U67 de-emphasis network

[Matt Nolan]

How efficient is a condenser capsule at energy conversion, in either direction?

My understanding is that condenser capsules put out a very low power signal, lower than ribbons', which are lower than typical dynamics; that's why condensers need some kind of head amp (in the mic) to put out enough current to drive a preamp.

That suggests to me that they're a lot less efficient than dynamic mic capsules, however efficient those are, so not very efficient.

That in turns suggests to me that the feedback-into-the-mic-connection trick is doing a lot less to oppose the diaphragm's motion than to electrically cancel its LF electrical output at the input of the head amp.

I may be speaking out of turn here. This is speculation and not really knowledge. To consider how efficient something is, you have to find where energy is lost. The first things that come to mind are accelerating the mass of the diaphragm and doing elastic work on the diaphragm and on the air behind it. Are there any other loss mechanisms?

 

[Matt Nolan]

Do you have any reason to think a capsule is not minimum phase? If a capsule is minimum phase, then the transient response can be directly derived from the frequency response.

Well, you already made some suggestions to answer your own question. I'll add a couple more from out of left field.

Coming from the world of drums, cymbals, and gongs (playing the former two, hand-crafting the latter two) I am aware that the harder you drive any kind of diaphragm / membrane the more non-linear it gets. Mic diaphragms are much more like a drum-head than a gong, and the excitation amplitude is much less high and the excitation is spread across the whole thing versus a drum, so it is bit tenuous, but if we are considering minutiae then maybe it suggests mechanisms for non-linearity at the extremes of performance - which is where all the significant differences are in audio equipment.

Gongs are a weird mix of idiophone and membranophone so even further away from a mic diaphragm. Their restoring force comes from stiffness, not tension, and this makes them very non-linear indeed. Once you get excitation amplitudes around the same order as the thickness of the material, things go very nonlinear very quickly. (bong turns to whoosh and frequency content evolves over time). While a condenser capsule diaphragm is almost entirely constrained by tension (ignoring the air behind for the moment) and thus ought to behave linearly, there has to be at least a tiny component which is constrained by stiffness. So there's another possible source of non-linear behaviour - and again, when it is being driven very hard.

In the physics of the real world, nothing mechanical or electrical is truly linear and equations do not completely model anything down to the smallest detail. They are just abstractions and approximations that work well enough, most of the time. I began to appreciate this more and more the more I read texts from Richard Feynman.

 

[GeorgeToledo]

Changing voltage to the capsule changes the amount of distortion the capsule generates.

As a side note to the discussion of stiffness/tension, in the world of capsules, this is what makes the pvc M7 different from mylar capsules. PVC has self rigidity. I believe the earliest styroflex CK12s (which apparently sucked) would have also had self rigidity, given styroflex that I have encountered.

 

[Paul W]

This is speculation and not really knowledge. To consider how efficient something is, you have to find where energy is lost.

The efficiency of a transducer is an end-to-end thing: how much energy comes out vs. how much energy went in. You don't need to know exactly where in between (or how) the energy was lost, although that would certainly be interesting.

Consider an SM57 and a condenser with a diaphragm the same size. If amount of acoustic energy hitting the diaphragm is the same, the electrical energy the capsule puts out will be very different. An SM57 puts out enough electrical power to satisfactorily drive a preamp, and a condenser capsule can't. The SM57 is therefore clearly more efficient. (In the forward direction.)

Given that the SM57's dynamic capsule can't be more than 100 percent efficient, the condenser capsule must be much less than 100 percent efficient.

Presumably in both cases, some of the lost acoustic energy is reflected, some passes through the capsule and into the mic body or back out into the air, and some is converted to heat, but however the energy is lost, it's not put out as electrical energy.

I don't know if it's a safe assumption that the capsule is necessarily a comparably efficient transducer in both directions, though. I wouldn't expect it to be more efficient in the inverse (electrical to mechanical) direction than the normal (mechanical to electrical) direction, but I really don't know.

 

[Paul W]

Again The the LF NFB connects to the capsule back plate and not directly to the amp so the capsule is part of the signal path.

Nobody disagrees with that. We all know that the (180 degree delayed) LF signal from the output goes back to the capsule.

there is no direct path between the LF NFB and the amp.

I don't understand that. The output of the capsule goes to the input of the amplifier, right? As I understand it, if you connect the LF feedback path to that point, you are connecting it to the input of the amp as well, since that's what the capsule connects to. They are the same electrical connection.

Here are the details for the 67 amp FR. The1500 should be 15k for the the HF roll off the lows are at 40HZ so dont extend that far up.

Regardless....
in simplest terms we have an AC voltage being fed in to a capsule if you don't think the capsule will wiggle from this. I don know...
It's a bit 2+2 =5

No, the problem is that you can't seem to understand that WE ALL AGREE that the capsule will act like a little electrostatic speaker, and that what we disagree about is the magnitude of that effect.

Just stop repeating that the LF feedback goes to the capsule, and that it will oppose the LF motion of the diaphragm. We know. We agree on that much.

THAT IS NOT WHAT THE ARGUMENT IS ABOUT.

 

[MagnetoSound]

The same technique is used in the U87. The resistor values in the NFB loop are quite low. I’d always assumed the reason the NFB loop is fed to the backplate is to avoid it acting as an electrical load on the grid circuit.

No doubt there will be some tiny interaction with the membrane, but it will be damping it, not exciting it, and with some time difference due to the mass of the diaphragm. This might result in what we all know as the U67 ‘softness’.

I seriously doubt that the design intention was to damp the membrane. As as I say, it seems much more likely it was to mitigate an impedance imbalance. The fact that it probably does interact with the membrane might be simply the serendipitous thing that gives this mic, and the U87, their character. Or I’m wrong, and Gerhardt knew all about it.

 

[OPR]

Just stop repeating that the LF feedback goes to the capsule, and that it will oppose the LF motion of the diaphragm. We know. We agree on that much.

THAT IS NOT WHAT THE ARGUMENT IS ABOUT.

ok I'll start by saying "that the LF feedback goes to the capsule, and that it will oppose the LF motion of the diaphragm" Ha ha!
And there's no need to yell! I can hear just fine Paul.

No, the problem is that you can't seem to understand that WE ALL AGREE that the capsule will act like a little electrostatic speaker, and that what we disagree about is the magnitude of that effect.

Do you "All"(Korg, Khron, Paul) also believe that this little electrostatic speaker effect that you agree causes mechanical movement of the diaphragm opposing the input signal, has zero measurable or audible effect on the sound besides purely passing though electrically?

If this is true, explain what brings you to that conclusion?... Besides IMHO...

 

[OneRoomStudio]

Consider an SM57 and a condenser with a diaphragm the same size. If amount of acoustic energy hitting the diaphragm is the same, the electrical energy the capsule puts out will be very different. An SM57 puts out enough electrical power to satisfactorily drive a preamp, and a condenser capsule can't. The SM57 is therefore clearly more efficient. (In the forward direction.)

Given that the SM57's dynamic capsule can't be more than 100 percent efficient, the condenser capsule must be much less than 100 percent efficient.

I think that assertion is frequency-dependent. At 25kHz, the condenser is likely more efficient. I don't think this is a simple mechanical equation. Frequency, impulse response, and other factors (i.e. polarizing voltage impacts output level, and therefore "efficiency") all change how efficient the capsule is. I think that's part of what OPR is getting at - the capsule itself is not a linear device, and therefore, the effects of negative feedback *through* the capsule will not be the same as the effect of purely electrical NFB.

As you mentioned though, this discussion is really about the magnitude of the effect. All I can say on that topic is that it can't be nothing. I currently have a U87 and a U67 built in the same bodies with the same capsules and similar output transformers, and they definitely sound different. There are other factors of course - tube vs SS, other circuit differences, etc...but the whole of the difference is the sum of it's parts.

 

[OPR]

No doubt there will be some tiny interaction with the membrane, but it will be damping it, not exciting it, and with some time difference due to the mass of the diaphragm. This might result in what we all know as the U67 ‘softness’.

I agree. And removing the S2 NFB also removes the softness that you speak of which you wouldn't expect from a high pass filter

Here's the old Marketing which talks about the anti pop properties of the circuit.

 

[ccaudle]

explain what brings you to that conclusion?

What is your analysis of the point I previously brought up about the feedback path being much lower impedance than the input impedance of the tube stage, so the entire diaphragm may move as common mode, with no differential voltage across the diaphragm?

 

[Paul W]

Do you "All"(Korg, Khron, Paul) also believe that this little electrostatic speaker effect that you agree causes mechanical movement of the diaphragm opposing the input signal, has zero measurable or audible effect on the sound besides purely passing though electrically?

No. None of us thinks that. All of us think that it will have a measurable but smaller effect (than the straightforward electrical cancellation), which may be audible.

AFAICT we all agree that the effect exists, and should be measurable, and may be audible.

We all think that there will be an electrical cancellation effect which is consderably more efficient and therefore much bigger and more audible than the diaphragm-motion-inhibiting effect.

We think that the high-pass filtering effect is mostly due to electrical cancellation of the signal being fed back from the output to the input, and some smaller effect is due to inhibiting the diaphragm motion with the electrical back pressure.

That raises interesting questions.

One question is how important the additional effect (of reducing diaphragm motion) is, as opposed to rolling off the bass in other ways, like with normal EQ.

Can you get much the same total effect just by using more electrical cancellation, with EQ? I don't know.

Are the different effects frequency-dependent in different ways, so that they sound different? I don't know.

What happens if the amp clips and feeds back the clipped signal (low passed and 180 degree delayed)? I don't know.

 

[OneRoomStudio]

Can you get much the same total effect just by using more electrical cancellation, with EQ?

This was basically the original question, and what I was trying to address. Because the level of cancellation through the capsule is impacted by dynamics to some degree, and may even have an element of resonance to it, I don't think you could get exactly the same effect with EQ alone. Could you get it close enough that it wouldn't be audible to the average person? Probably - but the average person probably thinks a Warm Audio and a Telefunken M251 sound the same too.

So how different is "different enough" to matter in this conversation? I think that's a matter of opinion, and what most of this thread has been rotating around.

 

[OPR]

No. None of us thinks that. All of us think that it will have a measurable but smaller effect (than the straightforward electrical cancellation), which may be audible.

AFAICT we all agree that the effect exists, and should be measurable, and may be audible.

We all think that there will be an electrical cancellation effect which is consderably more efficient and therefore much bigger and more audible than the diaphragm-motion-inhibiting effect.

We think that the high-pass filtering effect is mostly due to electrical cancellation of the signal being fed back from the output to the input, and some smaller effect is due to inhibiting the diaphragm motion with the electrical back pressure.

That raises interesting questions.

One question is how important the additional effect (of reducing diaphragm motion) is, as opposed to rolling off the bass in other ways, like with normal EQ.

Can you get the same total effect just by using more electrical cancellation, with EQ? I don't know.

Are the different effects frequency-dependent in different ways, so that they sound different? I don't know.

What happens if the amp clips and feeds back the clipped signal (low passed and 180 degree delayed)? I don't know.

Sounds like you all Kindá sort'a have you feet in both camps now yet at the same time kindá sortá minimising the effects of the NFB on the capsule as largly negligable... in your collective humble opinions of course...all sounds a bit vague really...

And still no expanation as to why you collectively think this is the case?

My initial response in this thread was to King Korgs assumption that there was absolutely no difference between the U67 negative feedback circuit and applying regular eq in post. which I disagree with. They may be similar but not exactly the same under every scenario close proximity vocal recording without a pop filter being a glaringly obvious one.

If you do belive the diaphram's state is chainged by NFB voltages? Then post EQ won't replicate that physical chainge.

 

[SubSpec]

Despite the debate, it is fascinating and impressive that Neumann engineers over a half century ago devised an approach that all this time later many very intelligent people are struggling to "define". The marketing materials shared a few posts back would seem to indicate it was an intentional design decision rather than a byproduct of attempting purely electrical cancellation.