Dear colleagues:

I have been proposed to develop an audio system to implement on an airplane (will never fly) for education purposes. I mean crew training and so on. To make it as real as possible they have given me some original handsets that have been removed from retired planes but are fully operational.
The perfect way to do it, is to use the original set of equipment (not only handsets but central unit or whatever it uses to work as a whole system) but for some reasons this is not an option, and that is why they have contacted me for.
I have found some information and have been able to make them work. They are state of art yet simple to understand. They have an audio line for the headphone and a PTT line to open the mic when speaking but the problem comes next.
The audio signal with the voice message from the mic comes on the same wire as the DC supply, also, they use this same wire for transmitting the DTMF signal to set the communication way, I mean, cockpit to PA, or the FWD attendant, etc. I have been able, as mentioned, to make all this work but as an audio technician that after more than thirty years have been using shielded lines, differential (balanced) lines,etc it is a bit weird to see how all this audio signal transmission is made through a non shielded wire that, also, carries other signals and voltages.
At this point, there was no trouble to get the mic signal from the dc with a decoupling capacitor and a two stage inverting amplifier, to make it work at a proper level but the problem is here, there is a noise, let's say a white noise, on the background.
I have tried to isolate it but I am a bit lost here. I have, for sure, found that it comes from the handset but I wonder if there is any solution to fix it. I have flown many times and I can't say the voice message from the captain or stewardess is a HiFi signal, but can't remember they carry this background noise.
I hope you can understand my explanation, sorry for any mistake in my writing, and if required I can give any extra information required.

Thanks

 

I don't know what constraints you have as to whether it must be analog signalling.
I would replace the analog signal with 1-bit digital signalling called delta modulation.

 

At this point, there was no trouble to get the mic signal from the dc with a decoupling capacitor and a two stage inverting amplifier

Are you sure the noise is not from this amp?

 

pull out signal generator and scope. compare input and output.

 

Are you sure the noise is not from this amp?

Thank you for your reply.
No, the amplifier does not generate the noise itself. As soon as I disconnect the handset the noise disappears. Also, as I have three different units, the noise level is different from one to others. I am trying to find a better way to couple both the handset and the amp. I have some ideas rounding my head, if any of them makes any improvement I'll publish it.

 

I don't know what constraints you have as to whether it must be analog signalling.
I would replace the analog signal with 1-bit digital signalling called delta modulation.

I am trying to keep it as simple as possible. Anyway I will find info about what you mention. At the end, the noise is not a problem I can't assume. But i'd love to get the better results as possible.

Thanks for your reply.

 

What type of microphone is in the handset? Is there any electronics? What does your input circuitry look like. Perhaps it is not properly matched to the microphone.

 

Welcome to AAC.

Can you provide the make and model of the interphone handset you are using, along with a sharp, will lit photo? That would be very helpful.

 

What type of microphone is in the handset? Is there any electronics? What does your input circuitry look like. Perhaps it is not properly matched to the microphone.

Thank for your reply
I will, soon, add the schematics of the handset I have. The ideas I have, and most agree or point to, is about improper impedance matching or something about.

 

Do you need me to draw the schematics of what you need?

Its noisy probably because you are not at the correct impedance.

Edit: Attached is the typical design. You should put in there a telephone matching transformer (about $2) Because the intercom phones rely on this transformer that is usually located in the cockpit intercom panel.
View attachment 315067

Do you need me to draw the schematics of what you need?

Its noisy probably because you are not at the correct impedance.

