Improvement on cd-players (Off Topic)
by madprofessor 
, 27777 Ganderkesee, Germany, Monday, October 15, 2007, 18:44 (6031 days ago)
I recently bought a Musical Fidelity Tri Vista CD-Player.
The sound was ok bt not great. Also the usual modifications brought no real big improvement.
Than I exchanged the internal oscillator for an external clock -in this case a PTS 160 ( http://www.programmedtest.com/pts160.html ) There was a real vast improvement!
Whoever aarround here is fit to find the oscillator in a cd player and the right connections should really try this version. Also any other low phase-noise, high stability generator, like HP 3325B, HP 8642A , HP 8662A could be used.
Depends which oscilator is taken , the price would be lower than one of the "High-End-Superclocks" - they are normally specified maximum 5 ppm.
A good source liehe the PTS 160 is 0,003ppm! And look at the signal quality also!
Good look with trying!
Best regards
Stephan
Improvement on cd-players
by Bert 
, Monday, October 15, 2007, 22:29 (6031 days ago) @ madprofessor
Hi Stephan,
I do not understand. Do you mean you have a whole box with buttons as shown on the pictures as external clock?
Bert
--
BD-Design - Only the Best!
Improvement on cd-players
by madprofessor , 27777 Ganderkesee, Germany, Monday, October 15, 2007, 23:24 (6031 days ago) @ Bert
Hallo Bert!
Yes, the whole box, the PTS 160 acts as external clox. Reason is, I can set it to different frequencies, in case I use another cd-player.
Ok, it´s a bit of overkill, but the PTS 160 is locked to the house frequency standard, consisting of several caesium and gps-controlled rhubidium standars.
that gives a frequency accuracy of 1 x 10E-13! but more important is: absolutely no jitter!
Best regards
Stephan
Improvement on cd-players
by Bert , Monday, October 15, 2007, 23:33 (6031 days ago) @ madprofessor
Yes, the whole box, the PTS 160 acts as external clox. Reason is, I can
set it to different frequencies, in case I use another cd-player.
Clear, you do plan to try different players each half year? Sorry, just kidding... :wink:
that gives a frequency accuracy of 1 x 10E-13! but more important is:
absolutely no jitter!
I am not a digital nerd but as far as I know is that there are many more things introducing jitter, not just the clock. I am sure that a stable clock does do very good things but the frequency needed already depends on temperature and the DAC itself...how do you know the exact frequency your DAC needs? And what about the clock of the transport...is that using the same external clock?
Bert
--
BD-Design - Only the Best!
Improvement on cd-players
by robmatthew , Monday, October 15, 2007, 23:45 (6031 days ago) @ Bert
$4500??? Is that correct?
Improvement on cd-players
by madprofessor , 27777 Ganderkesee, Germany, Monday, October 15, 2007, 23:51 (6031 days ago) @ robmatthew
Maybe new, but You get this units used a lot cheaper!
Improvement on cd-players
by madprofessor , 27777 Ganderkesee, Germany, Monday, October 15, 2007, 23:50 (6031 days ago) @ Bert
Hallo Bert!
I just want to be flexible...in case for example the transport dies...
Of course a lot of other components can cause Jitter.
Ok, a few components to change, like in the Tri Vista about 80 capacitors and 7 bridge rectifiers. Than keep the unit thermal stable, that means, not to switch it off and some work on the case. I´m using a lot of alloy foam... the unit is now about 5 kg heavier with all the applied damping material..
I use the same clock for transport and DAC.
I still have to change some digital ic´s for units better performing.
Also some work on the traGCsport to be done, to get rid of vibratins...
I think in one month time I´ll have the unit finished...
Best regards
Stephan
Improvement on cd-players
by Guido tent , Wednesday, October 17, 2007, 23:45 (6029 days ago) @ madprofessor
dear Stephan,
You are missing the point. ppm's are of no importance for playback quality of DACs.
