A mic that never was (a thought-experiment circuit for the RK-12 capsule)

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

OneRoomStudio

Well-known member
GDIY Supporter
Joined
Jul 8, 2004
Messages
966
Location
Minneapolis, MN, USA
The other day I was thinking about the design of the ELA M251, and how it differed from the C12 in ways that were intended to make it more Neumann-like - namely, the circuit was simplified and switched to cathode-bias, they changed the pattern control to an on-mic switch (like the U47), they used a larger, more Neumann-esque headbasket, and they added a cap from plate to ground to tame the top end (presumably to make the CK12 sound slightly more like an M7). The non-export version even used an AC701K, a tube that the M49 used.

Side note: does anyone know what the letters in the model name "ELA M251" stand for? I know the suffix "e" is for export (denoting a 6072 instead of the AC701K), but I have no idea what "ELA" or "M" stand for.

Anyway - I suddenly thought about the modern Chinese "CK12" capsules, and how they're just edge-terminated K67 capsules, and had the thought, "ha, seems like something Telefunken might have commissioned after the release of the U87i."

So I went down a rabbit hole...

The Neumann KM series was their first foray into solid state (the 'KTM' came out in 1965), and the U87i followed two years later, in 1967. As far as I can tell, the U87i became an instant hit, and Telefunken must have wished they had an answer for it. AKG didn't transistorize their C12a until ~1971 (calling it the C412).

There's a gap there, that could have been filled with my fictional mic...

In 1968, Telefunken approached AKG to design and manufacture a mic to go head-to-head with the U87i. It needed to share DNA with the ELA M251 that had become famous, but it had to be transistor-based (ideally a similar circuit to what Neumann used), needed to be powered by P48, and needed to be good for vocals (where microphones become famous).

So...the engineers at AKG got to work. They were impressed by Neumann's new K67 capsule and how much easier it was to produce than their own CK12, but they wanted to retain the comparative "flat" sound of their edge-terminated design, so they put an edge-terminated diaphragm on a K67-style backplate. Now to design a circuit to put it on...

While the U87i circuit was a good starting point, their new capsule didn't share the high-frequency lift of the K67, so the de-emphasis wasn't needed. They could use the KM84 circuit (a great new mic that had just been released two years prior), but they wanted to do something to keep a sense of "warmth" in the sound and voice it a little closer to the U87i - the solution was obvious - add a small capacitor from the drain of the FET to ground, similar to the plate-to-ground cap in the 251! The output transformer would be something of a cross between their beloved T14/1 and Neumann's Bv.13 (the best modern approximation of this might be a Cinemag 2480). The resulting circuit:


Screenshot 2024-10-13 at 1.32.33 PM.png
(Cardiod-only for now, more on that later)

The housing would be like the ELA M251, but smaller (something like the modern "GT-2B" body), and it would be painted in the 251 colors.
 
but I have no idea what "ELA" or "M" stand for.
ELA = "Elektroakustik"
M = I guess "Mikrofon" 😎

The abbreviation ELA often stands for an "Elektroakustische Anlage" (electro-acoustic system) that essentially corresponds to the PA system in English-speaking countries.

Edit: I had very similar ideas this weekend, inspired by another thread. My planned capsule is a "Rayking" and the schematic is almost identical. Your story is better though...;)
 
Last edited:
Edit: I had very similar ideas this weekend, inspired by another thread. My planned capsule is a "Rayking" and the schematic is almost identical. Your story is better though...;)
Great minds! Or at least...two people who think about microphones too much, haha.


To continue the idea, I thought that this might be the way they would switch patterns on the mic:

Screenshot 2024-10-13 at 2.39.14 PM.png
This has a few benefits:
1) The front diaphragm is kept at near-ground potential, and connected directly to the JFET gate - no need for a coupling cap.
2) The cardioid setting is "true" cardioid, with the rear backplate and diaphragm disconnected.
3) This avoids the need to use half-voltage (24V) on one capsule to develop figure-eight.

