New cheap portable player in the market

[sandreas]

But a touchscreen is IMNSHO worthless in a DAP, because without hard buttons, you have to be able to see the screen in order to effectively operate the device.  This makes it a non-starter for visually impaired folks, or in situations where looking at a screen would be unsafe.

Yeah this may be correct. But having a touchscreen AND buttons for play/pause, vol+, vol+ and possibly next, previous and hold would be the best approach in my opinion. Most people are familiar with phones these days and a clickwheel from the older iPods is pretty complex and way less flexible than a touchscreen, even when the touchscreen is small.

Looking at the Sony NW-A306, I like their basic approach (besides it is a bulky, power hungry android beast), the only thing that bothers me is the missing headphone remote support via audio jack. It works via USB-C to jack adapter with mic support from amazon, which made me laugh. I'm pretty sure OMTP via TRRS for headphone remote control could be implemented in linux pretty easily, because it is just detecting a click-pattern in the mic input... I wonder why nobody seems to care about this.

Apple prefers CTIA, but the click pattern works on ALL their devices from 2009 ipod classic (rewind and fast forward does not work on 2009 classics) till today, even on MacBooks with audio jack have support. On my iPod Nano 7G I can either use the EarPods remote and even the the devices buttons to control the whole playback scenario...

• Vol+
• Vol-
• Toggle
• Toggle (hold) - speak song title
• 2xToggle - next chapter / track
• 3xToggle - prev chapter / track
• Toggle+hold - fast forward
• 2xToggle+hold - rewind

That's all I need without looking at the screen or even touch the device. And it works flawlessly...

Speaking of hardware, I thought this writeup was pretty interesting:

https://jaycarlson.net/embedded-linux/

The Rockchip 3308 kind of won there, but looking at the power consumption chart, I would prefer a chip that is just slightly more than enough for modern audio codecs needing as little power as possible. I think the V3(s) is a good fit and the completely finshed hardware and firmware design of the FunKey S would be an interesting base to start. And it's cheap enough.

Another one would be the MangoPI:
https://mangopi.org/
Looks like they have an Audio-Player (Cyber Pad 1 mini - https://mangopi.org/cp1m) on the front page... maybe worth doing some further research

BTW: Here is a DIY Gaming Console also with a reference design for Allwinner S3 (Called X-Boy Plus) WITH Audio Jack, but I did not find much information about this and no purchase information:

Demo-Video: https://www.youtube.com/watch?v=TTBzunSkPDQ
Repository: https://github.com/hsinyuwang/X-Boy
Schematics (V3s): https://bbs.aw-ol.com/assets/uploads/files/1691901010607-schematic_allwinner-v3s-game-handheld_2023-08-13.png
More info: https://bbs.aw-ol.com/user/hsinyuwang

However, I'm not skilled enough to give this redesign a go (and I don't own a 3D printer!). Hopefully these guys see the potential and maybe fork the design for a dedicated audio player, maybe with rockbox support :-) I would buy it, even when rockbox does neither support OMTP nor CTIA nor Bluetooth nor Wifi... 

[speachy]

But having a touchscreen AND buttons for play/pause, vol+, vol+ and possibly next, previous and hold would be the best approach in my opinion. Most people are familiar with phones these days and a clickwheel from the older iPods is pretty complex and way less flexible than a touchscreen, even when the touchscreen is small.

Except that Rockbox sports a _very_ different UI paradigm from "phones these days".  The latter carries with it all sorts of functional user expectations that Rockbox will never meet -- it's not something that can be papered over with a new theme. 

This goes well beyond audio playback controls; It is a hard requirement that _every_ UI element be usable/available without looking at the screen. 

If what you want is effectively a smartphone with more buttons, there are countless DAPs out there that are all trying to outdo each others by (badly) copying smartphone design (anti)patterns from the era of crappy uninstallable crappy carrier/vendor apps.

Meanwhile.  The problem with building a new DAP isn't in the component selection or even things like PCB design/layout; it's the huge up-front cost of making a (robust) enclosure and obtaining enough components up front for sufficiently large production run -- economies of scale begin to really kick in at 10,000 units, but anything under 1,000 units is probably not even worth the effort, since anything created will have to be cost-competitive with not only the current crop of mass-produced DAPs but also the considerable secondary market for refurbished ipods.

