Go Back   Home > Forums > Source & Line > Digital Line Level

Digital Line Level DACs, Digital Crossovers, Equalizers, etc.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 21st November 2012, 08:30 AM   #4621
SSerg is offline SSerg  Russian Federation
diyAudio Member
 
Join Date: Mar 2012
Location: St.Petersburg
Hi, -ecdesigns-
Quote:
Originally Posted by -ecdesigns- View Post
I am now using 22KHz (post # 4615).
If DEM generator frequency constitutes 22 kHz, per time of the sample shaping will not happen the interleaving a current passive divisors of the current that is to say will not happen their averaging. Hereunder principle DEM is emasculated and it not works.

Are younot alarmed?


For correct working DEM generator frequency must be multiple 4Fs or close towards this value. In this case at sample shaping will take part all 4 currents of the passive current divider. Frequency of the standard anisochronous generator constitutes approximately 200 kHz that near towards 4Fs=176 kHz.



Last edited by SSerg; 21st November 2012 at 08:39 AM.
  Reply With Quote
Old 21st November 2012, 02:35 PM   #4622
oshifis is offline oshifis  Hungary
diyAudio Member
 
Join Date: Mar 2004
Location: Budapest, Hungary
Those are the active dividers. You would need exact 4x (or 8x, 16x, ...) DEM frequency if you did not use any external filtering capacitor, so that always 4 (8, ...) steps fall in a sample. Then perfect averaging would happen. But then the ripple would also appear on the output analog signal. In any other case, even at asynchronous DEM, the external capacitor filters the ripple. As long as this filtering is effective, it does not matter how is the DEM frequency related to the sampling frequency.
Not speaking here about on-chip interaction between the two signals, that have practical manufacturing and not theoretical reason, thoroughly explained by John in this thread.
With respect, Laszlo
  Reply With Quote
Old 21st November 2012, 11:39 PM   #4623
diyAudio Member
 
Join Date: Feb 2005
Location: Melbourne, Aust
Hi Laszlo, & John,

About this filtering, with the DEM freq down to about 44 or 22kHz ....

In any other case, even at asynchronous DEM, the external capacitor filters the ripple. As long as this filtering is effective, it does not matter how is the DEM frequency related to the sampling frequency.

If there was an extra R-C network added to each external cap (another 2 SMD units) would the increased filtering be of any benefit?

If so, what would the initial trial/estimated values?
__________________
... jh
  Reply With Quote
Old 22nd November 2012, 12:00 AM   #4624
diyAudio Member
 
Join Date: Oct 2009
Location: Brunei
Quote:
Originally Posted by oshifis View Post
Those are the active dividers. You would need exact 4x (or 8x, 16x, ...) DEM frequency if you did not use any external filtering capacitor, so that always 4 (8, ...) steps fall in a sample. Then perfect averaging would happen. But then the ripple would also appear on the output analog signal. In any other case, even at asynchronous DEM, the external capacitor filters the ripple. As long as this filtering is effective, it does not matter how is the DEM frequency related to the sampling frequency.
Not speaking here about on-chip interaction between the two signals, that have practical manufacturing and not theoretical reason, thoroughly explained by John in this thread.
With respect, Laszlo
So much is clear, however, wouldn't you need fDEM of at least 44khz to make sure the averaging is over 4 Dem steps during one word?

In my understanding, fDEM of 22khz gives averaging over 2 steps (inferior averaging compared to a 4 step cycle). As you say, it should not matter if fDEM is not an EXACT multiple (4,8,16) of fWS, as long as it is at least 44khz (so you get averaging over a full 4-step DEM cycle). For example, a 60khz fDEM would cycle through 5.4 steps per word (60/44 = 1.36 DEM cycle, consisting of 4 steps).

Am I Wrong?
__________________
--------
TDA1541a-S2 / 6072a / MediaplayerDac -- Michell Gyrodec -- Aikido Preamp -- Pass F5 -- Audio Physic Tempo 2SE -- Yamamura Cabling
  Reply With Quote
Old 22nd November 2012, 08:43 AM   #4625
galeb is offline galeb  Spain
diyAudio Member
 
Join Date: Apr 2008
Hi, after so many time.

