How to wire an electret microphone properly?

[staticV3]

I want to properly wire a Primo EM272 electret microphone capsule (specs) to use with battery-powered as well as mains-powered (mic->interface->PC->mains) recorders.
Which of these cable layouts would be the correct choice? And Why?

Spoiler: Method 1: single-conductor cable with shield used as signal return

Spoiler: Method 2: two-conductor cable with shield disconnected

Spoiler: Method 3: two-conductor cable with shield tied to GND

Spoiler: Method 4: two-conductor wire with shield as second signal return

Thank you so much for your help!

 

[AnalogSteph]

My order of preference would be either 4 > 3 > 1 > 2 or 4 > 1 > 3 > 2. #2 has poor shield coverage, and #3 may be susceptible to RF ingress at the capsule unless you connect shield to mic case. #4 has lowest shield resistance but also somewhat higher parasitic capacitance than #1, I imagine.

I can see a few potential problems here:
1. While the capsule is quite hot, you would still want to have a quiet bias supply. How do you safeguard against devices with noisy bias (more common than you might think, onboard audio for one is notorious)? Right now you don't. I would contemplate adding a battery and some active circuitry, either in the microphone body itself or as an inline box. This could enable driving substantially longer lengths of cable as well. Somewhere I had also bookmarked an article on distortion compensation for JFETs that may be applicable to electret capsules...

2. What sort of interface will take a consumer electret? It would pretty much have to be a consumer soundcard rather than the typical USB audio interfaces for home studio use with their balanced XLR mic inputs. With extra electronics, you have a chance of accomodating both.

This being an omni capsule, I am guessing it is for some sort of measurement application? Not sure how the recorder would be used in this context though.

 

[Cbdb2]

Most mic inputs, including the recorder in your pics are balanced so none of the above. Use a standard mic cable. + to +, - to -, and shield to ground.

 

[DVDdoug]

to use with battery-powered as well as mains-powered (mic->interface->PC->mains) recorders.

Most stand-alone recorders have a "pro" balanced XLR connection (with optional 48V phantom power) and they are not directly-compatible with "computer mics" or an electret element. Computers have a 5V "bias power" for an electret.

"Real" condenser mics work from 48V phantom power. (Dynamic mics don't require power.)

There are some 48V phantom-powered mics made with an electret element and I assume they have an op-amp (or two) inside to get the balanced output. And, there may be a few battery-powered mics with balanced outputs. (Those were popular in the past when most mixers & PA systems didn't have phantom power but they are rare now.)

There are some "cheap" electret-based XLR mics that look like studio mics and I'd guess they don't actually have balanced outputs. Sometimes they claim they'll work with a computer or with phantom power but the specs are vague and I don't trust 'em.

 

[staticV3]

My order of preference would be either 4 > 3 > 1 > 2 or 4 > 1 > 3 > 2. #2 has poor shield coverage, and #3 may be susceptible to RF ingress at the capsule unless you connect shield to mic case. #4 has lowest shield resistance but also somewhat higher parasitic capacitance than #1, I imagine.

I can see a few potential problems here:
1. While the capsule is quite hot, you would still want to have a quiet bias supply. How do you safeguard against devices with noisy bias (more common than you might think, onboard audio for one is notorious)? Right now you don't. I would contemplate adding a battery and some active circuitry, either in the microphone body itself or as an inline box. This could enable driving substantially longer lengths of cable as well. Somewhere I had also bookmarked an article on distortion compensation for JFETs that may be applicable to electret capsules...

2. What sort of interface will take a consumer electret? It would pretty much have to be a consumer soundcard rather than the typical USB audio interfaces for home studio use with their balanced XLR mic inputs. With extra electronics, you have a chance of accomodating both.

This being an omni capsule, I am guessing it is for some sort of measurement application? Not sure how the recorder would be used in this context though.

Thank you for your preference list and other tips.
Since I've got everything else figured out, I wanted this thread to focus soley on what the best wire configuration would be to connect an electret to a 3.5mm mono plug. To that end, your list was already a big help!

 

[staticV3]

Most mic inputs, including the recorder in your pics are balanced so none of the above. Use a standard mic cable. + to +, - to -, and shield to ground.

The recorder in my pics is a PCM-D100, which uses a 3.5mm TRS Stereo single-ended microphone input.

