hifi said:
Dear Hifi,
2 small remarks and a question for you
1. The input of the 1543 does not care how square the wave form is, the only thing which matters is the timing, this is the essence of digital logic. If the timing would have been bad in the first place, a VHC04 will not change this, even if it shows squareness in perfection.... A VHC74 with a new clock however will do the job. This is a proven solution, seen in many designs.
2. The reclocking in the circuit has nothing to do with driving the parallel DACs, it is to provide a lower jitter clock to the 1543
3. Earlier you say parallel DACs is asking for trouble. What kind of trouble would you suggest I have been looking for
? And if you would measure the clock signals with a scope, how would you judge and compare them (them = situations with different # of DACs) and what relation with lesser audio performance should we look for?thanks for enlightning me a bit here !
doede
datastream .. bla bla
Guys ,
Let me try to put most important thing clear :
Doede's paralleled 1543 dac has sonicaly overruled all my previously homebrew DA converters with great ease ( even without any system clock mods so far ) .
Witouth a doubt a major improvement in my audiosystem over years ..
I strongly believe that mainly the system clock from the used CD player or CD-drive has to ( could be ) modified in order to offer better datastreaming into the 1543's ....
At this point I fully agree with Doede's statement posted above .
Also batterie feed can bring magic into this project
On the other hand , keep in mind that this DAC can be realised at
a fraction in costs compared with factory build DA converters that in many cases performing quite lousy !
Happy Ears !
Guys ,
Let me try to put most important thing clear :
Doede's paralleled 1543 dac has sonicaly overruled all my previously homebrew DA converters with great ease ( even without any system clock mods so far ) .
Witouth a doubt a major improvement in my audiosystem over years ..
I strongly believe that mainly the system clock from the used CD player or CD-drive has to ( could be ) modified in order to offer better datastreaming into the 1543's ....
At this point I fully agree with Doede's statement posted above .
Also batterie feed can bring magic into this project
On the other hand , keep in mind that this DAC can be realised at
a fraction in costs compared with factory build DA converters that in many cases performing quite lousy !
Happy Ears !
Congratulations Mr. Happy ears with your achievement! Yep, those TDA1543 daccies are quite cool.
I am (still) busy with my newest non-os attempt, the Nonoz III. I think I'll release it in November. I waited because I was not certain whether the sound was completely balanced (sounded a bit too dry and thin in the beginning). But I really begin to like the DAC (and so did a friend of me). Anyway, it has gotten a bit warmer too after some burn-in. IMHO, the Nonoz III is much better and accurate than the first two versions...
Fedde
I am (still) busy with my newest non-os attempt, the Nonoz III. I think I'll release it in November. I waited because I was not certain whether the sound was completely balanced (sounded a bit too dry and thin in the beginning). But I really begin to like the DAC (and so did a friend of me). Anyway, it has gotten a bit warmer too after some burn-in. IMHO, the Nonoz III is much better and accurate than the first two versions...
Fedde
Have Fun !
Hi Fedde , thanks !
I just cut a few words from an earlier posting of you on this forum .
Of course I also start with a "single 1543" dac on Doede's PCB .
I didn't like the idea of simply starting with the "stacked version" .
Altough using one 1543 is sounding oke , and already with much greater imaging , better dynamics etc.
By adding #'s of 1543 dacs the sound became better and better .
I simply divided the Rload & Rref resistor each time I'v stacked another Dac on the 1543 Tower ..
With 4 pieces in parallel , 8.5 Volt feed , Rload 540 Ohm . R.ref counterpoint at 3.75 VDC at output . the sound reach that level of acuracy that cannot be achieved with only a single 1543 .
Let's speak in therms of : soundstage depth/wide/height (3D) ,
Huge dynamic response ( never heard like this before ) !
Also I find out that individual musicians and voices are realy steppin forward into the listening room , great expirience !!
Bass response in one word is spectacular !! not comparable with any other Dac .. massive , huge , and very natural .
I moved forward ........... up to 8 1543's ..........
In comparison : My latest version with 8 pcs. stacked is simply wonderful . with all previous improvements , but more authority and clearness .. a fraction better high freq. too . but relatively ......
4 pcs. are already super !!
My question to you Fedde ,
Are you considering the stacking methode in your MK3 NONOdac too ?? just curious !
Please keep us posted about your results too , perhaps you want to try I2S streaming too . I ommited the Cristal receiver in order to get rid of the extra electronics .... I strongly believe in shortest path .
Yours,
Happy Ears .
fedde said:
Hi Fedde , thanks !
I just cut a few words from an earlier posting of you on this forum .
Of course I also start with a "single 1543" dac on Doede's PCB .
I didn't like the idea of simply starting with the "stacked version" .
Altough using one 1543 is sounding oke , and already with much greater imaging , better dynamics etc.
By adding #'s of 1543 dacs the sound became better and better .
I simply divided the Rload & Rref resistor each time I'v stacked another Dac on the 1543 Tower ..
With 4 pieces in parallel , 8.5 Volt feed , Rload 540 Ohm . R.ref counterpoint at 3.75 VDC at output . the sound reach that level of acuracy that cannot be achieved with only a single 1543 .
Let's speak in therms of : soundstage depth/wide/height (3D) ,
Huge dynamic response ( never heard like this before ) !
Also I find out that individual musicians and voices are realy steppin forward into the listening room , great expirience !!
