Which flavour of phantom power? P12V, P24V or P48V (the commonest)?
All of them - I have no idea of the design process that starts with 48V or whatever and ends up with 6.81K - that's what I'm interested in. In general, if we have V volts phantom power, how do we get to the correct R ohms for those resistors?
You've got two choices here:
- Take the standard values as holy writ and just use them.
- Go through the whole process of figuring out how the standard values came to be chosen and end up using them anyway.
I suspect that you want to do the latter, which is fine, but it going to take some time*. There are two aspects to this, one is pure engineering and the other is the rather arbitrary process of how standards (that ipso facto have to serve many different people's requirements) get made.
We can go some way to starting you on the engineering side of that. I'm not gonna touch the standards side with a barge pole.
First off, let's go back in time to when phantom power was first used. Any microphone that might require phantom power was
guaranteed to have an audio output transformer - this was well before the point in time where a solid state differential output amplifier with adequate performance would have been practicable or affordable. There would have been a corresponding transformer at the amplifier end.
So we end up with a system that looks like this:
Microphone <=> transformer <=> cable <=> transformer <=> amplifier.
So why do we have a transformer there? It provides:
- Galvanic isolation
- Differential signals
- Floating signals
- Impedance transformation
Why are these desirable?
- Galvanic isolation - protects the microphone and amplifier (and their users) from supply voltage (mains and DC) that might be introduced under fault conditions.
- Differential signals - reject common mode noise picked up by cables e.g. mains hum, Radio 2 (i.e. RF interference, for some reason in the UK it always seemed to be the national Radio 2 LW service that broke through unless there was a taxicab company transmitter nearby)
- Floating signals - by dint of the differential signals being referenced off each other they need no ground reference. This mitigates problems from ground loops, especially on long cable runs when they are more likely to occur.
- Impedance transformation - permits a standard impedance to be presented to the microphone by the cable and by the cable to the amplifier. There's a whole bunch of other reasons, all the ones associated with impedance matching in general, but interoperability is the most important factor in this application.
Phantom power rides on the back of the transformers. One winding of the transformer will be connected to the two signal wires in cable, the other will be connected to the microphone (or amplifier). One signal wire is known as 'hot', the other as 'cold'; there is also an overall electrostatic screen in the cable wrapped around the outside, containing both hot and cold wires. The positive side of the phantom power supply is connected to a centre tap on the transformer winding, the negative side of the phantom power supply is connected to the electrostatic screen. The current from the phantom power supply rides on both the hot and cold wires as an identical positive DC offset on both of them. On the other side of the transformers this DC offset cannot be seen, so it is as if we have created an invisible ghost wire to carry the power, thus
phantom power.
Here's what it looks like as a schematic:
Image from Eargle's "The Microphone Book" p147In this setup a DC current is going to flow through the transformer windings. This is important because in general transformers don't like DC, it pushes the transformer core towards the point where it becomes magnetically saturated. When the transformer core is saturated it will distort the signal it is trying to transform, or in the worst case just stop acting as a transformer and pass no signal at all.
So the
principal requirement for resistors in series with a phantom power supply is to limit the DC current through the transformer windings. Here's where we get into engineering trade-offs. Ideally we want the DC current through the transformers to be as little as possible, the greater it is the physically larger and more expensive the transformer will have to be. On the other hand, we want the supply current available to a microphone's built-in preamplifier to be as high as possible, the more current available the better we can make this preamplifiers signal-to-noise ratio. Two opposite requirements.
So we have to pick a compromise maximum current. Arguing through the trade-offs here probably wouldn't shed too much more light, so let's skip to the compromise settled on which is 10mA maximum for a 48V supply. Calculating resistors to provide this current limit is, of course, just Ohm's law. For an output short circuit, that would be 48V/10mA = 4.8k
. There are two legs to this, so two resistors in parallel, which would each have to be twice that value = 9.6k
.
But of course we're not intending to drive a short circuit, we're intending to drive a preamplifier, which is going to have a minimum supply voltage requirement. What you choose that to be is somewhat arbitrary - you need enough for the maximum signal you want to push down the wire, which will be relatively low, a volt or two maximum, plus biasing overheads and also you need enough to generate a polarizing voltage for a condenser capsule via some step up arrangement. Compromises again, discussion skipped again, the value picked for a 48V phantom supply was 14V to be left available to the preamplifier. With a 48V supply and a 14V load at the preamplifier we have to drop 34V in our current limiting resistors. Again, there are two in parallel so each will have to carry 5mA. We go and see Herr Ohm again and get 34V/5mA = 6800
. The nearest standard (E96) series value is 6.81k
. Job done, rinse and repeat for P24 and P12.
* It did. Took me about an hour to write up, so you can see why people give terse answers on here.