Thanks guys....
Hi IY, many thanks for welcoming me in this thread. I'll keep posting in this thread about the Y grade play in my next kit. Your suggestions are really worth to try.
I just follow your links and also see a very nice thread started by spzzzzkt. Thanks IY.
Hi spzzzzkt,
You've a great job with D1V3. Well done!
Now, I will speed up my progress to solder the density kit of D1V3. I will take a chance to do it in the late nights. D1V3 is really interesting kit, even my 4 years daughter want to 'contribute' in my soldering session
Doing it by late night maybe better.
Regarding to VCXO, yes, I suggested it by Spencer also. But as you did before, I will leave it for this stage and want to try later using KC7. I know VXCO from Tent's will give us the simple solution but I just wonder how the KC7 gives contribution to this 'click' problem.
Thanks for sharing guys.
Nice building thread, spzzzzkt.
-ims-
PS: Frankly, I prefer to use your name than spzzzzkt, if you don't mind. Hard to type spzzzzkt, may give a typo
Hi IY, many thanks for welcoming me in this thread. I'll keep posting in this thread about the Y grade play in my next kit. Your suggestions are really worth to try.
I just follow your links and also see a very nice thread started by spzzzzkt. Thanks IY.
Hi spzzzzkt,
You've a great job with D1V3. Well done!
Now, I will speed up my progress to solder the density kit of D1V3. I will take a chance to do it in the late nights. D1V3 is really interesting kit, even my 4 years daughter want to 'contribute' in my soldering session
Doing it by late night maybe better.Regarding to VCXO, yes, I suggested it by Spencer also. But as you did before, I will leave it for this stage and want to try later using KC7. I know VXCO from Tent's will give us the simple solution but I just wonder how the KC7 gives contribution to this 'click' problem.
Thanks for sharing guys.
Nice building thread, spzzzzkt.
-ims-
PS: Frankly, I prefer to use your name than spzzzzkt, if you don't mind. Hard to type spzzzzkt, may give a typo
ims not sure what the kc7 is? tcxo?
the problem lies in the fact that the recovered clock and XO are not identical in frequency. The difference in frequency causes the data buffer in the filter chip to either overflow or under-run causing an interruption in the audio output. Feeding a fixed clock (XTAL, XO or TCXO) to the filter chip will have the same result, the two clocks have to be syncronised by some means - either by linking both dac and transport to the same clock, or using a secondary PLL and VCXO to match the recovered clock.
thanks for the compliments on the D1V3, spencer has done most of the difficult stuff it's really just join the dots
btw I've also ordered a set of the Y's after reading the rave reviews in this thread.
cheers
pj
the problem lies in the fact that the recovered clock and XO are not identical in frequency. The difference in frequency causes the data buffer in the filter chip to either overflow or under-run causing an interruption in the audio output. Feeding a fixed clock (XTAL, XO or TCXO) to the filter chip will have the same result, the two clocks have to be syncronised by some means - either by linking both dac and transport to the same clock, or using a secondary PLL and VCXO to match the recovered clock.
thanks for the compliments on the D1V3, spencer has done most of the difficult stuff it's really just join the dots
btw I've also ordered a set of the Y's after reading the rave reviews in this thread.
cheers
pj
good luck!
Paul,
If I do understand right then you are going to try the "Y" version of the PCMs 63 in the balanced mode, using Spencer’s “Pass” board and 4 “Y” chips. Wow!
This makes it to be a very exiting experiment, which is far from being a cheap try!!
I wish you a lot of success with it & am much looking forward to read your remarks!
IY.
Paul,
If I do understand right then you are going to try the "Y" version of the PCMs 63 in the balanced mode, using Spencer’s “Pass” board and 4 “Y” chips. Wow!
This makes it to be a very exiting experiment, which is far from being a cheap try!!
I wish you a lot of success with it & am much looking forward to read your remarks!
IY.
Hi IY,
That's the plan Unfortunately one of the Y's was dead on arrival and took out a resistor in the negative I/V power supply. A'af responded to email with in about 10 minutes (I was impressed) and is sending out a replacement.
At this stage the D1V3 is running with 3 Y's and a PCM63K (not PCM63P-K!!). The other major mod I've made is to implement the secondary PLL re-clocking circuit used in the original D1 using a Tent Labs VCXO.
