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

[LADave]

Embedded Linux ( https://www.yoctoproject.org/ ) is an alternative to replacing a cellphone's native O/S with Rockbox.  Then a RB port would have to be done, hopefully only once for all devices running this flavor of Linux.  Thanks to BenBrown for clueing me in!  Unfortunately Embedded Linux is still a work in progress.

[[Saint]]

One of the points of Lyre was to have truly open software, and hardware.

Something anyone with the source code and schematics could whip up on their own home desktop and fabrication table.

Using cellular phones defeats that entirely.

Now, before anyone rushes to say "But Android is open!", it isn't.

Period.

There is a project by the name of Replicant that aims to replace closed vendor binaries with open alternatives, but the outlook for this project is truly bleak. Off the top of my head there isn't a single device that isn't functionally crippled in some massive way.

But the main thing is the hardware. Using phones, the project cannot evolve as freely as a project like this should.


[Saint]

[john.cooper]

One of the points of Lyre was to have truly open software, and hardware.

Something anyone with the source code and schematics could whip up on their own home desktop and fabrication table.
Using cellular phones defeats that entirely.

The limitation isn't software nor hardware (effectively manufactured PCB) which throttles these efforts but ironically competent fabrication of the physical enclosure to the extent it is something you'd want to carry around.  So from that narrow perspective I can see the motivation to leverage existing cell phone product.  But doing so addresses only this single consideration and otherwise you're pounding a round peg into a square hole from a design suitability perspective.  Add to that there is no control of the repurposed product availability and the leveraged effort can be derailed at any time, and moreover will be derailed at some time, likely without notice.

But the main thing is the hardware. Using phones, the project cannot evolve as freely as a project like this should.

It is difficult to believe it has been 5 years since I'd designed this.  And it certainly could use some updating at this point.
The links earlier in this thread are obsolete so I've located the information here:

    http://www.third-harmonic.com/projects/nutshell/wip/0.02/

If someone is interested to move this project ahead feel free to PM me in this forum.

[wodz]

Would it be possible to release CAD/CAM and BOM as well?

[MakingIt]

Hi all,

I signed up in the forum because I want to contribute on this topic. These days you can get STM32 ARM CortexM for very low money, do any of them fullfills the RB min REQ?:

http://www.st.com/web/en/catalog/mmc/FM141/SC1169?sc=stm32

You could output the data through I2S to a decent DAC, then to a headphone output. I think the only we would be missing is native USB support for data transfer and charging.

Thanks

[MakingIt]

Ok, I found another thread about porting RB to a Cortex M4 or even a M7:

http://forums.rockbox.org/index.php/topic,33836.0.html

THis dates from a year ago. Has any work being done at all here? I could quickly do some quick PCBa design around it.

[saratoga]

THis dates from a year ago. Has any work being done at all here? I could quickly do some quick PCBa design around it.

I don't think so.  I don't even recall seeing anyone use Cortex M processors for a portable audio player yet.

[wodz]

saratoga: There was a guy which claimed to have preliminary rb port for some M3 eval board. From the technical point of view there is no problem other then time and persistance really.

[uhmgawa]

I could quickly do some quick PCBa design around it.

Unless you were planning on carrying a bare PCB around or the same
stuffed into a clunky enclosure, fabing a PCB really isn't the problem.
Rather it is the mechanical design of the device which torpedoes these
efforts.

I suppose if you could accept a 3D printed enclosure that may be a partial
solution.  But given the effective resolution of even commercial printers
it isn't something I'd want to brag about.  For small quantity runs it might be
feasible to CNC machine enclosures using a well fixtured blank.  You'll
still need to address surface finish quality but would be in a far better
starting point relative to printing.  Still you'd never reach the level of quality
achieved with large scale injection molded components from insanely
expensive dies.

