Hi, I'm the guy who made the Cd723's hacking thing.
I'm not very good in english, so I hope you will understand.
The page is actually in french, but I could translate it if you're interested.
Until then, let me explain a few thing about it :
As you understand well, the goal of the device is to modify the SAA7378 dac output interface format. To allow this, it changes some commands sended by the microcontroler which is in the front card of the cd player.
This commands are shortly described in the SAA737X datasheet. Althought, I didn't get the SAA7378 one, I got the SAA7377 and the SAA7376 datasheets. I found that the ony difference betwen them is the default configuration (mainly dac output interface).
Then I assume it is the same for the SAA7378 (OliverD is perfectly right) so I could use the same commands and it works.
I've got 2 evidences :
- I monitore the frequency of the SCLK signal and when you force 1fs you get about 2Mhz as it is written in the datasheet (and the sound is quite better too
)
- I plug my deply modified cd-4000 player (have a look on http://www.ifrance.com/robinetron/cd723/drive/drive.html sorry it's in french) on my home theather amplifier and put an DD5.1 cd. With the hacking, the signal is perfectly recognised and decoded by the amplifier so you really have a unmodified digital signal on SPDIF output (without this ugly 0.5db attenuation).
Anyway, thank you for "this guy is really clever"
I'm not very good in english, so I hope you will understand.
The page is actually in french, but I could translate it if you're interested.
Until then, let me explain a few thing about it :
As you understand well, the goal of the device is to modify the SAA7378 dac output interface format. To allow this, it changes some commands sended by the microcontroler which is in the front card of the cd player.
This commands are shortly described in the SAA737X datasheet. Althought, I didn't get the SAA7378 one, I got the SAA7377 and the SAA7376 datasheets. I found that the ony difference betwen them is the default configuration (mainly dac output interface).
Then I assume it is the same for the SAA7378 (OliverD is perfectly right) so I could use the same commands and it works.
I've got 2 evidences :
- I monitore the frequency of the SCLK signal and when you force 1fs you get about 2Mhz as it is written in the datasheet (and the sound is quite better too
)- I plug my deply modified cd-4000 player (have a look on http://www.ifrance.com/robinetron/cd723/drive/drive.html sorry it's in french) on my home theather amplifier and put an DD5.1 cd. With the hacking, the signal is perfectly recognised and decoded by the amplifier so you really have a unmodified digital signal on SPDIF output (without this ugly 0.5db attenuation).
You're right, I also read the old SAA7345 and TDA1301 datasheet but it didn't help much. The problem is the CD723 uses only the "4 wire bus" with only 3 wire to discuss with the SAA7378. The bus only allows the use of Servo command and the new Servo commands which have an effect on the output interface aren't described in the old datasheet.
Anyway, thank you for "this guy is really clever"
I didn't get it. Did you remove "nospam_" in the mail address ?
Hi Robinet,
good you came here to clear up some issues.
If I could only understand french, I'd be happy to assist you with the translation of your site - it is very interesting and should be available to more people!
So, to summarize about the CD 723:
1. It is a very cheap and very simple design which makes it interesting for DIY people as a base for modding. Used ones can be bought off ebay for around 40 € including shipping. They even play CD-R/Ws.
2. In the original state, it is not bit accurate due to some 0.5dB downscaling, and a very mediocre oversampling filter is further compromising sound quality. Thus, it is not a good base for modding if you just want to drop in some better caps, DAC or power supply.
3. The servo/decoder chip is a member of Philips' CD 7 family, and although the non-oversampling mode is not mentioned in the datasheet, it is available (and also disables the downscaling). Thus, the CD723 can be made a bit-accurate source of I2S data, but this involves hacking into the communication between the servo/decoder chip and the player's controller.
4. Robinet has a working implementation of this hack - to be seen on his homepage (scroll up to find the link).
5. Alternatively, wait until I designed my replacement drive and user interface PCB to basically build a CD player from scratch, only using the drive mechanics and the servo/decoder chip.
good you came here to clear up some issues.
If I could only understand french, I'd be happy to assist you with the translation of your site - it is very interesting and should be available to more people!
