advertisement


MDAC first listen (part XIII)

Status
Not open for further replies.
Plutox said:
I thought D3E was about making better use of the available data space when incoming word length was < 24bit ??
Click to expand...

haha there really needs to be more information on this one, i still cant find anything solid enough to fully understand, isn't it mostly for USB ?.
 
JohnW said:
Yep - appears to be the case.

I have now been able to "indirectly" measure the effect by performing an FFT on the DAC 1.2V core supply to the digital sections - you can see different spurie products caused by the digital processing - with the Phase flip resulting in different spurie patterns.

I say indirectly as there's bugger all measurable difference on the audio output.. Still nice to be able to measure a difference somewhere at least...

Noise on the DAC's substrate will invulnerably modulate the internal Clock tree, hence effect the Audio conversion - this Jitter might be "hidden" read de-correlated by the use of a Multibit DAC array, but it will still be there just "Spread" into "noise" which will no longer be our idealised white noise......
Click to expand...

So does flipping the phase of a particular sample just turn all the 0's into 1's? Is 0000000001 the phase inverted sample of 1111111110 is the code for +.1v the analog of -.1v? Because if the coding isn't mirrored and zero centred about the 0.000000v value then all you seeing is some form of signal correlated jitter and not a direct superimposition effect.

I don't know the answer to how the voltages are coded per sample, but I'm genuinely interested.
 
sq225917 said:
So does flipping the phase of a particular sample just turn all the 0's into 1's? Is 0000000001 the phase inverted sample of 1111111110 is the code for +.1v the analog of -.1v? Because if the coding isn't mirrored and zero centred about the 0.000000v value then all you seeing is some form of signal correlated jitter and not a direct superimposition effect.

I don't know the answer to how the voltages are coded per sample, but I'm genuinely interested.
Click to expand...

Does the polarity options actually change the analogue output or just internally on the digital domain ??.
 
The mdac is an excellent add-on for the ZP90 Sonos. :D

I like the instruments with 0.90 OT XD,
but would like to have the voice accentuated more in the foreground.

Is there a version with the instrumentsound of the 0.90 and a more presence voice?
 
sq225917 said:
So does flipping the phase of a particular sample just turn all the 0's into 1's? Is 0000000001 the phase inverted sample of 1111111110 is the code for +.1v the analog of -.1v? Because if the coding isn't mirrored and zero centred about the 0.000000v value then all you seeing is some form of signal correlated jitter and not a direct superimposition effect.

I don't know the answer to how the voltages are coded per sample, but I'm genuinely interested.
Click to expand...

On the later versions of software Phase Select "does it's business by inverting data before it hits the DAC chip as opposed to flipping the phase ahead of DAC array".

Its for sure that the SQ changes are due to internal signal correlated Jitter - the bit processing in the digital domain is modulating the internal clock tree - and hence affecting the conversions clock.

D3E attempts to add a more "Random" element to the input Data pattern and thus decorrelating the resultant internal Die clock modulation... the Digital sections are no longer "just" processing audio Data, but now also a significant "random" element to help decorrelate the internal PSU related currents... When you start thinking along these lines there's many things you can do - we don't want to say much more as understandably there's a lot of "unprotected" IP involved..

John
 
DANOFDANGER said:
Does the polarity options actually change the analogue output or just internally on the digital domain ??.
Click to expand...

The Phase invert is performed in the digital domain, and yes the Analogue output changes phase as a result - but the SQ effect you are hearing are not due to the systems absolute phase, but "second order" Data induced processing effects...
 
DANOFDANGER said:
haha there really needs to be more information on this one, i still cant find anything solid enough to fully understand, isn't it mostly for USB ?.
Click to expand...

D3E works on all inputs, we are deliberately being a little coy about it to protect our IP.
 
Plutox said:
I thought D3E was about making better use of the available data space when incoming word length was < 24bit ??
Click to expand...

