The circuit is a little more complicated than what you're assuming. See the schematic I attached. There are resistors internal to the mic (R3 & R4) that isolate the outputs further, and they are in series with the resistors in the mic pre. In addition the two sets of resistors are in parallel with respect to the voltage source, so their effective resistance is half. For example if the isolation resistors in the mic were 10k, then the total resistance used for calculating the voltage drop is (6.8k + 10k)/2 = 8.4k.

Interrupting the phantom supply in any way is going to cause loud pops on the output because you're changing the supply voltage for the mic. The trick is to effectively isolate the outputs of the mic without affecting the DC voltage.

EVERY console channel has a pair of 6.8K resistors. One end of each R is tied to 48VDC, and the other end of one resistor ties to pin 2, the other end of the second resistor ties to pin 3. Phantom can't work with out the pair of R's for each mic input.

Bri

There is a 48V supply that is connected to pins 2 and 3 via 6K8 resistors (usually) so when the mic consumes power the dc voltage at pins 2 and 3 does drop. Because both resistors pass current it is the same as if the 48V supply was fed to the mic through two 6K8 resistors in parallel i.e.3K4. 10mA flowing in 3K4 drops 34V so the voltage then available to the mic would be 48 - 34 = 14V. If the mic consumes 10mA it has to be able to work from this voltage. Most mics take no more than 5mA. |even so, some condenser mics have an internal switching supply to raise the voltage for the capsule.

Cheers

Ian

..can I respectfully re-suggest that there must be something else wrong if shorting 2+3 does not mute the microphone signal.? The signal left has to be common-mode only, and the third-most-important function of any balanced microphone input is keeping CMRR at a minimum

/Jakob E.

The OP quotes from Shure that the pin 2 to pin 3 jute should provide at least 50dB attenuation but they do not say why. Maybe it's just the typical CMRR of project mixers?

Cheers

Ian

Well IF you short the input of any mic pre only thing left is the noise of the pre..Regardless of whats before the short...
I would never short on a phantom mic, there are far better ways to provide a Mute..

I want to show you what I made so far. It was quite a lot of work by hand (about 30 holes with a manual thread-cutting tap ), because I don't have any machines, but I am satisfied with the result:

Top
https://i.prvy.eu/7ptmq9t.jpg

Bottom
https://i.prvy.eu/ddkisub.jpg

Side 1
https://i.prvy.eu/jxpxpnq.jpg

Side 2
https://i.prvy.eu/miciyhx.jpg

Inside
https://i.prvy.eu/8ifhj8t.jpg

(Unfortunately I am still not able to embed images, so here are the links)

From the image Inside you can see the switch I used. Pin 3 + Pin 3 into "4 NO" and Pin 2 + Pin 2 into "3 NO". Once the switch is pressed, they are shorted.

Aside from the issue that Shure points out that this does no mute, but only lowers the level, I would suspect that the short simply does not draw 100% of the signal. It may be an expected physical effect that a few information still gets back to pin 2 and 3.

I tested it with a RME Fireface UC and a SPL Track One. No difference, no infinite muting.

Is this a condenser or dynamic mic?

You mean what it is for? It is primarily for a condenser microphone, but shall in the best case also work for dynamic and electret condenser microphones.

No, the Mic you are using for your test, Is IT a dynamic or condenser?

Condenser (Oktava MK-102 / Rode NT1-A) I also tested it with an electret condenser (Oktava MKE-9).

Use a Dynamic..and see what you get.

I unfortunately forgot while writing my lines (post #39 ) that I sold my only dynamic microphone I had.

However, I would have left a ribbon microphone, the beyerdynamic TG V90r. If it is safe for a passive ribbon?

Turn +48V OFF and it will be, wait a few seconds to make sure the voltage is bleed off...
OK, I think I have an idea, I would bet that this would NOT happen with a transformer based mic pre, this could be a crosstalk issue..

I waited now for the 48V to bleed off and tested it also with a ribbon.

I used the SPL Track One + RME Fireface UC (A/D only over line, no additional preamp).

Here are two audios:

OK, wow you could hear it pretty good...
Do you have a transformer based pre you could use?
This is VERY surprising, Going to test this at my shop later on...

Well, I unfortunately only have the SPL Track One and the RME Fireface UC.

Indeed! It is so much audible. But once you look at the ratings at Amazon for mute buttons such like Rolls MS 111 the people complain about it not muting, but only lowering the level. I hope it is okay to quote it here:

The interesting thing for me is, that it even bleeds with a passive ribbon microphone where no phantom power at all is involved.

I did a short test by using only 2 x 2 x 330 Ohm resistors (= 660 Ohm on each pin 2 and 3) and as expected, that significantially lowers the level, but not perfectly. I also tested then with 2 x 2 x 680 (= 1,360 Ohm on each pin 2 and 3), but it is not perfectly muted.

