Subject
As you know digital output of SACD and DVD-A Players is restricted due to copyright reasons. This leads to the funny development that the players have to output the hires audio in analog via their built in D/A converters. So the quality of the integrated D/A converters determine the quality of the output. Typically the analog signal in the surround receiver is converted back to digital and then again back in the analog domain. All these conversions lead to a minor sound quality defined by the weakest element of your chain.
My goal is keeping the signal as long as possible in the digital domain eliminating the effects of intermediate D/A-A/D conversion.
Ingredients
Some time ago pioneers started to sell the DV-575A player in Germany which is – despite of the SCART-adapter – exactly the same unit as the DV-578A sold elsewhere in the word. This player is cheap (<200Euro) and has one UNIQUE feature. It converts the DSD-signal hosted on SACDs digitally to 88.2kHz PCM (DSD to PCM conversion). With this trick inexpensive D/A converters (PCM1742) can be used in the unit.
This feature is so remarkable because the DV-575A (578A) is the only player I know who sends ONLY PCM signals to the D/A converters. My idea was to convert this PCM signal with a hand of chips to S/PDIF to attain a true digital output independent from the media in the player.
Beyond the player as a first prototype I decided to use the digital transmitter of the Hoontech DSP24 Soundcard. Hoontech offered a so called "Value Bracket" a small PCB which could be connected to the soundcard. The Value Bracket accepts I2S digital input and outputs S/PDIF up to 24/96. It does not support 192kHz and so it is not the right solution for the final board but for first tests and those needing only stereo up to 24/96 it shall be sufficient. Currently the Value Bracket is out of production but with a little luck you can find it at Ebay or electronic stores. From its dimensions it fits perfectly inside the cabinet of the DV-575A.
Finally a bit of electronics has to be built to bring together the signals from the DV-575A (578A) and the Value Bracket.
Preparation
First take a look at the images I placed here (click them to increase)
http://freerider.dyndns.org/anlage/images-Pioneer-DV575A/index.html
At the first image you see the Hoontech Value Bracket. It has a lot of connectors for digital I/O and MINI. To insert the board in the DV-575A (578A) all the connectors have to be removed. The next image shows the result.
The Value Bracket is designed for I2S input but the Pioneer outputs a left justified signal. To accept this left justified signal the Value Bracket has to be modified. On the image you see two little red circles around pin 21 and 23 of the CS8404 transmitter. These two pins have to be disconnected from the PCB and reconnected to VCC and GND as shown. To disconnect the pins best push a needle under the pin and carefully heat the pin with the soldering iron. If the solder melts you can easily bend up the pin. Next image shows how to reconnect the two pins.
Having prepared the Value Bracket I inserted a new connector (golden one, see next image) at the back of the DV-575A. There is a hole in the metal rear panel of the DV-575A covered by plastic. If you cut a hole in the plastic the connector perfectly fits in there. There is no need to drill a hole in the metal rear panel of the DV-575A. I didn’t know if it is that easy with the DV-578A.
Next I drilled two holes in the bottom of the DV-575A chassis and inserted the Value Bracket with some screws and spacers (see next image). For the connection between the Value Bracket and the cinch-connector I used a 75 Ohm wire.
On that image you also see a ribbon cable which gathers up the signals from the DV-575A (578A) board. Looking at the datasheet of the PCM1742 it was not too difficult to figure out the connection points. Fortunately all signals are accessible on little vias. I inserted the wires of the ribbon cable in the vias and soldered them. For this task you have to use a magnifying-glass and a spiky soldering iron and a lot of care because the vias are very very small. You can see at the next images which wire has to be connected to which via. All even wires (except Pin2) of the ribbon cable are open on the side of the DV-575A (578A) and connected to ground on the other side. This is important to minimize capacitive influence between the signals.
Piggyback PCB
What still has to be done is a little PCB connecting the ribbon cable with the Value Bracket. In my research I found out a lot about the timing of the signals, those coming from the PCM1742 and those needed by the Value Bracket. There is no way to connect the ribbon cable directly to the value bracket. The reason is the timing of the system clock (sck). The CS8404 needs a system clock of 128 x Fs but the PCM1742 delivers quite different clock ratios depending of the type of medium in the player.
Looking at the schematics (last image) there is a HCT244 driver on the left side which attenuates the signals coming from the DV-575A (578A). This is important because the player works with 3.3V TTL and the Value Bracket with 5V. After the HCT244 the signals are routed to a HCT373 Latch which is needed to adjust the signals to the new system clock.
