- Joined
- Oct 3, 2016
- Messages
- 16
- CPU
- 6700HQ
- Graphics
- HD530
All,
Several issues I met with ALC668 after install OSX10.11.6 (without CC):
1. After cold start, speaker is working good, but headphone is working with high noise and Mic not working.
2. Then try sleep and wake up, all the speaker/headphone/Mic not working.
3. After restart from Win10, all speaker/headphone/Mic working good.
4. Then try sleep and wake up, all speaker/headphone/Mic not working.
Then I tried to put CodecCommander to EFI/C/K/10.11, the situation became to:
1. After cold start, speaker is working good, but headphone is working with high noise and Mic not working.
2. Then try sleep and wake up, speaker/Mic not working, headphone working with lower noise, then after unplug headphone, speaker working again, and Mic stays not working.
3. After restart from Win10, all speaker/headphone/Mic working good.
4. Then try sleep and wake up, speaker/Mic not working, headphone working good, then after unplug headphone, speaker/Mic stay not working.
So for simple the issues are:
1. Mic not working if without restart from Win.
2. Headphone noise if without restart from Win.
3. Speaker not working after sleep/wake.
From codec and hda-verb config_dump, I realized here is my node:
0x12 [Fixed] Mic at Int N/A
0x14 [Fixed] Speaker at Int N/A
0x15 [Jack] HP Out at Ext Right
I don't know if I have "Mic at Ext..." or "Speaker at Ext..." node, because other nodes are all showing not connected from config dump.
After all the above, I have been searching around and tried lots of solutions (with CC), but all failed:
1. Try to generate AppleALC with my own Codec info (Codec attached), and also tried all the five layout IDs:3, 20, 27, 28, 29. For sure in this step, I also injected DSDT with IRQ patch, and layout patch.
2. Try to put SSDT_ALC668.aml to ACPI patch.
3. Try to modify the command inside CC/Contents/info.plist with comparing widget_dump result.
4. With cold start, try to compare node_dump result for [After startup] and [After wake up before plug HP], they are different, but I also tried hda-verb command to align node state for [After wakeup before plug HP] to be the same with [After start up], also compared [After wake up and plug/unplug HP] and align, no effect.
5. Try to compare node_dump result for code start and restart from Win10, difference is few and only happened on other nodes. Also tried to align runtime, no effect.
6. Try to compare config_dump for cold start and restart from win, same result.
The links I've been through:
https://www.tonymacx86.com/threads/alc-668-only-works-well-after-restarting-from-windows.195934/
https://www.tonymacx86.com/threads/any-fix-to-the-garbled-distorted-headphone-issue.159031/
One question:
For SSDT patch of WIDGET Control, how is the execution SEQ?
Because the patch can exists in four places: ACPI/patched/SSDT-ALC668.aml; CC/Contents/info.plist; CC/Contents/Resources/SSDT-ALC668.dsl; Clover/config.plist.
Until here, I think all the above just proves the PIN_WIDGET or other node state control won't be the solution.
So following Rehabman's post in the above link, I attached all the info for reference (Clover/IOREG/ConfigDump/Codec).
Wishing for Giants's help, and thanks.
Several issues I met with ALC668 after install OSX10.11.6 (without CC):
1. After cold start, speaker is working good, but headphone is working with high noise and Mic not working.
2. Then try sleep and wake up, all the speaker/headphone/Mic not working.
3. After restart from Win10, all speaker/headphone/Mic working good.
4. Then try sleep and wake up, all speaker/headphone/Mic not working.
Then I tried to put CodecCommander to EFI/C/K/10.11, the situation became to:
1. After cold start, speaker is working good, but headphone is working with high noise and Mic not working.
2. Then try sleep and wake up, speaker/Mic not working, headphone working with lower noise, then after unplug headphone, speaker working again, and Mic stays not working.
3. After restart from Win10, all speaker/headphone/Mic working good.
4. Then try sleep and wake up, speaker/Mic not working, headphone working good, then after unplug headphone, speaker/Mic stay not working.
So for simple the issues are:
1. Mic not working if without restart from Win.
2. Headphone noise if without restart from Win.
3. Speaker not working after sleep/wake.
From codec and hda-verb config_dump, I realized here is my node:
0x12 [Fixed] Mic at Int N/A
0x14 [Fixed] Speaker at Int N/A
0x15 [Jack] HP Out at Ext Right
I don't know if I have "Mic at Ext..." or "Speaker at Ext..." node, because other nodes are all showing not connected from config dump.
After all the above, I have been searching around and tried lots of solutions (with CC), but all failed:
1. Try to generate AppleALC with my own Codec info (Codec attached), and also tried all the five layout IDs:3, 20, 27, 28, 29. For sure in this step, I also injected DSDT with IRQ patch, and layout patch.
2. Try to put SSDT_ALC668.aml to ACPI patch.
3. Try to modify the command inside CC/Contents/info.plist with comparing widget_dump result.
4. With cold start, try to compare node_dump result for [After startup] and [After wake up before plug HP], they are different, but I also tried hda-verb command to align node state for [After wakeup before plug HP] to be the same with [After start up], also compared [After wake up and plug/unplug HP] and align, no effect.
5. Try to compare node_dump result for code start and restart from Win10, difference is few and only happened on other nodes. Also tried to align runtime, no effect.
6. Try to compare config_dump for cold start and restart from win, same result.
The links I've been through:
https://www.tonymacx86.com/threads/alc-668-only-works-well-after-restarting-from-windows.195934/
https://www.tonymacx86.com/threads/any-fix-to-the-garbled-distorted-headphone-issue.159031/
One question:
For SSDT patch of WIDGET Control, how is the execution SEQ?
Because the patch can exists in four places: ACPI/patched/SSDT-ALC668.aml; CC/Contents/info.plist; CC/Contents/Resources/SSDT-ALC668.dsl; Clover/config.plist.
Until here, I think all the above just proves the PIN_WIDGET or other node state control won't be the solution.
So following Rehabman's post in the above link, I attached all the info for reference (Clover/IOREG/ConfigDump/Codec).
Wishing for Giants's help, and thanks.