Left and Right Signals Cancelling Each Other Out

filtrete867

Member
I'm working on fixing up a disused audio rack for the auditorium speaker system at my school. No one has used that system in years so nobody knows anything about it. On the rack there is a 1U mixer with about 7 or so inputs and an aux in. The mixer does not use standard connectors but rather has +,-,G terminal strips for each input.

The aux in is currently a 3.5mm stereo cable with one end stripped and wired into the aux terminal strip. I was running some tests on it the other day and noticed something odd. When I played some regular music through the aux in, there was not a problem. However, when I fired up audacity and played back a 440 sine wave, there was no signal on the mixer. When I panned the sine wave to the left or right in audacity, the signal came back. I left that first sine wave panned to the left and then created another 440 sine wave, panned to the right and offset 180 degrees from the first. The signal was back and louder than before.

Upon some further investigation, I found that the ground wire for that terminal strip had been pulled out.

Any ideas?
 
If it's really +/-/g that sounds more like a balanced input than a stereo one (and since you're putting the same signal into both parts of the balanced line, of course it will cancel, that means it's working). Fix up the ground and disconnect one of the signal lines on the "aux in" and the problem should go away, plus music should sound better :grin:
 
Sounds like you have a textbook case of a polarity reversal somewhere in the system that means the left channel is of opposite polarity to the right channel, creating the effect of a 180 degree phase shift but on a pure tone particularly the effect is one of complete cancellation.

If indeed the "aux" input is connected to a + - G input then do as cpf notes and change it's wiring.
However a number of mixers will have an aux input of L R G instead and usually surrounded by more normal balanced inputs just to make sure you get properly confused...
 
If it's really +/-/g that sounds more like a balanced input than a stereo one (and since you're putting the same signal into both parts of the balanced line, of course it will cancel, that means it's working). Fix up the ground and disconnect one of the signal lines on the "aux in" and the problem should go away, plus music should sound better :grin:

While this will correct the phase cancellation issue it will also eliminate either the left OR right signal. (this is like the cheap stereo minijack to single xlr you either get phase cancellation OR you disconnect one connection and only get one channel of the signal. SO if you input device is mono you are OK but if you input device is stereo you really need to get to use another input channel. OR convert the stereo signal to single channel mono first

Here is a link for a SIMPLY way to build a cable using some resistors that will make this work. It is for another system but It gives a good step by step process

http://www.intelix.com/media/tech_notes/152_Athena_how_to_wire_stereo_to_mono.pdf

Sharyn
 
Is the system itself mono or stereo? Do you know the actual mixer model? By the way, the Phoenix style terminal blocks are very common terminations for equipment intended for professional installations.

This reminds me of a system I worked on some years back where the Owner went with the low bid and a year later hired us, who had been the next lowest bid, to come in and completely redo the system, Along with things like blue 'bread bag' wires being used for tie wraps and poorly aimed speakers that if you sat in the middle of some rows of the seating actually made the sound seem to come from behind you, they had taken the left and right out of the mixer and the left and right into the amp and had tied together all four red wires, all four black wires and all four shields, simply twisting them together and wrapping them in a big ball of electrical tape.
 
Fun with phase cancellation. On my radio station, we used to air a music program produced in the host's home studio. One day, I got a call from a confused board operator. He was getting phone calls from listeners wondering why songs were missing vocals and bass. Everything sounded fine to him in the control room.

I turned on my stereo and it sounded fine. I then went to a mono table radio, and found it sounding weird. No vocals, except for some reverb, and no bass. The next two songs that played were fine, but the third one was again impaired.

I knew what had happend. The home studio had one CD player wired out of phase. Anything panned to the center in the recording cancelled. Anything panned left or right was OK. The following day, I called the host and, sure enough, he had installed new CD players before recording the show, and got one mis-wired.

In broadcasting, we stay very vigilant for phase reversal. There are still many mono radios and TVs out there, and most FM receivers blend to mono under weak signal conditions. I have even found CDs from major labels mastered with a phase error.
 
Fun with phase cancellation. On my radio station, we used to air a music program produced in the host's home studio. One day, I got a call from a confused board operator. He was getting phone calls from listeners wondering why songs were missing vocals and bass. Everything sounded fine to him in the control room.

I turned on my stereo and it sounded fine. I then went to a mono table radio, and found it sounding weird. No vocals, except for some reverb, and no bass. The next two songs that played were fine, but the third one was again impaired.

I knew what had happend. The home studio had one CD player wired out of phase. Anything panned to the center in the recording cancelled. Anything panned left or right was OK. The following day, I called the host and, sure enough, he had installed new CD players before recording the show, and got one mis-wired.