Bass response in one word is spectacular !! not comparable with any other Dac .. massive , huge , and very natural .
I moved forward ........... up to 8 1543's ..........
In comparison : My latest version with 8 pcs. stacked is simply wonderful . with all previous improvements , but more authority and clearness .. a fraction better high freq. too . but relatively ......
4 pcs. are already super !!
My question to you Fedde ,
Are you considering the stacking methode in your MK3 NONOdac too ?? just curious !
Please keep us posted about your results too , perhaps you want to try I2S streaming too . I ommited the Cristal receiver in order to get rid of the extra electronics .... I strongly believe in shortest path .
Yours,
Happy Ears .
Re: Have Fun !
I can't share your enthusiasm about stacking. I've build DAC with 4 chips in parallel, but quickly changed it to a single chip. While multi DAC version might offer more dynamics and solidity, it's a trade off for lack of immediacy and high freq. extention (air). It simply sounded dull in my system and I didn't like it.
Happy Ears said:
I can't share your enthusiasm about stacking. I've build DAC with 4 chips in parallel, but quickly changed it to a single chip. While multi DAC version might offer more dynamics and solidity, it's a trade off for lack of immediacy and high freq. extention (air). It simply sounded dull in my system and I didn't like it.
single or multi
Hi Peter , I think this is interesting !
at my set-up I'v noticed quite large differences in specs of 1543 chips ... especialy differantial DC voltage shifts at the L&R output legs .
I'm not that lucky to have acces to let's say 100 pcs. 1543 chips .. but I'm quite sure that carefully selecting each individual chip could offer better overall performance with stacked 1543 NONO's .
With all respect to the single chip version : it could be a matter of taste you prefer it , and of course I can only speak for myself here but honestly I couldn't live without the enormous soundstage and dynamics of the multi Dac version . honestly !
Happy Ears
Hi Peter , I think this is interesting !
at my set-up I'v noticed quite large differences in specs of 1543 chips ... especialy differantial DC voltage shifts at the L&R output legs .
I'm not that lucky to have acces to let's say 100 pcs. 1543 chips .. but I'm quite sure that carefully selecting each individual chip could offer better overall performance with stacked 1543 NONO's .
With all respect to the single chip version : it could be a matter of taste you prefer it , and of course I can only speak for myself here but honestly I couldn't live without the enormous soundstage and dynamics of the multi Dac version . honestly !
Happy Ears
I might try again, maybe this time I will hear it differently
But as I remember it's also opinion of other members, like Elso and J-P, that paralleling chips takes something away from the sound. Even Kusunoki in his 3rd version decided on a single chip.
But definitely you should try reclocking (if you didn't so far). In my DAC it brought quite a big improvement.
But as I remember it's also opinion of other members, like Elso and J-P, that paralleling chips takes something away from the sound. Even Kusunoki in his 3rd version decided on a single chip.
But definitely you should try reclocking (if you didn't so far). In my DAC it brought quite a big improvement.
A big advantage is the lowering of the output impedance. BTW: I think a 5V supply sounds dull anyway
My motivation to stay with a lonesome chippie in the Nonoz III is that I want to come close to the performance of the Shigaraki/Progression dacs. Maybe I'll try to raise the bar with the Nonoz IV, but don't estimate the power you can get from a single TDA1543.
Fedde
My motivation to stay with a lonesome chippie in the Nonoz III is that I want to come close to the performance of the Shigaraki/Progression dacs. Maybe I'll try to raise the bar with the Nonoz IV, but don't estimate the power you can get from a single TDA1543.
Fedde
fedde said:A big advantage is the lowering of the output impedance. BTW: I think a 5V supply sounds dull anyway
Fedde
Hi Fedde , I think you probably hit the spot here .
Doesn't lowering the output impedance automaticly means less problems with freq.range / bandwide of the analog signal output ?
I'v noticed better performance at both high and low freq.range after stacking 1543 daccies ..
You are absolutely right about the 5 volt supply .... between 7.5 and 8.5 volt soundstage increase dramaticly ... wow .
Despite of the cooling problem ... it's so easy to fit a small cooling fan ...
Happy Ears
Fan supply solution
Hi Mr. Picolo
Don't worry , the solution is on hand ...... simply use a separate power supply to avoid power supply influence ...... and as far as my knowledge reach , there are brushless fans available too .
For those who might be worried about noise effects : the 12 volt fan in my Marantz player works at 5 volt !! the cover of the player disable all audible noise .
To avoid possible resonance on the PCB I glued the fan with siliconkit .
The main reason I decided to use a cooling fan is the much better thermal stability of the 1543 stack . don't forget that thermal drifts automaticly means that bias and ref voltage of the dac chips change too . not recommendable IMO .
Happy Ears .
Hi Mr. Picolo
Don't worry , the solution is on hand ...... simply use a separate power supply to avoid power supply influence ...... and as far as my knowledge reach , there are brushless fans available too .
For those who might be worried about noise effects : the 12 volt fan in my Marantz player works at 5 volt !! the cover of the player disable all audible noise .
To avoid possible resonance on the PCB I glued the fan with siliconkit .
The main reason I decided to use a cooling fan is the much better thermal stability of the 1543 stack . don't forget that thermal drifts automaticly means that bias and ref voltage of the dac chips change too . not recommendable IMO .
Happy Ears .
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.