Latest DSD Decoder Project

What people are working on at the moment
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#1 Latest DSD Decoder Project

Post by Ray P »

A while back I started a thread on the Computer Audio subject area about 'no-dac decoding of DSD. At the time I was using a very simple low-pass filter to derive an analogue output from DSD. The sound quality was really good but there were some issued with pops and clicks.

http://www.audio-talk.co.uk/phpBB3/view ... =21&t=6168

Wind the clock forward a couple of years an things have improved and DSD is becoming established as an alternative to PCM in some quarters so I thought I would revisit DSD playback motivated by the sound quality I experienced previously. Amanero have now produced DSD specific firmware that addresses the pops/clicks and ther board can now process upto DSD512 data rates.

Signalyst has always been one of the main proponents of DSD and developed an open-hardware delta-sigma D/A convertor, the DSC1;

https://www.signalyst.com/hardware.html

This has been developed by others. I recently purchased a derivative of the DSC1 from a DIY-Audio member who lives in Russia. The DSC2 operates in differential mode and the output filtering is handled by transformers instead of op-amps. This morning I put the components together on a 'breadboard';

Image

Image

Feeding the Amanero DSD256, upsampled from CD quality FLAC rips, and I've just spent a couple of hours listening. The sound quality is outstanding, rich but very detailed and really engaging. It imediately caught Judy's ear and she came and listened too. None of the noise/pops/clicks of the previous experiments either.

If you're interested in a digital decoder project that is a little different the DSC2 is worth considering and not expensive.

I have some further tweaks planned, primarily to install an isolator/reclocker between the Amanero and the decoder board and maybe try a shunt power supply. Then I can think about boxing it up.

Could be a candidate for a future Owston.
Sorry, I couldn't resist!
User avatar
Nick
Site Admin
Posts: 15708
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#2 Re: Latest DSD Decoder Project

Post by Nick »

If I understand that DAC after a quick look, it seems to be summing the last 16 DSD "bits" so implementing a simple single pole low pass filter. I would expect it would still be dependant on the accuracy of the resistors used. Nice idea.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#3 Re: Latest DSD Decoder Project

Post by Ray P »

Nick, in case you didn't notice, there's a schematic for the orignal DSC1 available on a link on the open hardware page I included above. I also have the schematic for the DSC2 board.

BTW, another improvement beyond the differential mode of operation is muting which no longer uses a relay in favour of these devices;

http://www.ti.com/lit/ds/symlink/ts5a3167.pdf

Not sure if absolute accuracy of the resistors is that important (within reason of course), not like in an R2R ladder DAC. the DSC2 board doesn't have anything too esotoric and sounds incredible.

If anyone is interested the assembled board cost me about $200 delivered.
Sorry, I couldn't resist!
User avatar
Nick
Site Admin
Posts: 15708
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#4 Re: Latest DSD Decoder Project

Post by Nick »

Nick, in case you didn't notice, there's a schematic for the orignal DSC1 available on a link on the open hardware page I included above.
Yep, thats what I based my comments on. The DAC is (if I understanding it) taking the sum of a sliding 16 bit samples. My maths is not good enough to work out exactly what the response of that will be (and I don't have a copy of MATLAB), but if the resistors are not equal, it will in effect means what is a 16 tap FIR filter will have different coefficients for each tap. I suspect that resistor variation will alter the frequency response of the filter instead of the accuracy in the case of a R2R dac, so I can see how it would be less important.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#5 Re: Latest DSD Decoder Project

Post by Ray P »

The resistors used are 0.1% Sumiko RG.

There's more information about DSC2 here;

http://www.audiodesignguide.com/DSC1/index.html
Sorry, I couldn't resist!
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#6 Re: Latest DSD Decoder Project

Post by Ray P »

I've been progressing this project quietly and it has now moved on quite a bit for the better so it's time to put it into a nice chassis so that its fit for everyday use.

The DSD decoder board has been replaced with an updated version that uses Lundahl ll1684 transformers. In addition, the Amanero USB board has been replaced with a Beaglebone Black running as an audio appliance and the I2S output from the BBB is passed through a Twisted Pear Hermes/Cronus isolator/reclocker. Sound quality is really excellent.

Still some experimenting with the BBB software; I have a new version to install that has lots of the standard linux stuff stripped out so that is is getting more optimised for this specific application.

Today I've worked up drawings to drill the chassis metalwork.