Why discusing about the DEM?. It's very easy to try it. Just test some different values of capacitor on the DEM.

With a great capacitor, let's say 1 to 2uf, it sounds very fine, more relaxed, but at very low levels, it make strange sound and noise. On the other side, with 10pf for example, it sounds very fine, maybe a little bit less relaxed, but with an incredible low noise at extreme low levels. You can see on the scope the waveforms at pins 16&17 in both cases. Differences are notable.

I imagine that John is trying some thing more, appart of changing the capacitor value, but at the moment, until he doesn't end his tests, he will say nothing.

Kind regards to all.
__________________
galeb saleh
Sarte Audio Elite´s technical audio department
  Reply With Quote
Old 22nd November 2012, 11:37 AM   #4626
SSerg is offline SSerg  Russian Federation
diyAudio Member
 
Join Date: Mar 2012
Location: St.Petersburg
Quote:
Originally Posted by oshifis View Post
Those are the active dividers.
You have not understood me. I implied passive dividers, which have included in active dividers, in composition of DEM cells.

Quote:
Originally Posted by oshifis View Post
As long as this filtering is effective, it does not matter how is the DEM frequency related to the sampling frequency.
This untrue.
Quote:
Originally Posted by oshifis View Post
Not speaking here about on-chip interaction between the two signals, that have practical manufacturing and not theoretical reason, thoroughly explained by John in this thread.
John is not authority in the area of TDA1541 working. Many his glances are wrong.
Get acquainted with articles of family TDA1540-TDA1541 developers, and a great deal in TDA1541 working is made clear.
  Reply With Quote
Old 22nd November 2012, 12:08 PM   #4627
SSerg is offline SSerg  Russian Federation
diyAudio Member
 
Join Date: Mar 2012
Location: St.Petersburg
Here is reference to monograph of the developer.
http://repository.tudelft.nl/assets/uuid:283c827b-7257-461a-9fce-b0c2e3b6f99f/TR-diss-1781.pdf
  Reply With Quote
Old 22nd November 2012, 12:58 PM   #4628
diyAudio Member
 
Join Date: Feb 2005
Location: Melbourne, Aust
SSErg,
Perhaps you might like to enlighten us with your superior design - you are extremely critical of John's efforts and detailed explanations of his development progress that has made clear many aspects of this particular dac chip's behaviour to us "less technically skilled" people.
__________________
... jh
  Reply With Quote
Old 22nd November 2012, 01:56 PM   #4629
oshifis is offline oshifis  Hungary
diyAudio Member
 
Join Date: Mar 2004
Location: Budapest, Hungary
Quote:
Originally Posted by studiostevus View Post
So much is clear, however, wouldn't you need fDEM of at least 44khz to make sure the averaging is over 4 Dem steps during one word?

In my understanding, fDEM of 22khz gives averaging over 2 steps (inferior averaging compared to a 4 step cycle). As you say, it should not matter if fDEM is not an EXACT multiple (4,8,16) of fWS, as long as it is at least 44khz (so you get averaging over a full 4-step DEM cycle). For example, a 60khz fDEM would cycle through 5.4 steps per word (60/44 = 1.36 DEM cycle, consisting of 4 steps).

Am I Wrong?
I think that averaging a given number of DEM samples over an fWS is not the correct view. Rather, we are filtering the ripple (AC content) from the DC of the given bit current.
  Reply With Quote
Old 23rd November 2012, 05:49 AM   #4630
diyAudio Member
 
Join Date: Oct 2009
Location: Brunei
Refer to http://ftp.slivova.info/andro/DAC0/T...ntMatching.pdf

Page 2, "At the output terminals currents will then flow whereby the average values are exactly equal and have a dc value equal to I."
__________________
--------
TDA1541a-S2 / 6072a / MediaplayerDac -- Michell Gyrodec -- Aikido Preamp -- Pass F5 -- Audio Physic Tempo 2SE -- Yamamura Cabling
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 01:27 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2