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
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 8th February 2018, 08:13 AM   #571
Signalyst is offline Signalyst  Finland
diyAudio Member
 
Join Date: Jul 2014
Quote:
Originally Posted by zenelectro View Post
I have a question for you Miska - would it not be better to use differential OP flip flops for a balanced design?
I'm wondering if the bulk of non linearities are due to inherent non RTZ nature of DSD stream. As such I would think the fastest dif OP logic would give best results. Something like 16 x Potato 74G74 dual FF should work well.

Otherwise is it possible to integrate RTZ coding?
Now the DSD stream can be considered as fancy PWM generated by the modulator. For RTZ-like operation, one option I was considering to try out for DSC2 are fast analog mux chips. They have "break before make" behavior built-in. I'm a bit curious how it would perform in practice.

Using such differential FF is probably good option too.

At DSD512 speed, RTZ coding would be feasible with CPLD and provision for it could be easily added to HQPlayer (to take into account the 2x rate difference). So CPLD could "reset" every second sample.

Maybe I should put a small CPLD on DSC2? That would allow some experimentation...
  Reply With Quote
Old 9th February 2018, 02:51 PM   #572
mmerrill99 is online now mmerrill99  United States
diyAudio Member
 
Join Date: Aug 2014
Would a logic AND gate do the RTZ job?
DSD stream on A input & DSD clock on B input
Output 1,0 for 1 input & 0,0 for 0 input
Then use double clock speed for DSD clock in 595 clocking.

Anybody tried this or am I missing something?
  Reply With Quote
Old 9th February 2018, 04:02 PM   #573
Signalyst is offline Signalyst  Finland
diyAudio Member
 
Join Date: Jul 2014
Quote:
Originally Posted by mmerrill99 View Post
Would a logic AND gate do the RTZ job?
DSD stream on A input & DSD clock on B input
Output 1,0 for 1 input & 0,0 for 0 input
Then use double clock speed for DSD clock in 595 clocking.

Anybody tried this or am I missing something?
I don't think so, because you'd need a third symbol that is not naturally occurring in the data stream. Now '0' is full scale negative voltage and '1' is full scale positive voltage...
  Reply With Quote
Old 9th February 2018, 04:24 PM   #574
mmerrill99 is online now mmerrill99  United States
diyAudio Member
 
Join Date: Aug 2014
Quote:
Originally Posted by Signalyst View Post
I don't think so, because you'd need a third symbol that is not naturally occurring in the data stream. Now '0' is full scale negative voltage and '1' is full scale positive voltage...
Sorry, I don't follow - can you expand on this, please?
Do you mean that in the DSD stream 0 is fully negative & 1 fully positive but with RTZ we just want a return to halfway between full neg & full pos i.e to zero?

Last edited by mmerrill99; 9th February 2018 at 04:44 PM.
  Reply With Quote
Old 9th February 2018, 05:20 PM   #575
mmerrill99 is online now mmerrill99  United States
diyAudio Member
 
Join Date: Aug 2014
But is it not that a series of '0's is full scale negative, not just singular '0's interspersed between existing '1's & '0's?
I can see how it might bias the output more negatively i.e result in an different output offset - is this the issue you are alluding to?

Is this why the CS4303 DAC datasheet states
"Return-to-zero coding is used where each occurrence of a '1' is 75% high & returns low for 25% of the bit period"?

Last edited by mmerrill99; 9th February 2018 at 05:24 PM.
  Reply With Quote
Old 10th February 2018, 03:31 AM   #576
zenelectro is offline zenelectro  Australia
diyAudio Member
 
zenelectro's Avatar
 
Join Date: Sep 2009
Quote:
Originally Posted by mmerrill99 View Post
But is it not that a series of '0's is full scale negative, not just singular '0's interspersed between existing '1's & '0's?
I can see how it might bias the output more negatively i.e result in an different output offset - is this the issue you are alluding to?

Is this why the CS4303 DAC datasheet states
"Return-to-zero coding is used where each occurrence of a '1' is 75% high & returns low for 25% of the bit period"?
Just a quickie, WRT 25% - exactly. Which implies a much faster clock rate if everything is to be reclocked, AFAICS. This in itself carries some liabilities.
I'm wondering if you can do it with same clock and a delay line. Have to think about this.

More later, have to go.

T
  Reply With Quote
Old 13th February 2018, 07:01 PM   #577
ppy is offline ppy  Russian Federation
diyAudio Member
 
ppy's Avatar
 
Join Date: Aug 2015
Location: Russia
Quote:
Originally Posted by Signalyst View Post
Using such differential FF is probably good option too.
I assembled the DSC layout on the differential FF 74VHC175.
Works badly. I again hear noise (murmur) as in the original non-differential DSC1 scheme. The overall noise level also deteriorated compared to v2.5 and v2.6.
Attached Images
File Type: jpg DSCF2919.jpg (968.8 KB, 396 views)
File Type: png v3.png (133.9 KB, 373 views)
Attached Files
File Type: pdf dsc3.pdf (115.7 KB, 99 views)

Last edited by ppy; 13th February 2018 at 07:14 PM.
  Reply With Quote
Old 13th February 2018, 09:36 PM   #578
Signalyst is offline Signalyst  Finland
diyAudio Member
 
Join Date: Jul 2014
Quote:
Originally Posted by mmerrill99 View Post
Do you mean that in the DSD stream 0 is fully negative & 1 fully positive but with RTZ we just want a return to halfway between full neg & full pos i.e to zero?
Yes, zero would be between the two extremes. Which is now produced by the offset that happens when value distribution is 50/50[1]. In my design there's the decoupling capacitor to get rid of the offset. In this way, sure, DC latch-ups to either rail don't work. Which can be also good... So you don't get full rail voltage on the output in case there's a software problem and the DAC gets all 0's (or 1's which is more unlikely to happen though)...

But one could instead also apply static offset to have negative current. In fact I should have tried I/V setup that is sinking to virtual +2.5V reference...

With pulse trains the current setup works well though, but closer to DC it can get tricky.


[1] If you put HQPlayer to play silence file you get fancy version of such. If you just pause the playback, you get less fancy static pattern.
  Reply With Quote
Old 13th February 2018, 11:39 PM   #579
Lo_Tse is offline Lo_Tse  Canada
diyAudio Member
 
Join Date: Dec 2006
Just curious, does this DSC1 DAC only work with HQPlayer?

Thanks!

Regards,
  Reply With Quote
Old 14th February 2018, 12:21 AM   #580
mmerrill99 is online now mmerrill99  United States
diyAudio Member
 
Join Date: Aug 2014
Quote:
Originally Posted by ppy View Post
I assembled the DSC layout on the differential FF 74VHC175.
Works badly. I again hear noise (murmur) as in the original non-differential DSC1 scheme. The overall noise level also deteriorated compared to v2.5 and v2.6.
That seems odd? Any thoughts on what could be going on?
  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 12:22 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.79%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki