Connecting a capsule

[pasarski]

Oh, right, ok... I had no idea you were referring to "research" done by golden-ears, a decade ago... Confirmation bias, anyone?

Nevermind that Gus and ricardo called bs on his claims...

Like I wrote, I came to think this lately on my own and I rember I have seen it discussed here. I used the search to find that thread. Seems like I also participated, trying to tone down the aggressivenes. It was over ten years ago and I can't remember all the details.

I can't see where Gus and Ricardo called on his bs. On the he contary it seems like the one who was claiming it's not in the signal path admitted eventually that it actually is (the non linearities of the cap wouldn't influence the sound if I was not)

https://groupdiy.com/threads/measured-vf14-characteristic.53620/post-712139


Can you provide a link to a posts of Gus and Ricardo explaining it why it's not? I can only find Magnetos post where he thinks it is I the signal path.

 

[micolas]

If you remove the 33nF C5 from the circuit, it will still sound,but different, thinner,not the way you want (you already have a 4.7uF filter before it).

If you shunt C5, the polarization voltage will be 0.021v, so too low.

 

[Khron]

Like I wrote, I came to think this lately on my own and I rember I have seen it discussed here. I used the search to find that thread. Seems like I also participated, trying to tone down the aggressivenes. It was over ten years ago and I
can't remember all the details.

I can't see where Gus and Ricardo called on his bs On the he contary it seems like the one that was claiming it's not in the signal path admitted eventually that it actually is (the non linearities of the cap wouldn't influence the sound if I was not)

https://groupdiy.com/threads/measured-vf14-characteristic.53620/post-712139


Can you provide a link to a posts of Gus and Ricardo explaining it why it's not? I can only find Magnetos post where he thinks it is I the signal path.

I think this warrants a thread of its own - maybe we should ask the mods to branch off the conversation starting from this post, into a separate thread, so we don't further crowd this project with arguably unrelated things?

 

[pasarski]

If you remove the 33nF C5 from the circuit, it will still sound,but different, not the way you want (you have a 4.7uF filter before it).

I don't see how it could make sound if there would be no path to ground for the signal? Maybe through a parasitic capacitance a small signal could be present.

But like @Krhon said it's probably better to make new thread for the topic. If mods don't catch this then maybe I just start a new one.

This is a cool project and the way the capsule is connected is maybe not the most important aspect of it. All an all I'm delighted to see this kind of projects emerging, not just 1:1 clones.

 

[micolas]

I don't see how it could make sound if there would be no path to ground for the signal? Maybe through a parasitic capacitance a small signal could be present.

But like @Krhon said it's probably better to make new thread for the topic. If mods don't catch this then maybe I just start a new one.

This is a cool project and the way the capsule is connected is maybe not the most important aspect of it. All an all I'm delighted to see this kind of projects emerging, not just 1:1 clones.

Test and see if it sounds.
You still have (1.33M+4.7uF) at GND. Compare with 1G.
It will sound less bassy and I think 10...15dB lower level.

 

[abbey road d enfer]

I couldn't find where the posts have been moved, so I'll answer here.
There is one big difference between the two modes.
With a grounded capsule, bias has to be added in parallels with the signal, when in the case of a non-grounded capsule, bias and signal are in series.
The difference is that in the former case, the bias resistor and the gate resistor are in parallels, effectively halving the resistance seen by the gate, which results in 3dB more noise.
Another advantage of ungounded capsule is that the diaphragm does not attract dust particles (except in fig-8 mode).

 

[pasarski]

Test and see if it sounds.
You still have (1.33M+4.7uF) at GND. Compare with 1G.
It will sound less bassy and I think 10...15dB lower level.

I don't quite get it. To me it seems that C5 is the only cap providing a path to ground for the backplate. Without it isn't the capsule only a (variable) capacitor hanging in the air not contributing anything to to the circuit (other side of the capacitor not connected to anything)?

 

[Khron]

To me it seems that C5 is the only cap providing a path to ground for the backplate.

And what's wrong with that? Why must the capsule have (at least) one grounded terminal?

Without it isn't the capsule only a (variable) capacitor hanging in the air

And the problem is..?

not contributing anything to to the circuit

What are you basing that assumption on?

other side of the capacitor not connected to anything

That sounds like the other terminal of the capsule is indeed left hanging, unconnected to anything at all. That's very much not the case, and borders on being a strawman-argument. Why MUST at least one terminal of the capsule have a (low-reistance, and especially) DC path to ground???

 

[micolas]

It will sound less bassy and I think 10...15dB lower level.

@pasarski
Just test in the circuit with C5 not connected.
I guarantee it will work as I said above
I can't accurately estimate the signal level at the output, I approximated.

 

[pasarski]

I couldn't find where the posts have been moved, so I'll answer here.
There is one big difference between the two modes.
With a grounded capsule, bias has to be added in parallels with the signal, when in the case of a non-grounded capsule, bias and signal are in series.
The difference is that in the former case, the bias resistor and the gate resistor are in parallels, effectively halving the resistance seen by the gate, which results in 3dB more noise.
Another advantage of ungounded capsule is that the diaphragm does not attract dust particles (except in fig-8 mode).

