#12916 NORM Future : alc5631 pops at capture start
Zarro Boogs per Child
bugtracker at laptop.org
Thu Sep 10 18:32:10 EDT 2015
#12916: alc5631 pops at capture start
-------------------------------------+-------------------------------------
Reporter: Quozl | Owner: Quozl
Type: defect | Status: new
Priority: normal | Milestone: Future
Component: kernel | Release
Keywords: | Version: Development
Blocking: | source as of this date
Action Needed: review | Blocked By:
| Deployments affected:
| Verified: 0
-------------------------------------+-------------------------------------
Pops, or unplanned signal swings, are at the start of captures made
through the microphone or input socket on the XO-1.75 and XO-4.
See also #9640 for similar problem on XO-1.5.
See also #12289 for a test method for DMA that relies on the pops.
Is mostly solved on XO-1.75 by moving the register changes to the
''.startup'' operation, followed by a small delay of 340ms. The register
changes that cause pops are;
* turning on Vref,
* turning off Vref fast,
* turning on the record mixer block,
* turning on the microphone bias block,
* turning on the microphone boost gain block,
* turning on the ADC block, and;
* turning on i²s serial data port,
That pops are also caused by the microphone boost gain block and i²s
serial data port was not properly handled by the released driver in 13.2.5
and earlier.
Test case: present silence or a test tone to the laptop under test, then
run arecord at all sample rates, both mono or stereo, play back the files,
and examine the files using audacity.
{{{
for rate in 8000 11025 16000 22050 32000 44100 48000 88200 96000; do
for channels in 1 2; do
arecord --file-type=wav --channels=$channels --format=S16_LE \
--rate=$rate --duration=1 /tmp/x.$channels.$rate.wav
sleep 2
aplay /tmp/x.$channels.$rate.wav
done
done
}}}
--
Ticket URL: <http://dev.laptop.org/ticket/12916>
One Laptop per Child <http://one.laptop.org/>
One Laptop per Child bug tracking system
More information about the Bugs
mailing list