Did you search here? Done before a few times, this thread sums it all up:
http://www.diyaudio.com/forums/showthread.php?threadid=57719&perpage=10&pagenumber=1
As for the delay mentioned before, just think about how far sound travels in that time and then move one speaker a bit (...)
As for your schematic, you don't need the r and c on the clear.
you wont hear the one sample at switch-on being garbage.
As for the relevance of using L and -L / R and -R with two tda1541's. In theory it is an improvement if you know how the thing works. But how much of an improvement depends on the I/V stage and the powersupplies...
Search here, it's all there somewhere
And for the DEM reclocking, HtP's schematic works like a charm. Why re-invent the wheel???
http://www.diyaudio.com/forums/showthread.php?threadid=57719&perpage=10&pagenumber=1
As for the delay mentioned before, just think about how far sound travels in that time and then move one speaker a bit (...)
As for your schematic, you don't need the r and c on the clear.
you wont hear the one sample at switch-on being garbage.
As for the relevance of using L and -L / R and -R with two tda1541's. In theory it is an improvement if you know how the thing works. But how much of an improvement depends on the I/V stage and the powersupplies...
Search here, it's all there somewhere
And for the DEM reclocking, HtP's schematic works like a charm. Why re-invent the wheel???
Hi stefanobilliani,
The master clock of 8.467000 MHz has 118ns period time, since BCK is delayed 4 extra master clock cycles (QE instead of QA), total BCK delay with respect to Both DATA and WS will be 472ns.
BCK = 2.116800 MHz, and has 1 / 2116800 = 472ns period time.
Both WS and DATA can only be sampled correctly if the delayed BCK time stays well within half the BCK period time or 472 / 2 = 236ns (BCK not inverted).
BCK is delayed by 4 extra master clock period times (4 x 118ns = 472ns), this exceeds the maximum allowable BCK delay (with respect to WS and DATA) by 236ns. So it's very likely DATA corruption occurs.
Guido may have overlooked that R (2.2 Ohm) is actually just part of power supply decoupling (reset is simply tied to VCC).
The master clock of 8.467000 MHz has 118ns period time, since BCK is delayed 4 extra master clock cycles (QE instead of QA), total BCK delay with respect to Both DATA and WS will be 472ns.
BCK = 2.116800 MHz, and has 1 / 2116800 = 472ns period time.
Both WS and DATA can only be sampled correctly if the delayed BCK time stays well within half the BCK period time or 472 / 2 = 236ns (BCK not inverted).
BCK is delayed by 4 extra master clock period times (4 x 118ns = 472ns), this exceeds the maximum allowable BCK delay (with respect to WS and DATA) by 236ns. So it's very likely DATA corruption occurs.
Guido may have overlooked that R (2.2 Ohm) is actually just part of power supply decoupling (reset is simply tied to VCC).
Hi,
Bit switch decoupling caps
Having been pondering this for a few weeks and reading lots of different things from different people can someone give me a definitive answer regarding the best type of caps to use (film or electrolytic) :- Oscon SH, BG NX HiQ, PPS / PP film, COG/NPO ceramic or something else, perhaps a combination ?
I'm thinking of something like (MSB->LSB)
13,18 2.2uF Oscon SH , 0.47uF BG NX HiQ , 10nF ceramic
12,19 1uF Oscon SH, 0.47uF BG NX HiQ, 10nF ceramic
9-11,20-21 0.47uF BG NX HiQ, 150nF film, 10nF ceramic
7-8,22-23 150nF film, 10nF ceramic
I put the 150nF in because I have some Wima MKI2 PPS. Perhaps the ceramics are unnecessary.
Cheers,
Jon
Bit switch decoupling caps
Having been pondering this for a few weeks and reading lots of different things from different people can someone give me a definitive answer regarding the best type of caps to use (film or electrolytic) :- Oscon SH, BG NX HiQ, PPS / PP film, COG/NPO ceramic or something else, perhaps a combination ?
