I've been working for quite a while on a portable amplifier. I want to take audio from the headphone jack of a phone, connect it to an intercom system, and run it for 10+ hours. When amps are connectted directly to the headphones they work. When connected through intercom they fail - mostly. There seems to be some part of the intercom that is soaking up volume.

The intercom system is a black box...I don’t have/can’t get a schematic. Thankfully, the audio doesn’t actually go through the intercom on the way to the headphones I’m using. The headphone circuit essentially has a Y-connector. What I’m doing (any many before me with even less success) is sending an amplified signal into the headphones through the Y-connector while it is also receiving a signal through the intended route. Weird, I know. The other option is no music.

The headphones in question have two 19ohm speakers in parallel (~10ohm) at 90dB/mW. Without having my math in front of me I want to say that I need 100-350mW to power them at about 0.5A.

When testing on the bench (directly to the headphones), I can hear music from the phone through these headphones (faintly) without ampliphication. Off-the-shelf headphone amps increase the volume just fine.

In real life, installed through the Y-cord, the phone is inaudible - entirely. The off-the-shelf headphone amps (like a topping NX-1: http://www.tpdz.net/en/products/nx1/index.htm) are very faint. Class D amps tend to clip before they’re really loud enough to distinguish voices. Class AB tend to distort or fill with static. Granted, there is a great deal of background noise and none of these amps are really designed for this low of an impedance or sensitivity. Amps intended for 4 ohm or 8ohm speakers work well, but are too large. I’m also worried that they may somehow damage the intercom.

I had success AND failure with a cheap LM386 PCB with fixed gain of 200 https://www.petervis.com/Electronic...fier-module/lm386-audio-amplifier-module.html. With the stock 10k potentiometer acting as a voltage divider the amp was inaudible when installed. With a 50k pot, it was also inaudible. BUT, with a 100k pot the amp was the loudest, clearest amp that any of us had ever heard! I had 10 people asking for their own after a single use (the market is very small). When I attempted to use fixed resistors in the divider it failed again. Granted, the sum total of the resistors was only 5k. I had extremely limited resources while I was testing... I can’t figure out why the resistance of the voltage divider should matter this much since it is so much higher than normal headphone impedance. https://www.anandtech.com/show/8554/the-iphone-6-review/11 (Note: all of these configurations work on the bench - headphones only).

As long as I divide the voltage to <20% of that provided by an iPhone 4 (to meet the Lm386’s 0.4V limit) there is no distortion. As long as the resistance was high in the divider, the LM386 is audible. I could just use fixed resistors that added up to 100k, but I’d rather use a Class D amp to save battery power and unit size. Also, why is this an issue?

Does anyone have any suspicion as to what is soaking up my volume with the Y-cord setup? I’m trying to make a benchtop simulation so I can really nail down this design and create custom PCBs with a Class D amp and battery charging circuit together. Testing against the actual intercom is very difficult and infrequent.

 

Your explanation is difficult for me to follow, but is your goal just to have the audio playing through the intercom system? I don't know why you mentioned headphones so much if you're trying to play it through a different system.

If your goal is just to play it through the intercom system, how are you connecting it? You mention a Y cable, so does the intercom have an input jack? Do you know what kind of input is normally connected to it?

 

Yeah, sorry. Intercom is not a good description because there is only one person connected, but that is just what it’s called. It’s not a PA or anything.

So there is an intercom system (ICS) with a bunch of audio devices on one side (radios, computers, etc.) and a single person with a headset (headphones+mic) on the other. The side with the person connected has two output jacks connected in parallel. The jacks are supposed to be used separately. If they’re both used as outputs the volume drops by half.

The parallel connection is what we’ve been exploiting to play a separate source of audio. We’re basically using a y-connector as a summing box (bridged) - which I know is incorrect. But...that is what we have to work with.

 

I am reasonably sure that we can help you solve your problem. That is what we do here on AAC.

However, we have to be a bit more precise and concise with the communication.
Let us begin with the sound source.

What is the sound source, make and model? A photograph would help tremendously.
You are taking the sound signal from the headphone jack?
Is the output a mono or stereo jack?

 

The side with the person connected has two output jacks connected in parallel.
The parallel connection is what we’ve been exploiting to play a separate source of audio. We’re basically using a y-connector as a summing box (bridged) - which I know is incorrect. But...that is what we have to work with.

Sounds like you are attempting to connect the headphone amp to the output of the intercom amp.
SG

 

What is the sound source, make and model? A photograph would help tremendously.
You are taking the sound signal from the headphone jack?
Is the output a mono or stereo jack?

The source is headphone level from a phone or MP3 player. I use an old iPhone 4, but most people use their current smartphone.

(Wireless tech is not allowed and DACs would have to work with iOS and Android)

The headset is mono, so the source either outputs mono or is summed to mono. The signal into the Y-connector needs to be mono.

 

Sounds like you are attempting to connect the headphone amp to the output of the intercom amp.
SG

Exactly...