The improvements from re-clocking were quite substantial using the K chips, tightening up the bass and improving imaging. Installing the 3 Y has resulted in a very clear gain in richness and detail without compromising other qualities enhanced by re-clocking. Obviously a full complement of Y's will improve on the current situation.
I'd been a bit hesitant about the outlay for the 4 Y's but after hearing the level of refinement they bring to the DAC I'm not disappointed.
cheers
Paul
That's the plan
Unfortunately one of the Y's was dead on arrival and took out a resistor in the negative I/V power supply. A'af responded to email with in about 10 minutes (I was impressed) and is sending out a replacement.At this stage the D1V3 is running with 3 Y's and a PCM63K (not PCM63P-K!!). The other major mod I've made is to implement the secondary PLL re-clocking circuit used in the original D1 using a Tent Labs VCXO.
The improvements from re-clocking were quite substantial using the K chips, tightening up the bass and improving imaging. Installing the 3 Y has resulted in a very clear gain in richness and detail without compromising other qualities enhanced by re-clocking. Obviously a full complement of Y's will improve on the current situation.
I'd been a bit hesitant about the outlay for the 4 Y's but after hearing the level of refinement they bring to the DAC I'm not disappointed.
cheers
Paul
Ongoing
Hi Paul,
Good to know that it is going in such a positive direction for you. It should get better once all the 4 “Y” chips are being installed. I was trying once myself to listen with a mix ((one PCM63K (I had once the same; no “P”…) and one PCM63P-Y)) and there was no accurate stereo-image!!
You can easily experiment the above by simply having one channel working with 2x “K”, the other with 2x “Y”: I think that this will show the major differences between both Chips; might be interesting to do. I would say that it is quite strong and surprising sound effect, but not very appealing for the “long run”.
Sorry to hear about the dead “Y” but indeed, Juang (AA’F) is so reliable that there is should be no reason to wary.
Something that really stunned me about the “Y”, just as you yourself wrote - “…without compromising other qualities enhanced by re-clocking” , I felt much the same – there have been no any compromising that I was able to figure out by listening to my DAC once the Y are in.
I am truly happy to hear about the different measurement that you are taking & the all improvements, includes the using of the Tent. May be you could drop a line to Spencer about it once you are done.
Cheers
IY
Hi Paul,
Good to know that it is going in such a positive direction for you. It should get better once all the 4 “Y” chips are being installed. I was trying once myself to listen with a mix ((one PCM63K (I had once the same; no “P”…) and one PCM63P-Y)) and there was no accurate stereo-image!!
You can easily experiment the above by simply having one channel working with 2x “K”, the other with 2x “Y”: I think that this will show the major differences between both Chips; might be interesting to do. I would say that it is quite strong and surprising sound effect, but not very appealing for the “long run”.
Sorry to hear about the dead “Y” but indeed, Juang (AA’F) is so reliable that there is should be no reason to wary.
Something that really stunned me about the “Y”, just as you yourself wrote - “…without compromising other qualities enhanced by re-clocking” , I felt much the same – there have been no any compromising that I was able to figure out by listening to my DAC once the Y are in.
I am truly happy to hear about the different measurement that you are taking & the all improvements, includes the using of the Tent. May be you could drop a line to Spencer about it once you are done.
Cheers
IY
Paul,
YOu can just use two Y chips running the DAC in Single Ended mode and see how the Y is different from K. YOu will need 4 DAC chips only when you run the DAC in balance mode. I am using KY chips (not sure if real or not but are old used chips) and I feel the harmonics is full and sound better with normal PK chip.
Yes I would like to know how you modify and add the PLL in my DAC as I would like to try in futre.
Recent change in my D1V3:
1. C4/C15/C25/C36: Use Black Gate Non Polar 10uF 50V E-cap.
2. C24/C35/C223/C229: Change from 1000uf 35V back to 2200uF 16V. Pana FC grade.
3. C10/C20/C31/C41: Change from Panasonic FC to FM grade; 47uF 25V to 100uF 25V. These capacitors are the analogy supply decoupling just next to the Jfet IV +ve supply. 6x11.2mm
4. C100/C124/C128/C140/C149/C173/C177/C189: Change from Panasonic FC to FM grade; 47uF 25V to 100uF 25V. These capacitors are the analogy supply decoupling capacitors just next to the PCM63 ICs. 6x11.2mm
5. Add a decouple capacitors Panasonic FC 27uF 35 voltage to all 5V regulators ground resistors at R97, R98, R105, R109, R112, R113, R126 & R127 (300 ohm). 5x11.2mm
7. Use PCM63P-KY chips.
Spencer
YOu can just use two Y chips running the DAC in Single Ended mode and see how the Y is different from K. YOu will need 4 DAC chips only when you run the DAC in balance mode. I am using KY chips (not sure if real or not but are old used chips) and I feel the harmonics is full and sound better with normal PK chip.
