Connecting a capsule

[pasarski]

Hello! I haven't posted in a really long time, but I recently got bit by the mic bug again, so here we go!

In the project description, the first feature mentioned is "no capsule coupling capacitor." Additionally, there is talk about the capsule grounding cap acting as a high pass filter, which is true.
The thing I'm wondering is, besides that cap being a part of the filter circuit, isn't it also directly in the signal path? Doesn't AC "fluctuate around" the signal source (capsule) in both ways?

There is this idea that having the cap on the ground side is somehow better than it being between the capsule and the grid/gate. It's considered a beneficial feature. Can someone please explain to me why it matters on which side of the capsule the cap is?

Edit: I originally posted in the osr87 thread which was mistake. This thread was separated from it by the mods.

I attach this crappy picture here to make it more clear. Both pictures represent a single membrane capsule connected to a jfet in different ways. Both use one cap and are externally polarized.

Many say that the left option is better and the cap is not in the signal path. I don't understand why and want to learn.

 

[micolas]

Hello! I haven't posted in a really long time, but I recently got bit by the mic bug again, so here we go!

In the project description, the first feature mentioned is "no capsule coupling capacitor." Additionally, there is talk about the capsule grounding cap acting as a high pass filter, which is true.
The thing I'm wondering is, besides that cap being a part of the filter circuit, isn't it also directly in the signal path? Doesn't AC "fluctuate around" the signal source (capsule) in both ways?

There is this idea that having the cap on the ground side is somehow better than it being between the capsule and the grid/gate. It's considered a beneficial feature. Can someone please explain to me why it matters on which side of the capsule the cap is?

It would be good to post a diagram and name the exact components.
There were several changes, including in the name/value of the components.
The coupling capacitor(capsule >gate)has a completely different function than others.

 

[pasarski]

It would be good to post a diagram and name the exact components.
There were several changes, including in the name/value of the components.
The coupling capacitor(capsule >gate)has a completely different function than others.

I'm asking in a broader context. Could have maybe started a new thread but it's directly related to this project so I decided to post here.

I'm basically asking what is the actual difference between these hookups?

 

[Khron]

I'm basically asking what is the actual difference between these hookups?

Some designers prefer to have at least one element of the capsule grounded, to minimize dust attraction. Not sure there have been any studies done on the effectiveness of that, though.

In the second diagram though, you have the option of NOT necessarily needing to have the JFET gate (or tube grid) connected to ground (through a large-value resistor), but to a non-zero voltage.

 

[micolas]

I'm asking in a broader context. Could have maybe started a new thread but it's directly related to this project so I decided to post here.

I'm basically asking what is the actual difference between these hookups?

first of all, the phase of the signal is reversed
Yes, it must be discussed in another thread
I think the differences were discussed sometime but I don't know where, that's why a topic must be posted in the appropriate thread

 

[pasarski]

It's not clear? There's a (single membrane capsule), a cap, a ground, a jfet (could be any other active element) and a connection to the polarization voltage.

 

[pasarski]

first of all, the phase of the signal is reversed
Yes, it must be discussed in another thread
I think the differences were discussed sometime but I don't know where, that's why a topic must be posted in the appropriate thread

I remember it was discussed years a ago (when I was still active here). Can't find the thread/threads anymore. If the mods feel this needs a new thread I will do it. I just thought that this was a proper place because the omitting the capsule to gate cap is listed as feature. It's not explained though why it's beneficial. That's why I'm asking here.

 

[micolas]

I remember it was discussed years a ago (when I was still active here). Can't find the thread/threads anymore. If the mods feel this needs a new thread I will do it. I just thought that this was a proper place because the omitting the capsule to gate cap is listed as feature. It's not explained though why it's beneficial. That's why I'm asking here.

the configuration is more complex
C5 is not alone in the ors87 topology
*
It is always better for the quality of the signal a direct coupling than through a capacitor (of any type, quality)

 

[pasarski]

"It is always better for the quality of the signal a direct coupling than through a capacitor (of any type, quality)"

I don't understand why it's so? Isn't the cap in the signal path regardless on what side of the capsule it's located? I've build few mics and favored the "direct connection". But when I really started to think of it I don't see the difference. Also, can't hear it. There must be something I'm missing.

