LDC k67 capsule output

[kingkorg]

I was wondering about this for a while, couldn't find any info on how much LDC can produce (modulate pol.V. to be specific).

So i made this simple OPAlice circuit with opa1641 which is configured as buffer, and scoped output to see how large transient i can produce. For people not understanding what i'm talking about the output shows the same value as the input of the opamp - the output of the capsule, maybe tiny bit less. I polarized the capsule with 60v.

Here are the results, transient of a snare, and loud vocal scream in front of the capsule. One thing to note is that plosives of a vocal create easily peaks just as large as snare hit.

Notice that snare transient is asymmetrical due to the nature of the instrument, so a source capable of producing this kind of level with symmetrical waveform would give almost 12v peak to peak.

The opamp can take quite large input without distorting at all, so there is no clipping whatsoever. It can take much more than fet or a tube in plate out configuration.

 

[Khron]

That should be an incentive to use as great a supply voltage as feasibly possible, in the case of phantom-powered mics at least.

... But on the other hand, just how obnoxious is the sound of clipping some extraneous peaks, though?

 

[kingkorg]

A lot of times i clip those transients, however clipped LF transients sound horrible to me, i prefer to keep everything under 1K as clean as possible. Often if i distort complex sources with a lot of low end just by adding saturation to stuff above 1k, rarely under.

This stuff makes me think an ideal kick condenser would be using opamp instead of discrete fet with center terminated capsule with thicker diaphragm/backplate spacer.

 

[Khron]

Might be interesting to know how sE approached the electronics in the X1 D, in addition to the (alleged) "titanium diaphragm".

 

[rogs]

That should be an incentive to use as great a supply voltage as feasibly possible, in the case of phantom-powered mics at least.

That was my thinking when I tried out the OPA1641 instead of a FET.
I got the idea from Jules Rykenbusch's Alice OPA project, but I differed from his approach by using a single sided output using an OPA641 instead of the fully differential OPA1642 approach that Jules adopted. I also only used a zener for the voltage multiplier circuitry.
The lower current required by the 1641 and no zener gave me a c. 25V DC supply for the op-amp from a 48v phantom supply.
Initial results seem pretty good..
Some initial notes on the project here: www.opic.jp137.com

 

[kingkorg]

Awesome @rogs!

Btw, some recent posts at micbuilders group suggest single opamp version, impedance balanced might be actually performing better noise wise.

TL Sphere uses TLC070.

 

[rogs]

Btw, some recent posts at micbuilders group suggest single opamp version, impedance balanced might be actually performing better noise wise.

I can confirm that is true. (I'm 'Arjay 1949' on the micbuilders group).
The noise figure will also depend on the value of the resistors used to provide the inverting requirement of the second differential op-amp.
I replaced the 1G resistor with a 1K, removed the capsule and AC coupled a signal generator directly to the non-inverting input.
I then made a short single tone recording - followed by the same input with the tone source removed.
I then added the differential output and repeated the process.
Obviously the second recording gave an output that was 6dB higher.....
Adjusting the gain setting of one channel in my DAW - to match the two recorded tone levels - revealed that there is improvement of around 3dB in the noise floor of the single sided output.... Which makes sense, maths wise!
Using higher value resistors for the inverting amp - I tried 2k2 and 10k - progressively increases the noise level. Again it makes sense, maths wise.

There is a short recorded file of the difference in noise levels HERE . The single sided version is first.

 

[kingkorg]

I'll give a shot with your circuit for my kick mic.

 

[kingkorg]

An update. Here's what a plosive looks like straight into unprotected SDC capsule polarized at about 40v.

22.2Vpp. Add about 8-10% to that due to circuit loss.

 

[rock soderstrom]

An update. Here's what a plosive looks like straight into unprotected SDC capsule polarized at about 40v.

22.2Vpp. Add about 8-10% to that due to circuit loss.

Wow!

 

[ccaudle]

Here's what a plosive looks like straight into unprotected SDC

This could be a fun series. With and without pop filter, in front of a trumpet, symphonic bass drum, etc.

 

[rogs]

I bought myself a birthday present recently.... A Zoom F3 field recorder.

My first venture into the world of 32bit floating point recording. Not sure I understand all the maths involved, but essentially there is huge high resolution dynamic range. No need to set input levels.... It's very counter intuitive, but it seems to work well.
I really bought it for a new ultrasonic recording project I have planned for this summer.

As regarding your impulse tests, this F3 device has a maximum line input level of +24dBu. That's 34.7v p-p..... I might try and see if I can create a high enough input signal to check where some of my mics actually clip.
So far I've simply screamed as loud as I could about 10mm from the capsule. That didn't clip....although it did generate a signal with a peak some 115dB above the ambient level.

This is a whole new world for me !

 

[OneRoomStudio]

I wonder how much the output of the same capsule changes with polarization voltage (i.e. 48V vs 60V). Obviously there will be some sensitivity loss, but how much? This might be a good case for using lower polarization voltages on mics used for kick/bass, etc.

 

[Khron]

I wonder how much the output of the same capsule changes with polarization voltage (i.e. 48V vs 60V). Obviously there will be some sensitivity loss, but how much? This might be a good case for using lower polarization voltages on mics used for kick/bass, etc.

Math says about 2dB. Fwiw, padding in the Rode NT2-A is achieved by reducing the capsule bias voltage.

 

[OneRoomStudio]

The next question would be how low can the polarization voltage go on a K67 or K47 before there are audible consequences (other than sensitivity)?

 

[OneRoomStudio]

Some interesting and relevant info here: https://audioimprov.com/AudioImprov...Bias_Voltage_for_True_Condenser_Capsules.html

They go down to 26V polarization voltage with no obvious ill effects other than a decrease in sensitivity. That would be what, a ~7.25dB decrease from 60V?

 

[Khron]

They go down to 26V polarization voltage with no obvious ill effects other than a decrease in sensitivity.

They must go lower than that - NT2-A has a -5dB and -10dB pad setting.