Hi Soundcheck 
I'm glad you got it right! I think I missed your post over there...
Did you post schematics?
Now you should try an 8,5V TeddyReg for the DAC tower
That's what I will do as time allows...
Then I will add an active I/V with LM452 (Teddyregulated or ALWregulated), to get distortion low, like DIDAC does. That way the real benefit of the timing chain should show
Sorry for OT, John
Cheers,
M
I'm glad you got it right! I think I missed your post over there...
Did you post schematics?
Now you should try an 8,5V TeddyReg for the DAC tower
That's what I will do as time allows...
Then I will add an active I/V with LM452 (Teddyregulated or ALWregulated), to get distortion low, like DIDAC does. That way the real benefit of the timing chain should show
Sorry for OT, John
Cheers,
M
maxlorenz said:Hi Soundcheck
I'm glad you got it right! I think I missed your post over there...
Did you post schematics?
Now you should try an 8,5V TeddyReg for the DAC tower
That's what I will do as time allows...
Then I will add an active I/V with LM452 (Teddyregulated or ALWregulated), to get distortion low, like DIDAC does. That way the real benefit of the timing chain should show
Sorry for OT, John
Cheers,
M
Hi MaxLorenz.
No schematics yet. But it's pretty much the one as shown on Pinkfish with the extra 4002 diodes around the LM317.
Of course the resistor values I did change to get 5,5V unloaded.
Interesting try would be to get diodes in for setting the output voltage of the LM317. You might have read that this gets the noise even further down.
I tried to get low-noise Vishay or similar resistors. Unfortunately there is only a very limited selection of certain values available for most brands. I ended up using Beyschlag 1% resistors. Perhaps I am gonna put a BG cap in the ouput replacing the currently used tantalum type.
Regarding the DAC Tower supply. That's a good idea. But how much current can the Teddy-Reg deliver. What cooling would I need to get it going.
I am still not convinced about an active I/V stage. I would have to
try it first. I can't identify very much distortions caused by the DAC output.
Since I got rid of my coupling cap (Mundorf Silver/Gold), which is causing real problems, I am really happy about the DAC sound itself.
BTW: My last post related to my latest tweaks was reflecting my
sound impressions from Friday last week. (Didn't have time to do more
tests). Yesterday I was breaking the changed parts in over the entire day.
With the result that after listening to my Manger-CD last night I was asking myself : "What more do I want?"
The before mentioned distortions were almost gone.Cheers
Maybe you do want a DI16
aren't you curious about it?So, you do think that John's reclocker is an improvement?
Sorry, I was talking about the following tweak's schematics:
I believe the current max of the TeddyReg is limited by LM317 output. I will use an LM1085 Reg for my 16*tower (3A max). I bolt the units to the box wall and it stays cool.
End OT.
M
Back on topic...
I want to share the following experience:
I usually listen music for an hour before going to the job on the morning in one of my two main systems...
I put together the following humble system, just to experiment:
Sony VAIO laptop; Ubuntu (untweaked, I' m afraid)--> common type USB cable--> DI16 DAC with Doede's USB module (TeddyRegs here and there...)--> unbalanced IC's from Silver Sonics, "Pro Studio" type
( pseudobalanced config.)--> CS3310 based preamp--> 24AWG solid core braided OCC copper interconnects (4 wires per lead; the same material as my TVC)--> My-Reference monoblocks--> Supra Ply 2.0--> Paradigm M3 monitors...uff!
(I already commented how good I think it sounds)
Listening to Bach's English Suites, Glen Gould on pianoforte (Sony, remastered).
I assumed a relaxed sitting position and began a rhythmic, deep and slow type of breathing, concentrating my gaze on a fixed point (my unused CD PRO ). After a few seconds the periphery of the visual field begins to "melt" as one attains a relaxed state of conscience...at this point one can have one of two approaches: a passive listening, letting the notes strike you as they come; or an active one, actually "co-creating" the music at the same time that the system reproduces it. Obviously, the most satisfying is the later one...it produced me the impression of a true understanding of the piece, with a more-than-esthetic joy of a pure information flow (timeless, I should say). As you can imagine, it is very difficult to explain the experience...better try it for your selves
That's all...no more bothering from my part.
