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

[casainho]

I did put the first patch of Lyre on Rockbox tracker.

Hi Casainho,
Thanks, you answered my question. I was doing the same thing.

I think it's the sd_crc7 function.
Did you log a chart and want to just have a constant check at all times?
You must be trying to add a second card through the spi secondary. I think there is an e-book (free in pdf format) about this in sd/mmc subject.
I thought you were using definitions to use the atmel library?
You may want to make the data from the sd to be global so the other peripherals can use the data. Just be sure to kill the pointers after the other peripherals use the data.

I must do more tests on SD Card but for now, looks the main problem is on audio drivers, there is no audio out nor time of file do not advance because there is no audio data sent to the hardware.

[Bagder]

I went through and renamed all the wiki pages about this subproject on rockbox.org to now use names involving the new Lyre name. If you think any turned out wrong, just let us know and we'll adjust!

[casainho]

I am not finding free time for working on the audio... while that, here is the last video, for the first time I got Cabie V2 theme working on playing screen.


LyreProjectActualWorkingState video

[casainho]

I looked at your screen and saw that the disk size is at 7.59GB. I see why you need a different protocol. May I ask what kind of media are you using?

This may help for crc16 function. I didn't go though them yet.

SD Card. CRC16 is done and is working.

[casainho]

NEWS

We are in talkings with Matt, from AVRopendous, to produce and sell a hardware for Lyre.

Matt produce his boards on home, a bit DIY. He can solder BGA. He uses just as possible Free Software tools as KiCAD for schematics and board drawings (as we did before on this project).

Matt sells his boards on Ebay, here.

Matt wants to make a cheap hardware, because he needs to sell ;-) and because of that he choosed to use some Phillips LPC ARM9. Hopefully, we will have a new board with LPC ARM9 + color LCD (16 bit colors + 240x ??) + Lion battery chaged by USB + SD Card + wheel and buttons + stereo audio codec for playing and recording.

Target price is no more than 120 euros (?), which I think is good. And later he can design and make a plastic case ;-)

We didn't get good reception from big companies but we got from "small" business individuals like Matt... we should work together, grow together :-)

I believe that actual working state of Lyre, with the last video I did record, was important for "marketing", for prove that we can do it. It was good for convince Matt ;-)

---

I am being busy on preparing the Lyre patches instead of being able to work on audio :-(

The Lyre code were for the first time commited to Rockbox SVN :-) :-)

http://svn.rockbox.org/viewvc.cgi/trunk/firmware/target/arm/at91sam/lyre_proto1/

[Alex_Nsk]

Hi, Guys! Your project looks great... My congratulations!
But, if you work to make' em best - you must use best audio hardware at fist!!!!!!!!!
I look at short at the project and see that you plan to use TI TLV320 codec. I think that it is a waste of money and time! If doing so why just don't buying used Cowon X5/M5 - that dap USE SAME CHIP, looks great, price is as low as it can?
 By the way - it sounds cheap!
Right now I'm working at the X5 Xmod project: I m use a half of X5 hardware, rockbox software and almost rebuilt all sound path!
Audio path hardware would be:
IIS from CF5250 (I'm appreciate any help with rebuild in rockbox with up to 24bit word IIS) > PCM1708 > AD8620 i/u converters in balanced configuration (4mA inputs  into 2V outs)  > TPA6120 (2v inputs to the +/-4V outputs with max current capacity up to 300mA whit a slew rates up to 900V/uS!!!)
All digital-to-sound chips supplied with  +/-5V voltage from  high efficiency DC/DC converters fed from a single Li-Po 3.3V cell. Low-ESR Sanyo Os-COn caps, ferrite beads for filtering and so on..
 I think that novaday such signal path is THE BEST/SIMPLE EVER!
Right now I'm working at the printed circuit board layout, to suit best available quality and physical X5 compartment.
I think it could to  be a sound killer DAP!

P.S. I'm plan to use a software volume control that present in DSP module of  rockbox (precut item in EQ) So if one want to have a best quality sound he should to share IIS word length up to 24bits. Doing so all products of rounding in the out of the DSP modules must to lay below 16bit limit.

[casainho]

Hi, Guys! Your project looks great... My congratulations!

Thank you ;-)

But, if you work to make' em best - you must use best audio hardware at fist!!!!!!!!!

We would like to have the best of all, but we don't have all the knowledge nor possibilities... for now we are trying to have a working base hardware, on next iteration we can add better audio, but while then we should do as you are doing, small add on boards for developers test new features for finally add them on next iteration of hardware.

