This design runs from a 5Volt USB supply and attempts to minimise electrical noise.
I wanted to create a simple preamp for an electret mic insert using components from my bits box.
So component values and transistor types are not critical.
My immediate application for this preamp is to connect a cheap electret mic insert to a Raspberry Pi via a sound card, for use in recording bat calls.
Currently I have two bat call recording systems; the latest uses a specialised USB microphone, but the earlier system has evolved over a few years from a hand-held, battery powered, frequency division bat monitor, based upon an old design by Chris Eve. I would now like to rationalize this earlier system by replacing the analogue module with a simple amplifier powered from the same 5Volt supply powering the Raspberry Pi.
I may also want to run this system from a 5Volt battery 'power-bank', and mount the mic & amp remotely from the Pi, maybe at the top of a pole.
The clouds show diagnostic voltages across certain resistors.
The 5Volt supply is filtered by the two 100uF + 220 Ohm C/R networks. I also tried adding 0.1uF capacitors across the 100uF capacitors, and fitting a 100uF directly across the incoming 5Volt supply. But none of these options appeared to reduce noise any further.
The 10k resistor supplies approx 0.3mA to the electret mic's internal FET amplifier, developing approx 1.2Volts across the mic terminals. The datasheet for the Panasonic WM-61A microphone (the only microphone I have which is identifiable) gives 2Volts as the rated operating voltage and a current of 0.5mA. So the feed resistor should be 6.8k. I couldn't find one of those, but I have loads of 10k. Anyway, there didn't seem to be a huge difference in gain with resistors between the values of 4.7k - 10k.
The first transistor provides signal voltage gain, while the second provides current gain. This second stage (configured as an emitter-follower) creates a low-impedance output, allowing the signal to be connected to the Wolfson audio card on my Raspberry Pi via 2 or 3 metres of cable, without picking up unwanted noise or suffering attenuation at higher frequencies. So this second stage is surprisingly important for ensuring a good signal to noise ratio.
The value of the coupling capacitors have been selected so that frequencies below about 10kHz are attenuated. If you wanted to use this as an audio preamp, I'd suggest increasing the value of these capacitors by a factor of 100 (e.g. 2.2nF >> 220nF, 1nF >> 100nF & 10nF >> 1uF).
The choice of transistor was purely down to availability; I found a small packet containing 25 BC182L transistors inside a 'Man Tin'. Any small signal NPN transistor with a reasonable gain will probably be fine. (In fact I found all the components for this prototype preamp in various tins, cardboard boxes and other plastic containers).
The next step is to take my bread-board prototype and build it into a very small box, which may also house the electret mic. This will then be connected via 2 or 3 metres of audio co-ax to the Wolfson/Pi system, and tested alongside the Ultramic USB system.
Post a Comment