Yes I would like to know how you modify and add the PLL in my DAC as I would like to try in futre.
Recent change in my D1V3:
1. C4/C15/C25/C36: Use Black Gate Non Polar 10uF 50V E-cap.
2. C24/C35/C223/C229: Change from 1000uf 35V back to 2200uF 16V. Pana FC grade.
3. C10/C20/C31/C41: Change from Panasonic FC to FM grade; 47uF 25V to 100uF 25V. These capacitors are the analogy supply decoupling just next to the Jfet IV +ve supply. 6x11.2mm
4. C100/C124/C128/C140/C149/C173/C177/C189: Change from Panasonic FC to FM grade; 47uF 25V to 100uF 25V. These capacitors are the analogy supply decoupling capacitors just next to the PCM63 ICs. 6x11.2mm
5. Add a decouple capacitors Panasonic FC 27uF 35 voltage to all 5V regulators ground resistors at R97, R98, R105, R109, R112, R113, R126 & R127 (300 ohm). 5x11.2mm
7. Use PCM63P-KY chips.
Spencer
Attachments
Hi Spencer, IY,
I'm running a balanced system - D1V3 -> P1.7 -> Aleph 30 mono blocs so I'll leave as is. I think I should have the replacement early next week.
The PLL is pretty much as per the D1 service manual with some very minor changes. I'm using a 11.2896mhz vcxo vs the original 22.5xxmhz, so this requires use of the q8 (/256) output of the 4020 to give the 44.1khz to feed to the comparator (4046). D1 used q9 (/512) due to the x2 vcxo clock freq.
The PLL LPF values don't need to be changed. The PLL bias voltage via 2.2M works without problem with the tent VCXO but probably needs to be tweaked to get best performance.
The idea of the bias is to ensure that the 4046's comparator is always operating outside it's dead band to ensure lower jitter performance of the 4046. One of the few documents I found that mentioned this was a pdf titled chapter 10 "oscillators and synthesisers" (from the ARRL handbook??) which said (at length):
I'm powering the clock circuit from 8.3V digital supply, with 2 x 7805's on the clock board. This is non-ideal and it would work with 3 x 7805's - 1 for clock, 1 for PLL bias and 1 for logic chips.
Clock is picked up from FSYNC on the CS8412 and reclocked signal is wired to CLK pad on the edge of the board, with a couple of component positions jumper to send the signal back to the 8412/XO jumper. So it's exactly idea, but the clock looks cleaner than what's coming from the cs8412.
I've tested a couple of combinations of jumper settings: with the Y's installed the D1 settings of Jitter = Free, Dither = On seem to work very well with reclocking.
Spencer I'll check out your latest mods. I'm interested how you feel about the sound of the BG 10uf/50V? I'm thinking I might up the DAC PS decoupling caps from 47uF -> 220uF as per D1 schematic. Did you feel there were worthwhile gains from going 47uF -> 100uF?
Perhaps we are looking for different sound? I feel - with the PCM63K at least - the unclocked D1V3 is a bit too full/rich sounding. In comparison the clocked version is leaner and more precise sounding. It is still very musical, but the bass for example is more taut and better defined. My initial thoughts comparing clocked and unclocked were that the unclocked version presented everything as if wrapped in fine velvet and larger than life - beautiful and beguiling but ultimately obscuring the fine details. Some might quite understandably prefer that presentation.
cheers
Paul
I'm running a balanced system - D1V3 -> P1.7 -> Aleph 30 mono blocs so I'll leave as is. I think I should have the replacement early next week.
The PLL is pretty much as per the D1 service manual with some very minor changes. I'm using a 11.2896mhz vcxo vs the original 22.5xxmhz, so this requires use of the q8 (/256) output of the 4020 to give the 44.1khz to feed to the comparator (4046). D1 used q9 (/512) due to the x2 vcxo clock freq.
The PLL LPF values don't need to be changed. The PLL bias voltage via 2.2M works without problem with the tent VCXO but probably needs to be tweaked to get best performance.
