Building the ultimate NOS DAC using TDA1541A - Page 431 - diyAudio
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 7th April 2012, 03:34 AM   #4301
regal is offline regal  United States
diyAudio Member
 
Join Date: Jan 2004
Location: MD
Everyone has there favorites to work with which is great as long as they aren't PWM's . For exampe The PCM1704 sounds and measures good with 60-100 ohm passive I/V and post amplification, as long as you slow down the servo crap with over sized capacitors on the pins.

Nothing wrong with the PCM56 or the TDA1541 either. Anything is better than the sigma-delta RFI generators.

My only issue was when we weren't working on a D/AC in this thread, rather a SD player permanately attached to a TDA1541 so it wasn't really functioning as a DAC since it was tied to sharing the clock with the specialized transport. The latest iteration that allows a smart method to download files to the memory card is a big step toward the title or scope of the thread.

I would like to hear Benard's DAC, but buying 32 PCM56 chips seems a bit impractical. So my interest is TDA1541, AD1862, PCM63, and PCM1704. All good chips and better than the modulators.

Last edited by regal; 7th April 2012 at 03:48 AM.
  Reply With Quote
Old 19th April 2012, 08:18 PM   #4302
diyAudio Member
 
Join Date: Oct 2009
Location: Brunei
John (others?),
Is the maximum current through the MODPWR2 common mode regs limited only by the maximum collector current of the BD679 (4A) ? I would like to use one to power a squeezebox (roughly 1A draw)....

Thanks
  Reply With Quote
Old 23rd April 2012, 07:08 AM   #4303
diyAudio Member
 
-ecdesigns-'s Avatar
 
Join Date: May 2006
Hi studiostevus,

The maximum regulator output current (capacitance multiplier) also depends on BD679 base current / Hfe.

When higher output current is desired you need to increase base current (decrease resistor value between base and collector).

This in turn will increase ripple voltage on the capacitor (capacitance multiplier), so a larger capacitor value is needed.

I use a 2A capacitance multiplier for the modified Circlotron output stage that runs on a single power supply.

The latest power supplies consist of rectifier, capacitance multiplier, balanced passive RC filter and shunt regulator. This configuration offers very clean DC voltage even with highly polluted mains voltage (attached schematic).


There are some new developments with the TDA1541A-MK10. When using high resolution amplifier and speakers there were some slight problems with upper midrange and trebles that were traced down to the current buffer circuit.

In order to find a solution I tested most basic I/V conversion, resistor between DAC output and ground.The 2mA bias current (TDA1541A) creates an unwanted DC voltage on the DAC output. With say 39 Ohms passive I/V resistor we get 0.002 (bias current) * 39 = -78mV DC while the ac voltage equals 0.004 (full-scale) * 39 = 156mVpp. This means that the signal varies between -156mVpp and 0V instead of +78 and -78mV as desired.

So I added a compensation current by connecting approx. 2K5 resistor between +5V and DAC output. Now the signal swings between +78 and -78mV and DC at the DAC output is around zero. The compensation current now varies between (5-0.078) / 2500 = +1.96mA and (5+0.078) / 2500 = +2.03mA.

I was able to generate up to 500mVpp with low distortion using 2K5 and 120 Ohms. The issues with upper midrange and trebles were now gone.

In order to amplify the signal I was thinking of using an Aikido JFET or tube amplifier. Some tests with a 2SK389-based Aikido JFET amplifier gave good results.

There is also an interesting Aikido LV kit that runs on a single 24 or 48 volt DC supply (heaters included):

Aikido LV.
Attached Images
File Type: jpg MODSHUNT.jpg (197.2 KB, 1352 views)

Last edited by -ecdesigns-; 23rd April 2012 at 07:12 AM.
  Reply With Quote
Old 23rd April 2012, 10:54 AM   #4304
diyAudio Member
 
Join Date: Oct 2009
Location: Brunei
Thanks John

Why did you opt for a resistor to nullify dc offset at the dac output, over for example a (jfet based?) servo circuit (possibly fed from the dac +5v due to tda1541a internal circuit) ?