 

[Khron]

I'm asking in a broader context. Could have maybe started a new thread but it's directly related to this project so I decided to post here.

I'm basically asking what is the actual difference between these hookups?

In the first diagram, that "extra" capacitor (on the left side) is purely a filter capacitor. In conjunction with whatever resistors are "upstream" of it, it forms an RC low-pass filter, shunting AC noise to ground.

In the second diagram, the capacitor (only) plays a DC-blocking role. And, to a lesser extent, it forms an RC high-pass filter with the gate / grid resistor that goes to ground (or whatever DC bias voltage is used), but that's less consequential, since the corner frequency routinely ends up in the sub-1Hz region.

 

[micolas]

"It is always better for the quality of the signal a direct coupling than through a capacitor (of any type, quality)"

I don't understand why it's so? Isn't the cap in the signal path regardless on what side of the capsule it's located? I've build few mics and favored the "direct connection". But when I really started to think of it I don't see the difference. Also, can't hear it. There must be something I'm missing.

Let's forget for a moment, just for simplification, about filters, feedback circuit.
At the huge input impedance of the jFET the 'capacitor that puts the BP to ground' - of much higher value than the capacitance of the capsule, it behaves like a direct connection to gnd(forAC), you are right, but
the signal goes from the diaphragm FD to the jFET directly.
You don't have a 'coupling capacitor at the back' of capsule.

 

[pasarski]

In the first diagram, that "extra" capacitor (on the left side) is purely a filter capacitor. In conjunction with whatever resistors are "upstream" of it, it forms an RC low-pass filter, shunting AC noise to ground.

In the second diagram, the capacitor (only) plays a DC-blocking role. And, to a lesser extent, it forms an RC high-pass filter with the gate / grid resistor that goes to ground (or whatever DC bias voltage is used), but that's less consequential, since the corner frequency routinely ends up in the sub-1Hz region.f

But if it was just a filter cap you could omit it altogether. Omit/short it and the mic makes no sound. How I see it, it blocks the polarization voltage from going to ground and completes the loop for the audio signal, and thus is as much in the signal path as in the other hookup scenario.

So yeah I'm still puzzled but I don't want to derail this thread too much. Maybe I just accept that the direct connection is better or start a new thread if I can't get over it

 

[micolas]

But if it was just a filter cap you could omit it altogether. Omit/short it and the mic makes no sound. How I see it, it blocks the polarization voltage from going to ground and completes the loop for the audio signal, and thus is as much in the signal path as in the other hookup scenario.

So yeah I'm still puzzled but I don't want to derail this thread too much. Maybe I just accept that the direct connection is better or start a new thread if I can't get over it

If you remove the de-emphasis part(for non-k67 capsule ), then it will only block the polarization voltage (DC) and in AC it will be at ground.
Will it still be part of the signal chain?
Or use an electret capsule, which does not require polarization voltage, connecting BP directly to GND.
How do you compare the two situations, with and without the coupling capacitor?

 

[Khron]

In the first diagram, that "extra" capacitor (on the left side) is purely a filter capacitor. In conjunction with whatever resistors are "upstream" of it, it forms an RC low-pass filter, shunting AC noise to ground.

But if it was just a filter cap you could omit it altogether. Omit/short it and the mic makes no sound. How I see it, it blocks the polarization voltage from going to ground and completes the loop for the audio signal, and thus is as much in the signal path as in the other hookup scenario.

Somebody needs to brush up on their electronics theory

You ARE aware of what "power supply filtering" is, and why it's useful (and even necessary), right? And i sure hope you can at least imagine why you'd want as clean a DC supply as you possibly can (within reason), for biasing a microphone capsule...

