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

[funman]

server adds crlf to .txt files ..

[LiquidAcid]

Nope, doesn't work for me. Displays still turns white once the bootloader is through. Again tested both with and without the microSD card.

EDIT: Just logged into IRC, same nick as here. In case you want me to test some more without spamming the board.

[epithetless]

With the thought that this may not yet be known and may aid in development:

I am able to bypass the whitescreen on my 8GB Fuze v1 once it appears by repeatedly pressing the Play button. The time (and number of presses) it takes varies. The menu then displays normally. However, there is tremendous lag in player response time after this point, to the degree that it makes operation unfeasible. I don't know if this behavior is merely a result of the distressed loading process, but perhaps someone with more coding savvy will find the information useful...


[I've edited the above post to provide more detail. The procedure (repeatedly-pressing-Play-to-bypass-the-whitescreen) seemed more effective with r23199 than with the most recent build (r23203), but it still works after a fair amount of button pressing.]

[LiquidAcid]

At what point exactly do you press play? Does it matter how long the whitescreen is already displayed?

I'm asking because I can't seem to reproduce this. Pressing play doesn't give me the normal menu screen on my Fuze v1.

[FlynDice]

@ Those with fuzes getting white screens:

Would you please try changing line 369 in system-as3525.c from CVDD_1_10 to CVDD_1_20.
  This keeps the voltage scaling code in place we just don't lower the voltage to 1.10 volts, it "changes" it to 1.20v where it already is.

If this solves the white screen would you then try going back to CVDD_1_10 and then comment out line 342:
        while(adc_read(ADC_CVDD) < 480)
  This is the check that the voltage has returned to 1.20 before we boost the frequency.

If that works uncomment and I would be interested if changing the value from 480 to either 470 or 460 works.
  This cheats down just a little bit from the 1.20V we're looking for.

Thanks!

EDIT:
I've lowered the voltage check to 1.175 volts and it seems to do the trick.
Thanks liquidAcid for all the testing!

[kugel.]

I was wondering how this is possible if we use the same setting as the OF.

Are you sure the OF use 1.10V ?

That was according to you. Use looked at the disassembly and told that

[FlynDice]

I think I've figured out the CLKIDE link.  I think CLKIDE is MCLK for the pl180 controller for the internal SD.  I ran test_disk on the internal SD with IDECLK at 82 MHz, 62MHz, and 24 MHz and I think the results point strongly to this conclusion.  See the attached file.

I'm not looking for agreement here, I'm looking for reasons to disprove this hypothesis!
Fire away.


Edit:
Assuming this is correct I have been able to run the internal at 24.8 MHz and it works fine but feels a bit sluggish.  Once again I cannot run data without the bypass bit set so I can't run the controller MCLK at 49.6 and divide by 2 to get MCICLK = 24.8 MHz. I have to run the controller at 24.8 MHz and that slows things down...

Another thought is that we are overclocking the internal card much more than we thought.  The internal card is a v1(I tested on my e280v2 and I'm assuming the same for the rest of the as3525's) and , in fact will not accept the switch to HS timings. Which means the spec frequency is 25 MHz max.  We're currently running at 82 MHz MCLK with bypass meaning the card is being clocked at over 3 x the maximum!!

[funman]

The slowdown is a bit less than 50% so that looks OK.

We should compare to the OF transfer speed in USB mode

[Urutoraman]

Is this the right forum to post?
I was using Rockbox fine on my e280v2 for a week and decided to load the current build 2 days ago. Now I am getting this *PANIC* SD Transfer Error: 0x8 all the time. r23313. 

Have an A-Data MicroSDHC with 4GB.

Weird part is that if I navigate through the Database menu I can get songs to play from the microSD card but after a few minutes it skip and I get the SD Transfer error again.

Tried a chkdsk but it didn't found any issues on the SD card.

Any thoughts?

By the way... Good work.

[FlynDice]

Have I done this correctly?  I tried to add ISR INT_MCI1 and INT_IDE to demonstrate that MCI_MASK1(INTERNAL_AS3525) is masking for INT_IDE.  I made  MCI_MASK0(drive)=0 and MCI_MASK1(drive) = MCI_ERROR | MCI_DATA_END(ie I swapped them).  My aim was to run the code using the two Mask1 registers(replacing the current MASK0 registers) to show that the internal drive MASK1 is doing masking on INT_IDE.  Rockbox operates quite normally this way but is my logic sound enough to draw this conclusion?

Yes I know we don't need them right now but I'd like to understand how these things are put together better.....

see http://pastie.org/664619

EDIT:

@ Urutoraman

I would be interested in knowing what system/debug/View disk info reports for the Speed of your card.
The internal is displayed first then if you push the select button it should cycle through your disks.

[funman]

flyndice: my way to be sure my code runs is to add a panicf() in the isr

[Urutoraman]

EDIT:

@ Urutoraman

I would be interested in knowing what system/debug/View disk info reports for the Speed of your card.
The internal is displayed first then if you push the select button it should cycle through your disks.

Here it is

D% Rev 1.9
Prod: 0/2008
Ser#: 0x193025CB
M=1D, 0=AD
Blocks: 0x00775400
Speed: 50Mbits/s
Taac: 10.0ms
Nsac: 0 clk
R2W: *2
IRMax: 10.45mA
IWMax: 100.200mA

 
Update:
r23327-091024 Let me go into Files-><microSD 1> and browse the directories without any error.  But once I choose a song it play for some minutes and then BANG!! *PANIC* SD Transfer Error: 0x08.   Or sometimes it just a black screen and a player reset.

[FlynDice]

@ Urutoraman:

Would you be able to go back and try r23192?  That's just previous to the reimplementation of voltage scaling...


@funman:

Well that was too easy, thanks.  MCI_MASK1 for the internal does indeed mask INT_IDE.  I'm just trying to provide some comic relief with my Rube Goldberg coding tendencies...

[funman]

Well that works too

About IDE: in system-as3525.c:  system_init() we have this:

Code: [Select]

CCU_SRC = 0x1fffff0 & ~(1<<18); /* FIXME */

bit 18 is the reset bit for an IDE peripheral (there is 2), the internal SD doesn't work if we reset the peripheral, perhaps you have an idea of what we could do to configure it again properly after reset?

[Urutoraman]

@ Urutoraman:

Would you be able to go back and try r23192?  That's just previous to the reimplementation of voltage scaling...

If you can point me out where to download it from.  I overwrote mine by mistake.