Pictures in due course.
Sorry, I couldn't resist!
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#7 Re: Latest DSD Decoder Project

Post by Ray P »

I have too many projects active at the moment so decided to focus on getting this one done and dusted so I spend a few hours over the weekend assembling and testing the power supplies and installing the software onto the beaglebone. I now have the beaglebone running with power supplied by one of the new supplies and visible as a network audio appliance on my network.

Next up is to hook up the second power supply to the Twisted Pear Cronus reclocker and see if we can get the beaglebone working with automatic oscillator selection.

Pictures soon.
Sorry, I couldn't resist!
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#8 Re: Latest DSD Decoder Project

Post by Ray P »

I was off work today having finally succumbed to a stinking head cold but after a good sleep I felt a bit better this afternoon and made some more progress with my DSD decoder project - it's taken far too long to get this far but with a fair wind I should be getting noises from it this weekend. I just need to finish and test the last two power supplies and then then install the rest of the hook-up wires.

Image

Image

Image
Sorry, I couldn't resist!
User avatar
DSJR
User
Posts: 193
Joined: Mon Jan 11, 2010 11:11 am
Location: Suffolk Coastal

#9 Re: Latest DSD Decoder Project

Post by DSJR »

Starts off looking like a very few hundred quid and finishes off looking like thousands :D

What power supply boards are you using please and can possibly they be used for other things?
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#10 Re: Latest DSD Decoder Project

Post by Ray P »

Cheers. The power supplies are sourced via contacts over on DIY Audio,

The black power supply board is from Acko;

https://sites.google.com/site/ackodac/modules-and-kits

and have two of his voltage regulators. Mine are set to 5V and will power the actual DSD decoder sections.

The green power supply boards were picked up on DIY Audio's 'Swap Meet';

https://www.diyaudio.com/forums/swap-me ... -pcbs.html

They deliver 5V. One powers the Beaglebone Black and the other the Twisted Pear Cronus (seperate supplies because there is isolation between the BBB and the Cronus). The Cronus supply actually feeds two low noise 3.3V regulators.
Sorry, I couldn't resist!
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#11 Re: Latest DSD Decoder Project

Post by Ray P »

Finished assembling and testing the power supplies this evening. All tickety boo.

Now I can start testing the Beaglebone/Hermes/Cronus to check they're functioning correctly.
Sorry, I couldn't resist!
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#12 Re: Latest DSD Decoder Project

Post by Ray P »

I've left this project for too long after getting a bit disillusioned by problems when I tried testing it - just couldn't get it to play music.

Freshly motivated by the Owston affect, I decided it was time to crack it once and for all so I've spent a good few hours chasing myself in circles trying to get the BBB software working - that was a mistake as I was missing the bigger picture and not thinking rationally. Anyway, finally the penny dropped and I stripped it back and started testing the individual components and, lo and behold, I have found that the regulated power supply on the reclocker is only delivering 1.5V instead of 3.3V so no wonder I couldn't get the BBB software to lock onto a master clock input. Anyway, the nice chaps at Twisted Pear are sending me a replacement reclocker board so I'm hopeful that I'll get this working soon. Morale of the story - don't assume!
Sorry, I couldn't resist!
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#13 Re: Latest DSD Decoder Project

Post by Ray P »

I've just had a tracking update and the replacement Twisted Pear Cronus reclocker is on its way across the pond.

As its gratis, to replace the faulty board, it'll be interesting to see if HRMC try to extort any money out of me.
Sorry, I couldn't resist!
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#14 Re: Latest DSD Decoder Project

Post by Ray P »

The tracking data shows that the replacement Cronus reclocker arrived at Heathrow on Monday - it's still there! That's longer than it took to get from the US to the UK. Grrr...
Sorry, I couldn't resist!
User avatar
pre65
Amstrad Tower of Power
Posts: 21373
Joined: Wed Aug 22, 2007 11:13 pm
Location: North Essex/Suffolk border.

#15 Re: Latest DSD Decoder Project

Post by pre65 »

Ray P wrote: Fri Jun 21, 2019 3:11 pm The tracking data shows that the replacement Cronus reclocker arrived at Heathrow on Monday - it's still there! That's longer than it took to get from the US to the UK. Grrr...
That's because a multi national task force (US + UK + Parcelthief) are trying to work out how much to charge you for import duty and VAT on a FOC item. :lol:
The only thing necessary for the triumph of evil is for good men to do nothing.

Edmund Burke

G-Popz THE easy listening connoisseur. (Philip)
Post Reply