The idea of the bias is to ensure that the 4046's comparator is always operating outside it's dead band to ensure lower jitter performance of the 4046. One of the few documents I found that mentioned this was a pdf titled chapter 10 "oscillators and synthesisers" (from the ARRL handbook??) which said (at length):
I'm powering the clock circuit from 8.3V digital supply, with 2 x 7805's on the clock board. This is non-ideal and it would work with 3 x 7805's - 1 for clock, 1 for PLL bias and 1 for logic chips.
Clock is picked up from FSYNC on the CS8412 and reclocked signal is wired to CLK pad on the edge of the board, with a couple of component positions jumper to send the signal back to the 8412/XO jumper. So it's exactly idea, but the clock looks cleaner than what's coming from the cs8412.
I've tested a couple of combinations of jumper settings: with the Y's installed the D1 settings of Jitter = Free, Dither = On seem to work very well with reclocking.
Spencer I'll check out your latest mods. I'm interested how you feel about the sound of the BG 10uf/50V? I'm thinking I might up the DAC PS decoupling caps from 47uF -> 220uF as per D1 schematic. Did you feel there were worthwhile gains from going 47uF -> 100uF?
Perhaps we are looking for different sound? I feel - with the PCM63K at least - the unclocked D1V3 is a bit too full/rich sounding. In comparison the clocked version is leaner and more precise sounding. It is still very musical, but the bass for example is more taut and better defined. My initial thoughts comparing clocked and unclocked were that the unclocked version presented everything as if wrapped in fine velvet and larger than life - beautiful and beguiling but ultimately obscuring the fine details. Some might quite understandably prefer that presentation.
cheers
Paul
Hi,
The BG N series is really good and before I try them, I do not believe E-cap can sound so good! I think a 10uF 50V BG is much cheaper than a expensive BIG film cap which may also has microphonic effect. When I read service manual of SONY PCM-3348HR (48 track digital recorder), it also use 220uF 25V E-cap to couple the signal to output and thus E-cap is good in fact but I guess we need to choose the right one. Just like NP F4 also use E-cap to couple the input buffer to output buffer and it still sound very good to me!
Where can you find a VCXO at reasonable price?
Are you still using the 8412 as the receiver chip? I swith to DIR9001 and the receiver is really sing (relative to 8414 or 8412). 9001 givve much cleaner sound stage but still maintain the musical sound. 9001 also give more life to the music and although it is not crystal clear sound, it improve the separation of insturment. Once I go to 9001, I have no more interest on 8414 and 8412 receiver now. Yes I prefer full and rich sound and it is the Jfet IV that give lots of second harmonics distortion rather that odd harmonics.
Yes I think increase the deouble cap to even 220uF is a good experience and ultimately it is the diyer to decide what value to use and the sound he want. The pcb only provide a platform for diyers to tweak and play with it. I feel there is more base or deeper base if I increase the decouple cap just next to DAC and Jfet IV +ve supply. May be you should try the BG standard series to put in to the power supply!
Cheers
The BG N series is really good and before I try them, I do not believe E-cap can sound so good! I think a 10uF 50V BG is much cheaper than a expensive BIG film cap which may also has microphonic effect. When I read service manual of SONY PCM-3348HR (48 track digital recorder), it also use 220uF 25V E-cap to couple the signal to output and thus E-cap is good in fact but I guess we need to choose the right one. Just like NP F4 also use E-cap to couple the input buffer to output buffer and it still sound very good to me!
Where can you find a VCXO at reasonable price?
Are you still using the 8412 as the receiver chip? I swith to DIR9001 and the receiver is really sing (relative to 8414 or 8412). 9001 givve much cleaner sound stage but still maintain the musical sound. 9001 also give more life to the music and although it is not crystal clear sound, it improve the separation of insturment. Once I go to 9001, I have no more interest on 8414 and 8412 receiver now. Yes I prefer full and rich sound and it is the Jfet IV that give lots of second harmonics distortion rather that odd harmonics.
Yes I think increase the deouble cap to even 220uF is a good experience and ultimately it is the diyer to decide what value to use and the sound he want. The pcb only provide a platform for diyers to tweak and play with it. I feel there is more base or deeper base if I increase the decouple cap just next to DAC and Jfet IV +ve supply. May be you should try the BG standard series to put in to the power supply!
