Rockbox Player - Project to design and build a Free/Open hardware audio player

[alsaf]

Thanks. I thought it was possible but I didn't realise it would be possible with 3.5 inch drives although it would be external to DAP.

I noticed there are multimedia players like the MVIX MV5000 and IAMM NTD36HD would be ideal for my needs but AFAIK  they don't support lossless formats and from what I see of screenshots the UI doesn't impress me.  

[Rincewind]

An iRiver H1xx could be the basis for a high quality non-portable HDD-Player. I imagine using a huge standard 3.5" disc and building a custom case with buttons and the iriver LCD on the front and the audio and optical connectors at the back.
The result would be a perfect addition to a home stereo system (or even in a car). It would have lossless audio, all the disc space you want, digital input and output, line out and line in and of course Rockbox.

The only problem is, you have to sacrifce an iRiver for it. If I can get a broken one on eBay I might start this project.

[Kramdra]

It should be very possible to make a good player, which can support most if not all of the features in rockbox.

* Size - if the pcb was made by a prototyping company we could easly use very small components.
The big components would be the problem (harddisk and battery)

* Case/enclosure - plastic/aluminuim. wouldnt look as good as commercial products..

* Video - Not sure if commonly available ARM7 micros are powerfull enough for this, I dont know enough about them. video isnt THAT important, but whats the point in making a custom hardware if itcant do what the others can, yet still be more expansive?


I have done some very basic research into parts, and found:

LCD's: I thoguth this would be a problem since I have only seen large ones on electonrics shops however there are plenty of very cheap nokia 3310 ones on ebay which include a controller, it would be very easy to use on any micro and many people use them on pc serial ports.
There are also cheap colour ones. I found http://store.earthlcd.com/RNH942209R1A?sc=7&category=427 but there is probably a lot of cell phone replacements that would be better.

This one would be awesome, and its cheap too, http://www.sparkfun.com/commerce/product_info.php?products_id=712# (although Id rather have a gigabeat x30 screen...)

Processors:
I think an ARM core would probably be best. They are used in almost all embedded devices and seem to be cheap, easy to develope/program (usb/parallel/serial port JTAG programmers also cheap)


Decoder/DAC - Havent found out much about these. The  Wolfson WM8751L used in most players doesnt exist on their site and minimum order of 100pcs..
There is the VS1002 chip but I dont know enough about them, is it used in any commercial players?

It would be cool to use an accerlerometer for menus (like a Wii controller) but  maybe not a good idea for a hdd based player :/

[gnu]

If you want DACs: look into your old DVD player, and you'll surely find one. If you're lucky, you'll find documentation for it!

[Kramdra]

I cant see how a dvd player DAC would be much good for what we want...



I have ordered a STR710 FZ2T6 ARM to play with. If I can get far with it, then I have a spare wolfstone decoder from my karma which I could possibly reuse (its one of the few non-bga chips in there ), the hdd and I would like to use a nokia colour lcd (6610, controller included). Im going to play with the lcd first and try to do other cool stuff with it - it would be very difficult to get it all on a small enough pcb anyway.

[markun]

I was thinking we might save ourselves some trouble if we could use the screen, case, LCD and buttons of an existing DAP and just design our own mainboard for it.

The question is which DAP would be a good donor...

I guess, price and the quality of the components are important. And if the player by itself is very nice already it also doesn't make much sense.

[toffe]

For the case, you can make prototype easily with companies like this one:
http://www.emachineshop.com/
You can even know the price just doing the drawing.

For the electronic part, you can go with the new 16gb memory from Toshiba (no yet on the market but announced), so no need for a hard drive.

[gigabeatbox]

so i know this may sound like a dumb question im not sure if anyone asked it yet cuz i didn't have time to read pgs 2 and 3, but is this player going to be based around rockbox and all its features, like would the firmware still be rockbox, or what?

[Bagder]

"This player" is mostly talk and afaik no actual plans are yet made.

On a related subject, Neuros plans on making an open source portable audio player:

http://forums.rockbox.org/index.php?topic=11084.0

[gigabeatbox]

Oh I see, I'm sorry I wasn't aware of that.
I read all four pages, but I am not very 'technically aware,' so I had trouble understanding all the things people wrote, so has there been any progress? Where exactly in the conceptual stages is this project? It seems like it has quite a few supporters.

[bmordue]

Possibly of interest is this thread

http://www.head-fi.org/forums/showthread.php?t=246943

on head-fi, an audio site with emphasis on headphone listening and a very active DIY community.

[scharkalvin]

We REALLY do need to see an open source hardware player become available.   Soon all the 'hackable' players will be out of production and hard to find used and all the new players are locked down to keep them from running third party firmware (new ipods, zune, etc).

One development platform that I've seen is the NGW100 from Atmel.  This uses their AVR32 cpu.  This processor is an arm-like risc (not an arm instruction set) with lots of on board I/O.  It has a 16bit 50khz sample rate stereo DAC that is suitable for AC97 style audio.  It also has a build in color LCD driver that will do up to 2048x2048 pixels, dual 10/100 ethernet, high speed USB, external flash memory interface, pixel co-processor for video acceleration, etc.  It's supported by GNU tools (Atmel even has a version of their AVR studio jtag debugger that runs on Linux).

I've seen a hw example for using a PSP lcd display on this board: http://dma.elektroda.net/projects/ngw100_ext_lcd/ngw100_ext_lcd.html

Needless to say I'd like to look into this.  The NGW100 already runs Linux and the best part is the price, $90 from Digikey, Mouser, Arrow, etc...

What do you think?  Is this cpu fast enough?

[Bagder]

Fast enough for what?

And besides, getting a Rockbox-capable player is quite a lot more than just finding a dev board that can run Linux.

[scharkalvin]

I meant is the cpu up to the task of running the required codecs (I suspect so, since Linux on this platform has been running mplayer).  
I agree there is much more required to build a media player than a platform that runs Linux.  (The fact that it DOES run Linux means it is open source friendly and that the required documentation for a rockbox port IS available).

One of the first design choices required if someone were to design a player would be a choice of cpu platform.  The small volume involved would leave out any of the PortalPlayer processors.  The disadvantage of the AVR32 is the BGA package (for home brew hobby construction anyway).  A pc board house would have to do the assembly.

Anyway, sounds like a project to get ones feet wet in learning schematic capture and board layout.  I did install the open sourced gEDA tools on my Linux box (good excuse to get a larger LCD monitor....)

[linuxstb]

I agree there is much more required to build a media player than a platform that runs Linux.  (The fact that it DOES run Linux means it is open source friendly and that the required documentation for a rockbox port IS available).

That doesn't follow at all.  There are many devices out there which run Linux in conjunction with closed-source drivers and applications - the latter being where all the interesting device-specific code is.  The source code released by such manufacturers is rarely more than a standard Linux kernel source tarball, along with busybox and other standard utils.

Google for "tivoization".