How to loopback MIC and Earpiece?

[tony_lth]

ALL,
I had a design, which has a 16 ohms earpiece output, and a several kilo-ohms MIC input.
Now I want to loopback MIC and Earpiece so that I can test audio quality by CMU200 etc.
The question is how to loopback MIC and Earpiece? Directly linking MIC+ to EAR+ and MIC- to EAR- seems not impedance match, so how can I do?
Best,
Tony Liu

 

[Audioguru]

You NEVER connect the output of an amplifier like the MAX9719 to its input. It would probably cause oscillation. Instead you feed an electronic signal into the amplifier's input and measure the amplifier's output.
The datasheet for the amplifier shows extremely low distortion below output clipping and a wide frequency response so audio quality should be excellent. The audio quality will be affected by the mechanical mic and earpiece transducers.
Most audio circuits NEVER match impedances. Old vacuum tube amplifiers matched impedances because they had input and output transformers.
A 600 ohm dynamic mic feeds an amplifier input impedance of a few k ohms. An 8 ohm speaker is driven from an amplifier with an output impedance of 0.04 ohms or less.

A microphone amplifier usually has a gain of about 100 times but your circuit has a gain of only about 8 times. Is the gain enough?

 

[tony_lth]

Hi, Audioguru, and ALL,
Thanks.
The amplifier gain is enough, because after the amplifier there is a modem with CPU to process it.
With live test, the audio quality is very good on a GSM link.
But I want to loopback the MIC and Earpiece together so that I can test GSM RF impact on audio.
How can I do it? Should I use a transformer, like TR84, and how to link TR84 to MIC+- and EAR+-?
Best,
Tony Liu

- - - Updated - - -

I guess the link as following for TR84:
pin 0-->EAR-
pin 8Ohm-->EAR+
pin T-->MIC-
pin 3W-->MIC+
Shell-->GND
Right?

 

[Audioguru]

You show a TR84 transformer that is designed to change the 100VAC from the output of a PA amplifier to 4 ohms or 8 ohms so it can drive a speaker. It cannot be used for your "loopback" test.

A "loopback" test is never done with an audio amplifier circuit like the MAX9719 amplifier because it causes oscillation.

 

[tony_lth]

Hi, Audioguru and ALL,
I found some solutions for my requirement of audio loopback:
1. Setting your Sound Card as a loopback device, http://www.nch.com.au/streaming/kb/1079.html
2. Using Microsoft Loopback Adapter driver: http://driversupdatecenter.net/drivers/?driver=Microsoft+Loopback+Adapter+Driver&p=28
3. Using the following PDF.
Which solution do you think works for the loopback of #1 post circuit? Or any other comments?
Best,
Tony Liu

 

[Audioguru]

You said you want to test "GSM impact on Audio". Then feed a signal into the mic input then measure interference on the earphone output.
Then with no audio at the mic, measure interference on the earphone output.
Loopback is not used.

 

[tony_lth]

Hi, Audioguru,
GSM test station, CMU200, that is an instrument developed by RS, has a audio test function, which can test some audio performance, such as THD, etc.
So what I need is loopback the MIC and earpiece, how to do it?
Best,
Tony Liu

 

[Audioguru]

The instructions for the CMU200 never say "loopback".
To measure THD on a radio system you feed in a pure low distortion sinewave to the transmitter and measure the amount of distortion at the receiver. Since the receiver is far away then they can record the received signal and mail the recording back for you to measure. The mailing is what you call "loopback". You could also use a telephone line or another radio signal for you to get the received signal back to you (you will call it "loopback") but the return distortion will be added.

Maybe you want to transmit a signal, it is received at the far end then the received signal is transmitted back to you? Then you measure the distortion for the round trip of the signal. You can call that "loopback".

 

[tony_lth]

Hi, Audioguru,
Telephone is a good option, so what I need is:
1. Link my GSM module MIC and Earpiece to a telephone.
2. Put that telephone in an amber with noise reduction.
3. Setup a gsm RF link from my module to CMU200.
4. Test the THD or other audio performance in CMU200.
Right? Any comment?
Best,
Tony Liu

- - - Updated - - -

Hi, Audioguru,
Telephone is a good option, so what I need is:
1. Link my GSM module MIC and Earpiece to a telephone.
2. Put that telephone in an amber with noise reduction.
3. Setup a gsm RF link from my module to CMU200.
4. Test the THD or other audio performance in CMU200.
Right? Any comment?
Best,
Tony Liu

- - - Updated - - -

PS: And I need to use handfree speaker so that CMU200 audio signal can be picked up by telephone MIC.