What matters is jitter (at the moment of conversion).
The generator you refer to may be very precise in terms of absolute frequency (ppms'), this does not imply that it doesn't jitter.
On the contrary, the fact that it is programable and lockd to GPS makes me conclude that it is a frequency synthesiser; The inner PLL/VCO therefor is never based on crytals hence the jiter caNot be as good as that of a decently designed clock.
For a very short but clear insight in the difference between ppm's and jitter, please look here:
http://www.tentlabs.com/InfoSupport/Technology/page34/page34.html
best regards
-
Guido
Improvement on cd-players
by madprofessor , 27777 Ganderkesee, Germany, Thursday, October 18, 2007, 06:52 (6029 days ago) @ Guido tent
Hallo Guido!
I´m well aware, that the stability expressed in ppm, has no indication of the existing jitter.
Phase noise is a way also to express jitter, as also can be seen on the link You provided. The generator I use, has very low phase noise, which was a reason for the choice! The generator used is a decade generator, which gives less problems than a "normal" wide band synthesizer. Also the gps/rhubidium/caesium frequency standard has extrmely low phase noise. Of course all coupling is done by matched transformers, to avoid ground loops and harmonics, as well as rf-problems.
I agree, that a real good crystal based oscillator is the better solution. I think of the class of for example a HP 10811A, which has a phase noise down to -150dBc. Unfourtunately, I didn´t find a crystal oscillator of that quality in the required frequency.
I measured several tcxo´s in the range of the ones also offered by Trent labs. The performance in aspects of jitter, phase noise, and general noise floor are much worse, than the system I´m using now.
I don´t claim to have found the "final solution" - there will always be measures to get further improvements!
I´m interested in further discussins, regarding this topic!
By the way, thanks for mentioning that link - some very intersting articles can be found there! That´s one of the reasons, I like this forum so much - always to find new concepts to think about and some to put the into real life!
Best regards
Stephan
Improvement on cd-players
by p314 , Monday, October 22, 2007, 13:03 (6025 days ago) @ madprofessor
Stephan,
the idea of reducing the WAF of my stereo by adding another - preferably large, grey- box full of knobs and dials is extremely attractive. :-))
Now: At what frequency do you have to set the generator? If I understand things correctly, the frequency of an internal clock depends on the circuitry you are using the clock with. What about the "word clock"? Is there a given fixed frequency you must use for 44kHz or 96kHz sampling?
Regards,
Peter
Improvement on cd-players
by p314 , Monday, October 22, 2007, 13:06 (6025 days ago) @ p314
Bert,
why is "k-NO.b" in the above mail filtered? I was not aware that it is some sort of swear word...
Peter
Improvement on cd-players
by Bert , Monday, October 22, 2007, 13:39 (6024 days ago) @ p314
Hi Peter,
why is "k-NO.b" in the above mail filtered? I was not aware that it is
some sort of swear word...
DoNo...I copied a list with most swear words used and use that. I deleted the word knob from that list though (and edited your original message).
Sorry,
Bert
--
BD-Design - Only the Best!
Improvement on cd-players
by madprofessor , 27777 Ganderkesee, Germany, Tuesday, October 23, 2007, 10:21 (6024 days ago) @ p314
The basical idea is, to get a real good clock-signal instead of using the low quality crystal oscillator used by the manufactorer. OXCO´s in the required frequency and in good quality are not on the common market. So I used that decade oscillator, which gives a very good signal quality(Phase noise, jitter, noise floor and stability) cause it was on hand. I set, in the case of the Musical Fideluity Tri Vista the frequency to 16,9344 MHz. I still have to exchange some IC´s which create the frequencies internally needed.
In that configuration I need no extra oscillator for the word-clock. In the DA-converter I use two clock-oscillators - one for Bit-clock and one for word-clock.
By the way, I´ve the clock generator in the next room. Cable length is not critical, if proprly terminated.
Best regards
Stephan