Cheers
VCXO is from tent labs. a little more expensive than their xo but not too bad.
yes I'm still using the cs8412. At the moment my interest is in exploring what the original design was capable of, as it was clearly very well thought out and I've learnt quite a bit while trying to work out why things were done as they were.
The D1 had 220uF E-caps bypassed with 0.047uf film caps on the outputs. NP was pretty fond of Panasonic FC and ECQ-P caps so it's likely this is what was used. Not sure how that combination would stack up against the BG's but it wouldn't be too expensive to find out
cheers
Paul
yes I'm still using the cs8412. At the moment my interest is in exploring what the original design was capable of, as it was clearly very well thought out and I've learnt quite a bit while trying to work out why things were done as they were.
The D1 had 220uF E-caps bypassed with 0.047uf film caps on the outputs. NP was pretty fond of Panasonic FC and ECQ-P caps so it's likely this is what was used. Not sure how that combination would stack up against the BG's but it wouldn't be too expensive to find out
cheers
Paul
Hi Paul, Hi Spencer,
There are 3-4 subjects here which we should try to „isolate“ and then solve, if only possible. We should do this step by step.
Being somehow “non conservative” here, I do believe that there are some “ingredients” that one can basically trust. Even though I am not using the D1V3 yet, I do share several conclusions of both of you:
1.
First of all, I am happy that all 3 of us (or all 4 of us - when including Juang) do agree that the “Y” is an excellent chip, to be installed in PCM63 configurations (“PCM63P--------KY” is probably the same Chip, while being differently marked). The major discovery here is that it had no side-effects and that it contains only the positive characteristics. I find this to be already a major issue – and at least the topic of this tread seemed to be now solved (yes, we should still wait to hear from ims about his conclusions).
2.
Talking about the PCM63 family (J, K, K2, Y/KY), we can also agree now that it is remarkably sensitive to the input receiver: I am using newly the CS8414 adaptor and I am quite unhappy with it; I will have to go “backwards” to the YM3623b - at least for the comparison.
So I do agree here with you Spencer, that the combination of PCM63-K (or KY) with the CS8414 (8412?) is not 100% satisfying (I still have some hopes about the DIR9001 by the way, so I will need your help to make your adaptor works in my configuration, you know…).
Paul, if we are right and the PCM63 is so much influanced by the receiver, then may be you should re-consider moving from the 84xx family to something different. The DIR 9001 might well go your direction – at least by providing a cleaner sound. Very unfortunately I was not able to try it in my configuration, but it might well do for you that part of what you are trying to achieve by installing the clock(s) as a supplement to the CS8412.
So we are coming now to topics 3 (DAC’s capacitors) & 4 (Output capacitors). This is indeed unavoidable.
Here are my own thoughts & impression: I felt much the same like you Spencer - that there is more base or deeper base if I increase the decouple cap just next to DACs (even though my configuration is different) or – again my own impression – increase it anywhere else. BUT, I also don’t like this effect!
For me it is exactly this which is going in the direction of what you Paul calls “a bit too full/rich sounding”. My own experience however shows that part of the problem comes from using Panasonics Capacitors in the chain: In other words: I dare to question here their “Audiophile” qualities. Personally, I lost my trust in them.
An example: I have just put BG standard series to decuple the class A biased OpAmps (627 BP), first instead of the FC, then instead of the FM, and – now it sounds excellent!!
This brings me to think, that our wish combination of having the full and rich sound, detailed, transparent, balanced (etc) might not be possible to achieve with Panasonics FC or FM capacitors, neither inside the DAC nor in decupling the output stages.
Same feeling I had by using them for short time (3-4 weeks) at the I/V stage. No good.
For a change, I do have excellent experience with using Panasonics HA (M) 6800uf, 35V, 105° in the DAC’s & Preamp’s PSUs.
Paul, I can only partly share your thoughts, that unclocked version presents everything as if wrapped in fine velvet. I am not so sure about the “larger than life” sound – even if don’t like it, just like you do. Well, I can’t tell that the “unclocked” DAC provides the “ultimately obscure the fine details”; I do understand however what you mean with this.
Please don’t understand me not right: To my opinion, you are far from being wrong at all – and you might recall one of my earlier post, to always compare it to a live sound track being done on a good minidisk/DAT. The “true” might be found somewhere in between.
I think that we all will have to keep going step by step here, so that we will be able to decide sometime about the best input receiver and/or best capacitors (we might all agree on a good “compromising” like BG Standard), and and and...