This is assuming you have a design ready to manufacture.  Getting there is
no mean feat with interdependent tradeoffs between display technology and
available footprints; battery technology, footprint, runtime, interconnect, etc..
No doubt you could amass the engineering talent needed to get the job done
particularly in the context of a collaborative, open effort.  But successfully
executing the manufacturing task is usually where such projects fizzle out.

[MakingIt]

I agree with the comments about the case and real portability of such a product. We are not talking about an O2 DIY that will sit next to your PC and doesnt need to be taken out.

However, if someone could direct me to however did a port to the M4, please let me know, I would be really interested in give it a go.

Thanks

[betoqp]

Why not take the project to other sites/forums? After all, its supposed to be open. Maybe some subreddit? And once there is an initial design, Kickstart it? Or Indiegogo?

[nick_p]

Why not take the project to other sites/forums? After all, its supposed to be open. Maybe some subreddit? And once there is an initial design, Kickstart it? Or Indiegogo?

I can't think of a reason why you shouldn't do these things.

[dired]

Hey there!

Only registered here to share an -i think so- sophisticated idea to revive rockbox, maybe something similiar or even enable original rockbox to evolve.

We want a real free, open-source mp3-player that is not functionally crippled in some way like "smart"phones are etc., you know the point.
Nowadays, there are some promising single-board solutions that could fit our needs, after some investigation, I would recommend: https://www.olimex.com/Products/OLinuXino/iMX233/iMX233-OLinuXino-NANO+i/open-source-hardware
The things that would have to be done hardware-sides:

• 3.5 audio jack to be connected with the i/o, maybe fixated on the board else on the casing
• make a casing
• sd card for operating system (and some free space)
• get a 3.7v lipo battery and just connect it (built-in battery charger)
• a display (suggestion: https://www.olimex.com/Products/Modules/LCD/MOD-OLED-128x64/open-source-hardware)
• if not a touch display: some switches (anyone knows a good fit for an electronic part with some arrows and knobs?)

Software-Sides, I don't have too much experience, but I would guess that having an open-source single-board makes the porting action quite enjoyable with early results if you are into such things as I am. I guess it will be more like making a distribution fit for specific hardware. No custom firmware anymore, we need a "generic" firmware to make it fit the open and well-described hardware.

Another piece of hardware that would fit our needs is the raspberry pi zero but I don't think it's complete open-source so not that thrustworthy a thing to build our project on. Olimex says that as long as they "live", they will support and manufacture their hardware (-> forever). Not so sure about Raspberry Foundation.
Also I found http://nanopi.org/ about which i know as good as nothing about.

If I had time and stuff I would try this myself, a single humans effort might be enough to somehow get a specific rockbox for specific but genuine hardware done. Done correctly, with some guys into manufacturing (knowing what hardware fits the most and will be available forever), we might have the "best" version of rockbox soon

I hope there are some people out there sharing the vision, this could be real fun again.

Best Greetz from Berlin

[squarefoot]

I believe I'm not the only one who came to this thread after my beloved player went FUBAR, and noticed how this and similar Rockbox-able players prices skyrocketed during the years. I paid my refurbished Sansa Clip Zip less than €30 about 4 years ago, now it's over 2 times that price for a used one. That's crazy! If I needed a reason to consider making my own player now I have one.

What do you think about the ESP32? I did a quick search on the forums here and unless I missed something it is never mentioned anywhere.
That chip today could be an interesting platform to port Rockbox on; although it is commonly known for being a WiFi+BT "thing", it is actually a very fast and powerful microcontroller packed with peripherals and i/o ports, plus having wireless connectivity on board opens new possibilities.
http://esp32.net/

This project might be interesting as it uses an audio development board available for purchase with included external DAC, LiPo battery charging circuitry and OLED. https://github.com/possan/csid-esp32

[wodz]

IWhat do you think about the ESP32? I did a quick search on the forums here and unless I missed something it is never mentioned anywhere.

The main problem is RAM size. Rockbox needs more then 2MB of ram (I'd say 4 is bare minimum). Ah, and porting to new architecture (Xtensa) is *a lot of* work.