I look at short at the project and see that you plan to use TI TLV320 codec. I think that it is a waste of money and time! If doing so why just don't buying used Cowon X5/M5 - that dap USE SAME CHIP, looks great, price is as low as it can?
 By the way - it sounds cheap!
Right now I'm working at the X5 Xmod project: I m use a half of X5 hardware, rockbox software and almost rebuilt all sound path!
Audio path hardware would be:
IIS from CF5250 (I'm appreciate any help with rebuild in rockbox with up to 24bit word IIS) > PCM1708 > AD8620 i/u converters in balanced configuration (4mA inputs  into 2V outs)  > TPA6120 (2v inputs to the +/-4V outputs with max current capacity up to 300mA whit a slew rates up to 900V/uS!!!)
All digital-to-sound chips supplied with  +/-5V voltage from  high efficiency DC/DC converters fed from a single Li-Po 3.3V cell. Low-ESR Sanyo Os-COn caps, ferrite beads for filtering and so on..
 I think that novaday such signal path is THE BEST/SIMPLE EVER!
Right now I'm working at the printed circuit board layout, to suit best available quality and physical X5 compartment.
I think it could to  be a sound killer DAP!

Sounds nice your project... can you send the link so we can read more about it?

Right now we are working with a AVROpendous so he will design, make, assembly and sell online (on Ebay) the hardware. We don't have all the info updated online... there is time missing.

Don't you want to join us? we need all the knowledge and developers as possible...

And later we can work on Rockbox audio drivers for having the 24 bits PCM ;-)

[Alex_Nsk]

Sounds nice your project... can you send the link so we can read more about it?

Right now we are working with a AVROpendous so he will design, make, assembly and sell online (on Ebay) the hardware. We don't have all the info updated online... there is time missing.

Don't you want to join us? we need all the knowledge and developers as possible...

And later we can work on Rockbox audio drivers for having the 24 bits PCM ;-)

I' m from Russia, just haven't much possibilities for cooperation..
Just some thoughts:
At first - you must to decide for what and for whom is your project?
Make just another ugly MP3 player with decent sound at the price of all-in-one china SoC MP4 player - isn't a silly thing?
Second, I just HATE compressed sound! That is because I'll go with rockboxed device... Right now in my remodded X5 presents over 80 CD Images in lossless nor one in lossy!
So,  I'm not interested in MP3 player in a way you go. I don't want to waste a time and money in a thing with guaranteed poor sound potential without any simple way for growing...
Instead , I had searching the most "potential in sound" DAP available at the market and that is X5, i think...
Then I though:  what I could remode to reach the best sound available from the portable device? As usual a weakest thing is a sound path ! 
So right now I'm working on it.  Overall technical aspects I wrote already in the post above.
When it will be completed, I could offer some more information, or may be printed circuit boards, or assembly kits, or something else for people who want to reach the best with their X5...
Possibly my construction could be used as outer DAC/Headphone AMP for almost any DAP based on the processor AND outer codec chips (could to be connected except original codec via I2S bus)...that is just a thougths right now...
Unfortunately I'm not a programmer, and just cant share 16bit limits in rockbox...if you could help, i'll thank you!

That is not a "death theme" - just another sound degrading factor on a way to the high fidelity 
Let's go BEST!

[casainho]

I' m from Russia, just haven't much possibilities for cooperation..

We are all from various countries, for example, I am from Portugal.

Just some thoughts:
At first - you must to decide for what and for whom is your project?
Make just another ugly MP3 player with decent sound at the price of all-in-one china SoC MP4 player - isn't a silly thing?

Why not? we need also a cheap price as possible, but, with also good quality of sound. Did you read our objectives on project page? "Make a good quality hardware audio player and recorder;".

One advantage over that China SoC MP4 players is having Open Hardware.

Second, I just HATE compressed sound! That is because I'll go with rockboxed device...

That's the advantage of Rockbox firmware, Open Lossless codecs. That's why just seems ok to go with Rockbox firmware.

Right now in my remodded X5 presents over 80 CD Images in lossless nor one in lossy!
So,  I'm not interested in MP3 player in a way you go. I don't want to waste a time and money in a thing with guaranteed poor sound potential without any simple way for growing...
Instead , I had searching the most "potential in sound" DAP available at the market and that is X5, i think...

What is that X5? is that Open Hardware?

Then I though:  what I could remode to reach the best sound available from the portable device? As usual a weakest thing is a sound path ! 
So right now I'm working on it.  Overall technical aspects I wrote already in the post above.
When it will be completed, I could offer some more information, or may be printed circuit boards, or assembly kits, or something else for people who want to reach the best with their X5...
Possibly my construction could be used as outer DAC/Headphone AMP for almost any DAP based on the processor AND outer codec chips (could to be connected except original codec via I2S bus)...that is just a thougths right now...