By grounded capsule you mean the right hookup in this picture? I know the theory that polarizing the membrane is attracting more dust but it's not relevant for what I'm trying to understand. You can very well swap the backplate and membrane in both of the scenarios.

@pasarski
Just test in the circuit with C5 not connected.
I guarantee it will work as I said above
I can't accurately estimate the signal level at the output, I approximated.

Ok, now I see that in this schematic there sure is a path to ground for the signal even if you remove C5. It was defenetly a wrong thread for me to post and I apologize for that.

Like I said I'm interested in this in a broader scale. The question remains, is there really a sonic penalty to ground the other side of the capsule and take the signal from the other (polarized) side through a cap vs. connecting the other side directly to the gate/grid, polarize the other side and connect it to the ground via a cap?

 

[micolas]

Here FD is connected via C15=1nF to the Gate of jFET Q7 and BP is polarized directly from the DC/DC converter with positive voltage.

Why is the Backplate BP not connected to ground via a capacitor?

How can it work?

The scheme belongs to the B1 a good microphone, similar to another Neumann scheme.

 

[pasarski]

And what's wrong with that? Why must the capsule have (at least) one grounded terminal?



And the problem is..?



What are you basing that assumption on?



That sounds like the other terminal of the capsule is indeed left hanging, unconnected to anything at all. That's very much not the case, and borders on being a strawman-argument. Why MUST at least one terminal of the capsule have a (low-reistance, and especially) DC path to ground???

Now that I look the circuit I see the path to the ground even without C5. I apologize also for you bringing this question up in a wrong thread. What I don't apologize is that I came back here after ten years break to try to learn more about mics and ask a simple question which is still unaswered, it at least I don't get it.

Also, I really don't get your tone, talking about strawmans and me having a problem other than just trying to understand something. So please bear with me, we are not all EE's and I did not came back here to trying to annoy anyone.

 

[pasarski]

Here FD is connected via C15=1nF to the Gate of jFET Q7 and BP is polarized directly from the DC/DC converter with positive voltage.

Why is the Backplate BP not connected to ground via a capacitor?

How can it work?

The scheme belongs to the B1 a good microphone, similar to another Neumann scheme.

The other side is polarized and the other side is grounded via R15 and connected to the gate via a cap. You can hook up capsule like that also but then you don't have the direct connection from the capsule to the gate which is said to be beneficial.

It's a good mic btw. I like the capsule.

Edit: Actually there is path to ground for the AC from the other side of the capsule via C13 and C9.

 

[micolas]

The other side is polarized and the other side is grounded via R15 and connected to the gate via a cap. You can hook up capsule like that also but then you don't have the direct connection from the capsule to the gate which is said to be beneficial.

It's a good mic btw. I like the capsule.

Edit: Actually there is path to ground for the AC from the other side of the capsule via C13 and C9.

So if we disconnect C9 the circuit will stop working?

 

[micolas]

The other side is polarized and the other side is grounded via R15 and connected to the gate via a cap.

So, R15 grounds for the AC the capsule diaphragm, but allows the useful signal (which is also AC) to pass selectively to the Gate of the jFET.

 

[pasarski]

So if we disconnect C9 the circuit will stop working?

That's what I assume to happen. But I have to test. Have couple of these boards in the drawer.

So, R15 grounds for the AC the capsule diaphragm, but allows the useful signal (which is also AC) to pass selectively to the Gate of the jFET.

I think I was not making sense there. I didn't manage to edit that part, but yeah me thinks the AC's path to ground goes via C13 and C9. I will remove C9 and will report back!

 

[micolas]

That's what I assume to happen. But I have to test. Have couple of these boards in the drawer.



I think I was not making sense there. I didn't manage to edit that part, but yeah me thinks the AC's path to ground goes via C13 and C9. I will remove C9 and will report back!

I think removing C9 will still work, but it may increase the noise produced by the zenner diode a bit.

 

[micolas]

@pasarski

We have moved too far from ors87
I think you can edit the title of the thread and change it, for example:

' Grounding the capsule '

 

[pasarski]

@pasarski

We have moved too far from ors87
I think you can edit the title of the thread and change it, for example:

' Grounding the capsule '

Yes of course. I never ment to address ors87 specifically and was silly to intevene in to that thread.

I just want to understand why it's better to have a cap between the capsule and the gate/grid vs. cap between the capsule and the ground. And why is other consider to be in the signal path and the other not.

 

[Khron]

I just want to understand why it's better to have a cap between the capsule and the gate/grid

It's not necessarily better, unless you have an actual (technical) requirement to use that topology.

Also, i would very much NOT put those two topologies on the same footing, because apples vs. oranges. One is filtering AC noise to end up with clean DC, and the other is simply blocking DC between two circuit nodes while allowing AC to pass through (at least in theory, unscathed, but we don't have access to ideal components, so...)

And why is other consider to be in the signal path and the other not.

The former view i couldn't explain...