Honestly, what we really need is someone who can speak Mandarin and is willing to negotiate with a Chinese OEM so they'll crank out a batch of an exsiting design for us, on the condition that we get schematics and some reference source code along with it.

• Of course, Tockbox is purely GPL licensed, so anyone can port it to whatever hardware they want (and we'll probably accept it into the main repository) but if we're going to commission hardware intended for Rockbox, it's going to be something that enables our still-unique-in-the-market features, notably fully translated/voiced user interface that can be operated in situations where looking at the screen is impractical or impossible.

[sandreas]

This goes well beyond audio playback controls; It is a hard requirement that _every_ UI element be usable/available without looking at the screen. 

Did not know that... I like the approach.

If what you want is effectively a smartphone with more buttons, there are countless DAPs out there that are all trying to outdo each others by (badly) copying smartphone design (anti)patterns from the era of crappy uninstallable crappy carrier/vendor apps.

No, what I mainly would like to have is a working SOLUTION to my personal problem. However, I don't think this will ever happen, because it is way to specific.

• Small device - the smaller the better, but I would say that the maximum is 110x60x20mm
• readable display
• Usable offline
• Works with traditional headphones (Audio jack)
• Supports playback control via Headphone (at least volume+, volume- and play/pause)
• Supports modern codecs (m4a, flac, mp3, ogg, etc.) - especially m4b audiobooks with chapters
• Can be synced somehow - best case via Bluetooth / Wifi, but I would accept cable sync

All these are fulfilled by:

• iPod Nano 6 and 7
• Some Android phones / players (Unihertz Jelly 2, Fiio M6, LG G5 H850, Google Pixel 4a, ...)

Some optional nice to have features would be:

• Repairable (incl. a replaceable battery)
• Hackable / develop software for it
• Low power requirements
• Extendable storage
• Extra buttons (next, prev, hold)

What bugs me the most on the iPod nanos is that they are so hard to repair and have low battery life. That's why I'm looking for alternatives.


I would be happy with a 3d printed case... at least for now. The sophisticated industrial production would be not a criteria, but I understand that this would be the approach rockbox would like to go.

Thanks for your opinion.

[speachy]

• Small device - the smaller the better, but I would say that the maximum is 110x60x20mm
• readable display
• Repairable (incl. a replaceable battery)
• Hackable / develop software for it
• Extendable storage
• Extra buttons (next, prev, hold)

There are quite a few devices available right now that meet this criteria, including several still purchasable new -- The Eros Q/K (+ clones) clock in at 90x55x15mm, the xDuoo X3ii is 102x53x15, the classic ipods, and more.  My daily driver is an original xDuoo X3 (106x45x14) which even has dual SD card slots.

What bugs me the most on the iPod nanos is that they are so hard to repair and have low battery life. That's why I'm looking for alternatives.

It's also purely internal-flash-based, which means it _will_ eventually fail and become a (crappy) paperweight.  That's one of the best arguments for the HDD-based iPods; the spinning rust can be replaced with a CF or SD card, easily swappable when it fails.

I would be happy with a 3d printed case... at least for now. The sophisticated industrial production would be not a criteria, but I understand that this would be the approach rockbox would like to go.

3D printing is awesome for prototyping, but it's horrendously expensive (and slow) for serial production.  (And you have detail/tolerance/finish limitations which can cause problems for things like buttons, which is a really big deal here.

It's actually pretty straightforward to build a prototype DAP using (eg) an SoC devkit and a protoboard; but to build your own PCB you need to already have the enclosure.


Thanks for your opinion.

[sandreas]

Thanks for your opinion.

Thanks for providing so much information :-) I really appreciate the time it took to answer these long posts.

There are quite a few devices available right now that meet this criteria, including several still purchasable new -- The Eros Q/K (+ clones) clock in at 90x55x15mm, the xDuoo X3ii is 102x53x15, the classic ipods, and more.  My daily driver is an original xDuoo X3 (106x45x14) which even has dual SD card slots.

Yeah, I know that. I'm the proud owner of several iPod classic 2009, I modded myself with an iFlash Quad, 512GB storage via micro SD (more would be detected, but do not work as expected because of memory and cpu limitations) and a 2200mah battery.

