SanDisk Sansa c200v2, m200v4, clipv1, clipv2, clip+, and fuzev2

[kugel.]

I made a patch for e200v2 which makes the button driver more equal to the one of the fuze (which uses same repeat/acceleration logic since r21166):

• it renames the repeat variable to accel
• using same comments as on fuze (as they are describing the same)
• remove wait for fifo empty
• implement button_delay in the same way as on the fuze

Here is the patch: FS#10284

Just tested this! seems to work rather well, I definitely prefer it over the way it was.
Running this on my e280v2.
Edit: Also just made the hardware mod that gets rid of the "click" feel in the wheel, and tried it both with this patch and without, and both with the wheel mod and without, this patch is still preferable, to me anyway.
Nice work! thanks a lot!

Cool, so I guess it can be committed.

Also, if you don't have the hardware clicks, you can consider tweaking this line "if (++counter >= 2 && queue_empty(&button_queue))". That limits the wheel value posts to match the physical click to scroll 1 item per click in lists.

Don't know if this has been addressed yet, but the player randomly likes to turn the wheel light on during playback, and sometimes when browsing, it flickers.
I think the issue was addressed, but was it ever fixed previously?

Yea, that's known, but for now it's rather useful. I'm eventually looking into making the buttonlight act as disk indicator on all targets (as an additional option of course).

[kronflux]

This is a known "problem" also happenning on fuze, and which will eventually be looked at in the future.

Must not be all Fuze models.. I've had my 4gig Fuze running rockbox for a while now, keeping it up to date with the latest svn and occasionally applying patches, and I've never encountered the wheel light flashing, and I used it every day.. any idea which models it applies to? or am I just a lucky one?

Also, if you don't have the hardware clicks, you can consider tweaking this line "if (++counter >= 2 && queue_empty(&button_queue))". That limits the wheel value posts to match the physical click to scroll 1 item per click in lists.

any idea how what I'd tweak it to? I'm still very new to all this stuff
if this is the wrong place to ask, want to pm me if you don't mind making suggestions?

Yea, that's known, but for now it's rather useful. I'm eventually looking into making the buttonlight act as disk indicator on all targets (as an additional option of course).

that's a cool idea, the only thing I don't like about it is that the wheel light is very bright, so its very distracting for everyday use, but as you said, optional, so still a good idea.

[sko]

Cool, so I guess it can be committed.

I posted a second patch to FS#10284, missed one button_delay, but if it's fine it can be commited (from my position so far).

[FlynDice]

I did a couple of battery benches over the weekend and got some unexpected results.  I thought I would go for the longest lasting first so I tried svn with FS#10048 applied and modified the clocking to fastbus only, 62 MHz boosted & 31 Mhz unboosted.  I got 12:38 runtime.  Then I tried svn with FS#10048 which is 248/62 synchronous expecting the time to be slightly less.  No way Jose!  15:41 using the higher frequencies.  I'm attaching the text files to this post as It would take me too long to actually learn how to make the graphs....  I guess after this I'm reconsidering my thoughts on the whole lower PCLK frequency issue.  I see no advantage now.

Is there a good way to store a database or collect all the battery bench files in one location to get a better idea of what works best?

I was thinking of modifying clock-target.h to give a choice of pre-configured clocking schemes while also keeping the ability to use the present system if you need more options.  Does this seem like a good idea or unneeded complication?

[matsch]

Yeah playback sometimes stops on the Clip.

The problem is known, and specific to the Clip, but no cause has been determined yet.

Stable MP3 playing would be fine. I am not experienced enough to find the root cause, but
I can help testing some ideas to get stable MP3 playing.

[funman]

FlynDice : there is a wiki page for battery benches, and even a Sansa category.

To generate graphs I used gnuplot, thanks to a detailed howto

matsch: perhaps it would help enabling logf (there is an option when configure script is run) to see if codecs have to say something when playback stops

[matsch]

perhaps it would help enabling logf (there is an option when configure script is run) to see if codecs have to say something when playback stops

How can i handle logf? Compiled it with option logf, but going to logf results in a black screen, in case of logfdump nothing happens.

[funman]

codecs will write into the log, mostly if not always in error cases (FLAC codec seems to always write the number of decoded frames though)

[PouncePony]

SVN build after Revision 21228 freezes my 8GB Fuze on boot.  Used OF to format memory - still freezes at boot.

-Pony

[funman]

SVN build after Revision 21228 freezes my 8GB Fuze on boot.  Used OF to format memory - still freezes at boot.

Did you update the bootloader as well ?

[cyclon1978]

similar for me, build from 21232 freezes most times at startup screen (second one, Ver.), every time on video playback, not on mp3 playback. When it freezes, the Scrollwheel LED is on.

Device is an e280, no memory card inserted, fresh formated storage. Bootloader and Rockbox same version.

[PouncePony]

SVN build after Revision 21228 freezes my 8GB Fuze on boot.  Used OF to format memory - still freezes at boot.

Did you update the bootloader as well ?

No.  Will try later.  Thanks.

[funman]

similar for me, build from 21232 freezes most times at startup screen (second one, Ver.), every time on video playback, not on mp3 playback. When it freezes, the Scrollwheel LED is on.

Device is an e280, no memory card inserted, fresh formated storage. Bootloader and Rockbox same version.

Did you check / formatted the filesystem after rockbox crashes ? It could leave the file system corrupted if it crashes in the middle of a storage I/O operation.

When does it crashes on video playback ?

• At "loading" screen ?
• After some video has been displayed ?
• Always at the same position of the same video ?

When it's crashed, is the backlight thread still running ? (hold button will shut off the screen like rockbox normally does).

[cyclon1978]

after a crash the screen goes off after configured time and on with a key press

i had to delete nvram.bin after the last crash to enter rockbox again.

- it freezes entering a deeper folder (music/artist/album/), deleting new dated nvram recovers...
- it freezes on "loading" screen with elephantsdream and other videos are played, deleting new dated mpegplayer config recovers...
- Building database freezes at 142 found (number of files)
- now i can play a singe mp3 file from any folder - no issues for that.
- insert into playlist crashes with "inserted XX tracks" (XX counts up to number of selected files)

[rp]

on my e280 there is also a problem with the database.
Every time i try to initialize/update the database it "finishes"(without counting the files) within 1 sec and tells me to reboot.
but after a reboot, it still says that the database isn't initialized