I think as a rule it is OK to reproduce a part of a schematic in order to illustrate a point but not to publish the whole manual!
I don't know what Philips or other manufacturers think of all this but they should see it as having a huge (and unpaid) R&D department. Do you think that they would ask our permission to use a tweak suggest here in their next CD player?
resistors in data lines
In the English page a link in the beginning of this thread pointed to, there was a mod about reducing the resistors in the data lines to the DAC to reduce jitter.
Actually, there will be only one line that is jitter sensitive because it controls the timing. On older multi-bit DACs this is ususally wordclock (sometimes called latch enable), on more modern DACs it is MCLK or sometimes bitclock. It makes sense to place the clock close to the DAC and feed this line with a low resistance of below 100 R. On all other lines, you can get away with as much as 1 k. Higher resistance means less feedtrough, on the other hand the input stage in the DAC spends more time in cross-conduction, meaning it draws a current spike from the internal DAC supply.
It may be a better idea to put a digital buffer next to the DAC and feed it from a well-filtered supply. Something from the 74VHC series which outputs a clean waveform would appear suitable.
In the English page a link in the beginning of this thread pointed to, there was a mod about reducing the resistors in the data lines to the DAC to reduce jitter.
Actually, there will be only one line that is jitter sensitive because it controls the timing. On older multi-bit DACs this is ususally wordclock (sometimes called latch enable), on more modern DACs it is MCLK or sometimes bitclock. It makes sense to place the clock close to the DAC and feed this line with a low resistance of below 100 R. On all other lines, you can get away with as much as 1 k. Higher resistance means less feedtrough, on the other hand the input stage in the DAC spends more time in cross-conduction, meaning it draws a current spike from the internal DAC supply.
It may be a better idea to put a digital buffer next to the DAC and feed it from a well-filtered supply. Something from the 74VHC series which outputs a clean waveform would appear suitable.
Re: resistors in data lines
dear all,
The exact value of series resistors is a trade off between sufficient slope and current.
At a certain moment a higher slope is not anymore required, and when decreasing the resistor, the RF current increases as well, deteroriating the jitter performance once again.
The best thing to do is place a low jitter oscillator close to the DAC chip, directly clock the LE or MLCK from that oscillator, buffer that same clock and feed it back to the system (spindle motor control, decoders etc)
Needless to state that reclocking of all signals entering the DAC chip can be very fruitful as it reduces the amount of jittery currents that enter the chip.
An example of a DAC design with extensive reclocking can be found here:
http://members.chello.nl/~m.heijligers/DAChtml/dactop.htm
enjoy
capslock said:
dear all,
The exact value of series resistors is a trade off between sufficient slope and current.
At a certain moment a higher slope is not anymore required, and when decreasing the resistor, the RF current increases as well, deteroriating the jitter performance once again.
The best thing to do is place a low jitter oscillator close to the DAC chip, directly clock the LE or MLCK from that oscillator, buffer that same clock and feed it back to the system (spindle motor control, decoders etc)
Needless to state that reclocking of all signals entering the DAC chip can be very fruitful as it reduces the amount of jittery currents that enter the chip.
An example of a DAC design with extensive reclocking can be found here:
http://members.chello.nl/~m.heijligers/DAChtml/dactop.htm
enjoy
?Question for the new Digi-Out
Hi!
I have a question regarding the new digital out - this simple circuit (two resistors, one capacitor) is replacing the old one completely (three caps, one resistor and is it one "optical coupler"?)?
BTW, what are these muting transistors for - when is the output muted using them?
Thanks a lot,
Christian
Hi!
I have a question regarding the new digital out - this simple circuit (two resistors, one capacitor) is replacing the old one completely (three caps, one resistor and is it one "optical coupler"?)?
BTW, what are these muting transistors for - when is the output muted using them?
Thanks a lot,
Christian
@Coma
I wish you good luck with the mod. I tried it but did reverse
it. Sound got more grainy after all. I only removed the capacitor
to ground (2543) in the end.
To really improve it it would be worthwhile to create a reshaping
unit. Reclocking surely is the best but is much more work.
btw. i stopped playing with the 753 cause i got a Marantz 6000
for cheap. There isn´t much difference in the digital exit between
both and they use the same VAM1201 Laserunit. Sound from the
Philips in comparison is as descriped by Guido Tent some posts
before. There seems to be another problem - not the exit itself.
Ich bin Rheinauer
I wish you good luck with the mod. I tried it but did reverse
it. Sound got more grainy after all. I only removed the capacitor
to ground (2543) in the end.
To really improve it it would be worthwhile to create a reshaping
unit. Reclocking surely is the best but is much more work.
btw. i stopped playing with the 753 cause i got a Marantz 6000
for cheap. There isn´t much difference in the digital exit between
both and they use the same VAM1201 Laserunit. Sound from the
Philips in comparison is as descriped by Guido Tent some posts
before. There seems to be another problem - not the exit itself.
Ich bin Rheinauer
CD723/753 new digital output circuit
Nuuk,
The details you have posted are incomplete, so I will clarify.
Changes to transport board.
Replace 3896 (270R) with 390R.
Replace 3897 (180R) with 91R.
Remove 2890 (capacitor).
Changes to AF board.
Change C2541 to 0.47uF.
Remove transformer 5801 and fit links on the pcb between pins 3 & 4 and 1 & 6.
Replace R3556 with a link.
Remove C2543.
