I Think the next bit I'll talk about is the basic operation of a condenser microphone. Let's do it. m Um, let's just say we've got a fixed electrode here. M And we've got a flappy electrode here.

and just for the hell of it, what we're going to do is we're going to ground that one. Yep, and we're going to polarize that one with a polarizing voltage. We'll call it positive for the moment and we'll feed that onto there via a high value resistor. Mhm.

How high? How high? Really high? Hundreds of Megs Uh, quite often GS GS Yep, Reason being the capacit between that Flappy electrode and that one, there is of the order well for for a little electric microphone. And obviously this is not an electric microphone because we're externally polarizing it. Yes, it might be about five puff, right? Maybe seven puff, maybe four puff. That order of magnitude.

For a larger studio microphone, it might be between 50 puff and 100 puff mhm. low capacitances. Y And we need a high resistance here for reasons to be explained in a second. All right.

What we've done here is let's say that's about 100 volts. Uh, In a measurement microphone such as a Br and Cur microphone. Uh, which is externally polarized, they would typically use a 200 volt polarizing voltage, the Uh Studio condenser capsule which we had a little look at. uh, you might polarize it it with somewhere around about the 60 Vol, maybe even 90 volt.

Mark Okay, the amount of charge varies, but the whole point is we've got a capacitor here that we've put charge onto. It'll charge up over a period of some seconds because we've got gigs there and a few P of farads there, so it'll charge up take a couple of seconds after power on. Yeah, Uh, so somebody out there can calculate the time constant between. Well, no bugger it, Let's do it here.

Uh uh. let's say we got uh 10 9 Ohms there. Yep, and let's say we've got 50 * 10-2 farads there. Okay, that comes to about 50 * 10us 3 seconds I Think yeah.

Sounds about right. reasonable. So 50 milliseconds time constant? Uh, Incidentally, that's an Un that that. That would be an unrealistically low value, but it's some that old the microphone manufacturers use.

So we've got a 50 millisecond charge time constant there. mhm Uh, which incidentally will wind up corresponding to an electrical rolloff pole there of 20 HZ about 6 Herz. Okay, that's 50 milliseconds time constant. Yep, Uh 1.

Over 50 milliseconds is 20 HZ Mhm. and uh. well. actually the Uh with a 20 with a 20 Htz time constant, the actual minus 3db frequency will be uh, two, uh, 2 Pi.

Lower than that, which is about 3 Herz Got it? Yeah, fairly typical. Anyway, charges up. We got to charge there. What happens when we come along with a pressure wave on that diaphrag? It decreases.

the distance between the plates, increases the capacit. Yep. Now the thing is. charge on that capacitor has to be preserved.

There's nowhere for the charge to go in a hurry, so charge is preserved. If we've increase the capacitance mhm, the voltage has to fall to keep the same energy. And indeed, that's exactly what happens. If we get a pressure wave coming there, the voltage will go negative.

M If we have a rear fraction, the voltage will increase above the bias level. Yep, and those are the voltages that we're interested in gathering and amplifying. Is those AC voltages they they respond to? They correspond to audio. Yep, The other thing to keep in mind is a positive pressure wave here results in a decrease.

A negative going voltage? Something to keep in mind. Okay, uh, you can reverse that incidentally by putting a negative polarizing voltage there. Y In which case, if we have a negative polarizing voltage on the capsule, then a positive pressure wave will indeed, uh, respond to a positive going change in the output voltage. Yes, so that's one way of getting the polarity the right way around if you really want to.

Any other practical difference to the polarity. No, no, none, none, whatever. And uh, the other interesting thing is it doesn't matter whether you connect it as shown or right. let's put that over onto there and ground that still the same thing.

The signal on there because the capacitance is reducing will follow exactly that path. Yep, Okay, a positive going pressure wave results in a decrease in voltage. Negative pressure increase. Got it? It doesn't.

And what about noise? On our what about noise on our uh. bias voltage? Oh, we'll get into noise. Oh, we'll get into noise. Actually, the fascinating thing is high frequency noise coming out of here is low Pass filtered.