Looks like we have the same objectives :-)

Unfortunately I'm not a programmer, and just cant share 16bit limits in rockbox...if you could help, i'll thank you!

Don't worry, a team is composed of various people with different knowledges :-)

If we "give" people a good hardware base, then then will be motivated to work on 24bits audio PCM drivers for Rockbox.

[froggyman]

What is that X5? is that Open Hardware?

I think he is refering to Cowon's iAudio X5: http://www.dapreview.net/p/content/content.php?content.201

[casainho]

What is that X5? is that Open Hardware?

I think he is refering to Cowon's iAudio X5: http://www.dapreview.net/p/content/content.php?content.201

Thanks.

So,  I'm not interested in MP3 player in a way you go. I don't want to waste a time and money in a thing with guaranteed poor sound potential without any simple way for growing...

But right now you are putting your time and money to hack some hardware that is not different from the general other ones. You are hacking a closed source hardware, which I think it limit the acess to people for your work.

Don't you really think that a closed source hardware have more "poor sound potential" than an Open Hardware one which everyone can improve easily??

I would love to have your board schematics and documentation for future reference, who knows if we can reuse it? :-)

[Alex_Nsk]

But right now you are putting your time and money to hack some hardware that is not different from the general other ones. You are hacking a closed source hardware, which I think it limit the acess to people for your work.

Don't you really think that a closed source hardware have more "poor sound potential" than an Open Hardware one which everyone can improve easily??

I would love to have your board schematics and documentation for future reference, who knows if we can reuse it? :-)

Above I was talking about Cowon X5, of cause.
I choose this one also because of it's hardware was hacking almost to the end by rockbox guys. So we can call it "OPEN".

Just for $80 I buy a working thing which suits all my needs in hard&soft and is ready for the best audio hardwire modifications.
Thanks for the cowon & rockbox guys!

And you are just in the beginning...to make a thing that is ever equal to the X5 you'll must work really hard and spend more, more time...

Researching of Hardware..sotware..programming, design of the ergonomic and aesthetic body... Making the body at the end!

So why you must to do the same things again and again?
Especially if somebody was making a working platform just like "especially for us" already?

That is a question!

As for audio path schematic...just look at the TPA6120 datasheet and you'll see a complete BEST/SIMPLE modern audio path.
Some changes - I'll use PCM1798 as DAc chip & AD8620 as I/U convertors...There are not any mysteries... .
Good PCB layout and best passive components will do the best!

 There is Just one problem: 16bit PCM limitation from SCF5250 as I said earlier...

[casainho]

But right now you are putting your time and money to hack some hardware that is not different from the general other ones. You are hacking a closed source hardware, which I think it limit the acess to people for your work.

Don't you really think that a closed source hardware have more "poor sound potential" than an Open Hardware one which everyone can improve easily??

I would love to have your board schematics and documentation for future reference, who knows if we can reuse it? :-)

Above I was talking about Cowon X5, of cause.
I choose this one also because of it's hardware was hacking almost to the end by rockbox guys. So we can call it "OPEN".

No way that hardware is open.

And you are just in the beginning...to make a thing that is ever equal to the X5 you'll must work really hard and spend more, more time...

Researching of Hardware..sotware..programming, design of the ergonomic and aesthetic body... Making the body at the end!

The same as Rockbox firmware and many others Open Source hardware/firmware projects.

And I am not alone, we are a few developers, and there is one person that want to sell it which will help improving the system.

[Alex_Nsk]

Your selected components are fine for the exception of the power consumption. TPA6120 uses 10v or more, TDA2822 would be a better choice for me.  PCM1798 uses 5v analog and 3v3 digital, I like 3v3 not the 5v. AD8620 uses 5v, from the same company an op-amp can use 3v3. I prefer a codec with an op-amp. Your choices are very good ones, since they give schematics, I like that myself.

The booster DC/DC converter hardware from 3.7V LiPo to the high current +/-5V IS THE MUST for the trully HIGH END sound.
It is included in my schematic...

[casainho]

NEWS about the hardware

We are moving to an Phillips LPC ARM9, which is cheaper than the Atmel AT91SAM ARM9 but also lacks some functionalities. Although this LPC ARM9 should be powerful enough for decoding lossless codecs - I hope (there is no DSP instructions on this LPC as we had on AT91SAM ARM9).

We hope to sell the boards on early June.

Hopefully by July we could have a proper, encased, audio player/recorder.

We are working with another commercial partner and we can't reveal much more info until the board is on selling. Schematic and board will be Open Hardware, and designed with Free Software tool KiCAD.