This was the successful implementation of the new digital output circuit mentioned previously in this thread. An earlier attempt to replace the original circuit, which I installed on the AF board, and with the removal of the three components from the transport board, resulted in abnormal operation with CD/RWs. It was assumed that differences in the grounding arrangements between the two circuits were the cause, though this has not been firmly established. The circuit mentioned above preserves, as far as possible, the original grounding arrangements. I should mention that the earlier unsuccessful circuit had been introduced into the CD753, the successful results were obtained from the circuit (as mentioned above) when fitted into the CD723.
Tim.
Nuuk,
The details you have posted are incomplete, so I will clarify.
Changes to transport board.
Replace 3896 (270R) with 390R.
Replace 3897 (180R) with 91R.
Remove 2890 (capacitor).
Changes to AF board.
Change C2541 to 0.47uF.
Remove transformer 5801 and fit links on the pcb between pins 3 & 4 and 1 & 6.
Replace R3556 with a link.
Remove C2543.
This was the successful implementation of the new digital output circuit mentioned previously in this thread. An earlier attempt to replace the original circuit, which I installed on the AF board, and with the removal of the three components from the transport board, resulted in abnormal operation with CD/RWs. It was assumed that differences in the grounding arrangements between the two circuits were the cause, though this has not been firmly established. The circuit mentioned above preserves, as far as possible, the original grounding arrangements. I should mention that the earlier unsuccessful circuit had been introduced into the CD753, the successful results were obtained from the circuit (as mentioned above) when fitted into the CD723.
Tim.
Hi,
in the meantime I have made the new digital out like it has been suggested earlier: cutting out C2890, R3897 and shorting R3896 on the transport PCB and setting up the new parts on the AF board. Its working, and I have not encountered any issues yet.
@TimA: What problems with CD-RW did you observe when modding the 753 this way? I have not tried any CD-RW yet.
Christian
in the meantime I have made the new digital out like it has been suggested earlier: cutting out C2890, R3897 and shorting R3896 on the transport PCB and setting up the new parts on the AF board. Its working, and I have not encountered any issues yet.
@TimA: What problems with CD-RW did you observe when modding the 753 this way? I have not tried any CD-RW yet.
Christian
CD723 improvements
Here are the details and my listening impressions of some changes made to my CD723.
1. New digital output circuit and TDA1545 input resistor change (as mentioned earlier in this thread and according to chris Found's article).
2. Regulated (LM317) 9.5V supply fitted for the servo circuits etc.
3. Shunt regulated (TL431) 5.5V supply (fed from the regulated 9.5V supply) fitted for the digital circuits.
4. LM6172 fitted in the output stage, its output biased into class A using a FET constant current source and with improved supply decoupling including Elna silmic 220uF from +/- rails to ground and a 470nF polypropylene capacitor soldered directly across the op-amp supply pins.
5. Output 47uF electrolytics replaced with 2.2uF Ansar 'supersound' polypropylene type.
6. Muting transistors and the 47K resistor to ground (after the op-amp) removed.
The improvements have proved very worthwhile. Since fitting the regulators the digital output has lost its somewhat 'diluted' sound (previously I had found the digital output of a Marantz CD17KI preferable, being both more relaxed, and detailed). After fitting the LM6172, and only after dealing with its power supply decoupling requirements, and replacing a 10uF Oscon (which I had installed earlier) with a 2.2uF Ansar Supersound polypropylene capacitor in the analogue output, my preference is now for the modified CD723 over the Marantz CD17KI when listening to the analogue outputs of both machines.
Tim.
Here are the details and my listening impressions of some changes made to my CD723.
1. New digital output circuit and TDA1545 input resistor change (as mentioned earlier in this thread and according to chris Found's article).
2. Regulated (LM317) 9.5V supply fitted for the servo circuits etc.
3. Shunt regulated (TL431) 5.5V supply (fed from the regulated 9.5V supply) fitted for the digital circuits.
4. LM6172 fitted in the output stage, its output biased into class A using a FET constant current source and with improved supply decoupling including Elna silmic 220uF from +/- rails to ground and a 470nF polypropylene capacitor soldered directly across the op-amp supply pins.
5. Output 47uF electrolytics replaced with 2.2uF Ansar 'supersound' polypropylene type.
6. Muting transistors and the 47K resistor to ground (after the op-amp) removed.
The improvements have proved very worthwhile. Since fitting the regulators the digital output has lost its somewhat 'diluted' sound (previously I had found the digital output of a Marantz CD17KI preferable, being both more relaxed, and detailed). After fitting the LM6172, and only after dealing with its power supply decoupling requirements, and replacing a 10uF Oscon (which I had installed earlier) with a 2.2uF Ansar Supersound polypropylene capacitor in the analogue output, my preference is now for the modified CD723 over the Marantz CD17KI when listening to the analogue outputs of both machines.
Tim.
rbroer said:For all you guys tweaking the CD713 & CD723:
Built yourself s DIY Kwak-Clock.
Connect it at the crystals original location, BUT, also use it directly as BitClock on the TDA1545A
Since this player is 4* oversampling, BCK just happens to be 8.X MHz like the system clock.This will give you much more improvement than connecting at the servo/decoder chip only.
Hi,
I was looking through this thread and came across the above.
sounds good but when I checked in the service manual it seems to indicate that BCK is 4.23 Mhz and master clock is 8.46Mhz.
I am not that up on digital theory. Can anyone tell me if this idea is workable and if so advise me how to proceed
thanks
mike
Guido Tent said:
hi Guido
OK but does it work for TDA1545A as in my CD723 ?
does it matter that, apparently, I will be doubling the frequency of BCK.
where can I get details about your clock cct ?
do you do a cct board for it ?
all my cd psu's are batteries - sounds nice. Would you recommend any additional regulation after the battery for your clock cct ? ( apart from a good quality cap )
cheers
mike
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.