 

[Audioguru]

A telephone line has horrible audio. Its frequency response cuts low frequencies of voices and cuts high frequencies of spoken consonants. Its distortion is pretty high but its poor high frequency response cuts distortion harmonics.
Many telephone lines and telephone systems have an echo canceller that prevents a received signal from being transmitted back to where it came from. A cheap handsfree speakerphone uses half-duplex where it cannot receive and transmit at the same time.
Your idea will cause nothing or howling feedback.

 

[tony_lth]

So, Audioguru,
Could you suggest me some solution?
Best,
Tony Liu

- - - Updated - - -

PS:
My solution is using an audio transformer, to convert Earpiece several Ohms impedance to MIC K-Ohms.
SO how about 140NEX?

- - - Updated - - -

PS: or 119DA:
Built in response to requests from the "Collins Collectors Association" to match "classic" high impedance audio out
puts to modern low impedance speakers.

 

[Audioguru]

You do not understand feedback:
1) A sound goes in the mic and is amplified.
2) The amplified signal comes out the amplifier output and through a transformer (you do not need a transformer, use two resistors as an attenuator) to the amplifier mic input.
3) Then the signal goes around and around as an oscillator.

To measure the distortion (THD) then feed a low distortion sinewave into the mic input and measure the distortion of the earpiece output at the other end. No loopback and no transformer.

 

[tony_lth]

Sorry , Audioguru,
I didn't show you the whole design of my GSM module.
Its SCH as the following:
MIC+/- -->MAX9719 Amp--> GSM Modem--> GSM RF--> Antenna -->Radio Link --> CMU200
Earpiece+/- <--MAX9719 Amp <-- GSM Modem <-- GSM RF <-- Antenna <--Radio Link <--CMU200
Now CMU200 can test audio performance, so I need to loopback MIC+/- with Earpiece+/-, and CMU200 can test THD etc.
It can't self-osc when loopback MIC+/- and Earpiece+/- by some way, that is audio transformer what can match impedance between MIC and Earpicec.
Right?
Best,
tony Liu

 

[Audioguru]

Audio impedances do not need to be matched. They were matched 60 years ago when vacuum tubes were used because the output of a vacuum tube is a high impedance and a speaker is a very low impedance so a transformer matched them. The output impedance of a modern transistorized amplifier is 0.04 ohms or less and it feeds a speaker that is 4 ohms to 32 ohms. The output of an RF transmitter is matched to the antenna impedance for maximum power transfer without reflection.

If you match the input of a preamp to the impedance of a mic then the signal level is reduced to half. Therefore if the mic is an electret type with an impedance of about 2k ohms then the input impedance of its preamp should be about 20k ohms or higher to avoid throwing away half the signal level.

The MAX9719 is a stereo power amplifier with a voltage gain of only 2. Therefore its output level will be almost nothing when it is fed with the 0.01V signal from a mic. A mic preamp is needed with a voltage gain of about 100.

 

[tony_lth]

Hi, Audioguru,
I directly linked MIC+/- pins to earpiece+/- pins, and test audio with CMU200, the results are the same as MIC not linking to earpiece.
But with a live MIC and earpiece, I can talk to a mobile user with high audio quality, so it proves that my MIC and earpiece design is good enough.
And it proved that if linking MIC to earpiece, the impedance must be considered.
Best,
Tony Liu

 

[tony_lth]

Audioguru,
From my design, my MIC gain is 20log(45.3K/10K)=13dB.
Why you said my gain is only 2?
Best,
Tony Liu

 

[Audioguru]

The datasheet for the MAX9719D says that its gain is 6dB. Maybe you have a MAX9719A that has adjustable gain?

 

[tony_lth]

Yes, I used MAX9719A. And in #1 post, the circuit shows the adjustable gain design. If it's a fixed gain, the circuit should be different.

 

[tony_lth]

Hi, ALL,
I finally found the standard to test it.
The test system is very complex, but it is very clear how to do it.
Thanks.
Best,
Tony Liu

 

[c_mitra]

In good old days, the microphone signal was also fed into the earphone to control the sound level of the speaker in a telephone conversation. That acted like a simple AGC for the input microphone level. That is not possible with modern microphones and earphones and the two signals must be mixed in a mixer (a simple opamp can do) so that you know what you are talking about. With a simple op-amp you can accurately control the proportion of the two signals. A good op-amp is not really expensive and the phase shift should be small to prevent unwanted oscillation. Modern telephones still feed a part of your speech into your own ears!