Cheers,
M
The healer apprentice...
I want to share the following experience:
I usually listen music for an hour before going to the job on the morning
in one of my two main systems...I put together the following humble system, just to experiment:
Sony VAIO laptop; Ubuntu (untweaked, I' m afraid)--> common type USB cable--> DI16 DAC with Doede's USB module (TeddyRegs here and there...)--> unbalanced IC's from Silver Sonics, "Pro Studio" type
( pseudobalanced config.)--> CS3310 based preamp--> 24AWG solid core braided OCC copper interconnects (4 wires per lead; the same material as my TVC)--> My-Reference monoblocks--> Supra Ply 2.0--> Paradigm M3 monitors...uff!
(I already commented how good I think it sounds)
Listening to Bach's English Suites, Glen Gould on pianoforte (Sony, remastered).
I assumed a relaxed sitting position and began a rhythmic, deep and slow type of breathing, concentrating my gaze on a fixed point (my unused CD PRO
). After a few seconds the periphery of the visual field begins to "melt" as one attains a relaxed state of conscience...at this point one can have one of two approaches: a passive listening, letting the notes strike you as they come; or an active one, actually "co-creating" the music at the same time that the system reproduces it. Obviously, the most satisfying is the later one...it produced me the impression of a true understanding of the piece, with a more-than-esthetic joy of a pure information flow (timeless, I should say). As you can imagine, it is very difficult to explain the experience...better try it for your selves That's all...no more bothering from my part.
Cheers,
M
The healer apprentice...
Hi mikelm,
This depends on how noise affects a specific circuit. Clock circuits (crystal oscillator, reclockers and buffers) require very clean power supplies.
Techniques used to obtain clean power supplies for digital circuits, may not work for analogue power supplies.
I got an eureka today, after removing all ceramic decoupling caps from the TDA1541A modules (DI8T prototype), only leaving the "audio grade" Panasonic electrolytic capacitors in place. I expect another one after replacing all 78XX regulators with eight small TL431 shunt regulator modules.
I already use Vishay S102 bulk metal foil resistors on specific positions.
This depends on how noise affects a specific circuit. Clock circuits (crystal oscillator, reclockers and buffers) require very clean power supplies.
Techniques used to obtain clean power supplies for digital circuits, may not work for analogue power supplies.
I got an eureka today, after removing all ceramic decoupling caps from the TDA1541A modules (DI8T prototype), only leaving the "audio grade" Panasonic electrolytic capacitors in place. I expect another one after replacing all 78XX regulators with eight small TL431 shunt regulator modules.
I already use Vishay S102 bulk metal foil resistors on specific positions.
maxlorenz said:Back on topic...
I want to share the following experience:
I usually listen music for an hour before going to the job on the morning
I put together the following humble system, just to experiment:
Sony VAIO laptop; Ubuntu (untweaked, I' m afraid)--> common type USB cable--> DI16 DAC with Doede's USB module (TeddyRegs here and there...)--> unbalanced IC's from Silver Sonics, "Pro Studio" type
( pseudobalanced config.)--> CS3310 based preamp--> 24AWG solid core braided OCC copper interconnects (4 wires per lead; the same material as my TVC)--> My-Reference monoblocks--> Supra Ply 2.0--> Paradigm M3 monitors...uff!
(I already commented how good I think it sounds)
Listening to Bach's English Suites, Glen Gould on pianoforte (Sony, remastered).
I assumed a relaxed sitting position and began a rhythmic, deep and slow type of breathing, concentrating my gaze on a fixed point (my unused CD PRO
That's all...no more bothering from my part.
Cheers,
M
The healer apprentice...
Hi Max.
Good to see that somebody is able to enjoy the result of the all the DIY "hazzle" !