 

[staticV3]

Most stand-alone recorders have a "pro" balanced XLR connection (with optional 48V phantom power) and they are not directly-compatible with "computer mics" or an electret element. Computers have a 5V "bias power" for an electret.

"Real" condenser mics work from 48V phantom power. (Dynamic mics don't require power.)

There are some 48V phantom-powered mics made with an electret element and I assume they have an op-amp (or two) inside to get the balanced output. And, there may be a few battery-powered mics with balanced outputs. (Those were popular in the past when most mixers & PA systems didn't have phantom power but they are rare now.)

There are some "cheap" electret-based XLR mics that look like studio mics and I'd guess they don't actually have balanced outputs. Sometimes they claim they'll work with a computer or with phantom power but the specs are vague and I don't trust 'em.

Thank you for your comments, fortunately I know all this already.
The aim of this thread was to figure out the optimal wire configuration between an electret and a 3.5mm TS plug. Nothing else. Perhaps I should've been clearer about that.

 

[EHC]

Method 3 is the best, and if you can extend the shield around the capsule without touch it is even better. The parasitic induced currents will flow mostly through the shield.
Method 2 is the worst, almost like connecting the capsule without shielded cable.
Method 4 will give you similar results as 1, but is a shame to not proper use the cable if it has two inner conductors.
You can get away with method 1 if you will use the microphone outdoor and the cable is relatively short, like 1..3 meters. But if the medium is heavy parasitic and/or the cable is longer, use the third method!

 

[staticV3]

Method 3 is the best, and if you can extend the shield around the capsule without touch it is even better. The parasitic induced currents will flow mostly through the shield.
Method 2 is the worst, almost like connecting the capsule without shielded cable.
Method 4 will give you similar results as 1, but is a shame to not proper use the cable if it has two inner conductors.
You can get away with method 1 if you will use the microphone outdoor and the cable is relatively short, like 1..3 meters. But if the medium is heavy parasitic and/or the cable is longer, use the third method!

Thank you very much for your insight!
I think I'm confident enough to start making a bunch of cables with configuration 3 now.

 

[jkim]

This is excellent information. I wonder if the same principle applies to unbalanced RCA interconnect cable wiring.

I have some 2 inner-conductor & shield cable laying around and want to make RCA cables. The two cores will be wired to the signal and ground sections, respectively, on both ends. The question is whether the shield wire is connected to ground on both ends or on the downstream end only? Which is better? Or doesn't matter?

 

[Speedskater]

The shield is the most important part of an RCA interconnect cable.
While the 2 inner-conductors can carry the signal,
It's the shield that shielded the noise & interference.
Just use a coax with a braided shield for RCA interconnects.

 

[Speedskater]

1] All microphone cables need to have the cable shield connected at both ends.

2] Semi-pro mixers & consoles need to turn OFF Phantom power unless the microphone requires it.

 

[jkim]

The shield is the most important part of an RCA interconnect cable.
While the 2 inner-conductors can carry the signal,
It's the shield that shielded the noise & interference.
Just use a coax with a braided shield for RCA interconnects.

1] All microphone cables need to have the cable shield connected at both ends.

2] Semi-pro mixers & consoles need to turn OFF Phantom power unless the microphone requires it.

Please read the OP's original question and thread. I asked a very specific question. I just want to make use of some 2-conductor cable w/ braided shielding to make RCA cables. A general rule, in this case, is to use one core for the signal and the rest (the other core and shield) for ground on both cable ends. My question is, what happens if the shielding is wired only to the downstream/input end, not to the output end? Better, worse, or doesn't matter?

 

[Speedskater]

For microphone cables:
If the cable shield is not attached to the microphone case, it may pick-up interference.

For RCA interconnect cables:
Always connect the shield at both ends!
The shield is the most important part of a RCA cable.

 

[jkim]

For microphone cables:
If the cable shield is not attached to the microphone case, it may pick-up interference.

For RCA interconnect cables:
Always connect the shield at both ends!
The shield is the most important part of a RCA cable.

In the particular case of the OP's situation, do you mean the shield should literally contact the microphone capsule' case? What the OP ended up doing was not that.

And may I ask WHY, in THIS particular case (unbalanced RCA cabling with 2 conductors & shield), the shield should be connected to both ends? Not just an explanation like "the shield is the most important part of an RCA cable." In this particular case, the shield IS connected to a ground except only to the downstream end..