In broadcasting, we stay very vigilant for phase reversal. There are still many mono radios and TVs out there, and most FM receivers blend to mono under weak signal conditions. I have even found CDs from major labels mastered with a phase error.

Made this mistake just the other day. I was moving the air chain over to the backup studio at the station I work at, and patched over left, then right. Everything sounded OK on the mod monitor headphones output. Talent came on from the new studio and someone said they couldn't hear him talking. Took me a minute to realize that indeed I had reversed the phase in one of the two channels. I flipped the right phase, and everything worked great. Still need to chase that one down somewhere...I suspect I did it when I rewired that studio over the summer.
 
Interesting... I noticed this 'phase cancellation' last year when I was not the sound tech. Some songs, played through the system, would sound like what was described as phase cancellation, missing vocals and bass and just generally sounding bad. I just didn't know what it was called, until now.
 
CPF has it right. Everything gets cancelled in proportion to how closely to the center it is panned.

It's the same thing as happens when you use a 1/8" TRS to 1/4" TRS to connect an iPod or PC to a balanced line input. or a 1/8" TRS to XLR and connect that to a balanced mic or line input.
 
This all makes sense now that I think about it. Two in-phase signals sent down a balanced connector would cancel each other out, as is the design of balanced audio. When I flipped the second signal 180 in audacity, I was simulating the reversed polarity present on the cold conductor of the balanced line.

Because I'm not concerned about getting stereo sound, I will just disconnect one of the signal lines as suggested by cpf.

I'm guessing this is why the left and right channels of the CD player on the rack are wired into separate inputs.


Thanks Everyone!
 
This all makes sense now that I think about it. Two in-phase signals sent down a balanced connector would cancel each other out, as is the design of balanced audio. When I flipped the second signal 180 in audacity, I was simulating the reversed polarity present on the cold conductor of the balanced line.
A detail but you're actually referencing the affect of a differential input rather than balanced audio. The two commonly go together but balanced audio actually relates to having the same impedance for each conductor, thus making any noise induced in each more similar and via that making the noise cancellation provided by a differential input more effective.

Because I'm not concerned about getting stereo sound, I will just disconnect one of the signal lines as suggested by cpf.
This will lose more than having stereo, you may lose entire portions of the audio. Say you connect only the left channel then anything originally that was panned center, and thus equal in both channels, will be 3dB or so down in level while anything that was only in the right channel will not be there at all. You may want to consider a simple summing circuit like this, STD-150 Passive Audio Divider/Combiner - 150 Ω.
 
Also, most playback devices (iPods, VLC, iTunes) have a built-in option to output mono, but you'll need to remember to un-set it before returning that person's iPod or they'll wonder why their music suddenly sounds different.
 
Interesting... I noticed this 'phase cancellation' last year when I was not the sound tech. Some songs, played through the system, would sound like what was described as phase cancellation, missing vocals and bass and just generally sounding bad. I just didn't know what it was called, until now.

I had exactly the same problem with a dual C/D player on a prayer breakfast. I put a disk in the left deck and it was awful, changed the same disk to the right deck and it was good. There were two different types of adapters to get from phono to 1/4 inch. I assumed it was a bad cable so I got rid of it and bought new cables, both sides worked great. Just goes to show what happens when the light guy/rigger does sound, it's not pretty.
 
It's the same thing as happens when you use a 1/8" TRS to 1/4" TRS to connect an iPod or PC to a balanced line input. or a 1/8" TRS to XLR and connect that to a balanced mic or line input.

Aha! An 1/8" to 1/4" cable is exactly what we had. (in fact, we still have it, I may get rid of it after I leave to prevent the temptation to use it for this) It must be designed for stereo applications, and sends the left to the tip and the right to the ring, and in mono balanced audio, those connections end up being the signal + and -, and they cancel. I understand now. Thank you, the more I know, the better I can do my job.
This also explains why it went away when I took over, I split the right and left into the inputs of one of our two stereo channels, and let the board combine them to mono instead of connecting the device to a mono channel.
 
This will lose more than having stereo, you may lose entire portions of the audio. Say you connect only the left channel then anything originally that was panned center, and thus equal in both channels, will be 3dB or so down in level while anything that was only in the right channel will not be there at all. You may want to consider a simple summing circuit like this, STD-150 Passive Audio Divider/Combiner - 150 Ω.

Whoops, forgot about panning in stereo. In that case, because I have plenty of spare inputs on the rack mixer, I'll split the left and right channels into separate inputs, as was done with the CD player.
 

Users who are viewing this thread

Back