I'm thinking of something like (MSB->LSB)
13,18 2.2uF Oscon SH , 0.47uF BG NX HiQ , 10nF ceramic
12,19 1uF Oscon SH, 0.47uF BG NX HiQ, 10nF ceramic
9-11,20-21 0.47uF BG NX HiQ, 150nF film, 10nF ceramic
7-8,22-23 150nF film, 10nF ceramic
I put the 150nF in because I have some Wima MKI2 PPS. Perhaps the ceramics are unnecessary.
Cheers,
Jon
guido said:
I try to remind you that the air is the medium , not a calculation. Should we open a thread called " what is stereophony? " . Maybe.
I did read something on what are the geometries in stereo microphonic recording.
I dont think once you have a stereo track recorded , moving the speakers can change very much the things . Moving the microphones yes .
I mean , having wrong timing in a DAC between the channels , demage badly the stereo informations . It is not a matter of how far the sound travels . It is a matter of air , the medium - at that speeds .
And Listening to a stereo recording..., just - when things are correct you can *realize* or *see* how far the sound travels .
Hi all,
This thread has >180000 views and >1500 posts...but building enthusiasm is approaching 0
A pity because this is an excellent DAC.
As PCB only option it is quite affordable.
In fact this is the best source I own at present and the best I've heard (not that I know too many). NOS qualities plus enhanced resolution.
Now that I'm using my new PC based audio server with short IC's the sound improved further as EC predicted: faster transients. It is so clean that my former audiophile's impulse for wire and cable upgrades reappeared
I'm using large copper wire as PCords and common type mic cable or OCC copper solid wire as IC's. I should try some silver DIY IC cable and OCC copper Power Cord.
I own a cheap and highly modded Samsung multiformat player (not yet the clock upgrade: lost in cyberspace). With the mods it doesn't sound bad at all. DI16 retrieves more detail, texture and spacial info, while having a warmer, weighter presentation, on CD layer, than its SACD counterpart. The soundstage is even wider. So, no more "hi resolution format fever" for me now. I will wait for blueray.
These are the speakers that I use:
Height is 145cm
First varnish hand at first anniversary.
Sound update:
Bach's Goldberg variations; Bach partitas for hapsicord:
My active system makes the hapsicord, usually harsh and desagreeable, "nice" sounding (in a good way) with richness of high and low harmonics. Bass notes an its harmonics are not always heard on lesser speakers; not here. Bass harmonics reverberations travel through the recording room...
The attacks and the rhythm are strong points of DI16. As I said the "structure" and the "meaning" of the music is clearer: gestalt
Opera:
Madama Butterfly; London-Decca; Tebaldi-Bergonzi; Serafin cond.(?)
Rich Decca sound with low double basses, perfectly articulated on the short introduction and later. Voices full and round (well, Bergonzi's voice is round and velvety anyway) with great emotional impact (midrange).
Soundstage is immense This DI16-UCD180-active Autograph combination really gives a "live-concert hall" presentation.
Otello; Living Stereo; Vickers-Rysanek; Gobbi; Serafin-cond.
I love this recording. Totally theatrical. Here, most shocking is the lack of dynamic compression: my >100db sensitive monsters really like the workout. Slam!
The feeling of the recording room (Rome Opera House?) is superb: hugest ever, IME.
DI16 really separates the different section of the orchestra and makes you feel that there are groups of soloists instead of masses of instruments sounding together...
So, highly recommended for classical music fans!
But, anyway, the other day some friends came by and played non-classical music: Nirvana and Metallica, unplugged or something like that, The Police also. It all sounded very good, with impact and raw energy. Not my preference but...I played for them a Mingus CD in vengeance, free rhythm and virtuosity...they liked it
Well, hope this helps.
Cheers everybody.
M
PS: now I have to learn how to extract more juice from my PC.
This thread has >180000 views and >1500 posts...but building enthusiasm is approaching 0
A pity because this is an excellent DAC.
As PCB only option it is quite affordable.
In fact this is the best source I own at present and the best I've heard (not that I know too many). NOS qualities plus enhanced resolution.