Filed? Yes, yes with a Uh. the response of that low pass filter is of course flat to three HZ And we've got a 3 HZ Corner Frequency Sweet 20 HZ Divid by 2 Pi Yeah, nice. Yeah. I Like it.

So we do, in fact, get a a noise reduction technique there. Having said that, any sane designer of polarizing suppliers for these microphones would make bloody sure that it was as Noise free as they could get. Absolutely So Okay, That's the basics of how you run an externally polarized microphone. Yes, And the difference with an electric microphone just to cover it again, is that you don't apply an external charge.

The charge is embedded in the material either on the wobbly bit or on the fixed electrode. It doesn't matter, usually in a bit of Teflon or something like that cuz they hold charge really well. Yes, same principles apply though. a incoming positive pressure wave increases the capacitance and causes a decrease in voltage.

Excellent. All right. What do we do with them? Well, as we saw, we need to load these with a really, really high impedance buffer. Mm if you don't have a hugely High impedance, uh, you're not going to get much out of them.

And where do you get a high impedance from? Well, you can either use actually, there's there's two classes of components you can use. You can either use Jfets yep, or Jfets with pilot lights, right? J Fet as in tubes, Tubes Tubes Tubes tubes. Yanks All right. Incidentally, if we get old school about this, let let's uh, look at an externally polarized capsule studio microphone type stuff.

Okay, so we're looking at one of those without the electric material. Okay, we're feeding that beasty from a a polarizing Supply and a really, really high impedance resistor. Probably about uh. one gig.

Yep, and that's going to be grounded. Okay, that's where our audio is occurring. but we need to. Why does it nonse? Why does it have to be such high voltage? Uh, if it, uh, the the amount of charge Mhm there determines the sensitivity of the microphone.

Got it? But you can't make it arbitarily high and arbitarily? No, indeed, you get some problems. All right, the higher you make that voltage. A couple of things happen. First of all, because that's positive and that's Negative they attract So yeah, oops, yeah, you'll get that diaphragm trying to bow in pop.

Yeah, well. actually, you get two different phenomena, right? phenomena. Number one is that gets so close to that that it decides to Arc over. Yep.

Now let's face it. when you when it ARS over, you get a massive transient there, whack in the in whatever system you're listening to not not pretty and then all of a sudden of course once that's discharged, it kind of Springs back to where it was before and sucks in and whack it oscillates a couple HZ relaxation oscillator. Yeah, brilliant. The incidentally, that's about the nastiest thing that can happen.

Much more benign scenario is uh, when that gets attracted so far it gets so close and you still got that same voltage there that it actually de decides to just go stick and pretty much right across the surface of that diaphragm. Yep, it sticks and it won't unstick until you remove the polarizing voltage by turning off the microphone right. And of course, when it sticks, it can't vibrate. so the microphone just goes deaf.

Yep, so doesn't Does it physically damage them though? Uh, generally. generally. no. The reason being I've grossly exaggerated the ratio of that diameter to that distance.

We're looking at typical uh diaphragm to place distances of maybe 10 microns, 10 micrometers so it's really close. That defamation, uh, doesn't hurt that much, Got it? Or it shouldn't Okay, so that's a couple of the perils. However, ever prior to that occurring, you can go for your life and you can turn up the bias voltage M because the the, how do you put it? the sensitivity of the system in uh pressure to voltage is linearly dependent on that bias voltage. Yep, so turn it up until it Sparks or cracks right there you go.

Actually turn it up until it Sparks or cracks at the highest SPL You're going to be using it at yes because that will add to some movement. So is that typically an adjustable Um voltage that you could tweak in your system? Uh, or you generally wouldn't touch it, You would generally not touch it. Uh, if you were clever enough to dive into a microphone and start McKing about with the Bi voltage, Yes, you could tweak it. It's uh, generally not that adjustable though.

because uh, well. back in the good old days, that voltage there was simply derived from the same power supply that was used to drive your tubes. Got it And decreasing it would be easy. Just use pot.