The most sophisticated chip is the one which checks the input signals and outputs the new system clock. To avoid a lot of TTL logic I took a programmable logic device (PALCE22V10 or GAL22V10) and put the logic in there. This chip has not only an output for the new system clock but also an output for two LEDs. The white LED indicates a 192kHz media, the blue one 88.2kHz SACD or 96Khz DVA-A output. Once again: With this prototype only signals up to 24/96 can be output. The piggyback PCB operates the 192kHz but the Value Bracket could not output it.
These three logic chips were soldered together with a voltage regulator (7805) and connectors for the ribbon cable and the Value Bracket on the piggyback board. The piggy back board is plugged in the Value Bracket and …. cross fingers…. It works.
The programmable logic device
Well how to get this PAL or GAL chip? I have not decided yet if I might commercialize my development of the digital output board. But if, I guess it would be a good idea to protect my intellectual property which is inside the logic of the chip (sorry). On the other hand I would like to participate as much DIY people as possible in my development. So what I will do is selling you a functional chip or burn a chip for you if you send an empty one to me. If you are interested just mail and we will find a way.
Setup of the DV-575A (578A)
To get out a high resolution audio signal you have to DISABLE the standard digital output of the DV-575A (578A). This is necessary because with an enabled digital output the 575 only sends limited (48kHz) signals to the D/A converters which means limited signals to the Value Bracket.
What next?
Having finished this prototype now it will create a S/PDIF signal from 44,1 to 96KHz depending on the disk in the player. SACD will be output at 88.2kHz, DVD at 48 and DVD-Audio 48 or 96kHz. The DSP Value Bracket does not support 192kHz signals. This is the only drawback of the prototype.
I plan to build a PCB which also supports 192kHz and all three digital output channels for the whole 5.1 Sound but this is a bit far in the future maybe in spring …..
So have fun with my prototype and mail me if you need the chip.
Merry Christmas
Charly
As you know digital output of SACD and DVD-A Players is restricted due to copyright reasons. This leads to the funny development that the players have to output the hires audio in analog via their built in D/A converters. So the quality of the integrated D/A converters determine the quality of the output. Typically the analog signal in the surround receiver is converted back to digital and then again back in the analog domain. All these conversions lead to a minor sound quality defined by the weakest element of your chain.
My goal is keeping the signal as long as possible in the digital domain eliminating the effects of intermediate D/A-A/D conversion.
Ingredients
Some time ago pioneers started to sell the DV-575A player in Germany which is – despite of the SCART-adapter – exactly the same unit as the DV-578A sold elsewhere in the word. This player is cheap (<200Euro) and has one UNIQUE feature. It converts the DSD-signal hosted on SACDs digitally to 88.2kHz PCM (DSD to PCM conversion). With this trick inexpensive D/A converters (PCM1742) can be used in the unit.
This feature is so remarkable because the DV-575A (578A) is the only player I know who sends ONLY PCM signals to the D/A converters. My idea was to convert this PCM signal with a hand of chips to S/PDIF to attain a true digital output independent from the media in the player.
Beyond the player as a first prototype I decided to use the digital transmitter of the Hoontech DSP24 Soundcard. Hoontech offered a so called "Value Bracket" a small PCB which could be connected to the soundcard. The Value Bracket accepts I2S digital input and outputs S/PDIF up to 24/96. It does not support 192kHz and so it is not the right solution for the final board but for first tests and those needing only stereo up to 24/96 it shall be sufficient. Currently the Value Bracket is out of production but with a little luck you can find it at Ebay or electronic stores. From its dimensions it fits perfectly inside the cabinet of the DV-575A.
Finally a bit of electronics has to be built to bring together the signals from the DV-575A (578A) and the Value Bracket.
Preparation
First take a look at the images I placed here (click them to increase)
http://freerider.dyndns.org/anlage/images-Pioneer-DV575A/index.html
At the first image you see the Hoontech Value Bracket. It has a lot of connectors for digital I/O and MINI. To insert the board in the DV-575A (578A) all the connectors have to be removed. The next image shows the result.
The Value Bracket is designed for I2S input but the Pioneer outputs a left justified signal. To accept this left justified signal the Value Bracket has to be modified. On the image you see two little red circles around pin 21 and 23 of the CS8404 transmitter. These two pins have to be disconnected from the PCB and reconnected to VCC and GND as shown. To disconnect the pins best push a needle under the pin and carefully heat the pin with the soldering iron. If the solder melts you can easily bend up the pin. Next image shows how to reconnect the two pins.
Having prepared the Value Bracket I inserted a new connector (golden one, see next image) at the back of the DV-575A. There is a hole in the metal rear panel of the DV-575A covered by plastic. If you cut a hole in the plastic the connector perfectly fits in there. There is no need to drill a hole in the metal rear panel of the DV-575A. I didn’t know if it is that easy with the DV-578A.