This thing is absolutely great, here is why I do not use it on a daily basis (I use it as a vacation device to have all my audio books with me):

• Does not support fast forward / rewind via headphone remote - this is something I use pretty often
• Using the headphone remote wakes up the display - that woke up my children lying next to me more than once by accident
• A lot bulkier than the iPod Nano 6 / 7
• UI is not as easy to use (for me) - searching content is not really easy

I could use Rockbox and fix all this, but since there is lots of other things that do not work, it would be a huge amount of work... Rockbox whilst I appreciate all of the work that was put in is just not ready enough for listening to m4b audio books in my opinion :-)

It's also purely internal-flash-based, which means it _will_ eventually fail and become a (crappy) paperweight.  That's one of the best arguments for the HDD-based iPods; the spinning rust can be replaced with a CF or SD card, easily swappable when it fails.

Absolutely... I already thought of dumping the firmware and hacking it via reflow soldering and putting a cable between the flash chip and the soldering spots. I'm pretty sure the iPod 7g has:

• LGA60 or BGA60 nand flash (like iPhone 5s)
• and usb programmer https://de.aliexpress.com/item/4000512585123.html?spm=a2g0o.productlist.main.45.59095ef9PZMHCG&algo_pvid=3a0d1869-2d2e-4cef-9c26-207be00601b0&algo_exp_id=3a0d1869-2d2e-4cef-9c26-207be00601b0-22&pdp_npi=4%40dis%21EUR%214.25%214.25%21%21%214.44%21%21%4021038ed816951022284832975eb24d%2112000025080909361%21sea%21DE%214558788930%21S&curPageLogUid=MtjzYoAvJazK
• a long flexcable - https://de.aliexpress.com/item/32917136928.html?spm=a2g0o.productlist.main.51.2f515ef9bfwJRi&algo_pvid=5b632612-2928-4072-afbe-8922c02ddb37&algo_exp_id=5b632612-2928-4072-afbe-8922c02ddb37-25&pdp_npi=4%40dis%21EUR%217.28%217.28%21%21%217.60%21%21%4021038ed816951023340084053eb24d%2165942031885%21sea%21DE%214558788930%21S&curPageLogUid=wXpGKO5cPKSo

But I did not find the time.

3D printing is awesome for prototyping, but it's horrendously expensive (and slow) for serial production.  (And you have detail/tolerance/finish limitations which can cause problems for things like buttons, which is a really big deal here.

It's actually pretty straightforward to build a prototype DAP using (eg) an SoC devkit and a protoboard; but to build your own PCB you need to already have the enclosure.

Yep. But I don't need a polished one. I would need one that works. Would be great to "go in production", but that has some challenges that will also slow down the release to in two years from now :-) I would like to have one as soon as possible.

For now, I'm happy enough with my iPod nano 6/7 and my Unihertz Jelly 2e running Audiobookshelf app. You can read about my Audiobookshelf experience here.

[sandreas]

In the long term, this device might be interesting:

https://www.lilygo.cc/products/t-display-s3-pro?bg_ref=7LJo9u5wwo


Price about 40 bucks. Based on ESP32-S3.

It's not very powerful, but this is only a question of time... It could also be extended to have an audio jack, because there already is a DAC included... Touchscreen contols, decent form factor.

[cleansugar]

for references

The Cheapest MP3 Music Players from Ebay - Test and Teardown
https://www.youtube.com/watch?v=Q2DHw_EMJOc

Cheap Chinese MP4 Player Review
https://www.youtube.com/watch?v=QqbHN9hjygQ

DFRplayer mini
https://www.dfrobot.com/index.php?route=product/product&product_id=1121

https://github.com/Arduinolibrary/DFPlayer_Mini_mp3/raw/master/DFPlayer%20Mini%20Manual.pdf

XY-V17B Mini MP3 Player Module Audio Voice Board 8Bit I/O UART Control Support SD Card TF Card
https://www.icstation.com/mini-player-module-audio-voice-board-8bit-uart-control-support-card-card-p-13279.html

Ruizu x02 digital music player review and tear down
https://www.youtube.com/watch?v=uco3KGROhz8

ATJ2127 MCU
https://ko.datasheetbank.com/pdf-view/ATJ2127-ETC
ATJ2127 UPGRADE.HEX decryption
https://github.com/nfd/atj2127decrypt
https://github.com/Suber/PD196_ATJ2127

PFB29-12AL Flash
http://forums.xgecu.com/archiver/?tid-1168.html

XGecu Programmer Forums's Archiver
XGecu Programmer Forums » T56/TL866II Plus新增芯片支持 » Question for chip support
snipey 发表于 2023-8-16 03:42

Question for chip support
Is there any support for the following chip? If not, is it possible for me to add the chip manually some how?
PFB29-12AL
_
1438

It appears to be by spektek; here were the product specs I found.

https://www.spectek.com/pdfs/SPECTEK_4GB_NAND_m40a.PDF

It is a 16GB SD card with two of these chips and a microcontroller (So 2x8GB cards).

