Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

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

Signalyst DSC1
Signalyst DSC1
Signalyst DSC1 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 30th December 2019, 10:00 PM   #1711
gtose is offline gtose  United States
diyAudio Member
 
Join Date: Feb 2009
Location: Illinois
Signalyst DSC1
Default My DSC2 measurements

I took some time to make measurements of my DSC2 build (shown at Signalyst DSC1) before starting case construction. My results do not seem to be in-line with the measurements others are seeing. I'm wondering where I may have gone wrong and if there are any suggestions to get my THD measurements up to par with what other are seeing (Signalyst DSC1, etc).

While the measurements others have posted are around 0.007% THD, I am consistently seeing 0.013% in one channel, 0.017% on the other. I have attached my RTA spectrums.

Tests were run using a 1khz sine 384k/24bit wav file generated with SoX, -3db attenuation. This is upsampled to DSD512 or 256 in Roon (no meaningful differences in THD between the rates). Balanced output sends about 1.22V output to my Linear Audio Autoranger (which leaves attenuation to 0db) and then to my ESI Juli@'s balanced inputs.

Any thoughts on the cause of this? I have tried a few things to rule out some thoughts.

- Changed from a switch mode PSU with a cap-mx + regulator to a linear PSU + cap-mx + regulator. No changes in THD.
- Tried both SE and Balanced outputs, no change
- Tried HQPlayer instead of Roon for upsampling, no change
- Added 10k across the secondary of the output transformer (Lundahl LL1684), no change

Some suspicions I have:
- I ordered some 74AHCT595Ds from Taobao in an attempted to get old stock. It seems like they were based on NXP marking, but I wonder now if they are legit parts.
- It was stated by Vit123 that the Lundahls add significant distortion, but even for these transformers, is it expected to be this high?
- Any grounding issues? Right now its pretty isolated from anything, but I suspect this wouldn't be that big of an issue with balanced output. I also would assume any grounding issue wouldn't be spread across the harmonics as I'm seeing.
- Soldering or construction errors? Anywhere to start to check this? I read the previous posts about DC across the transformer primary. I measured and there is no substantial or constant DC voltage while playing music.
- Should I add additional snubbing or load to my output with the LL1684s?
- Perhaps my test setup is flawed.

Or... maybe these measurements are as expected with my build?

Any suggestions on where I can go next to try to lower the THD? Any other tests I should be running? Any other information that would be helpful?

Thanks all for your thoughts and expertise!
Greg
Attached Images
File Type: png dsc-bal-l.png (100.5 KB, 488 views)
File Type: png dsc-bal-r.png (99.1 KB, 486 views)
  Reply With Quote
Old 30th December 2019, 10:16 PM   #1712
randytsuch is offline randytsuch
diyAudio Member
 
Join Date: Jan 2003
Location: LA
Signalyst DSC1
Maybe measure another DAC, and see how it compares? Especially if you have (or can borrow) a commercial DAC with a data sheet number for THD.

Was wondering if the ladder resistors (and their tolerance, or maybe variance from each other) could impact THD?

There was a reason that Soekris used resistors with 0.02% or thereabouts in his ladder dac.

Randy
__________________
My system is here
randytsuch system
  Reply With Quote
Old 30th December 2019, 10:24 PM   #1713
gtose is offline gtose  United States
diyAudio Member
 
Join Date: Feb 2009
Location: Illinois
Signalyst DSC1
Quote:
Maybe measure another DAC, and see how it compares? Especially if you have (or can borrow) a commercial DAC with a data sheet number for THD.
I measured a small Roon / raspberry pi-based streamer using an ES9018Q2M with the same test file and setup. It measured 0.0061% THD as shown in the attached file. Therefore, I know at least my test setup can measure lower than what I'm seeing with the DSC2.

Greg
Attached Images
File Type: png es9018q2m-1.12v.png (103.6 KB, 477 views)
  Reply With Quote
Old 30th December 2019, 11:36 PM   #1714
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
Quote:
Originally Posted by randytsuch View Post
Was wondering if the ladder resistors (and their tolerance, or maybe variance from each other) could impact THD?

There was a reason that Soekris used resistors with 0.02% or thereabouts in his ladder dac.

Randy
As far as I know, the Soekris designs are multibit DACs while the Signalyst designs are single-bit FIRDACs. In a multibit design, resistor matching affects distortion a lot, in single-bit FIRDACs, it only affects stopband rejection (suppression of far-ultrasonic rubbish).
  Reply With Quote
Old 31st December 2019, 08:00 AM   #1715
gentlevoice is offline gentlevoice  Denmark
diyAudio Member
 
Join Date: Oct 2007
Location: Denmark
Signalyst DSC1
@gtose ... Hi ... Just a couple of brief comments on your measurements:

- A suggestion could be to disconnect the Lundahl and measure the DSC directly without transformers.
- Also, I notice that you describe a measurement setup that includes the Linear Audio Autoranger. Personally I would try measuring without this part in the signal chain.
- Given there's virtually no noise below 1 kHz I would not assume that you have ground loop issues ...
- Might the input impedance of any of the devices you connect after the DSC have low impedances so as to load the DSC?

Well, these were my first thoughts ... hope it may help ...

Cheers,

Jesper
__________________
"... It is always possible to be friendly ..." HH the Dalai Lama.
  Reply With Quote
Old 23rd January 2020, 09:34 AM   #1716
zaum is offline zaum
diyAudio Member
 
Join Date: Feb 2012
Where can i get the old 74AHCT595-s? Anybody has it for sale?
  Reply With Quote
Old 23rd January 2020, 09:42 AM   #1717
christian.lenci is offline christian.lenci  Italy
diyAudio Member
 
Join Date: Nov 2008
I may have some, I'll let you know in a few hours with a PM.

Chris
  Reply With Quote
Old 23rd January 2020, 10:57 PM   #1718
gionag is online now gionag  Italy
diyAudio Member
 
Join Date: Jun 2007
Quote:
Originally Posted by christian.lenci View Post
I may have some, I'll let you know in a few hours with a PM.

Chris
if you got philips... i am interested too...

anyway, whoever needs NXP ... i have them
  Reply With Quote
Old 1st February 2020, 11:19 PM   #1719
potstip is offline potstip  Turkey
diyAudio Member
 
Join Date: Oct 2013
Hello;

Regarding to version DSC2 with reclocking what i do not understand is dsd is always times of 44.1Khz
DSD 64 : DSD 2.8 MHz = 2 822 400 Hz = 44100 Hz x 64 times;
DSD 128 : DSD 5.6 MHz = 5 644 800 Hz = 44100 Hz x 128 times;
DSD 256 : DSD 11.2 MHz = 11 289 600 Hz = 44100 Hz x 256 times;
DSD 512 : DSD 22.6 MHz = 22 579 200 Hz = 44100 Hz x 512 times;
DSD 1024 : DSD 45.2 MHz = 45 158 400 Hz = 44100 Hz x 1024 times;

So why recloking circuit includes 49.152Mhz?
Attached Images
File Type: png Ek Ašıklama 2020-02-02 022018.png (63.2 KB, 169 views)
  Reply With Quote
Old 2nd February 2020, 12:04 AM   #1720
laserscrape is offline laserscrape
diyAudio Member
 
Join Date: Dec 2017
48k multiple PCM rates converted to DSD?
  Reply With Quote

Reply


Signalyst DSC1Hide 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

Forum Jump


New To Site? Need Help?

All times are GMT. The time now is 12:13 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Copyright ©1999-2020 diyAudio
Wiki