BTW: I am running Voodoo-Cable (silver solid core all over the place - internal and external)
Try to get the TVC out of the loop. I got one (S&B from Bent Audio) sitting idle since a year by now. You can not imagine how much
better and faster a direct amp connection is.
This way you can get rid of the DAC output caps, if your amp got
coupling caps on the input.
Perhaps you start trying XMMS with internal volume control.
Install yourself an Ubuntu Studio this comes with a stock realtime
kernel. The realtime-features you can use for XMMS by selecting its
realtime option.
The USB Input tweak: Since you use the DDDAC receiver it is "almost"
not applicable for your setup, since the DDDAC has its own powersupply for the receiver. Of course you're still not galvanically isolated. This I achieve with an Opticis Cable http://industrialcomponent.com/opticis/optusbext.html
The downstream Opticis receiver is powered with 5V. Mine is now battery powered with a 5V Teddy-Reg. I just cut the old 5V supply cable ( You can't see it on the Opticis pictures , but it comes with an external power-supply for the receiver side) and connected it to the Teddy-Reg. The receiver is directly connected to the
ec-design input - no cable in between. SInce the re-clocker is fired via USB, it'll be fired by the 5V Teddy-Reg over the Opticis receiver-interface. There are no extra tweaks necessary.
Another advantage: You could even use now a 6m or longer Opticis without having the USB limitations.
And yes - well powered and isolated - IMO the ec-design re-clocker is a clear improvement over the DDDAC receiver. Though Off-The-Shelve without these tweaks I wouldn't call it improvement. Even though you gain some details, you also catch more distortions on the receiver side - if powered with 5V PC-Power.
Cheers
Hi -ecdesigns-
First of all, sorry for all the OT's (my people use to say sorry before speaking ).
Now that I am a fan of the tantalum + SMD ceramic combo (after the TeddyRegs) for powering labors, I can speculate that maybe your Pana and ceramics were resonating?
Nah! you surely thought of this before.
I only got 0805 ceramics but I search the smaller ones...
Dear Soundcheck,
Arrrggh! At those prices I'd better learn how to use ultra fast optocouplers...
Thanks again for your Ubuntu tips. I remember them and would try them ASAP (meaning "who knows when"...)
My other DAC is DC coupled to my other amp through my TVC
Isn't any digital (software) volume control supposed to degrade resolution???
Cheers,
M
First of all, sorry for all the OT's (my people use to say sorry before speaking
).Now that I am a fan of the tantalum + SMD ceramic combo (after the TeddyRegs) for powering labors, I can speculate that maybe your Pana and ceramics were resonating?
Nah! you surely thought of this before.
I only got 0805 ceramics but I search the smaller ones...
Dear Soundcheck,
Arrrggh! At those prices I'd better learn how to use ultra fast optocouplers...
Thanks again for your Ubuntu tips. I remember them and would try them ASAP (meaning "who knows when"...)
My other DAC is DC coupled to my other amp through my TVC
Isn't any digital (software) volume control supposed to degrade resolution???
Cheers,
M
Hi soundcheck,
The USBDI2S 48 MHz masterclock is powered by the filtered USB power supply, this was necessary to split receiver and I2S output GNDs, by using differential buffers. However, depending upon the quality of USB bus power supply quality, masterclock power supply could get polluted, this will then result in higher jitter, but jitter will always be much lower than the PCM2706 BCK output. Bad USB interlinks could also degrade USB power supply quality (USB bus power supply wire resistance).
So you can indeed improve performance by using an external clean USB power supply. The PCM2706 always produces sufficient jitter, despite clean power supply. The PCM2706 VLSI suffers from on-chip ground-bounce, simply preventing low jitter (effects of ground-bounce increase jitter).
This problem was solved by designing the LRM (local reclocker module), it can be fed by a separate clean power supply, and can also be connected to the USBDI2S module when the reclocker on the USBDI2S module is bypassed. I will post some information about this later.
