Basic microphone troubleshooting techniques

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Measure DC voltages at non-high-impedance points like capsule bias voltage dividers (R8/R9 junction below). When measuring DC at a capsule, tube grid, etc, your voltmeter will load the circuit, giving incorrect voltage readings. Most voltmeters have a 10M Ohm input impedance which will heavily load a 100M or greater high impedance sections.

ExampleMeasuring M49 voltages.png
 
Most of the vintage schematics i've seen, quite clearly indicate the expected voltages at the few important nodes around the circuit. And that's neverminding the... let's call it, forgiving nature of tube mic circuits.
 
Most of the vintage schematics i've seen, quite clearly indicate the expected voltages at the few important nodes around the circuit. And that's neverminding the... let's call it, forgiving nature of tube mic circuits.
But since the pcb's different people make often divert from the original schematics, it possible that the voltages might differ?
 
Pcb's only mechanically support the components. In theory at least, even if you soldered the circuit together "in 3D" / deadbug style, with no pcb, as long as you used the same component values, the voltages you measure should be largely the same (allowing for component tolerances, of course).
 
Measuring Signal to Noise ratio with a DAW and a test tone (relative).

Connect mic to low noise preamp. Compare the tube mic to a dynamic mic as reference, like SM57.
Use a constant volume test tone, such as a 440 Hz test tone for tuning, a set distance from the mic (~10"). Adjust preamp gain. Measure level in DAW. This is signal level.
Turn off test tone (and any other background noise: quiet room) and measure level in DAW. Do not change the preamp gain. This is noise level.
SNR = level/noise.
High noise can be from the tube, improper grounding, or contamination of the high impedance areas.

When I did this with my tube mic builds and some dynamic mics for reference, using a low noise preamp (appx. -120 dB EIN), I measured:
SM57: SNR = 27 dB
SM7: SNR = 22 dB
Tube mics: SNR = 18-22dB (tube noise dominated - builds had proper grounding).
 
Pcb's only mechanically support the components. In theory at least, even if you soldered the circuit together "in 3D" / deadbug style, with no pcb, as long as you used the same component values, the voltages you measure should be largely the same (allowing for component tolerances, of course).
It's not the pcb itself, but the components.
 
I think what they mean is different mics have different components and designs - so measuring the voltages for a M49 might not apply to a Elam.
Comparing the schematics, the main differences are:
1) in the Elam the tube is cathode biased with a capacitor. The M49 stabilizes the bias with the heater current. Both should have a grid-cathode voltage of a few volts. As shown above for measuring the M49, for the Elam, measure the grid-cathod voltage from cathode to the other side of pull down resistor (ground)
2) the 251uses a dual triode, so half is grounded.
3) capsule pattern switching

But overall, tube mics are highly similar, so checking the DC voltages as above is applicable.
 
I'm... not quite sure what you mean now.
In two different u67's I have, there are different values of resistors, one pcb has 150m, another has 400m as an example, one U67 has many parts, another has lesser.
 
Last edited:
Those uber-high value resistors shouldn't (and likely won't) affect anything you can measure without really specialized equipment. Case in point, one recent thread just here:

https://groupdiy.com/threads/neumann-m367-polarization-mystery.82294/#post-1063069
But if the (other) components used differ greatly from an already established circuit schematic, then all bets are off.

Do those two U67's of yours sound dramatically different? Do they have the same capsule, or different ones? Are the differences in component values restricted to those over-100meg ones, or others as well?

PS. "Pcb" usually refers to the actual fiberglass substrate and copper traces, not to any particular set of components / values that may be installed.
 
Maybe an obvious for all of you, but when doing my first built it cost me quite a bit of time and patience ;-):

when trying out the mic i measured voltages and everything was fine but when connecting the mic it was quite noisy. tried a lot and checked grounding and tubes before i realized that just closing its metal body did the job.

to my little knowledge, the grounded metal body prevent rf interference of the circuit. but for sure there are tons of people her who can lay out a better and more scientific explaination.

anyway - maybe it spares a few minutes of your trouble shooting time … 🙂

best regards
michael
 
Measure DC voltages at non-high-impedance points like capsule bias voltage dividers (R8/R9 junction below). When measuring DC at a capsule, tube grid, etc, your voltmeter will load the circuit, giving incorrect voltage readings. Most voltmeters have a 10M Ohm input impedance which will heavily load a 100M or greater high impedance sections.

Example
Measuring M49 voltages.png
As I was recently reminded in that recent thread about Neumann M367, don't get fooled by those "non-high-impedance points" !
1 MΩ is indeed much lower than 100 MΩ or higher, but still, 1 MΩ is merely a tenth of a standard 10 MΩ input impedance. In this exemple, if you measure the "backplate 58V" accross R8 with a 10MΩ multimeter, you'll read only 55V...

Axel
 
Wear an earthed ESD-strap?

Other than maybe JFETs, and perhaps less so BJTs, there's not much that's ESD-sensitive in most microphones.
 
As I was recently reminded in that recent thread about Neumann M367, don't get fooled by those "non-high-impedance points" !
1 MΩ is indeed much lower than 100 MΩ or higher, but still, 1 MΩ is merely a tenth of a standard 10 MΩ input impedance. In this exemple, if you measure the "backplate 58V" accross R8 with a 10MΩ multimeter, you'll read only 55V...

Axel

Might even be wise to quickly whip up an LTspice sim with the involved resistors and a 1M or 10M resistor (see own meter's input impedance specs) added between the relevant node(s) and ground, to get a better idea of what readings to expect, skewed as they may be.
 

Latest posts

Back
Top