Increasing it. well. you'd have to have the voltage there to increase it in the first place. Yeah, uh.

In later times, uh, various designs involving Uh oscillators and step up Transformers were involved. Uh, the method that I tended to use Iz at Road microphones was a Seos oscillator. Really, really dumb. Uh yeah.

Seos oscillator in a US usual resistor resistor capacitor type. Arrangement Yep, Just so that you got a fairly well determined frequency of oscillation, whack that through another couple of stages of that, and then feed that into a modified variant of a Uh voltage multiplyer. Yes, and the thing is, you can actually go two phase on there. if you you can, you can be sneaky bu it.

Yep, because we've got alternate phases on that output and that that output, uh, some them might that keep on going. and I typically use a number of stages of multiplication there to step, say 12 or 15 DC Supply to these 4,000 series Coss mhm up to to 60 Vols 90 Vols whatever. And how would you clean that up? Oh, by the time you come out here, it's remember what's what's the actual current consumption out of there nothing. It's zero, Zero zero of course.

So in principle, it requires absolutely no clean up at all. But uh, let's face it. uh. one.

Meg of resistance and a couple of nanofarad capacit Bob's your uncle. All that's required. Yep, so it's intrinsic. In fact, the only real source of noise over there would be any, uh, relatively low frequency variation to the DC Supply there because that will cause that to wander and that represents no.

mainly be temperature dependent or something like that. Uh, we zenas. Bloody zenas. Quite often this will be a Xena derived Supply right? Okay, from a high voltage, you crude bastards.

Yeah, absolutely Well you? you? you wouldn't Chuck a regulator in there because actually Regulators are probably going to be noisier than Xena Cool, Yeah. And the issue here is if you've got a Zena with a soft knee, it's actually going to be quieter than a hard knee one. And by knee I'm talking about if we're talking about uh I can't remember. Let's call that uh, current versus voltage? uh uh.

soft knee xener would look a bit like that M So that's your regulated voltage there. y a hard knee Xena would be much more likely to be like that. The hard knee low impedance one seem to be way noisier than the soft knee ones. Well, that's what we found at the time.

Right after we got some, uh, hard knee, Zenas inadvertently put them in and discover that they are noisy, noisy, noisy brute. There you go. They didn't make them in production, did they? Uh, actually they made it in. We did about 100 and then kind of realized, hang on, these aren't passing muster on noise because we were testing for noise.

What the hell's going on? Why are these different to the last ones and got it down to the zenas. We still because the hard knee zeners have a low impedance MH than lower impedance than the soft knee ones. It doesn't matter how much capacitance you put across them, you're You're not really going to shut them up because you've got a low imp in that capacitor and it just doesn't clean up the soft knee ones. clean up beautifully.

Mhm, Go figure you go. So that was simply a lesson learned along the way in low noise. Electronics Uh, if you're going to Xena regulate, pick something that's got a fairly soft knee. And the other thing is, if you're going to filter your Zena don't filter it there.

Yes, duh, that's a low impedance Point add some impedance and decouple it over here. y with can actually form an RC absolutely low pass f d Nois it. Yeah, cuz that Seos uh Su Saad is going to take bugger all anyway. yeah, that's right.

So yeah, so you can use kilom. Yeah I know you can use couple of K it just sits there and boogies. Excellent! So that was a typical circuit inside a A road production y M and still used to this day at the best of my knowledge. When I first came there though, they were not based on this kind of Seos oscillator.

They were a transistor oscillator with three discrete inductors hung around it and basically relied on getting a 12 or 15 volt. Supply and look, I actually can't remember the circuit topology for this oscillator, but it was a real nasty thing and not at all deterministic. There you go, and it required on the particular transistor having a fairly specific gain and VC and so they production is going to be all over the shop well and what varies as a result. mhm the sensitivity of the microphone.

