each node with the netlist. The symbols GND denote the same node; they do not prescribe a long common wire traversed by all currents.

Locate the notch of the socket and the pins. Leave U1 out of the socket during continuity and polarity checks.

Wire J2, GND, C4, and C5. Check for the absence of short circuits and the polarity of the capacitors before connecting the power supply.

Wire the input J1, C1, and RV1. Verify that the minimum volume connects the slider to GND and that the source does not reach the output.

Add R1, C3, C2, and J3 exactly following the schematic nodes. R1 and C3 are in series; the load is connected after C2.

Add C6, positive to pin 7. Check the connections of pins 2 and 4 to GND and leave pins 1 and 8 free.

Power off and capacitors discharged, insert U1 in the correct orientation. Connect the test resistor to J3, then prepare the multimeter and measurement points before powering on.

First commissioning protocol

These limits are prudent criteria proposed for this prototype, not performance guaranteed by TI. Note the values actually observed. A test resistor of 8,2 Ω replaces the speaker; it is not added in parallel.

Audio source disconnected, minimum volume, and S1 open: adjust 6 V and an initial limit of 50 mA, then close S1. Disconnect in case of incorrect polarity, smell, rapid heating, or persistent limiting.

After stabilization, record the idle current. More 20 mA without signal justifies a stop and a diagnosis. Measure TP1 relative to TP0; aim for the set voltage, without persistent drop.

Record the DC voltage at TP3: it should be close to half of TP1. A value outside 2 to 4 V with a 6 V power supply requires checking for a fault. At TP4, wait for the C2 charge then check that the residual DC approaches zero.

After these checks, a limit of 200 mA is suitable for the planned low-level tests. Inject a sinusoid of 1 kHz, gradually increase the volume, and actually measure 20 mV effective at TP2. Take into account the load setting of the generator.

Compare the alternating signal at TP3 and TP2. A calculated nominal value is 0,40 V effective at TP3 for 20 mV to TP2; it is not a measurement reading. Check the signal shape, the absence of clipping, and oscillation before any increase.

Measure TP4 on the known test resistor and limit the test to effective 1 V without clipping. Turn off before replacing the resistor with the speaker. Note the conditions; listening alone does not qualify stability.

Connection of instruments

Use an isolated power supply compatible with the ground of the instruments. Connect the generator returns and the oscilloscope ground clips only to TP0/GND. TP3 has a DC bias: choose the appropriate measurement coupling. An AC multimeter must be suitable for low level and frequency; its display alone does not prove the absence of oscillation.

Calculations and assumptions

The examples are analytical and use nominal components. They do not represent a SPICE simulation or measurements. The speaker power cannot be exactly deduced from its nominal impedance alone.

Nominal internal gain; the gain from J1 also includes the attenuation of the potentiometer and the capacitors.

Input, supposed low-impedance source: at maximum volume, the typical 50 kΩ model loads the 10 kΩ potentiometer. The values are orders of magnitude. The 100 nF alternative further attenuates low frequencies and does not replace C1 in the reference setup.

Output: estimation of the cutoff with a pure resistor. A 8,2 Ω load and a nominal 8 Ω speaker remain two distinct cases.

Power calculable on the known test resistance, with an unclipped sine wave and an appropriate effective voltage. The example of 0,40 V effective does not constitute measured power.

Maximum output power depends on the variant, voltage, load, and accepted distortion. This file does not promise either the commercial power of the LM386 family or a maximum specific to the prototype.

Diagnosis by symptoms

No sound: turn off, check U1, the power supply, continuity GND, the cursor, and the connection after C2. Then resume with a known signal and measure before replacing components.

Buzzing: compare with the source disconnected and volume at minimum; check decoupling, the path of returns, and loops created by the instruments.

Whistling, high current, or overheating: turn off; examine wire lengths, local decoupling, and series connection R1/C3. An oscillation may be outside the audible range.

Clipping: reduce the level, increase the supply voltage under load, and check the limitation. Do not compensate for a defect with a voltage higher than that of the design.

Rotation or low attenuated noise: check the source, C1, the potentiometer, and the mounted values. The input model assumes a low impedance source.

Results sheet to be filled out

The form CSV provided is empty. Keep the date, instruments, the actual measured load, settings, measurement points, and observations, including defects. Only report a result after the corresponding test.

Folder files

Manufacturer documents and method

Versions other than the French source use a public automatic translation with completeness checks and technical terms. No professional linguistic proofreading or electronic measure is claimed. The diagram references and netlist values remain common across languages.

Av = 20; 20 log10(20) ≈ 26.0 dB

Rin = 10 kΩ || 50 kΩ ≈ 8.33 kΩ; fc(C1 = 1 µF) ≈ 19 Hz

fc(C1 = 100 nF) ≈ 191 Hz; R = 10 kΩ: ≈ 16 Hz / 159 Hz

fc(C2 = 220 µF; R = 8 Ω) ≈ 90 Hz; R = 8.2 Ω: ≈ 88 Hz

TP2 = 20 mVrms → TP3_AC ≈ 0.40 Vrms

P(RTEST = 8.2 Ω; TP4 = 0.40 Vrms) ≈ 19.5 mW

Texas Instruments LM386 · SNAS545D · Rev. D · 2023-08
https://www.ti.com/lit/ds/symlink/lm386.pdf

Texas Instruments · Guidelines for Measuring Audio Power Amplifier Performance · SLOA068A
https://www.ti.com/lit/pdf/SLOA068

schematic.svg

design.json

calculations.json

bom-en.csv

measurements-en.csv

manifest.json

generate-lm386.mjs

Disclaimer

These dossiers are supplied as is for educational and experimental purposes. Calculations, simulations and file checks are neither measured performance nor certification or qualification of the assembled device.

The person building or using the device must check the schematics, components, connections and protective measures, follow manufacturer instructions and applicable regulations, and have the completed device validated before use. Depending on the device, electrical voltages, radiofrequency energy and heat can cause electric shock, burns or fire. Tests must be conducted by a competent person, starting at low power with suitable protective measures.

To the extent permitted by law, ON4CRD and the contributors do not guarantee fitness for a particular purpose or results, and disclaim liability for damage arising from construction, modification or use. This notice does not limit any mandatory rights or any liability that the law does not allow to be excluded.
