[Trac #977] Suspend/resume causes a popping sound

[Zarro Boogs per Child]

#977: Suspend/resume causes a popping sound
-------------------------------+--------------------------------------------
 Reporter:  wmb at firmworks.com  |       Owner:  wad      
     Type:  defect             |      Status:  new      
 Priority:  blocker            |   Milestone:  Untriaged
Component:  hardware           |    Keywords:           
-------------------------------+--------------------------------------------
 When the suspend-to-RAM code turns off power to the CPU, and again when
 resuming, a brief "pop" is heard from the speaker.

 This pop is quite useful when debugging the suspend/resume code, but it
 will be very annoying when we do automatic short-interval suspend/resume
 via the OS.

 The amplifier has a shutdown function that can be activated by a CODEC
 output bit (EAPD, net name AMP_SHDN).  I tried enabling that (i.e.
 shutting down the amp) prior to suspending, but it made no difference.
 The pop was still present both when going down and coming back up.

 The amplifier is powered from +5V (via AVDD_AMP).  According to the
 voltage rail chart on page 2 of the schematics, that rail is off during S3
 state, so I suppose the pops could be caused by the amp power supply going
 off and on.  The data sheet for the amp claims that the part is supposed
 to suppress pops, but I think that is true only when the amp is shutdown
 via its SD# pin.  Supposedly, the amp draws very little current (20 nA) in
 shutdown mode, so perhaps we could leave the amplifier powered during
 suspend and gate it on and off with SD#.  The CODEC also has software-
 controlled shutdown capability, so perhaps it could be left powered too.
 If  not, we would need to control the amp SD# via some pin other than EAPD
 from the CODEC.

-- 
Ticket URL: <http://dev.laptop.org/ticket/977>
One Laptop Per Child <http://laptop.org/>