Next I drilled two holes in the bottom of the DV-575A chassis and inserted the Value Bracket with some screws and spacers (see next image). For the connection between the Value Bracket and the cinch-connector I used a 75 Ohm wire.
On that image you also see a ribbon cable which gathers up the signals from the DV-575A (578A) board. Looking at the datasheet of the PCM1742 it was not too difficult to figure out the connection points. Fortunately all signals are accessible on little vias. I inserted the wires of the ribbon cable in the vias and soldered them. For this task you have to use a magnifying-glass and a spiky soldering iron and a lot of care because the vias are very very small. You can see at the next images which wire has to be connected to which via. All even wires (except Pin2) of the ribbon cable are open on the side of the DV-575A (578A) and connected to ground on the other side. This is important to minimize capacitive influence between the signals.
Piggyback PCB
What still has to be done is a little PCB connecting the ribbon cable with the Value Bracket. In my research I found out a lot about the timing of the signals, those coming from the PCM1742 and those needed by the Value Bracket. There is no way to connect the ribbon cable directly to the value bracket. The reason is the timing of the system clock (sck). The CS8404 needs a system clock of 128 x Fs but the PCM1742 delivers quite different clock ratios depending of the type of medium in the player.
Looking at the schematics (last image) there is a HCT244 driver on the left side which attenuates the signals coming from the DV-575A (578A). This is important because the player works with 3.3V TTL and the Value Bracket with 5V. After the HCT244 the signals are routed to a HCT373 Latch which is needed to adjust the signals to the new system clock.
The most sophisticated chip is the one which checks the input signals and outputs the new system clock. To avoid a lot of TTL logic I took a programmable logic device (PALCE22V10 or GAL22V10) and put the logic in there. This chip has not only an output for the new system clock but also an output for two LEDs. The white LED indicates a 192kHz media, the blue one 88.2kHz SACD or 96Khz DVA-A output. Once again: With this prototype only signals up to 24/96 can be output. The piggyback PCB operates the 192kHz but the Value Bracket could not output it.
These three logic chips were soldered together with a voltage regulator (7805) and connectors for the ribbon cable and the Value Bracket on the piggyback board. The piggy back board is plugged in the Value Bracket and …. cross fingers…. It works.
The programmable logic device
Well how to get this PAL or GAL chip? I have not decided yet if I might commercialize my development of the digital output board. But if, I guess it would be a good idea to protect my intellectual property which is inside the logic of the chip (sorry). On the other hand I would like to participate as much DIY people as possible in my development. So what I will do is selling you a functional chip or burn a chip for you if you send an empty one to me. If you are interested just mail and we will find a way.
Setup of the DV-575A (578A)
To get out a high resolution audio signal you have to DISABLE the standard digital output of the DV-575A (578A). This is necessary because with an enabled digital output the 575 only sends limited (48kHz) signals to the D/A converters which means limited signals to the Value Bracket.
What next?
Having finished this prototype now it will create a S/PDIF signal from 44,1 to 96KHz depending on the disk in the player. SACD will be output at 88.2kHz, DVD at 48 and DVD-Audio 48 or 96kHz. The DSP Value Bracket does not support 192kHz signals. This is the only drawback of the prototype.
I plan to build a PCB which also supports 192kHz and all three digital output channels for the whole 5.1 Sound but this is a bit far in the future maybe in spring …..
So have fun with my prototype and mail me if you need the chip.
Merry Christmas
Charly
converting DSD to PCM for conversion...
didn't they say that DSD was far better because it involves less stages in the AD/DA conversions?
http://www.superaudio-cd.com/technology_explained/sacd_in_plain_english/
didn't they say that DSD was far better because it involves less stages in the AD/DA conversions?
http://www.superaudio-cd.com/technology_explained/sacd_in_plain_english/
You are right. DSD to PCM conversion may lead to a decrease in signal quality.Bricolo said:
The best alternative is using a very good SACD player and a pure analog chain which has to be good enough to get all reproduced the SACD player delivers. The limiting factor of this setup is the quality of the D/A converter in the SACD player. This means in this scenario you have to invest a lot in the player to get really good results.
Using any type of digital gear in your chain always leads to a conversion to PCM. In my chain I use a Behringer DCX2496 as digital x-over.
See: http://www.diyaudio.com/forums/showthread.php?s=&threadid=15943
Playing SACD on a player with analog output will lead to some signal conversions:
- in the Player: DSD to Analog
- in the DCX2496 Analog to PCM and PCM to Analog
To avoid these conversions it is a lot better to convert DSD to PCM in the digital domain and send the digital signal to the DCX2496. Here it will be processed in the digital domain and transformed back to analog. With this configuration it looks like this:
- in the player: DSD to PCM (conversion in the digital domain)
- in the DCX2496 PCM to Analog
The only drawback with the DC-575A is that it outputs SACD only with 88.2kHz. This is a bit less than my DCX2496 (24/96).
