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

[kugel.]

linuxstb, I already added the Fuze to the build system. funman added my patch to the git repo back then. I've somewhere the patch around, hopefully I find it.

For the keymap, it's similar to the e200 one yes. It just lacks the REC button, but has Power split up in a Power and Home button.

Maybe I'll give it a shot in the tiny bit of free time.

[linuxstb]

kugel,

It would be ideal if you could create a patch against current SVN and upload it to flyspray.

It unfortunately sounds like the Fuze is going to need it's own set of key mappings.

I had a quick look at the e200v2 sim myself, and it turned out to be very quick to get working, so that's now committed to SVN.  It's not actually very useful (it is identical in use to the e200v1 sim), but it's another small box to tick in the porting process.

Dave.

[jesterday]

i like the idea of "everyone does his part" to get things work (hmm. sounds a bit like socialism... *lol* ).

So, as i'm not firm in "low-level hacking" and also i'm al lousy writer and speaker of english, i would be glad, if i could help in any way to archive this.

To speak clearly: i anyone,who speaks german, is working on this, i would be proud to help. i own a E260 and got basic knowledge on coding "machine code" (hey, i know the dez. value of the unconditional return on Zilog Z80 *lol* ...201 btw.). No fear of "dirty work", if have time to. But i think time is not the thing we sould care about..

bw

jd

[mc2739]

We now have a semi-working display on the e200v2. Still some issues to be worked out.

Sorry for the poor quality of the pic.

[kugel.]

Looks very nice! I wonder if you can estimate how similar the LCD driver will be for the Fuze (given that the Fuze display is just "rotated").

[funman]

Great news !
Now the e200 is only left behind for the buttons, but hopefully the display will make it easy to figure them.

@kugel: the FuZe doesn't look like it uses the same controller (based on disassembly)

In firmware V4.0.45 the dbop_init routine is at 0x3428 and the lcd_init at 0x354C if there is any volunteer

[atomikpunk]

Hi everyone,

it's been a while since I wrote but it's just hectic at work wince 2 weeks. Anyway, I'm still following the forums and (unsuccessfully) try out stuff for the SD.

For the fuze, I can have a look when I'll have some spare times, but meawhile, if someone would volunteer to get me all serial numbers that could appear on the LCD and/or its flat cable, I could have a look if I can find the related LCD controller and thus help out coding the driver for it.

Good work everyone!

[kugel.]

When I took my Fuze apart I didn't find any numbers on the lcd or the cable. But ask saratoga, his is open at this time AFAIK.

[linuxstb]

I'll probably have a go at writing an LCD driver, similar to what I did for the e200v2.  Hopefully once we start having a C version of the init function, someone will be able to identify the controller.

It would be useful if in the meantime, someone could add the Fuze to SVN (or post a patch doing so).

[funman]

Hi

Just to let you know the mirrored/rotated display on the e200v2 is fixed : you should have perfect display / colors.

Also mc2739 told us the SD code works on the e200v2, so if you have one you can help writing the SD code to finish the bootloader.

Unfortunately for Clip / m200v2 the SD card power-up phase doesn't complete and I have no idea why

What the OF does after the reset vector is:

• Set up the stacks (clear them, etc ..) # not interesting for us
• Select the priority 1 on the AHB bus ( 1/test 2/dma 3/usb 4/ide 5/cpu )
• Reset all the peripherals but the memory
• Setup the clocks (PLLA, peripheral clocks ..)
• Setup and enable interrupts
• Setup DMA
• Setup I2C Audio Master
• Modify some registers on the i2c bus:
     Setup interrupts for battery charging
     Modify CVDD/DCDC3 register
     Modify Audioset2 & Audioset3 registers
     Modify System register
  
• Enable SD/NAND
• Enable SDRAM, LCD, USB, ...

I tried to reproduce this behaviour, but still no luck ..
Of course you're welcome to look at the code and see if I have forgotten some bit.

If we can't find what we miss by disassembly, then perhaps a JTAG connection to check the registers statuses will help: I have ordered an adaptor which should ship in 2/3 weeks

[Darkstylist]

Seems like you're far into your sansa build why not release an alpha that we could try out with sources included ? i wont mind testing an unfinished rockbox on V2

[funman]

Seems like you're far into your sansa build why not release an alpha that we could try out with sources included ? i wont mind testing an unfinished rockbox on V2

There is nothing to test at this moment : the effort is ONLY for developers.

When there will be something usable, builds will be available on the download page.

[mc2739]

Also mc2739 told us the SD code works on the e200v2

I'm not sure what I saw to make me think that the SD code was working, but it is not.  The test loops with "OP COND: 0xFF8000" displayed. It never gets to "now - card ready !".

But, some good news, I did get the buttons working (at least the ones already documented on the SansaV2HardwareMappings wiki). The problem was the dual function GPIO's. GPIOB and GPIOC are used for buttons and LCD. The LCD code blocked access to the buttons.

[funman]

About the SD, fragilematter just tested the code on his e200, but for the SD slot with a 128MB SD card in it, and here it succeeds: proof .

That means a bootloader which loads rockbox from the SD card could be developed until we find what's missing to use the embedded SD card.

[mc2739]

It also detects my 4GB card and correctly identifies it as "High Capacity".