Edit: Attached is the typical design. You should put in there a telephone matching transformer (about $2) Because the intercom phones rely on this transformer that is usually located in the cockpit intercom panel.
View attachment 315067

Thanks for your reply:
Well, yes we all agree, in similar ways, where the problem may come from. I appreciate your information. I did try with an impedance trafo previously, as you show, but in a simpler way. I will try your design and let you know.
I have found, that not only filtering the signal, which is obvious, but also, improving the dc filtering to better than usual standards, improves the noise level (defeat it) in a very reasonable way.
As mentioned before, this is a shared way to both provide DC and audio+dtmf signals, so, if there is any noise on the DC this will, then, affect to the sound quality.
On the next post I 'll attach the schematics of the handset.
Again, thank you four your support

 

This is the block diagram

And this is the schematic diagram

 

The background noise seems to be generated in the headset, So if the microphones are electret types there can be a spread in noise outputs, because the internal FET amplifiers do vary a bit. If they are carbon microphones , they vary and can sometimes be cured by gentle bumping to break up carbon granule clumps.
The external fix will be a "noise-gate", which works a bit like a volume expander, except that the gain is lowered for inputs below some level. Noise gates are used in some sorts of PA systems to do just exactly that, which is to hold the volume/gain low until an intentional signal is present.
This is not a cure for the noise, BUT IT IS a solution to the audible noise problem.
I have not yet looked at the circuit of the original amplifier to see if it has an internal noise gate. That would be the circuit to copy, if it is included.

 

THe TS seems to have ignored my request for make/model and photos of the handset. The system is called “interphone” and depending on its vintage, there is a non-zero chance the handsets are using a carbon transmitter like a telephone instrument handset.

Carbon elements have a lot of self noise, some of us will remember banging the handset against something (a hand, a desk, a coworker) to stop the hissing and increase sensitivity by freeing the carbon granules—particularly on older sets. If it is a carbon type, appropriate bias is needed to make it work well, and impedance matching may be required.

 

It sounds to me like it is using some other type of mic but emulating the interface requirements of a carbon mic for backward compatibility.

 

It sounds to me like it is using some other type of mic but emulating the interface requirements of a carbon mic for backward compatibility.

Yes, that could be. And many later hand/headsets used electrets with impedance matching. But the classic interphone system used a handset like a standard telephone with a PPT switch and a ¼“ 3-circuit plug.

very vintage (L), vintage (R)
​

The cost for these parts varies wildly depending on whether the fit current planes. For example, the left-hand part is about 50 bucks but the right-hand part is for a 747 and it will cost you about 500 bucks for a used, working part.

 

OK, "Y" concurs regarding the carbon mic noise and recoveries. Certainly a noise gate function would reduce the problem no matter what the source. "Noise Gate" was part of a speech compression that AD offered for a while, not sure if it is part of the present product line up. But certainly it could be achieved with a classic "volume expander" circuit just by adding a clamp to the expansion control line. (This might be going far over the heads of a lot of folks) But the concept is certainly valid. Really, it could use a selective filter on the noise signal to be suppressed.

 

Attached is the typical design. You should put in there a telephone matching transformer (about $2) Because the intercom phones rely on this transformer that is usually located in the cockpit intercom panel.

In your schematic, you can eliminate R1 and C2. They form a Zobel network that is needed only if the source amplifier has a phase margin problem. Also, since both sides of the transformer are referenced to the same ground, I don't see any need for it.

ak

 

In your schematic, you can eliminate R1 and C2. They form a Zobel network that is needed only if the source amplifier has a phase margin problem. Also, since both sides of the transformer are referenced to the same ground, I don't see any need for it.

ak

A transformer should provide isolation, and long wires often pick up common mode noise that an isolation transformer will block completely. But of course the incoming side must NOT be tied right to the common.

 

One might think that but its a filter for 400Hz full wave rectified power suppression in the audio.

10 ohms and 0.047 uF is a corner freq of 339 kHz, so ... no.

A high-frequency Zobel network like that is common (and often required) for just about all IC audio power amplifiers, from the small LM386 up to the higher-power TDA parts. It should be located close to the device output pins.

ak

 

The preamp is AC coupled which rules out an electret Mic.
Could it be a magnetic mic?

The audio amp looks fuzzy to me. Noise is most likely from the PS.