[LAU] lost sound (it had been working)

[Reid Vail]

The Mic was a really generic stand-alone 6-year-old component.  A mic at 
one end, a 5-foot cord and a jack at the other.  Worked fine on a Dell 
laptop (with an integrated socket) but never got it to work with my 
ubuntu machine.  Nothing suspicious about it whatsoever (that I know of).

But that is a good idea about the a live-linux CD or knoppix, which I 
also have and might be good for a test.

Thanks

Reid
 
James Cameron wrote:
> On Tue, Jan 20, 2009 at 04:29:01PM -0500, Reid Vail wrote:
>   
>> I get nothing at all from the head-phone jack either.
>>
>> The funny thing is, this all started when after I tried plugging in a 
>> mic. And that's it.
>>     
>
> That might relate.  Was the plugging in successful?  What sort of
> microphone?  Was the cable attached to any other device?  Have you tried
> plugging it in and out again?  Was the plug tip undamaged?  Was it
> difficult to plug in?
>
> A story: this might not be your problem, but it was fun to fix, and it
> might give you some ideas to explore.
>
> Friend of mine bought a laptop, and used it for showing audio and visual
> content at church.  It is an Intel HDA sound device, with three 3.5mm
> TRS sockets at the front.  One is marked as an input, the other two as
> output.  He connected the output to the church sound system, which might
> have had phantom power.
>
> After this, the integrated speakers would not work, and the left-hand
> headphone channel would not work.
>
> Booting and old Linux on CD, Knoppix 3 something, could make the
> speakers work, but not Windows, and not a modern Linux.  Therefore this
> was a software related problem.
>
> The Intel HDA sound system has a Conexant 20549 codec attached, and the
> codec has electrical sensing of the presence of the headphone plug.  Not
> physical sensing like a switch, but electrical.
>
> Since we know the headphone jack is electrically damaged, it goes to
> show why the speakers are not working ... unless an older driver is
> installed that doesn't enable the "mute speakers when headphones are
> inserted" feature.  This mute is not an ALSA control.
>
> Here is my write-up:
>
> "Success.  I have the speakers playing and I understand the problem
> better.  I made a change to the Linux kernel source code to ignore the
> headphones.  Here is what seems to have happened;
>
> 0.  there are two headphone sockets on the front, one is also an SPDIF
> optical output, and there is a set of speakers on the body of the
> laptop,
>
> 1.  electrical damage to the headphone circuit, caused by either
> manufacturing fault or something peculiar about the church sound system,
> or static discharge, or power surge between power supply of laptop and
> earth conductor of church wiring (would be almost impossible to prove
> either way),
>
> 2.  the left headphone amplifier section no longer functions, the right
> headphone amplifier section works fine, (both headphone sockets exhibit
> this symptom),
>
> 3.  the left headphone amplifier is used as the method to detect whether
> headphones are present, since it is connected to the sharp end of the
> headphone plug, and so a half-inserted plug won't normally trigger it,
>
> 4.  because the left headphone amplifier section has failed, the
> headphone detection operates as if the headphones are present,
>
> 5.  the driver for the sound device notes the headphones are present and
> turns off the amplifier for the speakers."
>
> References:
>
> http://www.alsa-project.org/main/index.php/Help_To_Debug_Intel_HDA
> (developer section),
>
> cxt5045_hp_master_sw_put() function in patch_conexant.c source file,
> comment reads:
>
>         /* toggle internal speakers mute depending of presence of
>          * the headphone jack
>          */
>
> http://quozl.linux.org.au/2008-12-19-hp-present/
>
>