The one thing you don't want to vary is the sensitivity of the mic. Oh yeah, oh boy. All righty. so we're back to this microphone circuit where we've created some audio superimposed on DC there, right? What are we going to do with that? Of course we're going to capacitor couple it out to our amplifier MH which also needs to have its bias condition set by a resistor going to whatever bias voltage we want.

Yep, that might be ground, might be a few volts. Depends whatever depends on what. Supply you got here for your amp and whatnot. Yeah, things to remember: that capacitor has to be, you know, way larger than that capacitance there.

Mhm talking a couple hundred Nan few few. Given that that's say, around about 50 puff y a nanar do just fine, right? And typically 1 to 10 nanofarads is what you'd see there. Yep, what value does that need to be? Well, at least the same order of magnitude as that, because as far as audio there is concerned that thing and that thing appear in parallel. they do.

and uh, back in the good old days of valve microphones and not a lot of knowledge, Uh, typical resistances? there might have only been about 100 Meg Ohms. Yep, Okay, that order of magnitude, the amplifier here might have been a tube. Yep. and that would go to supply output.

Transformer And that goes off to the rest of the system. And in fact, the very first microphone that I designed at Road microphones was a tube microphone. It actually had a had a tube in there. Y J J fit with a pilot light, right? Yeah, Stuck in the end of the yeah incident, there are real advantages in in having a valve a tube up in the head with the microphone.

Why is that? why? Because you're providing it with filament power. and it's the filament power that makes that thing warm. Yep, and it's that warmth that keeps the whole thing dehumidified. Ah, you're dealing with high impedances here so you really want to keep it.

uh, dry, not humid. you want to discourage. Interesting side benefit. And indeed, when I think that this might be an anecdote.

Going back to maybe the 60s or 70s when Brw and K decided to go solid. St instead of valve they discover that they were having all kinds of problems in their measurement marks with humidity buildup. So they because they' taken out the tube so they actually had to put heater resistors in there. they put a resistive heater.

Nice. Oh terrific. One of the uh, ugly things about this kind of circuit uh is the fact that the output impedance of valves is really quite high. That means that whatever load you put over here here you know, whatever your mixer console or what whatever it is that you're using over there, its load impedance will change the gain of the whole thing.

And if you've got a load over here that's nonlinear with frequency for example, if it's you know, heavily capacitive or something like that, your frequency response is going to droop because it's being fed by quite a significant impedance in this first mark that I designed at Road the classic 2 I heretical. I was a bad bad boy. First of all, do that, then do that. then capacity.

couple. add into your Transformer and guess what? You get a nice low anded strive point? You do cuz it's an amid follower. Yeah, and it. It doesn't do anything to interfere with the tubish linearities or nonlinearities of that.

It's just a dumb voltage follower. Did you get the tube Sound? Oh yes, yes. Excellent. Excellent.

Interestingly enough, on this particular microphone, part of the what I reckon everybody called the tube sound was created by the fact that in be instead of being a skinny little run of a microphone like this damn thing was about that kind of diameter. It was hugely meaty and had this quite large uh, mesh covering over the capsule. so you you were never up close and personal to capsule even if you were a kuner. Uh, you were always kind of at least the the radius of this shell away from it and it made it sound all Spacey and Airy And yeah, and that was just the physical construction of the mic forcing people not to.

Yeah, they couldn't get their gobs up to it so it sounded more better. Brilliant. Incidentally, that follower there y also allowed me to. put some filter circuitry in there so we could either get it uh, flat as possible or bit low pass filtering or a bit more.

uh sorry High Pass filtering Swit Swi High s but uh yeah, the fact that it had a bit of bipolar in there was heretical. Yes, absolutely. Oh, Unbelievable. Incidentally, uh, transistor.

just I think it was a a 2541 or one of those 120 volt 150 volt class transistors because the supply was only about 120 to 150 volts. Yep, so we could get away with u a fairly High Gain fairly low noise transistor there. and yeah, it just worked. a treat.

Brilliant. Um.