A carbon microphone turns speech into a varying line current by letting the diaphragm squeeze loose carbon granules, whose electrical resistance falls as they are pressed together, while a battery supplies all the energy. Because the signal is the fractional change in a loop’s total resistance, the same granule swing gives about 34 times more signal current at 48 V through 1,000 ohms than at 3 V through 1,460 ohms. An electromagnetic microphone needs no battery but delivers only millivolts.
Part of Telecommunications
Converting sound pressure into an electrical signal for a telephone line, with the carbon-granule and electromagnetic designs compared by what they actually deliver.
A Speaking Voice Pushes On The Diaphragm With About 25 Micronewtons
Ordinary conversation is about 60 dB SPL, and the corpus gives the reference pressure as 20 µPa (see Sound Wave Basics). Sound pressure level is 20 times the base-10 logarithm of the pressure ratio, so 60 dB is a ratio of 10^(60/20) = 1,000.
- Pressure: 20 µPa × 1,000 = 0.02 Pa, the figure Sound to Electricity also gives
- Take a 40 mm diaphragm, inside the 30 to 50 mm that Carbon Microphone gives. Area = π × (0.020 m)² = 1.26 × 10⁻³ m²
- Force = 0.02 Pa × 1.26 × 10⁻³ m² = 2.5 × 10⁻⁵ N, about 25 µN, the weight of roughly 2.6 mg
That tiny force moves the diaphragm less than a micrometre. A transducer must turn it into something a wire can carry. The two buildable designs do it in opposite ways.
Carbon Granules Convert Pressure Into Resistance And Borrow Their Power
Loose carbon granules touch at few points, so resistance is high. Press them and the contacts multiply and flatten, so resistance falls. The diaphragm rests lightly on the pack, the pack sits in series with a battery and the line, and the resistance swing becomes a current swing. The energy in that current comes from the battery. The microphone is a valve on someone else’s power, which is why Electromagnetic Microphone contrasts the two: a carbon transmitter on a 48 V circuit swings the current by about ±5 to 20 mA, a signal power of milliwatts, whereas an electromagnetic one generates about 1 to 10 mV, microwatts.
The Signal Is A Small Fraction Of The Loop, So Loop Resistance Decides It
Signal current is the change in current caused by a change in the microphone’s resistance. For a battery V in series with a total loop resistance R, I = V / R, and a small change ΔR in the loop gives:
ΔI = V × ΔR / R²
Two loops with the same assumed granule swing of ΔR = 10 Ω (an illustrative value, not a measured one):
Salvage loop on a 3 V battery. Microphone 100 Ω (inside the 30 to 300 Ω that Carbon Microphone gives), receiver coil 1,000 Ω (inside its 400 to 2,000 Ω), and 2 km of 0.5 mm wire at 180 Ω per km of loop (from Wire Selection for Telephony) = 360 Ω.
- R = 100 + 1,000 + 360 = 1,460 Ω
- I = 3 V / 1,460 Ω = 2.05 mA
- ΔI = 3 × 10 / 1,460² = 30 / 2,131,600 = 1.41 × 10⁻⁵ A, about 14 µA
Exchange loop on 48 V. Using the 1,000 Ω loop from Battery and Powering:
- I = 48 V / 1,000 Ω = 48 mA
- ΔI = 48 × 10 / 1,000² = 4.8 × 10⁻⁴ A, about 480 µA
The ratio is 480 / 14.1 ≈ 34. That is the engineering reason for central-battery exchanges: the microphone is the same, but each ohm of granule swing moves far more current. It assumes the same ΔR at both bias currents, which a real granule pack will not do exactly. Treat 34 as the size of the effect, not a prediction.
What You Need
- Granules from the crushed centre rod of a zinc-carbon cell (not alkaline), hard charcoal or crushed pencil graphite; sieve out dust and oversize pieces
- A diaphragm, thin metal 0.1 to 0.3 mm or mica, clamped at the rim only
- A chamber with a back electrode, filled to the depth below, and a milliammeter with a 1.5 V cell
- The full construction is in Carbon Microphone; this article is about why the settings matter
Method: Setting And Testing A Carbon Transmitter
Step 1: Choose the granule size by testing, because the corpus disagrees. The carrier article says 0.5 to 1.5 mm; Carbon Microphone says 0.1 to 0.5 mm; fill depth is quoted as about two-thirds, 60 to 70% and 80% in different places. Sieve two batches and build one chamber with each, then compare on the same bench.
Step 2: Measure the rest current. With a 1.5 V cell and a milliammeter in series, a filled chamber should read about 5 to 50 mA (30 to 300 Ω, as 1.5 V / 300 Ω = 5 mA and 1.5 V / 30 Ω = 50 mA). Far outside that range means a bad contact or a packed or bridged pack.
Step 3: Tap, then speak. The needle should flicker on a tap and follow speech smoothly. A steady current that does not move means the diaphragm is stuck or the granules are bridged.
Step 4: Loosen the pack when it ages. Granules compact with use and sensitivity falls. Shake the dry granules in a closed container and reassemble.
Common Mistakes
- Damp granules. Water films bridge them and the pack stops responding. Dry them completely and keep them sealed.
- Preloading too hard. A diaphragm pressing hard on the pack lowers resistance and sensitivity together; the swing needs room to compress.
- Ignoring the loop. A good microphone on a high-resistance line still gives a tiny ΔI, as the 3 V example shows.
- Judging a salvaged coil microphone by current. It produces a voltage, so measure millivolts, not milliamps.
Safety
Dust, heat and cells
Crushing carbon makes fine dust that should not be breathed; wear a dust mask and work outdoors. Roasting charcoal granules makes smoke and carbon monoxide, so do it outdoors in an open metal container, never indoors. Open dry cells outdoors with gloves, since the paste inside irritates skin and eyes. The test circuit is 1.5 to 9 V, but a microphone later connected to a central battery carries 48 V and up to the 100 mA the carbon-microphone range allows.
What’s Next
- Learn next: Speaker, the same transduction run backwards
- Combine with: Diaphragm Design for the resonance of the plate you are pressing on
- Combine with: Impedance Matching for coupling a low-voltage electromagnetic microphone to a 600 Ω line