At least I feel that one can reach a coherent and established opinion if only keep working systematic.
Regards, IY.
There are 3-4 subjects here which we should try to „isolate“ and then solve, if only possible. We should do this step by step.
Being somehow “non conservative” here, I do believe that there are some “ingredients” that one can basically trust. Even though I am not using the D1V3 yet, I do share several conclusions of both of you:
1.
First of all, I am happy that all 3 of us (or all 4 of us - when including Juang) do agree that the “Y” is an excellent chip, to be installed in PCM63 configurations (“PCM63P--------KY” is probably the same Chip, while being differently marked). The major discovery here is that it had no side-effects and that it contains only the positive characteristics. I find this to be already a major issue – and at least the topic of this tread seemed to be now solved (yes, we should still wait to hear from ims about his conclusions).
2.
Talking about the PCM63 family (J, K, K2, Y/KY), we can also agree now that it is remarkably sensitive to the input receiver: I am using newly the CS8414 adaptor and I am quite unhappy with it; I will have to go “backwards” to the YM3623b - at least for the comparison.
So I do agree here with you Spencer, that the combination of PCM63-K (or KY) with the CS8414 (8412?) is not 100% satisfying (I still have some hopes about the DIR9001 by the way, so I will need your help to make your adaptor works in my configuration, you know…).
Paul, if we are right and the PCM63 is so much influanced by the receiver, then may be you should re-consider moving from the 84xx family to something different. The DIR 9001 might well go your direction – at least by providing a cleaner sound. Very unfortunately I was not able to try it in my configuration, but it might well do for you that part of what you are trying to achieve by installing the clock(s) as a supplement to the CS8412.
So we are coming now to topics 3 (DAC’s capacitors) & 4 (Output capacitors). This is indeed unavoidable.
Here are my own thoughts & impression: I felt much the same like you Spencer - that there is more base or deeper base if I increase the decouple cap just next to DACs (even though my configuration is different) or – again my own impression – increase it anywhere else. BUT, I also don’t like this effect!
For me it is exactly this which is going in the direction of what you Paul calls “a bit too full/rich sounding”. My own experience however shows that part of the problem comes from using Panasonics Capacitors in the chain: In other words: I dare to question here their “Audiophile” qualities. Personally, I lost my trust in them.
An example: I have just put BG standard series to decuple the class A biased OpAmps (627 BP), first instead of the FC, then instead of the FM, and – now it sounds excellent!!
This brings me to think, that our wish combination of having the full and rich sound, detailed, transparent, balanced (etc) might not be possible to achieve with Panasonics FC or FM capacitors, neither inside the DAC nor in decupling the output stages.
Same feeling I had by using them for short time (3-4 weeks) at the I/V stage. No good.
For a change, I do have excellent experience with using Panasonics HA (M) 6800uf, 35V, 105° in the DAC’s & Preamp’s PSUs.
Paul, I can only partly share your thoughts, that unclocked version presents everything as if wrapped in fine velvet. I am not so sure about the “larger than life” sound – even if don’t like it, just like you do. Well, I can’t tell that the “unclocked” DAC provides the “ultimately obscure the fine details”; I do understand however what you mean with this.
Please don’t understand me not right: To my opinion, you are far from being wrong at all – and you might recall one of my earlier post, to always compare it to a live sound track being done on a good minidisk/DAT. The “true” might be found somewhere in between.
I think that we all will have to keep going step by step here, so that we will be able to decide sometime about the best input receiver and/or best capacitors (we might all agree on a good “compromising” like BG Standard), and and and...
At least I feel that one can reach a coherent and established opinion if only keep working systematic.
Regards, IY.
Hi IY,
Hope I will confirm your conclusion about the quality of this dac chip soon, so there are 5 can confirm that matter. I got the chip last week but never get the enough time to solder them to Spencer's kit. I will come to contribute to this thread soon.
Thank you guys, for sharing your findings and opinions.
-ims-
Attached dac chips from A'af but sorry for the blur image.
Hope I will confirm your conclusion about the quality of this dac chip soon, so there are 5 can confirm that matter. I got the chip last week but never get the enough time to solder them to Spencer's kit. I will come to contribute to this thread soon.
Thank you guys, for sharing your findings and opinions.
-ims-
Attached dac chips from A'af but sorry for the blur image.
