Can someone propose a solution to share this type of headset microphone (most likely electret condenser) with multiple devices without creating ground loops. Headset example: https://www.amazon.com/Sonku-Isolation-Headphones-Microphone-Cellphone/dp/B081ZZ63RL

I know there are isolation transformers but they are mostly for very low impedance input (150-300 ohm). Electret mic has typically 2200 Ohm impedance. For example this one, could it be used? https://www.jensen-transformers.com/wp-content/uploads/2014/08/jt-mb-cpca.pdf

 

Your diagram really does not show what you are trying to do. Do you want to connect the headset and mic to three different devices at the same time? If so, why?
How do you propose to supply the energizing source for the mic, and especially if you use a transformer?
Regards,
Keith

 

There is one headset, and three devices. The operator selects on control panel (switches) where to send the microphone signal.
The control panel and also mic is powered off the same source as one of the devices (lets say radio1).
An analog switch is used to route the audio to different ouputs. IE MAX4738.

So I need to send isolated signal to radio2 or PC without making any modifications to the radio2 or PC hardware. it needs to see this signal as if regular microphone.

For simplicity lets omit speakers for now.
EDIT: I updated above schematic

 

It would simplify the task if you used a buffer op-amp for your microphone with an integrated supply for the mic. Then all you have to do is route the mic and phones signals with the analog switches. Switching the mic signal along with its energizing supply could get a little complicated and unpredictable.

 

So I should add opamp for mic, ok sounds reasonable, but do you mean I should ditch the transformers? How to acheive isolation then?

 

Why do you need isolation? It is not usually required on a setup like this. Just use shielded conductors.
The analog signal voltage to be switched must be within 0 to VCC (5V?). What is the amplitude of the audio signal to the phones?
It may exceed the limits. If it does, you will need to attenuate each input and then amplify the signals after switching to drive the headset.

 

Well the fact is I don't know if isolation is needed or not, but the system needs to work the first time, there is absolutely no way to test every possible situation beforehand. There may be some noisy equipment installed on the power lines. Additionally it turned out the radio 2 mic minus is not grounded at all, so I would be making an extra GND connection to existing system which in this case is not acceptable.

 

The audio signal should be around 2.2Vpp. But the speakers will not be switched. All speakers will be mixed together.

Is there a standard at witch voltage level the audio signal should be to feed the microphone input? And at which output impedance?

What should be the parameters of transformer (turn ratio, impedance...)

 

A domestic line input electrical signal typically has a voltage ranging from 0,3 to 2 Volts rms, while a microphone level signal is more often in the range from 5 to 50 mV (millivolts) rms. A typical line out connection has an output impedance from 100 to 600 Ω, with lower values being more common in newer equipment. Line inputs present a much higher impedance, typically 10 kΩ or more.
The signal isolation transformer windings ratio should be 1:1 as it is just isolating the signals and not modifying the voltage levels.
Regards,
Keith

 

A better way is to treat the ground loops and eliminate them.

Start with individual, heavy ground wires from each device back to the - battery terminal.

Next, run a separate power wire to the microphone and regulate/filter it locally.

Next, use a double-pole switch to connect the mic signal and the mic ground to each device.

ak

 

Ok my design is coming along and I did a test circuit on a breadboard and it seems to be working fine.
Thanks for the suggestions so far.
One question that I have is how much AC current can I load this transformer with?
http://catalog.triadmagnetics.com/Asset/SP-70.pdf

There are some confusing parameters. Is it correct to calculate the max current from the given "power level" of 50mW?
So if I output a signal at 1V rms, the load current at 50mW would be 50mArms?
So making the minimum allowed load resistance 20 ohms?

 

Yes, but without more information, I'd say that is the current level at which the transformer core starts to saturate. For clean audio, you want to stay below 10% of that.

Please post your latest schematic.

Are you sure that you need 2.2 Vp-p? That is nowhere near microphone level, by about 500x.

ak

 

Are you sure that you need 2.2 Vp-p? That is nowhere near microphone level, by about 500x.

Well that is what I measured at the mic output when placing a smartphone with 1kHz sinus tone generator close to the microphone. It started clipping at 2.7Vp-p. Im not sure how that relates to human speech though.

Well here is the microphone side of schematic. A little background info: The headset is aviation headset with about 200ohm output impedance. The load will be connected to input varying from 150 ohms to 1k ohms depending on equipment used. The R5,6,7 was used to make output impedance match the headphone output to perfectly simulate headphones.
The signal is attenuated a little first (RR3, R27, R28 and R32/33/34) G=0.8. Then the potentiometers are used to adjust the gain to fine tune output level.(so you cold attenuate/amplifi the signal using the same pot)

So my original concern was if a short would cause any damage to the transformer?(also other comments are welcome)

 

R40 must be returned to the 4.5 V reference, not GND. That is the DC reference potential for IC1A. As it is, the output will try to make +18 V DC plus some audio.

Same for R3 on the output. It shifts the audio centerline away from 4.5 V to 3.0 V, or 1.5 V below VGND. The three output amplifiers will try to amplify this and saturate.

ak

 

Yes, well noted, this was the only part I had yet to test on breadboard. I have actually revised the level adjustment. which gives me better range. 2x amplification is now done with IC1A and then later attenuated with pots if needed.
[reuploaded schematic]

 

If you use linear pots for R29, 30, 31, you will get pseudo-audio-taper performance.

ak

 

And here is the mixer part(X2, X3. X4 is connector for volume pot). However there is a problem with the TPA6120 part. When I load the output with 16 ohm resistor the bottom of the signal is clipped. All should be as is in the schematic and according to datasheet. Any idea why this happens?

Unfortunately I only have single supply to work with, could it be power supply issue?