I am using the LV aikido as a preamp, sounds very good indeed!
__________________
--------
TDA1541a-S2 / 6072a / MediaplayerDac -- Michell Gyrodec -- Aikido Preamp -- Pass F5 -- Audio Physic Tempo 2SE -- Yamamura Cabling
  Reply With Quote
Old 23rd April 2012, 01:29 PM   #4305
oshifis is online now oshifis  Hungary
diyAudio Member
 
Join Date: Mar 2004
Location: Budapest, Hungary
Quote:
Originally Posted by -ecdesigns- View Post
In order to find a solution I tested most basic I/V conversion, resistor between DAC output and ground.The 2mA bias current (TDA1541A) creates an unwanted DC voltage on the DAC output. With say 39 Ohms passive I/V resistor we get 0.002 (bias current) * 39 = -78mV DC while the ac voltage equals 0.004 (full-scale) * 39 = 156mVpp. This means that the signal varies between -156mVpp and 0V instead of +78 and -78mV as desired.

So I added a compensation current by connecting approx. 2K5 resistor between +5V and DAC output. Now the signal swings between +78 and -78mV and DC at the DAC output is around zero. The compensation current now varies between (5-0.078) / 2500 = +1.96mA and (5+0.078) / 2500 = +2.03mA.

I was able to generate up to 500mVpp with low distortion using 2K5 and 120 Ohms.
John, if you use 120 ohms passive I/V resistor, you have +240mV and -240mV. This is OK for the audio output level, but you far exceed the +/-25mV voltage compliance of the TDA1541A. As I remember, there are some internal diodes that get forward biased and cause distortion above that level (although it is most likely not a hard limit). Have you taken any special measures to avoid this distortion?
  Reply With Quote
Old 23rd April 2012, 04:26 PM   #4306
diyAudio Member
 
Bernhard's Avatar
 
Join Date: Apr 2002
Location: Munich
Dear ec

Quote:
There are some new developments with the TDA1541A-MK10. When using high resolution amplifier and speakers there were some slight problems with upper midrange and trebles that were traced down to the current buffer.

So I added a compensation current by connecting approx. 2K5 resistor between +5V and DAC output. Now the signal swings between +78 and -78mV and DC at the DAC output is around zero. The compensation current now varies between (5-0.078) / 2500 = +1.96mA and (5+0.078) / 2500 = +2.03mA.

I was able to generate up to 500mVpp with low distortion using 2K5 and 120 Ohms. The issues with upper midrange and trebles were now gone.

In order to amplify the signal I was thinking of using an Aikido JFET or tube amplifier. Some tests with a 2SK389-based Aikido JFET amplifier gave good results.
Still I have to answer your last post...

To this one:

You could not hear the problems before ?

Don't you think that a 2,5k resistor will inject noise into the circuit ?

Did you measure the distortion with decent equipment or did you look at the waveform with a scope ? Remember the 25mV voltage compliance.

Now post amplification ?

Last edited by Bernhard; 23rd April 2012 at 04:31 PM.
  Reply With Quote
Old 23rd April 2012, 05:02 PM   #4307
diyAudio Member
 
Bernhard's Avatar
 
Join Date: Apr 2002
Location: Munich
ThorstenL

Quote:
A number of parameters, including 0dBfs and -60dBfs FFT.
Would it be big trouble for you to post -60dB FFT ?
  Reply With Quote
Old 24th April 2012, 10:37 AM   #4308
diyAudio Member
 
-ecdesigns-'s Avatar
 
Join Date: May 2006
Hi Bernhard,

Quote:
You could not hear the problems before ?
No, the performance of my audio set gradually improved and these marginal distortions became more clearly audible after these improvements.

The distortion is mainly audible on specific vocals on some audiophile recordings like from Stockfish.

