I believe it's for the common-mode rejection:
"The following pairs of resistors must be matched to a tolerance of ~0.1%:
RD22-RD24
RD26-RD28
R604-R602
RD21-RD23
RD25-RD27
R603-R601
R602 and R601 can be replaced with trim pots to fine-tune the CMR but this is probably not necessary.
In order to adjust, disconnect RD22 RD24 from the DA
converter.
Apply a sine wave to both resistors and adjust R602 for lowest
output (pin 1). Then do the same for RD21 and RD23 adjusting R601."
It made a big difference in sound and especially depth...
"The following pairs of resistors must be matched to a tolerance of ~0.1%:
RD22-RD24
RD26-RD28
R604-R602
RD21-RD23
RD25-RD27
R603-R601
R602 and R601 can be replaced with trim pots to fine-tune the CMR but this is probably not necessary.
In order to adjust, disconnect RD22 RD24 from the DA
converter.
Apply a sine wave to both resistors and adjust R602 for lowest
output (pin 1). Then do the same for RD21 and RD23 adjusting R601."
It made a big difference in sound and especially depth...
Common-mode rejection
The common-mode signal being digital switching noise etc. I think I'll probably go with matched 0.1%. Adding a trimpot is fiddly, and I'm not sure how much room there is. We shall see.
Next step is to look up the AD825 and LM6172 data sheets. I'm surprised nobody suggested OPA2134.
The common-mode signal being digital switching noise etc. I think I'll probably go with matched 0.1%. Adding a trimpot is fiddly, and I'm not sure how much room there is. We shall see.
Next step is to look up the AD825 and LM6172 data sheets. I'm surprised nobody suggested OPA2134.
I use the 2134 and 2604 and can't make my mind up. Seems to me that the 134 sounds softer and more detailed (contradiction?) 
When I get around to it I'll throw in a 627 (I only use the first opamp) on a veroboard and see what happens, but I guess the 12V supply needs to be a cleaner first...
When I get around to it I'll throw in a 627 (I only use the first opamp) on a veroboard and see what happens, but I guess the 12V supply needs to be a cleaner first...
miguel2 said:
The service manual doesn't mention CD48, but it might only be tiny differences. It's far more likely to be similar than not, so why not download it and compare? If you can read a circuit diagram and trace circuits, the manual should give you enough clues. All these circuits are likely to be pretty similar since they all perform the same function.
Ergo suggested above to use the alst two FET of the HDAM to improve the output capabilities:
Is there any reason for not cutting the trace AFTER RH17/18? What about the "feedback" to QH01/02 from before RH23/24?
While I am at it:
I have read about the 5V noise in the CD63 and now I wonder if anyone out there has any concrete ideas for a post-regulator (or alternative supply) for the 5 and 12V rails? Veroboard would be convinient as I don't have any good lab facilities.
I have been thinking about using the holes for the remote control rca contact on the back as an input for separate supplies. I believe there are voltage input contacts that have a built-in switch that breaks a circuit when a plug is inserted. In this way the on-board rails are used when no plug is inserted and switched off when an out-board supply is connected. What do you think about that arrangement?
Cheers,
Tom
Is there any reason for not cutting the trace AFTER RH17/18? What about the "feedback" to QH01/02 from before RH23/24?
While I am at it:
I have read about the 5V noise in the CD63 and now I wonder if anyone out there has any concrete ideas for a post-regulator (or alternative supply) for the 5 and 12V rails? Veroboard would be convinient as I don't have any good lab facilities.
I have been thinking about using the holes for the remote control rca contact on the back as an input for separate supplies. I believe there are voltage input contacts that have a built-in switch that breaks a circuit when a plug is inserted. In this way the on-board rails are used when no plug is inserted and switched off when an out-board supply is connected. What do you think about that arrangement?
Cheers,
Tom
Not concrete, but thick custard
LM317 (+ve) and LM337 (-ve) are adjustable regulators that produce less noise than most of the fixed regulators. Do a search, and you will find from the manufacturer's data sheets that they are very easy to use, and Veroboard is fine. You can't regulate a 5V rail from another 5V rail, so what you do is to find the existing regulator, and pick up the voltage going into it, and regulate that for your quiet requirement.