Now that I'm using my new PC based audio server with short IC's the sound improved further as EC predicted: faster transients. It is so clean that my former audiophile's impulse for wire and cable upgrades reappeared
I'm using large copper wire as PCords and common type mic cable or OCC copper solid wire as IC's. I should try some silver DIY IC cable and OCC copper Power Cord.
I own a cheap and highly modded Samsung multiformat player (not yet the clock upgrade: lost in cyberspace). With the mods it doesn't sound bad at all. DI16 retrieves more detail, texture and spacial info, while having a warmer, weighter presentation, on CD layer, than its SACD counterpart. The soundstage is even wider. So, no more "hi resolution format fever" for me now. I will wait for blueray.
These are the speakers that I use:
Height is 145cm
First varnish hand at first anniversary.
Sound update:
Bach's Goldberg variations; Bach partitas for hapsicord:
My active system makes the hapsicord, usually harsh and desagreeable, "nice" sounding (in a good way) with richness of high and low harmonics. Bass notes an its harmonics are not always heard on lesser speakers; not here. Bass harmonics reverberations travel through the recording room...
The attacks and the rhythm are strong points of DI16. As I said the "structure" and the "meaning" of the music is clearer: gestalt
Opera:
Madama Butterfly; London-Decca; Tebaldi-Bergonzi; Serafin cond.(?)
Rich Decca sound with low double basses, perfectly articulated on the short introduction and later. Voices full and round (well, Bergonzi's voice is round and velvety anyway) with great emotional impact (midrange).
Soundstage is immense
This DI16-UCD180-active Autograph combination really gives a "live-concert hall" presentation.Otello; Living Stereo; Vickers-Rysanek; Gobbi; Serafin-cond.
I love this recording. Totally theatrical. Here, most shocking is the lack of dynamic compression: my >100db sensitive monsters really like the workout. Slam!
The feeling of the recording room (Rome Opera House?) is superb: hugest ever, IME.
DI16 really separates the different section of the orchestra and makes you feel that there are groups of soloists instead of masses of instruments sounding together...
So, highly recommended for classical music fans!
But, anyway, the other day some friends came by and played non-classical music: Nirvana and Metallica, unplugged or something like that, The Police also. It all sounded very good, with impact and raw energy. Not my preference but...I played for them a Mingus CD in vengeance, free rhythm and virtuosity...they liked it
Well, hope this helps.
Cheers everybody.
M
PS: now I have to learn how to extract more juice from my PC.
Re: Connecting D-I DAC to squeezebox using I2S
Hi -ecdesigns-,
Thanks a lot for a detailed and informative answer (post #1479)! Nice re-housing for the squeezebox you've got there.
I was a bit surprised by your statement that I2S output from the Squeezebox2 was jittery. I was under the impression that the Squeezeboxes has lower jitter on their digital outputs than most silver disc spinning transports out there. Is there perhaps a difference between the SB2 and the SB3 in this regard?
Using a mac mini and controlling it remotely with a cellphone is obviously an option, but I'd prefer to stick with my Squeezebox and the simplicity and ease of use that it offers. (Having a mac mini with a TV screen of some sort is out of the question.)
Would it not be an option to modify the SB so that it is slaved off the clock in your DAC? (Like LessLoss does it) Or is S/PDIF with a seperate masterclock connection inherently inferior to using USB? I would think that a S/PDIF input and master clock output board as an option for your DAC would make it compatible with many more different digital sources.
Looking at your website, the DACs offered seem like finished designs, but from this thread I get the impression that they are still work in progress. Any news in this regard?
I have three 1541A R1 chips that seem to come from the same batch. Are there negative effects from mixing chips from different batches?
cheers,
-coolbiz
-ecdesigns- said:Hi coolbiz,
...
Well Actually I dug out my squeezebox 2 player, bought it years ago... The attached photograph shows it driving a DI 16 directly trough I2S
The I2S signals were located quickly: R5 BCK, R6 DATA, R7 WS. Then compatibility problems occurred as expected, they were solved by adding complicated glue logic....well, it only took 1 D flip-flop to delay the DATA signal 1 BCK period, that's all.