Quote:
Don't you think that a 2,5k resistor will inject noise into the circuit ?
Yes it will, but it also depends on power supply noise levels. The power supplies in the latest MK10 (schematics in my previous post) are very clean, and I applied additional RC filtering between 5V supply and bias resistor.

Of course I also tested (JFET) CCS. This resulted in slight grain and more agressive upper midrange and trebles. So it basically re-introduced some of the problems I experienced with the current buffer circuit.

Quote:
Did you measure the distortion with decent equipment or did you look at the waveform with a scope ? Remember the 25mV voltage compliance.
Yes I did, Marconi TF2331A distortion meter indicates approx. 0.02% THD + noise (DAC + Aikido line amplifier). The TF2331 also has an output (extracted THD + noise). This output signal can be used to analyse the distortion + noise.

The first oscillogram shows the 1 KHz output signal (3Vpp) at the Aikido line amplifier output. Voltage across the 120R passive I/V resistor equals 480mVpp and positive bias current equals 2mA.

The second picture shows THD measurement setup. The TF2331A indicates approx. 0.02% THD & noise. The scope shows the distortion meter output signal. The distortion mainly consists of quantization noise as can be seen on the oscillogram.

When using oversampling / reconstruction filtering the quantization noise can be significantly reduced and THD + noise can be lowered to approx. 0.002%.

The DACs presented at the beginning of this thread did just that, significantly reducing quantization noise and thus reducing THD for bass-midrange. Since the midrange is very important for audio this wasn't such a bad idea after all.

It is obvious that linear interpolation causes increased THD at higher frequencies as you already pointed out. So why not use the best of both worlds, linear interpolation (4 ... 8 DAC chips in parallel) for bass / midrange and a second NOS DAC just for upper midrange and trebles?

When using a 2-way speaker each DAC could drive a dedicated speaker chassis, filtered interpolation DAC feeds bass / midrange chassis and filtered / corrected NOS DAC drives the tweeter chassis. Trebles boost can be used for the tweeter only in order to compensate for trebles roll-off without affecting bass-midrange. The delay circuits (shift registers) can be used to time-align both OS and NOS DAC output signals.
Attached Images
File Type: jpg piv120r.jpg (102.6 KB, 1177 views)
File Type: jpg thd.jpg (107.7 KB, 1070 views)
File Type: jpg thd1.jpg (74.9 KB, 999 views)

Last edited by -ecdesigns-; 24th April 2012 at 10:44 AM.
  Reply With Quote
Old 24th April 2012, 02:03 PM   #4309
diyAudio Member
 
Join Date: Mar 2008
Quote:
Originally Posted by -ecdesigns- View Post
I use a 2A capacitance multiplier for the modified Circlotron output stage that runs on a single power supply.
Can you post some more info and maybe a schema of this Circlotron?

This would be much appreciated
  Reply With Quote
Old 26th April 2012, 10:24 PM   #4310
diyAudio Member
 
Join Date: Apr 2007
Quote:
Originally Posted by -ecdesigns- View Post

When using oversampling / reconstruction filtering the quantization noise can be significantly reduced and THD + noise can be lowered to approx. 0.002%.

The DACs presented at the beginning of this thread did just that, significantly reducing quantization noise and thus reducing THD for bass-midrange. Since the midrange is very important for audio this wasn't such a bad idea after all.

It is obvious that linear interpolation causes increased THD at higher frequencies as you already pointed out. So why not use the best of both worlds, linear interpolation (4 ... 8 DAC chips in parallel) for bass / midrange and a second NOS DAC just for upper midrange and trebles?
Hello John

Most of the signals for the imaging/soundstage and engagement in a music recording are roughly between 2 Khz and 6 Khz

Using linear interpolation could be good in that frequencies range, but at
6 Khz linear interpolation start to have higher THD

Up to which frequencies you whould use linear interpolation ?

Thank

Bye

Gaetan

Last edited by gaetan8888; 26th April 2012 at 10:31 PM.
  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 11:41 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