pstamler
Well-known member
[quote author="skipwave"]Is there really any value in using an input trafo with a 150 ohm primary? I don't have any mikes that require this low load impedance, and most recommend a 600-1k load.[/quote]
It doesn't really have a 150 ohm input impedance. In some ways, the standard nomenclature for transformers is confusing, and is read sort of backwards, or right-to-left. Think of a "150:600" transformer as one that transforms a source impedance of 150 ohms (the microphone) into a source impedance of 600 ohms as seen by the opamp. It has a 1:2 turns ratio, so it steps up the source impedance by a factor of 2^2, or 4.
If you look at it the other way, it presents a load to the microphone that equals the loading resistor (6.19k for this tranny) divided by the turns ratio squared, or 6.19k/4, or about 1.55k, a good load for many mics.
Yes. (Here's an oversimplification) A circuit with a bipolar input has a source impedance at which it will add the minimum amount of noise. Anything less or more will be noisier than optimum. That source impedance is en/in, where en = the input voltage noise in V/(sqrt(Hz)) and in = the input current noise in A/(sqrt(Hz)). For an LT1115, for example, that would be (mumble, mumble) 0.9nV/rtHz / 1.2pA/rtHz, or 750 ohms. So the ideal transformer for that chip would be a 150:750 one. Lower or higher ratios would have more noise. 150:600 isn't a perfect match, but it's close enough.
Switch to a 600:600 tranny, and you'll really be showing the opamp a 150 ohm source impedance (a 1:1 reflection of the microphone), which will be noisier than Jung's design.
With FET inputs or tubes, it's simpler: the higher the source impedance, the less noise the FET or tube circuit will contribute. (More oversimplification.) So you want a nice high-ratio transformer. Unfortunately, "nice" and "high-ratio" don't usually go together, so one has to compromise.
Peace,
Paul
It doesn't really have a 150 ohm input impedance. In some ways, the standard nomenclature for transformers is confusing, and is read sort of backwards, or right-to-left. Think of a "150:600" transformer as one that transforms a source impedance of 150 ohms (the microphone) into a source impedance of 600 ohms as seen by the opamp. It has a 1:2 turns ratio, so it steps up the source impedance by a factor of 2^2, or 4.
If you look at it the other way, it presents a load to the microphone that equals the loading resistor (6.19k for this tranny) divided by the turns ratio squared, or 6.19k/4, or about 1.55k, a good load for many mics.
Yes. (Here's an oversimplification) A circuit with a bipolar input has a source impedance at which it will add the minimum amount of noise. Anything less or more will be noisier than optimum. That source impedance is en/in, where en = the input voltage noise in V/(sqrt(Hz)) and in = the input current noise in A/(sqrt(Hz)). For an LT1115, for example, that would be (mumble, mumble) 0.9nV/rtHz / 1.2pA/rtHz, or 750 ohms. So the ideal transformer for that chip would be a 150:750 one. Lower or higher ratios would have more noise. 150:600 isn't a perfect match, but it's close enough.
Switch to a 600:600 tranny, and you'll really be showing the opamp a 150 ohm source impedance (a 1:1 reflection of the microphone), which will be noisier than Jung's design.
With FET inputs or tubes, it's simpler: the higher the source impedance, the less noise the FET or tube circuit will contribute. (More oversimplification.) So you want a nice high-ratio transformer. Unfortunately, "nice" and "high-ratio" don't usually go together, so one has to compromise.
Peace,
Paul