 

A mixer or combiner amp between the ICS jack and the MP3 player may work.This can be done with the LM386 module but will need some info on the wiring of the two ICS jacks.
SG

 

For a start, you cannot hook 2 outputs together. The iPhone headphone output needs to connected to the "intercom" input somewhere. Do you really just want to have an audio mixer running an amplifier so a number of inputs can drive a speaker?
Please post pictures and information on the actual devices you are trying to use. "Intercom" is not at all helpful!

Once again, I wonder why it is so hard to get some posters to give us real info. It does get a bit frustrating when we are trying to help but cannot get clear info. Sorry for the gripe but it is very early morning here and I have woken up with a chest bug and not much sleep. At least that is my excuse

 

This is what I propose but is dependent on the ICS jack wiring. You indicated that the user headset included a mic so that may be an issue, again depending on the wiring. Assuming it's a three wire jack then the mic connection can be passed through the amp as in the layout shown.
SG

 

The source is headphone level from a phone or MP3 player. I use an old iPhone 4, but most people use their current smartphone.

(Wireless tech is not allowed and DACs would have to work with iOS and Android)

The headset is mono, so the source either outputs mono or is summed to mono. The signal into the Y-connector needs to be mono.

Now that we have identified that the source is a smart phone, we'll consider the consequences and then look at the next step.

All phones I would assume output stereo. Hence you need to make connections to a stereo jack with the appropriate load impedance. I don't know how the Y-connector fits into the picture.

The next step is you want to feed the signal into an "intercom" system. What is the make and model of the intercom? What is the input to the intercom? A photograph would be very helpful.

Where does the headset come into the picture?

 

A mixer or combiner amp between the ICS jack and the MP3 player may work.This can be done with the LM386 module but will need some info on the wiring of the two ICS jacks.
SG

So, this is a great idea. It has been done with great success in the past. There are a two big issues, though...and you wouldn’t have guessed them because I’m being intentionally vague.

1) the connectors to tap into this connection cost $1200 each and you would need two (M/F). I’m trying to keep this under $20 production cost. You can get decommissioned connectors once in a blue moon but I’m trying to make ~100 of these.

2) it’s very difficult and risky to disconnect this connection. Difficult because it is hard to reach. Risky because not getting it reconnected means that you’re radio-out...and it’s very hard to reach.

I’ll try to clear a few things up...
-I’m going to use this on an aircraft. (More at the bottom of this post)
-The ICS has no input jacks.
-We can’t really inject a mixer between the ICS and Headset (cost and risk)
-the ICS as two parallel output jacks intended as an either/or connection (different plug types)

...
I’m fully aware that there is a potential safety issue if something were to cause a radio-out situation...it’s a burden for sure. Pilots were using audio devices connected to this jack before I started flying this plane (which I’m not going to name). Pilot’s have been doing it for years with no ill effects and we have a green light from the aircraft’s manufacturer. Needless to say, it is not officially approved or condoned.

I’m hoping to devise a single, effective, cheap solution so that pilots stop searching eBay and Amazon for amps with unknown effects. Not to mention unknown batteries...yikes! Recently a couple guys started using amps intended for 4 ohm cabinet speakers because those were the first to make a podcast understandable. I’m really concerned that the amount of power they produce is going to hurt something.

 

Well then as they say on Shark Tank: "and for that reason I'm out"
SG

 

For the same reason, I'm out of here.

Do you want this thread closed?

 

This is what I propose but is dependent on the ICS jack wiring. You indicated that the user headset included a mic so that may be an issue, again depending on the wiring. Assuming it's a three wire jack then the mic connection can be passed through the amp as in the layout shown.
SG
View attachment 148109

What did you use to make your block diagram? Was it on a mobile device?

 

For the same reason, I'm out of here.

Do you want this thread closed?

Well, sorry to scare people off. I was just trying to find some expertise out there.

I was really just hoping to find out if a resistor or capacitor on the ICS output might cause the behavior.

 

Thanks for the help, anyway. Should’ve stayed vague...

 

I think I finally understand what you are trying to do. If I am correct, here is what you need.

The headset must be plugged into a jack (A) on a black box. You need to separate the mono connection to the headset from the connection to the mic. The ICS jack must be plugged into a similar jack (B) on the black box. The mic signal is fed through from connector A to connector B.
The mono output from the ICS jack (B) must be terminated with the proper impedance.

The stereo output signal from the smart phone must be plugged into a stereo jack (C). The Left and Right channels of the stereo signal must be terminated with the proper impedance.

Next, you need to sum three signals

1) from ICS mono output
2) from smart phone Left channel
3) from smart phone Right channel

using three summing resistors and feed the sum signal into a mono audio amplifier. The mono output of the amplifier is fed to jack (A).

You probably need a level control setting on the smart phone signal as well as a mute switch to cut the muzak when desired.

 

What did you use to make your block diagram? Was it on a mobile device?

Windows Paint on my desktop.
SG

 

MrChips you are beating a dead horse here. I think you missed the part about the 1200 dollar connectors needed.
SG