I am not sure if it would be safe to use e. g. 10K or 33K Ohm to absolute block anything, even in pin 2 & 3 shorted mode.

Photo
https://i.prvy.eu/3hvsz3h.jpg

Was curious, did a test at my shop; shorted it is NOT NO output..I measured around -70dB...

If I knew that earlier

Thus, we should not call shorting pin 2 & 3 at all muting (also due to security concerns), but refer to as something like –50 dB padding or finite level reducing. I also heard the term caugh-button, so such a pin 2 & 3 short-button is not a mute-button, it can only be considered as a button to temporarily reduce the level significally whenever full-mute is not required.

Sorry guys, like my signature says... I’m wrong. I was working from memory... the console does have them. My bad.

On the other hand, looking at these single channel pres that are racked from a different console, the phantom was added later; It looks like, in this case, phantom is sent to an input via two wires and the Rs are added at the power supply end... so a different schematic entirely! I never noticed this! I guess this is an attempt to not modify the original pcb while creating a power supply with ‘isolated’ 48v outputs?

It seems like you could then create a +48v power supply with 16 outputs with a 6k8 in series on each and call them ‘isolated’ and power all 16 pins (pins two and three of each xlr) of an eight channel pre... that looks like what is going on here, but with fewer channels.

Anyway, you’re right!

-B

But the "cough button" On a broadcast console is NOT on the mic input, its after the mic pre and remote ctr..The way I would do it..Clean & pop free..

No problem!

When you refer to Image: Inside 1 or Image: Test 1 the very only thing this box shall do is muting. Any pop or clicks sounds are not a big issue, but they are if they are too distracting. So it would be optimal to reduce them a little bit.

On the left side (see Images) XLR is going into the Fireface UC (48 V, Preamp) or the SPL Track One (48 V, Preamp). On the right, the XLR is going to the condenser microphone. Or do I need to split a cable here?

Isn't it it in my image? It is after the Microphone Preamp, already getting phantom power between pin 1 and 2/3, but before the microphone input.

EDIT: Here is an image with labels of the current inside of the box: https://i.prvy.eu/5nfuoxe.jpg

I also did this test (see this gorgeous photo):

Microphone into the SPL Track One (Anlog Input). Then Analog Output from the SPL Track One into the Analog Input of the Fireface UC while reducing the Preamp of the Fireface UC to 0 dB + disabling 48 V and keeping 38 dB for the SPL Track One.

See audio (+6 dB digitally). Doesn't work

Your box is between the Mic & Mic Input right?
If so its NOT after the mic pre, plus if it was After the Pre you would NOT have any +48V..

I think I see now. My previous post in the meanwhile, that was after the microphone pre, wasn't it?

EDIT: It seems it may be simply not enough to lower the gain of the Fireface UC to 0 dB. I would have to bypass its Preamp at all by using preamp-less Balanced Line instead of any of the two preamp-able Balanced XRLs. But even in that case I don't understand why the signal is still audible, because 0 dB should be almost nothing.

Is this supposed to be able to be used with a phantom-powered mic, or a line-level output? It does matter, as there are different trade-offs in these two applications when working out a variation to improve the level of "offness".

Also, a good place to start would be to figure out the target level of attenuation that would be acceptable, and measure a little more precisely what you have now, perhaps play a 1KHz tone through a speaker and into the mic, so you can take a proper reading. From there, if you know the source impedance of the microphone, you can get an idea of where the problem may be. For instance, if you really are getting about -50dB attenuation, and the mic has a 150 ohm source impedance . . . this means that the switch has a around 0.45 ohm of impedance across the audioband when it's closed . . . and this is pretty high! If it's rated to have a low contact resistance, then perhaps it's defective . . . or perhaps you're really getting more like -70dB and it's still not enough, or your mic (or other driving source) has a lower source impedance . . .

In any case, nailing down some of these variables would help immensely. With a design goal of "-infinite dB" from "any source", then the only way to solve it is with "crazy overkill".

56 posts, and the good info was in the first three. Schnarbel, there are good ways to do things, and bad ways. Yours is a bad way. Stop, take the good advice you have been given, and do the job the good way.
First, amplify the signal to line level, and then switch it correctly with a breaking switch, and no phantom involved. And that big switch you have? Looks totally wrong for the job.

With that plan, you're "hauling" the balanced mic line through the chassis to connect to the resistors being fed from 48V at the PSU. Much simpler wiring is to solder the 6.8k resistors right at the mic input socket and run a 48V line to the resistors.

Bri

Imperfect common mode rejection, which might be happening the mic, the interconnects, the preamp, or some combination of the three.