USBDI2S module signal flow:
PCM2706 > reclocker > RS422 driver > RS422 receiver > output
USBDI2S + LRM signal flow:
PCM2706 > (reclocker bypassed) > RS422 driver > RS422 receiver > LRM > output
UTOS2 + LRM2 signal flow:
PCM2706 > buffer > high-speed opto-coupler > LRM2 > output
Note that the USB GND and RS422 receiver GND on the USBDI2S module are not connected, and the I2S signals are transported through RS422 differential signals.
The problem with optical insulation between both source and DAC was already solved by UTOS2 module.
You can also read these posts:
DI8M block diagram (including UTOS & LRM2) post #1878
LRM info: post #1773
LRM2 info: post #1839, #1876, #1894
UTOS2 info: post #1789, #1838
Apple airport express info (bit perfect playback with windows + iTunes): post #1884
The USBDI2S 48 MHz masterclock is powered by the filtered USB power supply, this was necessary to split receiver and I2S output GNDs, by using differential buffers. However, depending upon the quality of USB bus power supply quality, masterclock power supply could get polluted, this will then result in higher jitter, but jitter will always be much lower than the PCM2706 BCK output. Bad USB interlinks could also degrade USB power supply quality (USB bus power supply wire resistance).
So you can indeed improve performance by using an external clean USB power supply. The PCM2706 always produces sufficient jitter, despite clean power supply. The PCM2706 VLSI suffers from on-chip ground-bounce, simply preventing low jitter (effects of ground-bounce increase jitter).
This problem was solved by designing the LRM (local reclocker module), it can be fed by a separate clean power supply, and can also be connected to the USBDI2S module when the reclocker on the USBDI2S module is bypassed. I will post some information about this later.
USBDI2S module signal flow:
PCM2706 > reclocker > RS422 driver > RS422 receiver > output
USBDI2S + LRM signal flow:
PCM2706 > (reclocker bypassed) > RS422 driver > RS422 receiver > LRM > output
UTOS2 + LRM2 signal flow:
PCM2706 > buffer > high-speed opto-coupler > LRM2 > output
Note that the USB GND and RS422 receiver GND on the USBDI2S module are not connected, and the I2S signals are transported through RS422 differential signals.
The problem with optical insulation between both source and DAC was already solved by UTOS2 module.
You can also read these posts:
DI8M block diagram (including UTOS & LRM2) post #1878
LRM info: post #1773
LRM2 info: post #1839, #1876, #1894
UTOS2 info: post #1789, #1838
Apple airport express info (bit perfect playback with windows + iTunes): post #1884
Bernhard said:
Who cares about "lots of people"!
If these people are interested I gotta a nice Bent-Audio TVC for sale at a very fair price! Just drop me a mail.
I reported some time ago an A/B test we did at Bert Doppenberg
premisis. BD-Design is running also S&B TVC. I was running
my brutefir volume control. We were running a TwinDac directly
coupled to Bert's amps. The actual idea was to compare my Linux
setup with XXHE. The difference with or without TVC was so obvious. Transients are so much fatster. You'll loose some kind of muddyness! Actually the same what I experienced in my own setup.
People are saying the S&B sounds like a peace of wire. I respond to that: Have you ever listenend to a piece of wire?
One thing which of course might happen:
Since the S&B is softwashing the sound you might hide the deficiences of your upstream chain. If you take the TVC out
you need to optimize your chain a bit further. After that - you don't look back.
Cheers
soundcheck said:
I love those postings, they imply that the "chain" of the poster stands above all others.
Similar to postings like that it depends on proper implementation, which implies that the poster has it and the others don't
By the way, did you implement your TVC properly ?
My own experience is limited at the moment because I have not finished the second TVC yet, but from transformers in general I can say that they introduce less errors than caps or the very best opamps.
Any op amps left in your chain ?
Which DAC are you using ?
Bernhard said:
I love those postings, they imply that the "chain" of the poster stands above all others.