D flip-flop: clock connected to BCK, D input connected to the I2S data signal of the squeezebox, Q output connected to the DI 16 DATA input.
The I2S BCK signal was a bit jjiitteerryy. It could be corrected using the shiftregister reclocker.
But there are other elegant options for music player. I use a 17" iMac G5 with 500Gb HD, connected to the DI 8M / DI 16 through USB. The USBDI2S module provides exceptional sound quality by providing a very low jitter timing signal.
If you want a hi-tech remote control, use a cellphone and install Salling Clicker remote control software (Bluetooth). Now you can remote-control your computer, that can be located in another room (fan noise), and use the full colour screen of the cellphone to browse through your CD collection. The USB connection between computer and D-I DAC can be up to 5 meters (passive USB interlink) or longer (active USB interlink).
The mac mini with a 500Gb external storage drive could serve as a beautiful compact music player, no screen or keyboard needed, just remote-control it with a cellphone.
Hi -ecdesigns-,
Thanks a lot for a detailed and informative answer (post #1479)! Nice re-housing for the squeezebox you've got there.
I was a bit surprised by your statement that I2S output from the Squeezebox2 was jittery. I was under the impression that the Squeezeboxes has lower jitter on their digital outputs than most silver disc spinning transports out there. Is there perhaps a difference between the SB2 and the SB3 in this regard?
Using a mac mini and controlling it remotely with a cellphone is obviously an option, but I'd prefer to stick with my Squeezebox and the simplicity and ease of use that it offers. (Having a mac mini with a TV screen of some sort is out of the question.)
Would it not be an option to modify the SB so that it is slaved off the clock in your DAC? (Like LessLoss does it) Or is S/PDIF with a seperate masterclock connection inherently inferior to using USB? I would think that a S/PDIF input and master clock output board as an option for your DAC would make it compatible with many more different digital sources.
Looking at your website, the DACs offered seem like finished designs, but from this thread I get the impression that they are still work in progress. Any news in this regard?
I have three 1541A R1 chips that seem to come from the same batch. Are there negative effects from mixing chips from different batches?
cheers,
-coolbiz
Hi Onnosr
Thanks for the Foobar tip. I will try it since I still cannot make Ubuntu my main OS because I cannot connect to the net throught it. (it seems I need an ethernet type of modem and my internet provider is very slow at answering )
From memory:
Goldberg variations: Ralph Kirkpatrick.
Partitas: Ch. Hogwood.
I am a fan of Bruno Walter (also a Virgo ) but I lack his version. I own Claudio Abbado's.
To me, the most satisfying aproachs to Brukner (as a fan, not as an expert
) are Eugen Jochum's and Wilhelm Furtwängler's. Of course, there are lots of conductors that I have not heard.
FYI, I opened a thread to talk about historic recordings on music forum.
Cheers,
M
Thanks for the Foobar tip. I will try it since I still cannot make Ubuntu my main OS because I cannot connect to the net throught it. (it seems I need an ethernet type of modem and my internet provider is very slow at answering
)From memory:
Goldberg variations: Ralph Kirkpatrick.
Partitas: Ch. Hogwood.
I am a fan of Bruno Walter (also a Virgo
) but I lack his version. I own Claudio Abbado's.To me, the most satisfying aproachs to Brukner (as a fan, not as an expert
) are Eugen Jochum's and Wilhelm Furtwängler's. Of course, there are lots of conductors that I have not heard.FYI, I opened a thread to talk about historic recordings on music forum.
Cheers,
M
Hi Tubee
Did you like the R-core Tx link?
This one is more tube oriented, it seems:
http://www.vt4c.com/shop/program/main.php?group_id=7
Bye,
M
Did you like the R-core Tx link?
This one is more tube oriented, it seems:
http://www.vt4c.com/shop/program/main.php?group_id=7
Bye,
M
Hi Maxlorenz,
I am very gratefull that so many diyAudio members viewed this thread. I agree that building enthusiasm is a bit low right now, but that might change.