Attachments
IY,
I will try out the DIR9001 at some stage soon but I think you are missing an essential point in your presentation.
The DIR9001 has a jitter spec of 50ps vs the CS8412 200ps so I suspect a key aspect of the difference you and Spencer are hearing is the lower intrinsic jitter of the receivers. Note this is not an absolute jitter figure but the jitter contributed by the receiver which is added with jitter from your source if stereophile and wolfson amongst others are to be believed.
The secondary PLL goes along way to eliminating these causes of jitter by driving the filter chip from the clock generated by a low jitter vcxo. In this case the benefits of the low jitter 9001 appear to be moot.
I'd also make the observation that the d1 implementation is extremely finessed. For instance it isn't immediately obvious why WC & NP used the FSYNC output of the CS841* but close reading of datasheet reveals that FSYNC is not derived from the internal PLL/VCO of the CS841* but recovered from incoming data stream at times when "intersymbol interference is at a minimum" to provide "a sample frequency clock that is as spectrally pure as the digital audio source clock". The DIR9001 derives the equivalent clock signal from the PLL/VCO recovered clock and thus imparts it's intrinsic jitter on this clock. So in this case the CS841* is actually preferable to the DIR9001.
I don't think it's conservative to prefer a well designed solution to the problem of jitter to "chip rolling" that only toys at the edges of the real issues.
cheers
Paul
I will try out the DIR9001 at some stage soon but I think you are missing an essential point in your presentation.
The DIR9001 has a jitter spec of 50ps vs the CS8412 200ps so I suspect a key aspect of the difference you and Spencer are hearing is the lower intrinsic jitter of the receivers. Note this is not an absolute jitter figure but the jitter contributed by the receiver which is added with jitter from your source if stereophile and wolfson amongst others are to be believed.
The secondary PLL goes along way to eliminating these causes of jitter by driving the filter chip from the clock generated by a low jitter vcxo. In this case the benefits of the low jitter 9001 appear to be moot.
I'd also make the observation that the d1 implementation is extremely finessed. For instance it isn't immediately obvious why WC & NP used the FSYNC output of the CS841* but close reading of datasheet reveals that FSYNC is not derived from the internal PLL/VCO of the CS841* but recovered from incoming data stream at times when "intersymbol interference is at a minimum" to provide "a sample frequency clock that is as spectrally pure as the digital audio source clock". The DIR9001 derives the equivalent clock signal from the PLL/VCO recovered clock and thus imparts it's intrinsic jitter on this clock. So in this case the CS841* is actually preferable to the DIR9001.
I don't think it's conservative to prefer a well designed solution to the problem of jitter to "chip rolling" that only toys at the edges of the real issues.
cheers
Paul
Paul, I agree with you logic.
I don’t know the reason for me being somehow “not happy" with the CS8414, at least not at this stage. There might be different reasons for this. Indeed, I might well adapt your vcxo solution myself and will probably become more positive about the topic.
Many thanks for sharing this important work which you are doing here.
Yours, IY
I don’t know the reason for me being somehow “not happy" with the CS8414, at least not at this stage. There might be different reasons for this. Indeed, I might well adapt your vcxo solution myself and will probably become more positive about the topic.
Many thanks for sharing this important work which you are doing here.
Yours, IY
IY,
it's all in the context - the CS8414 is not going to be the best solution if your dac is relying on recovered clock. There are receivers with far better jitter performance and the DIR9001 will clearly offer better performance than CS8412 if compared using the recovered clock.
BTW I'm not clear how you replaced YM3623B with the CS8414? Was this on an adapter board of some kind?
Note: I found your posts re an audiotuning CS8414 adapter which appeared to be hardwired to a specific configuration. Did you also upgrade to one of the audiotuning digital filters the board was configured to work with or confirm the existing filter was configured in a way that would work correctly with this input?
cheers
Paul
it's all in the context - the CS8414 is not going to be the best solution if your dac is relying on recovered clock. There are receivers with far better jitter performance and the DIR9001 will clearly offer better performance than CS8412 if compared using the recovered clock.
BTW I'm not clear how you replaced YM3623B with the CS8414? Was this on an adapter board of some kind?
