Back in February, I set out to compare two versions of the AKG C 414 microphone series, and that column evolved into a “history of” piece. I felt the historic trail was necessary to add depth and background to the comparison. I am still on that trail.

With this, I gained some insight into the AKG C 12 and its made-for-Telefunken Ela M 250/251, the heart of which is AKG's CK 12 capsule. The Euro versions have an AC-701/k miniature triode (with soldered leads) while the “E” (Export) version, such as the C 12, uses half of a 6072 (a standard 9-pin tube). The circuitry of the C 12 is nearly identical to that of the Ela Ms.

The C 12A “detour” also took me into Nuvistor territory. Introduced by RCA in 1959 for its New Vista color TV tuners, this transistor-sized vacuum tube was quickly embraced wherever a high-performance miniature tube was required. And though it may not have been the best retrofit for the VF-14 (used in the Neumann U47/U48), it was during the research process that I found Neumann's conversion documentation. This solved a mystery that I had described in the May '06 issue — a U47 power supply with un-German amounts of extra juice.

After that column was published, I realized that the power supply in question had not been mis-repaired, but had been modified to drive a Nuvistor U47. The fact that it was driving a stock U47 meant that somewhere along the way, it had become separated from its mod-mate. This “discovery” finally gave closure to what was fortunately not a catastrophic failure. If you own a U47 with a VF-14, then make sure your power supply is delivering 105 volts (with the mic connected and warmed up). Full documentation of the U47 and its power supply is available at www.tangible-technology.com.

Most tube-based audio gear is fairly easy to repair. Vintage valve mics, in particular, are so simple that a vacuum tube data manual — plus pencil and paper — will do in a pinch. There is typically only one gain stage, and the external power supply generates all of the required voltages. A solid-state mic must derive all needed “juices” from phantom power, a finite resource that must be efficiently managed.

Early solid-state circuit designs were nearly as simple. (See Fig. 1, the AKG C 414, circa 1970. For Figs. 1 and 2, the pad [C 414] and highpass filter [P48] have been removed for clarity. Between the capsule and the output transformer (aka, the head/output amplifier) are one FET and one transistor. In addition, the DC-to-DC converter comprises a single transistor oscillator, a transformer and three diodes configured as a voltage “tripler.” Now, a surface-mount Hex “Schmitt Trigger” IC is used to generate the capsule-polarizing voltages (the RØDE NT-1A), and for many modern mics, the count of active components — transistors and FETs — is well beyond a dozen.

The most versatile mics tolerate the widest range of available phantom power (9V to 52V), the C 414 EB (circa 1977) being one example, with the output amp consuming most of the available current. However, the C 414 EB-P48 (circa 1982) has no DC-to-DC converter, so it must have 48V phantom power (via R7 in Fig. 2) to deliver a usable polarizing voltage.

Having amassed several versions of the 414 (some for repair, others for parts), my quest for schematics began. The introduction of the C 412 in 1970 inspired evolutionary changes that were intended to solve known headroom issues (a pad), add features (low filter, more polar patterns) and take advantage of new technology. Remember that many of these features previously resided at the power supply and became more vital as the mic moved closer to the source and rock 'n' roll SPLs were on the rise.

The C 414 (circa 1971) looks like a C 12A. Of the pair in for restoration, one had poor low-frequency response. Capacitors degrade over time — interstage caps “lose” capacitance, the LF output suffers — so the C 414 problem (C3 in Fig. 1) was typical, obvious and quickly resolved.

A C 414 EB-P48 initially had no output because the output transformer (U54) was damaged. Once the output transformer was replaced, it was again possible to “hear the grille” when it was scratched. The P48 version has two PCBs on each side of the mic body: one for the head amp and polar patterns, the other for the output amp and highpass filter. Conveniently, the interconnecting wires between the two allow you to interrupt the head amp signal so a test signal can be injected. This initially gave the impression that the output amp was okay, leading me to suspect (and troubleshoot) the head amp (a dead end).

Electronic components can be drawn differently; check out the FET (TR-1 in Fig. 1 and T3 in Fig. 2). The P48 version had part values on the schematic, so scouring the Web for data sheets and available stock revealed that all but T3 was available. The C 414 “documentation” comprised only a schematic — it had become separated from its parts list. Fortunately, no critical parts were required.

In the product literature, the P48 consumes a mere 1 mA of current, but this broken P48 had a 20V drop across R7 — almost 10 times the specified current draw! I pulled the output amp (T4), put the 'scope on the head amp PCB and — voila — signal! All of the noises in my shop were now clear. I was at once overjoyed with progress and determined to bring this journey to a happy conclusion.

Comparing the C 414 with the P48, the output-transistor TR-2 is an emitter follower, meaning it does not make the signal “bigger” (as would be the case with voltage gain) but does make it more powerful (current gain). The load (emitter) resistor is R7, C3 blocks the DC voltage but passes the audio signal on to the output transformer.

By contrast, the P48 output transistor is T4, and its load/source resistor is T3, a FET configured as a current source (a dynamic impedance that optimizes the load on T4). After replacing T4 and T3, I finally checked the output cap (C11) and it was shorted. (Boy, did I feel stupid!) This caused the current to flow from T4's emitter straight through the primary of U54, the output transformer — hence the additional current consumption.

With the cap replaced, everything returned to normal. Then I cleaned all the flux off the PCB because over time, flux absorbs moisture and begins to conduct — something that high-impedance circuits in mics (and 1176 limiters) do not like. After listening for noise and sorting through a few components, the amplifiers were suitably quiet.

Perhaps the saddest of my AKG experiences is that none of these mics had their original “brass” CK-12 capsules. All had the “nylon” version, part number 2072-Z-0005 (or 0009), including one sold on eBay claiming an original CK-12 capsule but arriving with a nylon capsule with the diaphragm hanging loose. To my surprise, after cleaning the debris from the backplate and the backside of the diaphragm, I popped all the pieces together and got sound, including the figure-8 null.

Despite my bone-headedness, this journey taught me a lot. Had I found the bad cap straight away, I might not have scrutinized any of the schematics as closely. Now I have greater appreciation for how these mics work and for the technical evolution from version to version. I can't tell you which one sounds better, but the older styles definitely have a softer overload characteristic.