So, to summarize about the CD 723:
1. It is a very cheap and very simple design which makes it interesting for DIY people as a base for modding. Used ones can be bought off ebay for around 40 € including shipping. They even play CD-R/Ws.
2. In the original state, it is not bit accurate due to some 0.5dB downscaling, and a very mediocre oversampling filter is further compromising sound quality. Thus, it is not a good base for modding if you just want to drop in some better caps, DAC or power supply.
3. The servo/decoder chip is a member of Philips' CD 7 family, and although the non-oversampling mode is not mentioned in the datasheet, it is available (and also disables the downscaling). Thus, the CD723 can be made a bit-accurate source of I2S data, but this involves hacking into the communication between the servo/decoder chip and the player's controller.
4. Robinet has a working implementation of this hack - to be seen on his homepage (scroll up to find the link).
5. Alternatively, wait until I designed my replacement drive and user interface PCB to basically build a CD player from scratch, only using the drive mechanics and the servo/decoder chip.
Carondimonio said:
Hi Carondimonio,
I guess my way of writing doesn't make the job easy (I'm really impressed by the fact babelfish is able to make it quite readable !) so thank you for translating and of course I don't mind you post it here. It is a good idea !
Anyway I can also add it on my web page (if you allow me
).Hi Robinet
Just checked back to this thread - great that you are here -
Excellent work
I tried to do the same thing about a year ago by using DSA commands and a PC program downloadable from the DaisyLaser site but failed to get anywhere - this seemed like the most direct way for initial comparison of non-oversampling settings(controlling CD4000 using PC)
Your method , however is permanent and more suited to long term use.
Are you sure that the CD4000 (CD723) is bit accurate once 1fs is enabled?
Than you again for openeing up this work to the public - tyhis will make the CD723 a very good platform for modding , etc
John
Just checked back to this thread - great that you are here -
Excellent work
I tried to do the same thing about a year ago by using DSA commands and a PC program downloadable from the DaisyLaser site but failed to get anywhere - this seemed like the most direct way for initial comparison of non-oversampling settings(controlling CD4000 using PC)
Your method , however is permanent and more suited to long term use.
Are you sure that the CD4000 (CD723) is bit accurate once 1fs is enabled?
Than you again for openeing up this work to the public - tyhis will make the CD723 a very good platform for modding , etc
John
Hi
I forgot to say importants things about my hacker :
- I never use the volume control of the cd-player and thus I don't know the effect it has with the hacker. I think the hacker won't work but there also might be little trouble (don't worry it won't break anything).
- Sometime (let say 1 out 1000), the cd-player read the cd 2 times faster. It happens only with old cd so the problem may come from the mechanic but it must be said.
- The source code is avaible so feel free to improve it !
Beware DSA bus is only available with cd-pro mechanics. The ECO-MKIII (used in CD-723) use "4 wire bus" which is quite different from DSA.
Yes I am, because I'm able to play DD and DTS cd with it.
I forgot to say importants things about my hacker :
- I never use the volume control of the cd-player and thus I don't know the effect it has with the hacker. I think the hacker won't work but there also might be little trouble (don't worry it won't break anything).
- Sometime (let say 1 out 1000), the cd-player read the cd 2 times faster. It happens only with old cd so the problem may come from the mechanic but it must be said.
- The source code is avaible so feel free to improve it !
I used"ultra low cost AVR programmer" and ICprog software. Yes, I used AT90s1200 and it was also my first time with microcontrollers. Because everything was very well described it took me just a few hours from building the "programmer" to finised project.
You could use that display.
Anyone needs a spare display for their CD723?
I decided to use a Crystalfontz PLED display (see image below).I will only use a few parts from the original player, some of which are not easy to source - it's best to buy them in a cheapo player of some sort (You get an ESD protected packaging and built-in self-test, too
)- VAM1201 drive mechanism
- Drive Foil Connector
- Infrared Remote Sensor
- SAA7378
- TDA1302T
- Maybe some of the drivers (I think I'll go balanced to keep the ground clean)