The resistor you talk about may also regulate harmonic distortions in some designs.
The best thing would be finding a schematic with expected voltage points so you could pin point the probleme.
If this model has a transformer, a drop of level can come from there too.
No expert here, I'm just starting to be more aware of those things after builing a microphone-parts.com kit...

PhantomBox, post: 445627, member: 29061 wrote: I have an AKG C414B-ULS with low output.
Please correct me if I'm wrong: these condenser mics have FETs that need to be calibrated to a certain voltage that polarizes the capsule. If this voltage is too high by just a few volts, it looses low frequency response, and if it's to low, the output is reduced drastically.
There are 1 or 2 FETs that regulate this voltage with a fixed resistor between gate (G) and source (S).
This resistor(s) might have drifted from its original value.

I have the following questions:
Which FETs do I have to check and what is the voltage that should be present?
Does anyone know how to take the PCBs off the frame in order to get to the underside for measuring? I don't want to force anything without wrecking the mic.

There are a number of things that are not correct in your post.

Firstly, you need to separate the concepts of polarising the capsule from buffering the audio output. In a conventional capacitor (condenser) mic, the nominal 48V phantom power (PP) determines the voltage across the transducer element whose capacitance value varies with the air pressure. Since the amount of charge Q stored on the capacitor stays approximately constant, the instantaneous voltage V across the element changes in inverse proportion to the capacitance C according the to formula Q = CV. Amongst other things, the polarisation of the element allows the instantaneous voltages to go negative relative to the mean value. The design figure for mean polarisation voltage at the capsule can be in the range 30 -50V. Allowing for voltage drop across the PP resistors at the mic socket due to supplying other operating current inside the microphone, the PP supply voltage is normally specified as 44 - 52V, i.e. a tolerance of +/-4V on the nominal value. It's important to note that the designers specify the performance characteristics of the microphone over this full 8V range without any loss of low frequency response that you allege. There will indeed be some change of output voltage per unit of pressure input according to the formula as the PP voltage is changed, but it's measured in centiBels (tenths of a dB) and goes normally unnoticed in a recording setup.

Note that most valve (tube) condenser mics use higher polarisation voltages derived not from PP but from a separate power supply. However, a large proportion of condenser microphones sold these days employ permanently-charged elements (e.g. electret), and so the analysis above does not apply to them, and the PP is used only for powering the output buffers.

Secondly, your strange tale of FETs regulating the polarisation voltage by using a gate-source resistor. There certainly has to be a resistor connected to the floating plate of the capsule, else it would never get charged, but the resistor value is not that important as long as it's in the GOhm range (10^9 Ohm), since, once the capsule is charged, it's only the signal current that flows in it plus any small leakage current. This means that drift in the resistor value is immaterial as long as it does not change by more than something like an order of magnitude. The FETs play no part in the regulation of this voltage, as their use is as impedance converters to buffer the GOhm impedances at the capsule to the 100 Ohms or so needed to drive the output transformer or the cable. The drain voltage on these buffer FETs varies across designs: some use the full 48V PP voltage down to those that can run from a 9V battery.

While some microphones have complete differential-output amplifiers in them, others use simple FET buffer designs. There is no concept of FET calibration, although a few microphones I have had in for repair do have adjustment pots to set the d.c. operating levels.

Unless you really know what you are doing, I would strongly advise against your removing the circuit board from your C414.

Hello Boswell.
So it seems that I got it completely wrong, way off target. Maybe I should have googled it first. On the other hand, I wasn't able to find much on this particular AKG mic anywhere on the internet.
I got the information from a friend who happens to professionally fix vintage Neumann U47 / U48 / U67 / U87 (and any modell between) mics.
In the past, I send him an AKG C414EB with a messed up capsule, which he managed to rescue.
Unfortunately, he lives in a different town. When I spoke with him on the phone, I could hardly understand what he said.
And to be honest, I'm no mic guru, either, hence the "... correct me if I'm wrong...".
There is no special "vintage microphone repair guy" where I live (or even in this country), and those guys who say they do repair those mics, probably also are veterinarians...
That being said, do you have any idea what could be wrong with the C414B-ULS? Maybe the at the output transformer or the other trafo (input matching?).
You seem to know a lot about these mics, vintage or modern.

Sorry that I was a bit heavy about what you said in your post and didn't comment much on your actual question. It's usually better to present all the detail of your problem first before describing what you have tried to do to fix it.

Nowhere do you say what you mean by "low output" of your C414. Is it, for example, 6dB less than expected, which could be accounted for by a poor connection or an open-circuit transformer winding affecting one of the output signal lines?

If the output is considerably more than 6dB below the expected level and is also distorted, then it's likely that the fault is elsewhere in the circuit.

There have been many different internal circuit designs in the history of the C414. The C414B-ULS has a FET+transistor amplifier driving the output transformer, together with a more complex section for delivering elevated voltages for capsule polarisation; these are changed around depending on the settings of the attenuator and the pattern switches.

A failure of the dc-dc converter section could be a cause of low output. If you have access to the rear of the attenuator switch, you could check the voltages on the pins. Take great care, as a casual short circuit of the multimeter probe to the microphone case while probing these would likely destroy the dc-dc converter if it's working.

Hi Boswell.

The output is reduced by more than 6dB, but without distortion, as far as I can tell.
On the last page of the service manual it says the following:

From point "6" to point "2" (ground) I measure 24VDC (52VAC?). R117 is covered with resign. R09 is a trim pot with a red dot on it to prevent changing its value.

Do you get the same low audio output in all the polar patterns?

R9 and R117 are d.c. level adjustment pots and I should avoid changing their settings. As long as you are using a high-resistance multimeter, the voltage reading you report of 24V is well below the 62V expected on the capsule when switched to cardioid pattern. It may be what would be expected if measured with a medium or low-resistance meter.

There are three diodes in series in the charge pump section: D101, D102, D103. It would help diagnosis to know the d.c. voltage reading on the 4 nodes of this chain, although the cathode of D101 will show the 24V that you measured previously.

D101 does in fact show 24V on its kathode side. D102 has 0.9V and D103 has 8.0V. The anode of D103 is on the underside of the PCB, and up to now I con´t get to it.

I meant to ask in my previous post: what make and model of meter are you using to measure these?

I use a UNI-T made UT33C model, made in China. I use it for anything, from professional audio to automobiles, household appliances to 110/220VAC power installations. I never had a problem with it before.

After a bit of hunting I found a specification of that multimeter. It shows an input resistance of only 10M Ohm on the d.c. ranges, which is not high enough to get reliable readings when dealing with the charge pump circuits found in a condenser microphone. You would probably need a meter with at least 1GOhm resistance so as not to affect significantly the voltages you are attempting to measure.

With that in mind I'm temped to say that the voltages are likely to jump to their correct values when the meter is removed, so we should look for a problem elsewhere. Are you in a position to do the check I alluded to earlier about whether the low output problem is the same in cardioid and omni patterns?

Voltage readings in cardioid and omni are exactly the same.

Yes, the voltage readings would be the same, but what about the low output problem?

In omni the out put is actually lower than in cardioid hypercardioid or figure 8. Nevertheless, I have to set the input gain at 12 o'clock to hear something coming out of it, nothing compared to a Neumann TLM170