Once those aspects are covered, surely you can see why you WOULDN'T want to omit a power supply filter capacitor...

Moving on - are you implying that omitting and shorting out a power filter capacitor is the same thing, or at the very least has the same result? If that's the case, and you wanna test out this claim of yours, grab a 100u to 1000u 400v capacitor, attach it (the correct way) to a bridge rectifier, and plug that into the mains, then unplug it. Next, borrow a welding helmet and welding gloves, and a nice big chunky screwdriver that you're willing to sacrifice, and touch both capacitor terminals at the same time. Sparks fly, big "POP!..". You don't want to "short" power supply filtering capacitors.

If you "short it", referring to that filter capacitor between the polarization voltage and ground, of course you get no sound. No voltage difference between backplate & diaphragm => no capacitance => no capacitance TO change => no signal for the JFET to amplify => no sound.

"blocks the polarization voltage from going to ground" - well, if we're talking DC, sure. Why would that NOT be desirable? and what completing "the loop for the audio signal" do you mean?

 

[pasarski]

Somebody needs to brush up on their electronics theory

You ARE aware of what "power supply filtering" is, and why it's useful (and even necessary), right? And i sure hope you can at least imagine why you'd want as clean a DC supply as you possibly can (within reason), for biasing a microphone capsule...

Once those aspects are covered, surely you can see why you WOULDN'T want to omit a power supply filter capacitor...

Moving on - are you implying that omitting and shorting out a power filter capacitor is the same thing, or at the very least has the same result? If that's the case, and you wanna test out this claim of yours, grab a 100u to 1000u 400v capacitor, attach it (the correct way) to a bridge rectifier, and plug that into the mains, then unplug it. Next, borrow a welding helmet and welding gloves, and a nice big chunky screwdriver that you're willing to sacrifice, and touch both capacitor terminals at the same time. Sparks fly, big "POP!..". You don't want to "short" power supply filtering capacitors.

If you "short it", referring to that filter capacitor between the polarization voltage and ground, of course you get no sound. No voltage difference between backplate & diaphragm => no capacitance => no capacitance TO change => no signal for the JFET to amplify => no sound.

"blocks the polarization voltage from going to ground" - well, if we're talking DC, sure. Why would that NOT be desirable? and what completing "the loop for the audio signal" do you mean?

I mean that if you omit the cap the mic makes no sound. Also, if you short it there's no sound (for a different reason) buy that was not the point. If you omit a PSU filter cap the mic would still make sound.

 

[Khron]

I mean that if you omit the cap the mic makes no sound.

https://groupdiy.com/threads/open-source-diy-mic-project-ors-87-stripped-down-u87.86814/post-1156809
We're still talking about the connection example in the left-side drawing in THIS post, correct?

 

[pasarski]

https://groupdiy.com/threads/open-source-diy-mic-project-ors-87-stripped-down-u87.86814/post-1156809
We're still talking about the connection example in the left-side drawing in THIS post, correct?

Yes. I found at least one older post where a member is saying that cap is directly in the signal path. (It's the same hookup as in U47) https://groupdiy.com/threads/measured-vf14-characteristic.53620/post-711935

 

[micolas]

I mean that if you omit the cap the mic makes no sound.

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).

 

[Khron]

Yes. I found at least one older post where a member is saying that cap is directly in the signal path. (It's the same hookup as in U47) https://groupdiy.com/threads/measured-vf14-characteristic.53620/post-711935

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

Nevermind that among others, "even" ricardo called bs on his claims...

 

[micolas]

Yes. I found at least one older post where a member is saying that cap is directly in the signal path. (It's the same hookup as in U47) https://groupdiy.com/threads/measured-vf14-characteristic.53620/post-711935

I believe that the differences were due to the tolerances of the different values and not to the quality of the dielectric.