Thanks.

FNNM40A
https://13.209.45.252/?q=FNNM40A

A strategy to emulate NOR flash with NAND flash
https://dl.acm.org/doi/10.1145/1807060.1807062
This paper quotes "Spectek. 2007. FNNM40A 16Gb NAND Flash memory data sheet."

[rockcandy]

Great discussions, insights, and information you have all provided above.
 
Seems like this thread lost momentum over the uncertainty and variability of the hardware inside various brands/models/iterations of the media players.
 
So I just wanted to point out, there exists a HUGE and consistent supply of complete finished devices, in new condition, which are ALWAYS powered by a Spreadtrum SC6531
 along with 4MB of RAM. You DO NOT need to order large quantities, you can buy just one, no problem.

They are CHEAP. They regularly go on sale for less than seven USD shipped, and their usual retail price is still under ten dollars shipped. Another good news: they even have
REMOVABLE rechargeable Lithium-ion batteries.

They are LEGIT, and readily available on Aliexpress. Furthermore, they're "Choice" items, meaning they're ready for immediate shipment, and will therefore arrive much faster. 
   
The catch is, they come in the form of a feature phone.

Yes, I'm talking about the 2G GSM phones. Typically they look like the kind of candy bar style Nokia handsets you may remember from the 2000's. Probably not the most attractive,
 nor very light, nor so small. But acceptable, super durable, and portable enough IMO.
 
I've bought a few of those before, set them up and given them away as FM radios. They were some of the cheapest FM radios I could find which came with a stereo 3.5mm
output jack and a rechargeable (800 to 1800 mAh) Li-ion battery that is actually managed by a proper power management unit.

I myself have been using Rockbox on a Sansa clip+, now with its battery health in decline. My search for the next Rockbox player has led me here.

------------------------------------------------------------------------------------------------------------------------------------------
One example: MKTEL OYE 3 - $6.22 sale price for the Red color (there's an Aliexpress "anniversary sale" going on now through March 27, 2024)
(save another 2% using Aliexpress coins) = $6.10

(Spreadtrum SC6531EFM), 1800 mAh battery. Rubberized plastic back plate. The main body looks and feels like ABS plastic. Large quantities (1200+ units available as of this writing, and they do get replenished regularly) between the 3 different colors

https://www.aliexpress.us/item/3256805015455298.html

You might get them for even less if it's offered to you as part of a personalized "Super Deals" on the homepage (search for it, add to cart, show some interest in the item etc., and eventually the algorithm might offer it to you). If you're new to Aliexpress, they may be bought as a "Welcome Deal" for as low as $1.24 (or about five dollars off). If you miss the anniversary sale, being Choice items, they'll be on sale again at the beginning of every month.

I never opened up the devices physically, but dialed the "secret" code below and it reported SC6531.

#*8378#4# (#*test#4#) - uptime/charge/IMEI/chipset information

<CHIP TYPE>
SC6531EFM_AB

The same seller (Desoon) also has MKTEL M2023 with a smaller (but not that much more portable) 800 mAh battery, for around the same price, in large quantities (1600+ units available). M2023 also reported the identical string for its chip type as OYE 3.

I have no affiliation with Aliexpress/Mktel/Desoon, other than having bought from them before. They just seem widely available on Aliexpress. There are also other brands such as PLUZZ and more, although it's probably wise to stick with the most available brand/model for consistency.

------------------------------------------------------------------------------------------------------------------------------------------

Another way to look at it: 2G feature phones must have sold in the hundreds of millions by now globally. It's ancient technology, to be honest. By market share, MediaTek and Spreadtrum/Unisoc appear to have a strong duopoly when it comes to 2G handset SoC. MediaTek MT626x family are ARM7, and Spreadtrum SC6531x family are ARM926EJ-S. These phones usually advertise either 32 "Mb" or "mb" or "MB" of RAM and ROM, more likely than not, it's the small b for bits, not Bytes. (32Mb = 4MB).
 