To compare the SACD sound of both configurations I used my SONY DVP-NS930V (quite good gear) with the first configuration (analog) and the Pioneer DV-575A with the new digital output in the second configuration. Both players were directly connected to the DCX2496 (no preamp!). The difference was small but audible. Bass was the same. The digital configuration has more precision in the mid and high regions and gives a better room impression. The stage gets a bit wider and deeper.
The difference between these two configurations is not big but the Pioneer is much cheaper than the Sony so I guess it was a good idea to mod it to a digital transport.
I did not now exactly how. I guess it is integrated inside the big chip of the DV-575A.Bricolo said:
As far as I kow there is only one diy chip on the market who does this conversion. It is from NPC and is called SM5819AF. To make this mod for other players than the DV-575A (578) you have to use this chip or a self programmed DSP ;-))
BUT this chip is very difficult to get. I guess due to copyright reasons they are not allowed to sell it to diy-ers. I know a dealer in italy where I could buy them (qantity >15). So some time ago I asked for interested people in the group buy forum
http://www.diyaudio.com/forums/showthread.php?s=&threadid=41191
but got no reply.
This means: Take the DV-575A (578) or die.....
jwb
I've been reading some of your posts and figure you would be a good person to comment on the "El Cheapo" DVD mod candidate thread. From what I've been reading, the quality of the result is highly dependent on the tuning of the OPU to the accompanying chipset. If this is so, then is it impractical to even consider replacing an OPU on a premium player of yesteryear even if it were possible? The way I figure it, the transport and power supply is worth more on these older players than the OPU and supporting logic since the latter will be going through at least 50 more iterations before all seems to settle.
I've been reading some of your posts and figure you would be a good person to comment on the "El Cheapo" DVD mod candidate thread. From what I've been reading, the quality of the result is highly dependent on the tuning of the OPU to the accompanying chipset. If this is so, then is it impractical to even consider replacing an OPU on a premium player of yesteryear even if it were possible? The way I figure it, the transport and power supply is worth more on these older players than the OPU and supporting logic since the latter will be going through at least 50 more iterations before all seems to settle.
Re: Interested but unable to see images
Sorry my server was offline for some days because my router and DSL modem had died. Currently it is up again and you can visit my pages between 8:00 am an 10:00 pm German time. I am sorry about this time restriction but this is a tribute to my wife's power saving mania ;-))whodiini said:
Actually there are more units
Charly
You mentioned you thought the 575 was unique... Well not quite. It turns out that Sony has a bunch of "dream machines" that are sacd receivers with digital amplifiers. (for example AVD-S500es, AVD-C700ES,...) The digital amplifiers take in PCM, and Sony converts the DSD from SACD to 88/24 PCM prior to sending the PCM to the digital amp, bypassing any analog conversion. In any case, the pioneer is a universal player and does DVD_A, whereas the Sony does not.
BTW, do you have a schematic of the 575/578 that you could post? Havent found that anywhere on the web.
Charly
You mentioned you thought the 575 was unique... Well not quite. It turns out that Sony has a bunch of "dream machines" that are sacd receivers with digital amplifiers. (for example AVD-S500es, AVD-C700ES,...) The digital amplifiers take in PCM, and Sony converts the DSD from SACD to 88/24 PCM prior to sending the PCM to the digital amp, bypassing any analog conversion. In any case, the pioneer is a universal player and does DVD_A, whereas the Sony does not.
BTW, do you have a schematic of the 575/578 that you could post? Havent found that anywhere on the web.
Re: Actually there are more units
Well, I am also looking for the schnematic with so success but I am strongly interested. If you got it please let me know.
You are right but the digital signal is not accessible in any way. With all units I know it is transferred to the receiver via Fire Wire or any proprietary protocol. Any way it is encrypted. This means no chance getting out an S/PDIF.whodiini said:
Well, I am also looking for the schnematic with so success but I am strongly interested. If you got it please let me know.
For anyone that might still be interested in this project (I am going to attempt soon) I found the SB Digital I/O bracket II from:
Pete Larson
Audience Digital Products
pete@audiencedp.com
http://www.audiencedp.com
He said he had some left. His website is undergoing revision right now so no online shopping but you can shoot him an email and he is responsive.
Pete Larson
Audience Digital Products
pete@audiencedp.com
http://www.audiencedp.com
He said he had some left. His website is undergoing revision right now so no online shopping but you can shoot him an email and he is responsive.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.