The only downside is the use of C1 from the rear diaphragm in figure-eight. I suppose this could lead to slight phase and frequency discrepancies between front and rear diaphragms, but I doubt it would be a problem in practice.
 
But for the zener regulation and the polarization stuff, that looks a lot like the 2N5457 based C12 topology that I built a few weeks ago. The bias resistor on mine is 820R, if that's of any help to you - a 25k pot may be difficult to dial in if what you really need is that small relative to the pot size.
 
ELA = "Elektroakustik"
M = I guess "Mikrofon" 😎

The abbreviation ELA often stands for an "Elektroakustische Anlage" (electro-acoustic system) that essentially corresponds to the PA system in English-speaking countries.

Edit: I had very similar ideas this weekend, inspired by another thread. My planned capsule is a "Rayking" and the schematic is almost identical. Your story is better though...;)
I was “going to say the same thing” after reading the post you are responding to.

I will add that Phillips was using “ELA” prior to the “ELA M” branding on Telefunken 251s.
 
Hey, I am adapting this design to a Rode NTK housing and a Neutrik NTE 10/3 transformer. It would be my first attempt with a circuit, that includes one. Is my wiring correct here?
1738330632885.png
Manufacturer diagram for reference:
1738330734216.png
 
Hey, I am adapting this design to a Rode NTK housing and a Neutrik NTE 10/3 transformer. It would be my first attempt with a circuit, that includes one. Is my wiring correct here?
View attachment 144743
Manufacturer diagram for reference:
View attachment 144744

No, you want the transformer back-to-front.

Red & black towards the circuit, white & yellow towards the output connector.
 
Hey, I am adapting this design to a Rode NTK housing and a Neutrik NTE 10/3 transformer. It would be my first attempt with a circuit, that includes one. Is my wiring correct here?
View attachment 144743
Manufacturer diagram for reference:
View attachment 144744
You must reverse the wires of the transformer, i.e. the primary should be Red/Black and the secondary White/Yellow.
You have no green wire at nte10/3.
Important: If the Output Signal will not be in phase, the primary wires must be inverted between them, or the secondary wires between them or the XLR pins 2 and 3 between them.
I would put C6=22uF for more low frequencies and gain.
C9 should be tested for some treble attenuation to taste, it can be even more if you use a bright capsule. You may want to put a 33v zener instead of C4 (or in parallel with it) to stabilize the supply voltage of the jFET.
 

Attachments

  • Screenshot_2025-01-31-16-35-10-041_com.android.chrome.jpg
    Screenshot_2025-01-31-16-35-10-041_com.android.chrome.jpg
    552 KB
And my final question for search of confirmation. I want to add a DCDC converter to the circuit. Would the point below be adequate (see blue net point):
1738337084984.png
 
Regarding the transformer would this be correct:
View attachment 144748

Or this instead?

View attachment 144749

If you swap BOTH pairs of wires, it doesn't change anything.

And my final question for search of confirmation. I want to add a DCDC converter to the circuit. Would the point below be adequate (see blue net point):
View attachment 144752

I'd take it from before (not R13, but) R3, so there's even less chance of conducted interference from the DC/DC reaching the JFET stage.
 
Last edited:
SolidState.png

I want to use a UTM0547 as OPT. I have used other ratios with a similat design in the past. J2 can be even be an BJT, but I'm just using parts I have laying around. Disregard R17, I forgot to pull it out, lol.
 
Last edited:
And my final question for search of confirmation. I want to add a DCDC converter to the circuit. Would the point below be adequate (see blue net point):
View attachment 144752
Yes, voltage can be taken from here, or from the two 2.2Kohm/C3 through an Rdcdc resistor that must be calculated/tested, so it would be good to post the DC/DC converter diagram as well.
R3=10Kohm seems a bit small to me, I put 22..33 Kohm, it also depends on DC/DC.
 
Back
Top