• Unbalanced or balanced mono input (3.5mm mono/stereo socket)
• Unbalanced mono output (3.5mm mono socket)
• Very low distortion and noise
• Small and easy to build
• Runs off a 5-20V DC plugpack or battery
• Adjustable gain over a wide range
• Line level output to at least 1.5V RMS
• Provision for electret microphone bias (approx. 390µA)
Supply voltage: 5-20V DC (operates at 2.8-5V with reduced performance)
Supply current: typically below 6mA
Voltage gain: 3-111
Input sensitivity (line level output): 14mV RMS
Input sensitivity (1V RMS output): 18mV RMS
Input impedance: 50kΩ (8.3kΩ with bias enabled)
THD+N ratio: 0.0035%
THD+N ratio (10mV RMS in): 0.014%
Signal-to-noise ratio: -90dB (-93dB A-weighted)
CMRR* (1% resistors): -55dB
CMRR* (0.1% resistors): -88dB
Frequency response: 20Hz-20kHz ±0.01dB
Signal handling: >1.5V RMS output
Signal handling (3.0V supply): >1.0V RMS output
Note 1: CMRR = Common Mode Rejection Ratio
Note 2: all specifications relative to 50mV RMS input, 775mV RMS output, 20Hz-22kHz bandwidth and a 6V supply, unless otherwise stated.
The reason that a microphone preamplifier is necessary is that most microphones, especially unpowered types, have a low output signal level. A typical microphone will deliver 10-200mV RMS at maximum volume. Audio “line level” is around 775mV RMS (0dBu) or higher but a great deal of audio equipment can actually handle 1V RMS or more. Higher signal levels usually mean more dynamic range.
So to interface a microphone to a mixer, computer sound card, amplifier etc, we need to insert a preamplifier in-between to boost the signal level. Otherwise it may be impossible to get enough volume.
Some such devices contain internal amplifiers but they don’t always perform well. Their internal microphone preamplifiers can be noisy and may not provide enough gain for some microphones (ie, those with very low output levels). Many, if not most, computer sound cards do not use high-quality analog components.
Adding a microphone preamplifier does not guarantee good results, as the line level circuitry can still introduce noise and distortion but it certainly improves your chances of getting acceptable sound quality. On the other hand, a preamplifier is a necessity for connecting a microphone to any gear which only has line level inputs.
Fig.1: this graph plots the frequency response of the Mini Microphone Preamplifier. Note that the vertical scale is greatly magnified, as the frequency variation is within just ±0.01dB. This figure is at the limit of our Audio Precision System One test gear's resolution – the response is about as flat as it gets.
Fig.2: this graph shows the total harmonic distortion (THD) with respect to frequency. Distortion levels are higher than quoted because this is measured over a wider bandwidth (10Hz-80kHz), so more noise is registered.
The slight drop at high frequencies is due to the 80kHz cut-off. Again this is essentially a flat measurement.
As can be seen from the specifications and graphs, this preamplifier has very good performance despite its low supply requirements. Signals below 50mV RMS will result in worse performance while higher level signals will provide better performance. For a 25mV RMS input, the signal-to-noise ratio will be reduced by 6dB, for 12.5mV by 12dB and so on. With a 100mV RMS input, the S/N ratio goes up to 94dB and THD+N improves to below 0.002%.
The performance doesn’t vary with signal frequency. The frequency response is very flat with -3dB points at around 1Hz and 1MHz (see Fig.1). The total harmonic distortion plus noise (THD+N) level is the same across the audible band (see Fig.2) and at typical microphone levels consists mostly of noise.
Under our test conditions with 50mV RMS input and 775mV RMS output, harmonic distortion accounts for just 12% of the total distortion measurement and is primarily second harmonic.
Such a wide frequency response is not necessary but is the result of making this project as small and simple as possible. There is no low-pass filter except for the internal compensation of the op amps. We are assuming that most devices which accept line level signals will have their own bandpass filters to remove frequencies outside the audio spectrum.
The Common Mode Rejection Ratio (CMRR) is a measure of how well a device with a balanced or differential input is able to reject a signal that is common to both inputs. In other words, if the same amount of 50/100Hz hum is coupled into both signal conductors in the cable, this is the amount by which that hum is attenuated.
For our first prototype, we used standard 1% resistors throughout and measured a CMRR of -55dB. Our second prototype used more expensive 0.1% resistors in the differential amplifier which improved the CMRR to -88dB. In practice, -55dB is perfectly adequate unless you have a very long microphone cable run.
To get the best performance, either the power supply ground or signal ground should to be connected to earth. This reduces the possibility of mains 50/100Hz hum entering the circuit. However, you should avoid earthing both so that an earth loop cannot be created.