Similar to postings like that it depends on proper implementation, which implies that the poster has it and the others don't
By the way, did you implement your TVC properly ?
My own experience is limited at the moment because I have not finished the second TVC yet, but from transformers in general I can say that they introduce less errors than caps or the very best opamps.
Any op amps left in your chain ?
Which DAC are you using ?
Perhaps you missed some parts of my post. I proved it
on more then one chain, even with other people involved.
BD-Design must have installed his TVC the same wrong way
than as I did.
You could have saved some money for your 2nd TVC if we would have had this discussion earlier.
I am not lifting my chain above others. I am just talking about my experiences about a small part of the chain. You can buy as many
TVCs as you want. It's your own free decision.
I am running a DDDAC with passive I/V Riken-Resistor. The USB-receiver is the ec-design reclocker. I removed the
output cap of the DDDAC. On my amp input I am running a very small (0,1uf) "Glimmer"-type cap from Thel. No active elements! Just 1inch of wire.
Cheers
soundcheck said:
Perhaps you missed some parts of my post. I proved it
on more then one chain, even with other people involved.
BD-Design must have installed his TVC the same wrong way
than as I did.
You could have saved some money for your 2nd TVC if we would have had this discussion earlier.
I am not lifting my chain above others. I am just talking about my experiences about a small part of the chain. You can buy as many
TVCs as you want. It's your own free decision.
I am running a DDDAC with passive I/V Riken-Resistor. The USB-receiver is the ec-design reclocker. I removed the
output cap of the DDDAC. On my amp input I am running a very small (0,1uf) "Glimmer"-type cap from Thel. No active elements! Just 1inch of wire.
Cheers
1, the TVC needs to be terminated properly on the secondary with a defined impedance that is built by a termination resistor plus the parallel input impedance of the power amp. A very small cap might be needed additionally in some cases,
2, failure to do so results in rising frequency response plus bad transient / square response ( ringing ) , giving smeared sound.
3, TDA 1543 is a cheap & low performance DAC that sounds smeared already, no matter how many chips are put in parallel.
Maybe not the first choice to judge other components.
4, Non os without a proper reconstruction filter...
5, Fortunately I built my TVC myself, no need to spend money.
Dear Soundcheck,
Now that you are happy with the sound, try a low noise (low Temp. coef.) resistor instead of the Riken which are über-noisy...(I was once fond of them, though).
With that and the TeddyReg based tower PS I bet you will get more detail...or more HF noise? Who knows...
With a TVC you could connect a balanced output DAC DC coupled to the amp...even DIDAC on balanced mode
I have postponed the square wave analysis of mine...it seems RC is a must with some amps...thanks Bernhard for the reminder.
Cheers,
M
Now that you are happy with the sound, try a low noise (low Temp. coef.) resistor instead of the Riken which are über-noisy...(I was once fond of them, though).
With that and the TeddyReg based tower PS I bet you will get more detail...or more HF noise? Who knows...
With a TVC you could connect a balanced output DAC DC coupled to the amp...even DIDAC on balanced mode
I have postponed the square wave analysis of mine...it seems RC is a must with some amps...thanks Bernhard for the reminder.
Cheers,
M
Handling settling time
Hi Mauricio,
I've been following for what you've done with ps and other matters for the DAC and it's very interesting indeed. I knows that three leg regs 7xxx series which were used in ecdesigns DAC has good performance, but the you also mentioned that you get better performance in every aspect (post #1875) and that's tempting.
TeddyRegs has a lack of settling time. How this affect to the DAC and how did you handle this problem?
Thanks for sharing.
-ims-
Hi Mauricio,
I've been following for what you've done with ps and other matters for the DAC and it's very interesting indeed. I knows that three leg regs 7xxx series which were used in ecdesigns DAC has good performance, but the you also mentioned that you get better performance in every aspect (post #1875) and that's tempting.
TeddyRegs has a lack of settling time. How this affect to the DAC and how did you handle this problem?
Thanks for sharing.
-ims-