Yes - we now attempt to make "better use" of the "unused" portions of the Data path, we are trying to use this otherwise "free" and wasted Data BW to reduce second order effects...
 
sam_cat said:
No panic, its giving me a chance to refine my DIY PCM2702 solution.. PSU tweaks make a big difference!

Sam
Click to expand...

Just a quickie...

Are you using the a simple crystal with the PCM2702 or are you injecting an external clock?

Either way, try adding a 100R to 470R resistor in series with the Xin pin (as close a possible to the Pin), the PCM series of devices seem to have an issue with there internal oscillator design, it looks like its being overdriven and this cause HUGH low frequency jitter - adding the series resistor as close a possible to the Xin pin reduces the LF jitter levels very significantly...

It appears as if the internal Oscillator clamps or locks up maybe an internal AGC circuit (this would help explain the LF element)... hard to say without seeing a die level schematic...

I just want to complete a circuit section on the BDAC layout, then I can take a break and modify your units- sorry for the delay,

John
 
TimR said:
Got to agree, Winamp would easily be my preferred music library interface - among other things it's got the most elegant and powerful search engine I've come across. If only it could be made to control the SBT, that'd be a dream come true.
Click to expand...



Tim, what is the SBT ?
( Hope you don`t mind my curiosity.. :) )
 
JohnW said:
Are you using the a simple crystal with the PCM2702 or are you injecting an external clock?

Either way, try adding a 100R to 470R resistor in series with the Xin pin (as close a possible to the Pin), the PCM series of devices seem to have an issue with there internal oscillator design, it looks like its being overdriven and this cause HUGH low frequency jitter - adding the series resistor as close a possible to the Xin pin reduces the LF jitter levels very significantly...
Click to expand...

Thanks very much, will give it a try.
Current implementation is as per the datasheet with a simple crystal, I have a PFM Flea external clock board with a TentLabs XO almost ready to try.

JohnW said:
I just want to complete a circuit section on the BDAC layout, then I can take a break and modify your units- sorry for the delay,

John
Click to expand...


As I said John, no panic.

Sam
 
hluga said:
Me too. The main problem is that after tasting mdac my maudio appears to sound so bad....
Click to expand...

Kept meaning to ask, as I have your unit here - shall I "Sovereign" your unit?

I'll have to donate the components I have set reserved for my own unit but I'm sure your would do the same for me?... :p
 
DANOFDANGER said:
Yep there it is, WASAPI in output devices, looked their but didn't click through all the options xD. Thanks sam.
Click to expand...

I guess you both use Windows?
I didnt think XBMC 11 Eden was capable to output a bit perfect signal. As I understood it that was one of the reasons for creating the new sound engine (wich is not in the Eden version)
Either way, I think I am shit out of luck since I am running XBMC on MacOS.
I need to get hold of someone who is more skilled with MacOS!
 
JohnW said:
Kept meaning to ask, as I have your unit here - shall I "Sovereign" your unit?

I'll have to donate the components I have set reserved for my own unit but I'm sure your would do the same for me?... :p
Click to expand...

I have reserved a bottle of fine single malt which was reserved for my consumption, for when you 'Sovereign and MPAX' my unit :cool:.

See you in the spring ??
 
JohnW said:
Kept meaning to ask, as I have your unit here - shall I "Sovereign" your unit?

I'll have to donate the components I have set reserved for my own unit but I'm sure your would do the same for me?... :p
Click to expand...

It's difficult to resist...
but i dont want to borrow your components if you are not sure that you can find the equivalent parts during your next trip to asia.
Anyways if you come back to visit france you will be welcomed in Toulouse !
 
hluga said:
It's difficult to resist...
but i dont want to borrow your components if you are not sure that you can find the equivalent parts during your next trip to asia.
Anyways if you come back to visit france you will be welcomed in Toulouse !
Click to expand...

If you were me that would read "yes please John, the Malt is en route"

I'm just listening to Ry Cooder-Bop Till You Drop, on UKPD's Sovreigned unit and it sounds like the proverbial mutts nutts. :cool:

(A reference recording-to date)
 
Status
Not open for further replies.


advertisement


Back
Top