Frequently Asked Question
There are a variety of professional lavaliere microphones that are well liked. Every user has their own opinion about which ones are the best for different situations. The problem is that these mics are radically different in their output levels, bias currents and in some cases the voltages that they will tolerate. In addition some are wired as three wire microphones (bias + audio+ ground) but others are two wire microphones with bias and audio on one lead plus a ground lead. Variations in output levels from different manufacturers can be more than 30 dB and bias currents can range from 20 uA to 800 uA. In the movie industry, the mic may be required to pick up a whisper in one scene and a scream in the next. It is no wonder that microphone and transmitter design is always a series of compromises. The input to the SM transmitter tries to overcome these compromises.
The bias voltage in the SM input is set by a servo loop that regulates the DC voltage at the microphone to a user selectable choice of 2 or 4 Volts. This is in contrast to the typical 5 Volts plus series resistor bias circuit that can result in a mic voltage that can vary from 1 Volt to almost 5 Volts. The lower voltage range can result in reduced headroom and the higher voltage can result in internal Zenering (overload) in some microphones. The SM input can handle mic bias loads from 1uA to 2000uA while still maintaining full bias voltage regulation. The servo loop also incorporates a filter that causes it to servo out frequencies below 20 Hz and rolls off the response of the lavaliere itself to wind noise, thumps and breath pops. These low frequency excursions are stopped right at the mic FET and then do not overload early audio stages in the transmitter.
At audio frequencies, the servo bias looks like an extremely high impedance resistor (constant current source) so that none of the output of the microphone is wasted in a 1k to 4k bias resistor. To prevent large voltage swings, the input to the first amplifier is a virtual ground input. This input is very low impedance so that the current developed by the mic FET is used entirely to drive the virtual ground input. Since the virtual ground input sees a high impedance source made of the mic FET's drain and the servo bias, the virtual ground input has very little loop gain noise. Since the mic's FET is operating into a virtual ground, there is very little voltage swing on the FET drain which reduces distortion on the FET compared to a conventional input.
The new input has the advantages of low noise since the noise is determined by the noise of the mic's FET and not by a bias resistor. It has the advantage of a well defined bias voltage that is not dependent on a compromise choice of transmitter bias resistors and mic current drain, i.e., two different manufacturers trying to guess what the other one is going to do. The input also has the advantage of very low voltage modulation on the FET drain reducing distortion. Finally, the input does not run out of voltage or current headroom since the bias voltage is well defined, DC current is supplied by the servo loop and AC current is "supplied" by the virtual ground amplifier. At minimum gain, the input will handle 240 uA of peak input current without engaging the limiter.
The most important advantage has to do with the limiter circuit that we have in all our transmitters since we can make it work better in the SM. Our standard limiter is a shunt circuit that shunts excess audio signal to ground when input levels are too high. In the past we have had to buffer this low impedance limiter circuit from the relatively high impedance input circuit for the mic bias supply. The amplifier that we had to have between the mic input and the shunt limiter was subject to overload at high input levels. Generally, the lavaliere mic overloaded before the buffer amp but not in all cases. Some high current mics could overload the buffer. The buffer amp also had to have unity gain so its output didn't overload and this meant this low gain amp added at least 3 dB of noise. With the new input circuit, the shunt limiter can be right at the input. No buffer amplifier is needed. This is because the virtual ground input circuit is very low impedance and is just what the shunt limiter is looking for. The advantage is that the limiter range is at least 30 dB no matter what the transmitter gain setting or input level from the lavaliere mic. There is no other transmitter that has anywhere near this limiting range for high input levels.
Some careful design went into this circuit and it is compatible with almost all of our previous mic wiring recommendations including line level inputs. Some microphones can benefit from a slightly different wiring scheme and that is noted in the SM manual. Old wiring, new wiring and compatible wiring is listed. About the only thing that doesn't work is the 40 dB attenuator wiring for very high signal level line level inputs. This can still be accomplished by putting a single 25k resistor in series with pin 5 of the TA5F input connector.