The mp4 players at the start of this thread may have been Shenzhen's attempt to leverage existing/surplus feature phone parts into creating newer sleeker media players based on the same 1.77 inch screen. Unfortunately their components were all over the place. I just thought it's probably worthwhile to target the phones themselves instead, given how abundant/available/affordable they are.

Thank you!

[osaka]

SC6531 DOOM gives a bit more insight. Yet there's someone to actually explore the platform :-)

[jpegqs]

I bought three cheap players, two were YP3 and one from Actions. Took apart one YP3, it has a Jointbees MP3 S91H13B6A0 chip, no flash chip.

Both YP3s can enter firmware update mode (ID 301a:2800) if a certain button is pressed while connected to USB, in one case it was the "left/previous track" button, in the other the "next track" button.

This brand uses the same platform, and supplies firmware updates for its devices: http://www.eroshifi.com/download/firmware/

The often mentioned "SetupFlashloaderSL-USR(6.X.X).exe" is a program for updating the firmware, SL seems to stand for Smartlink. Firmware for the SL6801 chip from the archive "固件升级-解决801读取音频文件最后几秒没声音的问题(6.7.7).zip" uses ARM Thumb2 instructions (ARMv6T2 or ARMv7-M architecture).

Someone might take the risk of using these updates to reverse engineer the update protocol.

SC6531 DOOM gives a bit more insight. Yet there's someone to actually explore the platform :-)

This is my project (FPDoom). But I don't share your optimism, I tried to do part of the work, hoping that someone would join me. But everyone only says that it's cool, but no one wants to participate in creating free software for this platform (which requires a lot of work on reverse engineering device drivers).

This github repository provides video conversion scripts for these players:
https://github.com/fdd4s/portable_mp3_player_video_converter_tools

Sadly, it says that players on Spreadtrum chips are discontinued: "MP4 Format - Portable MP3 Player Spreadtrum UNISOC 1.8 inch 128x160 branded as Rockstar Audio Player (2022 discontinued)"

[Bilgus]

I had code execution on on of these spreadtrum devices it wasn't a great Dev experience as i bricked the first one and had very little readable documentation and after some investigation it probably wouldn't be a great end user experience either due to lack of ram and terrible sound quality

[jpegqs]

I made a firmware dumper for yp3 players. The firmware update program is written in python and is quite easy to decompile.

The chip's bootloader has the ability to load binary code into memory and execute it (just like in feature phones).

[Bilgus]

But everyone only says that it's cool, but no one wants to participate in creating free software for this platform (which requires a lot of work on reverse engineering device drivers).

^we have a similar issue with finding people for things that don't even require that much effort

That said it takes a lot of work getting a port together so the more legwork towards that goal the more likely you'll get someone interested/hooked
but like I said I don't think you'll end up with something worthy of the work after some investigation of the platform

However I'm not going to tell you its not worth it and we'd gladly accept more devices to run our labor of love on..

[bahus]

I made a firmware dumper for yp3 players.

Nice! Did I get it right that it dumps firmware from mp3 to disk? Can this dump be flashed to the player somehow? I do have a bricked YP3 player (1.8 inch screen with round button), which isn't a big loss, but it's interesting to try to recover it.
It's still able to go into bootloader mode, so I believe it's still flashable.

[jpegqs]

The YP3 chip looks promising and convenient. There is a ROM, which contains basic functions from libc (like malloc and string functions) and other (specific to chip), you just need to figure out where what is.

The bootloader mode supports reading and writing flash, as well as writing and executing code from memory, but doesn't support reading memory. I managed to write a payload that reads the chip's memory, so I can analyze the ROM.

Floating point instructions are supported.

The capabilities of the chip (SL6801) are still unknown, what frequency it has and how much memory. 2M flash memory on two players that I have.

Nice! Did I get it right that it dumps firmware from mp3 to disk?

Yes.

Can this dump be flashed to the player somehow? I do have a bricked YP3 player (1.8 inch screen with round button), which isn't a big loss, but it's interesting to try to recover it.
It's still able to go into bootloader mode, so I believe it's still flashable.

Try to read the remains of the old firmware, maybe it still recoverable. Unless you were "lucky" to find the command to clear the entire chip (0xc0, 0x06, ...).

I know commands to write flash memory, but I haven't tried it. So my tool doesn't support this yet.

You can try to flash some firmware from the "eroshifi" website using their updater.