I appreciate your detailed description of sound quality produced by both the Autograph speakers and the DI 16. It's interesting that you noticed similar improvements using your audio set as I have with mine.
When audio set performance increases, interlink quality becomes audible too. I noticed that both RCA and speaker interlinks can significantly affect sound quality. The interlink materials used (copper, silver) have an effect too.
Operating system?
I am very gratefull that so many diyAudio members viewed this thread. I agree that building enthusiasm is a bit low right now, but that might change.
I appreciate your detailed description of sound quality produced by both the Autograph speakers and the DI 16. It's interesting that you noticed similar improvements using your audio set as I have with mine.
Now that I'm using my new PC based audio server with short IC's the sound improved further as EC predicted: faster transients. It is so clean that my former audiophile's impulse for wire and cable upgrades reappeared
I'm using large copper wire as PCords and common type mic cable or OCC copper solid wire as IC's. I should try some silver DIY IC cable and OCC copper Power Cord.
When audio set performance increases, interlink quality becomes audible too. I noticed that both RCA and speaker interlinks can significantly affect sound quality. The interlink materials used (copper, silver) have an effect too.
Operating system?
Hi coolbiz,
Thanks for your reply [post #1552]
The housing isn't annodized yet, and it has a different design from the housings I am using now. This project was put on halt after experimenting with iTunes.
Yes, both BCK and DATA originate from the same chip, so there is crosstalk, resulting in DATA related BCK jitter (BCK phase varies with DATA content). If I am correct both SB2 and SB3 have this problem. However, the SB internal DAC uses the SCK (masterclock signal) for timing.
BCK jitter is a problem with TDA1541A / TDA1543 based NOS DACs when connected to the SB2 I2S outputs.
SB2 has a separate master clock circuit (SCK), so this can be used for synchronous reclocking of BCK.
The small DATA delay circuit enables direct I2S connection to both a DI 8 or DI 16
The DI DACs only have a 48 MHz master clock for the USB interface and shiftregister reclocker, the external DI2S signal goes straight to the timing-chain / DAC chips. The SB can be placed very close to the DI DAC, so a slaved clock isn't necessary. DI DACs are also less sensitive to jitter than OS DACs.
The designs are basically finished, but there is always room for improvement. The added modifications have significantly improved sound quality. All modifications I made so far, can be made using the existing PCBs.
The selection (A, R-1, S-1, and S-2) is based on specified tolerances. So it's basically not strictly necessary to use chips from the same batch.
Thanks for your reply [post #1552]
The housing isn't annodized yet, and it has a different design from the housings I am using now. This project was put on halt after experimenting with iTunes.
Yes, both BCK and DATA originate from the same chip, so there is crosstalk, resulting in DATA related BCK jitter (BCK phase varies with DATA content). If I am correct both SB2 and SB3 have this problem. However, the SB internal DAC uses the SCK (masterclock signal) for timing.
BCK jitter is a problem with TDA1541A / TDA1543 based NOS DACs when connected to the SB2 I2S outputs.
SB2 has a separate master clock circuit (SCK), so this can be used for synchronous reclocking of BCK.
The small DATA delay circuit enables direct I2S connection to both a DI 8 or DI 16
The DI DACs only have a 48 MHz master clock for the USB interface and shiftregister reclocker, the external DI2S signal goes straight to the timing-chain / DAC chips. The SB can be placed very close to the DI DAC, so a slaved clock isn't necessary. DI DACs are also less sensitive to jitter than OS DACs.
The designs are basically finished, but there is always room for improvement. The added modifications have significantly improved sound quality. All modifications I made so far, can be made using the existing PCBs.
The selection (A, R-1, S-1, and S-2) is based on specified tolerances. So it's basically not strictly necessary to use chips from the same batch.
maxlorenz said:Hi Tubee
Did you like the R-core Tx link?
This one is more tube oriented, it seems:
http://www.vt4c.com/shop/program/main.php?group_id=7
Bye,
M
Like them both!