FS#12354 - WM8985 DAC voltage tweak for Cowon D2

Attached to Project: Rockbox
Opened by Stanislav Chizhik (anotherche) - Thursday, 27 October 2011, 07:40 GMT
Last edited by MichaelGiacomelli (saratoga) - Tuesday, 15 November 2011, 23:39 GMT
Task Type Patches
Category Settings
Status Closed
Assigned To No-one
Operating System Cowon D2
Severity Low
Priority Normal
Reported Version Daily build (which?)
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No


Some tech details. Audio codec Wolfson WM8985 uses 4 power supplies: DCVDD - for digital core logic, DBVDD - for digital I/O buffer, AVDD1 and AVDD2 - for analogue output. Allowed range for them is 1.7 to 3.6 V (1.8 to 3.3V recommended for DCVDD and DBVDD, while 2.5 to 3.3V is for AVDD1/2). Rockbox configures these voltages as following DCVDD=1.8V, DBVDD=AVDD1/2=3.3V, which is optimal for quality and power consumption according to wm8985 documentation.

There is an opinion that rising the DCVDD voltage can give some additional improvement of audio quality for the cost of power consumption (up to ~10% higher consumption). The idea is first implement by user stridervvv and discussed at (Russian).

The patch develops the idea by allowing Cowon D2(+) users to regulate the DCVDD voltage in the menu (sound settings menu). Voltage range: 1.8 - 3.3 V. Changes are applied immediately. The patch is for r29795.
This task depends upon

Closed by  MichaelGiacomelli (saratoga)
Tuesday, 15 November 2011, 23:39 GMT
Reason for closing:  Rejected
Additional comments about closing:  We don't want user customization of voltages, and it seems very likely that these changes are not beneficial. If some of the D2(+) voltages should be adjusted, feel free to start a new task.
Comment by MichaelGiacomelli (saratoga) - Thursday, 27 October 2011, 16:34 GMT
I don't think putting voltages under user control is desirable. If this setting really provides an improvement, you should measure it as well as any battery life impact and then decide which setting is ideal.
Comment by Stanislav Chizhik (anotherche) - Friday, 28 October 2011, 02:47 GMT
I'd rather agree with you. This ability is for enthusiasts only. As for myself, I cannot hear the difference nor have a possibility to measure it. The patch was inspired by persons who heard a clear difference in sound quality at DCVDD=3.0V (compared to default 1.8V) . I used my D2+ with DCVDD 3.0V during a month and noticed about 10% shorter battery life, as I mentioned in the patch description.
Comment by Thomas Martitz (kugel.) - Friday, 28 October 2011, 07:07 GMT
How can the digital power supply possibly increase sound quality?
Comment by Stanislav Chizhik (anotherche) - Friday, 28 October 2011, 10:08 GMT
I cannot answer this question. Moreover, it is said in the wm8985 documentation that dcvdd doesn't affect sound quality. But as I told there are persons who think that it affects. Precision of da conversion? I don't know..
Comment by Thomas Martitz (kugel.) - Friday, 28 October 2011, 10:42 GMT
I would expect that DCVDD doesn't affect sound quality at all, and that only analog components can possibly have an effect.

I'm thinking that's hoax by audiophiles.
Comment by Stanislav Chizhik (anotherche) - Sunday, 30 October 2011, 14:34 GMT
May be. By the way, what 's about  FS#12352  & 12353? At least the charger detection patch should be very useful one for D2+. I can't stop enjoying its working in the car-adapter-mode. And that car-adpter-mode improvement patch makes its usage just far more handy.
Comment by Thomas Martitz (kugel.) - Monday, 31 October 2011, 00:42 GMT
Well, we have no active maintainer for that port who can review and commit stuff (doing it not blindly).
Comment by Stanislav Chizhik (anotherche) - Monday, 31 October 2011, 03:48 GMT
So let it remains in the community storeroom and that's it :)
Rockbox patched with  FS#12352  works smoothly in my D2+ (3 weeks of usage with out any (visible) problem). Thank you for commenting, Thomas.