Note: I found your posts re an audiotuning CS8414 adapter which appeared to be hardwired to a specific configuration. Did you also upgrade to one of the audiotuning digital filters the board was configured to work with or confirm the existing filter was configured in a way that would work correctly with this input?
cheers
Paul
Paul,
Indeed I am using the two “Audiotuning” adaptor-replacements: The CD8414 (replacing the YM3623b) and the DF 1704 Digital Filter (for the DF1700). Both products are of course very good, no doubt about this. All configurations are correct.
I just met the “Audiotuning” people when I have been to Berlin last week and also purchased some Black Gates from them, for any case…
Back to the receiver: I was experimenting at the time with 3 different YM3623b but I liked really only 1 of them, which gave me the excellent mixture of light and strong, transparent but centred, airy but realistic sound. The other 2 picks were a bit more precise, but then part of the special atmosphere was gone.
There is no wonder therefore that I am still questioning the input receiver. Moreover, I can’t even tell you if I will like having a separate clock, or if the DIR9001 will be satisfying for me.
I will need few months to clear things out. Alone the using of socket x or socket y under the receiver might as well “contribute” to the success. I will probably will have time for this not before the summer. I will however keep in touch.
Cheers, IY.
Indeed I am using the two “Audiotuning” adaptor-replacements: The CD8414 (replacing the YM3623b) and the DF 1704 Digital Filter (for the DF1700). Both products are of course very good, no doubt about this. All configurations are correct.
I just met the “Audiotuning” people when I have been to Berlin last week and also purchased some Black Gates from them, for any case…
Back to the receiver: I was experimenting at the time with 3 different YM3623b but I liked really only 1 of them, which gave me the excellent mixture of light and strong, transparent but centred, airy but realistic sound. The other 2 picks were a bit more precise, but then part of the special atmosphere was gone.
There is no wonder therefore that I am still questioning the input receiver. Moreover, I can’t even tell you if I will like having a separate clock, or if the DIR9001 will be satisfying for me.
I will need few months to clear things out. Alone the using of socket x or socket y under the receiver might as well “contribute” to the success. I will probably will have time for this not before the summer. I will however keep in touch.
Cheers, IY.
ongoing.
Paul,
Meanwhile I had some further thoughts about the input receiver, the new CS8414 adaptor-replacement from “Audiotuning”. By the way, the adaptor contains a small (but not cheap..) Black Gate, 47uf, 6.3 V, BG-NX capacitor..
I was trying some different brands and types of capacitors in the DAC - I am using newly all OsCons in the DAC section.
Once everything “Broke in”, the DAC sounds now even better then I was dreaming about
I don't feel anymore that the input receiver is of any particular problem or significance, al least not at this moment. I will surely stay now with it will not go back to the YM3623b.
I placed some Black Gates (standard) behind the Bursons Discrete OpAmps (by the way - the brand new Bursons are sounding even more transparent (!) - a couple of resistor has been updated to a different type - and my DAC (CS8414 + DF1704+PCM63P-Y,s + OsCons + Bursons) sound is fabulous.
About your remark on NP and the Panasonics FC - I don't know if NP was wishing to invest so much additional money in a line-product being already so expansive. I might be wrong here, of course. At least to my opinion, there are much better Capacitors then the Panasonics. I personally will never use them anymore. I had enough of it.
Hope you will get the missing “Y” soon.
Have a nice weekend!
IY.
Paul,
Meanwhile I had some further thoughts about the input receiver, the new CS8414 adaptor-replacement from “Audiotuning”. By the way, the adaptor contains a small (but not cheap..) Black Gate, 47uf, 6.3 V, BG-NX capacitor..
I was trying some different brands and types of capacitors in the DAC - I am using newly all OsCons in the DAC section.
Once everything “Broke in”, the DAC sounds now even better then I was dreaming about
I don't feel anymore that the input receiver is of any particular problem or significance, al least not at this moment. I will surely stay now with it will not go back to the YM3623b. I placed some Black Gates (standard) behind the Bursons Discrete OpAmps (by the way - the brand new Bursons are sounding even more transparent (!) - a couple of resistor has been updated to a different type - and my DAC (CS8414 + DF1704+PCM63P-Y,s + OsCons + Bursons) sound is fabulous.
About your remark on NP and the Panasonics FC - I don't know if NP was wishing to invest so much additional money in a line-product being already so expansive. I might be wrong here, of course. At least to my opinion, there are much better Capacitors then the Panasonics. I personally will never use them anymore. I had enough of it.
Hope you will get the missing “Y” soon.
Have a nice weekend!
IY.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.