As a result of topics discussed here
http://www.diyaudio.com/forums/showthread.php?s=&threadid=24593&highlight=
and here
http://www.diyaudio.com/forums/showthread.php?s=&threadid=33041&perpage=15&pagenumber=1
and here
http://www.diyaudio.com/forums/showthread.php?s=&threadid=30016&perpage=15&pagenumber=1
...
My DAC shows first signs of live
The power supply board and the DAC with reciver works, no I/V stage ubtil now but jumpers set for Vout from AD1865
I feed from the CD670 i got from Jean Paul via SPDIF BNC cable.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=24593&highlight=
and here
http://www.diyaudio.com/forums/showthread.php?s=&threadid=33041&perpage=15&pagenumber=1
and here
http://www.diyaudio.com/forums/showthread.php?s=&threadid=30016&perpage=15&pagenumber=1
...
My DAC shows first signs of live
The power supply board and the DAC with reciver works, no I/V stage ubtil now but jumpers set for Vout from AD1865
I feed from the CD670 i got from Jean Paul via SPDIF BNC cable.
Attachments
No listening music, Vout from AD1865 driving 5m XLR cable to BZLS. The DACs first half is loosely based on http://www.geocities.com/yury_g/dac.htm, next will be D1 stage instead of Vout.
Power supply gives 5V with one 7805 for reciver, one for 74HC04, one for digital supply of DACs, the circuit form the zehner regulator thread above for +-5V DAC analog supply, +-30V for D1 stage with regulators similar to BZLS supply.
When i manage to set the right DAC mode in the CD PRO2 with a microcontroller i will connect the transport direct to the DACs instead of CS8412.
Power supply gives 5V with one 7805 for reciver, one for 74HC04, one for digital supply of DACs, the circuit form the zehner regulator thread above for +-5V DAC analog supply, +-30V for D1 stage with regulators similar to BZLS supply.
When i manage to set the right DAC mode in the CD PRO2 with a microcontroller i will connect the transport direct to the DACs instead of CS8412.
Till,
On the power supply pins of the DAC. In addition CBS has a test CD that has some low level signals so you can test the lowest order bits for noise. Of course most of that noise will come from the power supply and switching noise on ground when you connect the CD transport with the 8412. Of course you should monitor the DAC output while you look at the power supply noise.
You might want to sync on the master clk or SPDIF to see if your noise is related.
On the power supply pins of the DAC. In addition CBS has a test CD that has some low level signals so you can test the lowest order bits for noise. Of course most of that noise will come from the power supply and switching noise on ground when you connect the CD transport with the 8412. Of course you should monitor the DAC output while you look at the power supply noise.
You might want to sync on the master clk or SPDIF to see if your noise is related.
Till,
The noise on all of your scope shot’s is just too high.
Do you have separate ground for analog and digital and you DAC's. What I use on prototypes breadboards is 3M copper tape, it make a great ground plane. In addition, some OSCON's caps close to the DAC will help. If you using 78& &79 series regulators they need more filtering a PI filter will help. However, all is lost if the ground is not a low impedance.
You should have pulse transformers at both ends on your SPDIF. If you have a CAP to ground at your CD player or on the DAC that connects to your shield that will introduce noise, big time into the DAC. It causes circulating ground currents.
Make sure that your scope probe has a very short ground lead. A longer lead will make the reading look higher.
The noise on all of your scope shot’s is just too high.
Do you have separate ground for analog and digital and you DAC's. What I use on prototypes breadboards is 3M copper tape, it make a great ground plane. In addition, some OSCON's caps close to the DAC will help. If you using 78& &79 series regulators they need more filtering a PI filter will help. However, all is lost if the ground is not a low impedance.
You should have pulse transformers at both ends on your SPDIF. If you have a CAP to ground at your CD player or on the DAC that connects to your shield that will introduce noise, big time into the DAC. It causes circulating ground currents.
Make sure that your scope probe has a very short ground lead. A longer lead will make the reading look higher.
I made ground like Analog Devices told me in the datasheet: separate at the DAC and connected at the power supply board.
The ground clip of my probe is 150mm long. Its a 150MHz Testec probe, i have no better one. All the noise shown is only visible with the 7904, the slow 25MHz bandwith 7313 does not shows this stuff. In the CD player i did not change anything, only a BNC socket instead of the corroded old RCA for SPDIF. I have no pulse transformers, so i have to use what they call "consumer interface" in the CS8412 datasheet until i find something better.
I use 7805 for the digital supplys only, not the DACs analog. All supplys are filtered. The Analog with the circuit as shown in the thread for the power supply board linked above. At the DAC board with 100nF + 100uF at connector, 100uH/ 1.7Ohm series, 10uF + 100nF at DAC.
The 7805 at PS board with 330nF at in and out blocked, some electrylytic at in, at DAC board with 100uF, 100uH+1,7Ohm series, 10uF+100nF at each IC pin. Of corse i can make some additional RC into this.
The CDP has no real ground, only two pin mains plug. The DAC i also have not connected to real earth ground - only 2 pin mains plug.
I understand i will have to make a groundplane from all unused area on the board with some copper foil, but i will do this when everything works and i´m sure i don´t change parts anymore.
The ground clip of my probe is 150mm long. Its a 150MHz Testec probe, i have no better one. All the noise shown is only visible with the 7904, the slow 25MHz bandwith 7313 does not shows this stuff. In the CD player i did not change anything, only a BNC socket instead of the corroded old RCA for SPDIF. I have no pulse transformers, so i have to use what they call "consumer interface" in the CS8412 datasheet until i find something better.