LM317 (+ve) and LM337 (-ve) are adjustable regulators that produce less noise than most of the fixed regulators. Do a search, and you will find from the manufacturer's data sheets that they are very easy to use, and Veroboard is fine. You can't regulate a 5V rail from another 5V rail, so what you do is to find the existing regulator, and pick up the voltage going into it, and regulate that for your quiet requirement.
Veroboard
Stuff like this works well in the CD63...
http://www.acoustica.org.uk/t/63/63hacks_vreg.html
Andy.
Stuff like this works well in the CD63...
http://www.acoustica.org.uk/t/63/63hacks_vreg.html
Andy.
Tant's
Up to about 100-150kHz that may be true, but above that tant's win.
Modern tants are lower ESR than older parts too, and very reliable, providing you don't reverse bias them. The last tantalum I had fail was in 1980 something
Lifespan of a solid cap is likely to be better than a wet-electrolyte type too.
I do use a LOT of Rubycon ZA's though...
Andy.
Up to about 100-150kHz that may be true, but above that tant's win.
Modern tants are lower ESR than older parts too, and very reliable, providing you don't reverse bias them. The last tantalum I had fail was in 1980 something
Lifespan of a solid cap is likely to be better than a wet-electrolyte type too.
I do use a LOT of Rubycon ZA's though...
Andy.
God, just forgot about that acoustica link, thanks ALW!
Just arrived from the local component shop with some veroboard and stuff...here we go!
I'm going to build two separate 5V regs for the DAC - one digital and one analogue. or
What about taking the supply from the +12V rail at the headphone amp jumpers instead?
They had smth called LM317LZ, I bought some of those, too. They are low current ones (100mA), looking like small signal transistors. What's the opinion about those?
Can't see if they use the protection diode on the acoustica.org.uk pic, what do you guys think?
Do I win anything by bypassing the elcos with 0,1uFs?
Lot of silly questions, please bear with me
Cheers,
Tom
Just arrived from the local component shop with some veroboard and stuff...here we go!
I'm going to build two separate 5V regs for the DAC - one digital and one analogue. or
What about taking the supply from the +12V rail at the headphone amp jumpers instead?
They had smth called LM317LZ, I bought some of those, too. They are low current ones (100mA), looking like small signal transistors. What's the opinion about those?
Can't see if they use the protection diode on the acoustica.org.uk pic, what do you guys think?
Do I win anything by bypassing the elcos with 0,1uFs?
Lot of silly questions, please bear with me
Cheers,
Tom
Protection
Those veroboard installs were done by me, but I didn't use the diodes.
It is good practice though...
EC8010,
take a look at http://www.avxcorp.com/docs/catalogs/tps3.pdf
SMD only though...
Andy
Those veroboard installs were done by me, but I didn't use the diodes.
It is good practice though...
EC8010,
take a look at http://www.avxcorp.com/docs/catalogs/tps3.pdf
SMD only though...
Andy
How many regulators, really?
Hello ALW!
Maybe you see futher than me nver thought about the termic stuff 
Anyway, I am a bit confused over about how many new regs the DAC need.
As far as I understand the circuit, there's three different voltage feeds into the chip. What I have done now is to separate the 5V rail from the X'TAL power supply (pin 27).
Now there's one analogue feed and one digital left. Should I make a separate for each one, or can I use the X'tal feed to the rest of the digital circuit?
Or maybe the analogue is more important?
Eagerly awaiting the expertise on diyaudio.com!
Cheers,
Tom
Btw, is 5,4V Vin damaging the 5872? It's close to the maimum...
Hello ALW!
Maybe you see futher than me
nver thought about the termic stuff Anyway, I am a bit confused over about how many new regs the DAC need.
As far as I understand the circuit, there's three different voltage feeds into the chip. What I have done now is to separate the 5V rail from the X'TAL power supply (pin 27).
Now there's one analogue feed and one digital left. Should I make a separate for each one, or can I use the X'tal feed to the rest of the digital circuit?
Or maybe the analogue is more important?
Eagerly awaiting the expertise on diyaudio.com!
Cheers,
Tom
Btw, is 5,4V Vin damaging the 5872? It's close to the maimum...
I posted the link again to be able to save by right clicking... 
www.sprogs.ic24.net/audio/CD43_53_63 Service Manual.pdf
www.sprogs.ic24.net/audio/CD43_53_63 Service Manual.pdf
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.