I use 7805 for the digital supplys only, not the DACs analog. All supplys are filtered. The Analog with the circuit as shown in the thread for the power supply board linked above. At the DAC board with 100nF + 100uF at connector, 100uH/ 1.7Ohm series, 10uF + 100nF at DAC.
The 7805 at PS board with 330nF at in and out blocked, some electrylytic at in, at DAC board with 100uF, 100uH+1,7Ohm series, 10uF+100nF at each IC pin. Of corse i can make some additional RC into this.
The CDP has no real ground, only two pin mains plug. The DAC i also have not connected to real earth ground - only 2 pin mains plug.
I understand i will have to make a groundplane from all unused area on the board with some copper foil, but i will do this when everything works and i´m sure i don´t change parts anymore.
Till,
If you plan to integrate the DAC into the player maybe this is not such a big deal since the DAC and Player will be command or stared to the same ground.
If you do not then will should consider changing to 75-ohm SPDIF interface that has transformer at both ends. Burr Brown makes some isolation chips that you can put between the digital hardware and the DAC to isolate the noise I will have to look for the part number. It is kind of like using an isocoupler but has lower noise
What kind of transport are you using and does it have a transformer? It really not to hard to change the receiver the output player may be a little harder. A lot of this noise is common mode noise and can be reduced with a differential inductors or a common mode filter for the supply and ground.
If you plan to integrate the DAC into the player maybe this is not such a big deal since the DAC and Player will be command or stared to the same ground.
If you do not then will should consider changing to 75-ohm SPDIF interface that has transformer at both ends. Burr Brown makes some isolation chips that you can put between the digital hardware and the DAC to isolate the noise I will have to look for the part number. It is kind of like using an isocoupler but has lower noise
What kind of transport are you using and does it have a transformer? It really not to hard to change the receiver the output player may be a little harder. A lot of this noise is common mode noise and can be reduced with a differential inductors or a common mode filter for the supply and ground.
thanks jewilson,
At the moment i use a Philps CD670 spdif output, i have to look inside how spdif out is made.
I did not find oscon at the parts supply i ususaly buy, whats the speacial thing about oscon?
I use the spdif at the moment because its less money to kill a CS8412 while prototyping than kill a CD transport. My plan is to build a CD player with balanced DAC and CD PRO2 or any DSA transport. I don´t see the advantage of external DACs , and integrated i will be able to make short connections, do not use the reciver IC , and have a hopefully better clock than spdif. To me all the reclocking is some kind of dubios, i would prefer to use only one clock that is allready there (the transports), and have everything near to each other. The reason for all this is i want a balanced out CD player as without a balanced source it doesn´t make to much sense to me to go for balanced preamp and amplifiers.
I will use this unit,
http://www.daisy-laser.nl/homeoptics/
and try to tell it to give me the data in the format the analog devices dac wants. So i don´t need reciver chip for converting the signals format anymore.
I split this "project" into some parts, first power supply, now DAC working, next I/V stage. I will need to get a controller for the DSA CD transport working,some assembly language coding necessary, and then mate CD transport with the DAC.
The isolation amplifiers i found at TI had not much bandwith?
At the moment i use a Philps CD670 spdif output, i have to look inside how spdif out is made.
I did not find oscon at the parts supply i ususaly buy, whats the speacial thing about oscon?
I use the spdif at the moment because its less money to kill a CS8412 while prototyping than kill a CD transport. My plan is to build a CD player with balanced DAC and CD PRO2 or any DSA transport. I don´t see the advantage of external DACs , and integrated i will be able to make short connections, do not use the reciver IC , and have a hopefully better clock than spdif. To me all the reclocking is some kind of dubios, i would prefer to use only one clock that is allready there (the transports), and have everything near to each other. The reason for all this is i want a balanced out CD player as without a balanced source it doesn´t make to much sense to me to go for balanced preamp and amplifiers.
I will use this unit,
http://www.daisy-laser.nl/homeoptics/
and try to tell it to give me the data in the format the analog devices dac wants. So i don´t need reciver chip for converting the signals format anymore.
I split this "project" into some parts, first power supply, now DAC working, next I/V stage. I will need to get a controller for the DSA CD transport working,some assembly language coding necessary, and then mate CD transport with the DAC.
The isolation amplifiers i found at TI had not much bandwith?
The OSCON or a Organic type of electrolytic cap, they are very low ESR, good for removing switching noise. Of course, you should have some .01 ceramics to decouple the higher frequencies.
One of the main reasons to get the DAC out of the CDP box is due to digital switching noise. However, if the SPDIF interface is implemented poorly the performance will be bad, as you seen.
You can implement the SPDIF 75ohm interface with transformers it is not that big of a deal. I use a PS audio Lambda transport, which has a balance digital drive output using 74ALS74's with some resistors and pulse transformer.
If you want to check out the OSCON the one's I use are make by www.vishay.com and can be found with their electrolytic caps. The only better cap for this application is a low ESR tantalum which cost about 5x the price of the OSCON's also Sanyo makes the too.
One of the main reasons to get the DAC out of the CDP box is due to digital switching noise. However, if the SPDIF interface is implemented poorly the performance will be bad, as you seen.
You can implement the SPDIF 75ohm interface with transformers it is not that big of a deal. I use a PS audio Lambda transport, which has a balance digital drive output using 74ALS74's with some resistors and pulse transformer.
If you want to check out the OSCON the one's I use are make by www.vishay.com and can be found with their electrolytic caps. The only better cap for this application is a low ESR tantalum which cost about 5x the price of the OSCON's also Sanyo makes the too.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.