New DAC with ASRC and USB - diyAudio
Go Back   Home > Forums > Source & Line > Digital Source

Digital Source Digital Players and Recorders: CD , SACD , Tape, Memory Card, 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 March 2012, 07:01 PM   #1
Arius is offline Arius  United States
diyAudio Member
 
Join Date: Mar 2004
Location: California
Smile New DAC with ASRC and USB

I was looking for a well designed compact DAC and surprisingly, I just couldn't find anything decently priced. It must have USB and hi-rez capability.

Ebay - Junk designs with TE or AK chips, no ASRC.
HRT Music Streamer - Good USB, basic DAC, hamstrung by power and layout.
MF M1DACA - Exactly what I wanted, poor design especially power delivery, basic parts quality.
Diyaudio - Nothing close to what I was looking for.
Aussie SmartDAC - Close but USB audio support dubious.

So I went and build my own. My main inspiration came from the Benchmark DAC1. It's based on TAS1020B for USB with a AD1896 96kHz ASRC. Next is the Bel Canto DAC3.5VB. Stereophile rated that as a class above the Benchmark DAC1. The DAC3.5VB has a SRC4392 ASRC running at 192kHz feeding PCM1792 DAC with OPA1632 diff opamps for the output. Alas, it has no USB. Both products are also a bit pricey while just falling short of my requirements.

My goals are:
1) Compact (for possible inclusion in a future DIY integrated amp - that seems to be a trend with many amps featuring crappy PCM27XX 16-bit/48KHz DAC).
2) High rez USB, at least 24-bit/96kHz but with no proprietary drivers for Windows.
3) ASRC (AD1896 was initial choice but SRC4392 seemed easy enough to use plus it allowed for more features).
4) PCM1794A DAC. I didn't like the Wolfson stuff. ESS is not diy-friendly.
5) Well designed power delivery with individual high performance regulators for each supply, no shared supplies, latest TPS7A4901/3001 15uV regulators for the analog section. Ample and properly laid-out decoupling.
6) No signal coupling caps.
7) No exotic but good quality parts throughout. 10ppm crystal oscillator for digital. Metal MELF resistors for the analog section with film PPS caps. LME49720 for I/V and OPA627 for line driver.
8) Hand solderable, so nothing smaller than 0603 and no BGA's.

The first prototype has been completed and initial tests complete. It is working and I'm most pleased with the results.

Features:
1) 24-bit/96kHz USB Class 1 audio with asynchronous interface using TAS1020B.
2) 24-bit/192kHz digital coax input via transformer isolation.
3) USB is fully isolated via ISO7240C.
4) All inputs upsampled to 192kHz via SRC4392 ASRC.
5) PCM1794A DAC with LME49270 I/V and OPA627 for single-ended outputs.
6) Ten (yes 10) low noise LDO regulators.
7) PSOC for auto-input selection, LED sample rate display and ASRC configuration. PSOC idles in sleep state with internal clock and CPU core shutdown.
8) PCB layout that meets my own rules and goals, which are inline with industry high-speed digital and analog expectations.

Some pictures below. I guess it shows that I'm not so good with a camera.

Please note that this is a completed design. So you can make your own if you're looking for stuff like Crystek clocks, Burson opamps, Bybee filters, 22000uF filtering, FET I/V, etc. I have realistic goals to meet (and a family to feed).


PS: PCB size is 142mm X 50mm
Attached Images
File Type: jpg topview_loaded2_med.jpg (855.4 KB, 740 views)
File Type: jpg blank_pcb_med.jpg (815.9 KB, 693 views)

Last edited by Arius; 7th March 2012 at 07:05 PM. Reason: Added board size
  Reply With Quote
Old 8th March 2012, 10:53 AM   #2
diyAudio Member
 
Join Date: Jan 2008
Location: Virginia
Some bad choices of parts.
AD1896 performs a way better job that SRC4393. Luckly they are pin-compatible.
Why use OPA627 as line driver and LM49720 as I/V??? 627 is better I/V than 49720... if you already spend the money on it.
  Reply With Quote
Old 8th March 2012, 11:51 AM   #3
diyAudio Member
 
Join Date: Apr 2003
Location: Eastern Pennsylvania
Interesting PCB you have there. Did you hand solder all of those SMD parts yourself, or did you find an PCB assembly vendor willing to populate a small one board project? If it was an assembly house, may I inquire as to who it was and how much they charged?

Thanks.
__________________
Ken
  Reply With Quote
Old 8th March 2012, 03:37 PM   #4
Arius is offline Arius  United States
diyAudio Member
 
Join Date: Mar 2004
Location: California
Thank you for your feedback.

Hi Sonic,

I am aware of the better jitter filtering spec of the AD1896 ASRC compared with the SRC4192 which is pin compatible. However, the part I used is the SRC4392 which has better specs and is NOT pin compatible. My initial design actually used the AD1896 (easy to solder SSOP28 package and no firmware required). I settled on the SRC4392 because it offers more features (namely the 4-port input digital receiver) and it can do 192kHz. I know Benchmark amongst others say 192K upsampling is no good but hey, Stereophile rated the Bel Canto DAC3.5VB which uses the SRC4392 running at 192kHz as a class above the DAC1. Can't really call that a bad choice now, eh? Even the best-of-the-best M1 Bricasti uses AD1955 with opamps for I/V.

If you run the calcs, you'll find the LME49720 to have more than sufficient specs for I/V duty here (192kHz X4 for margin, so calc for 1MHz). Now I know this is subjective, some will say you shouldn't even run opamps at all. You'll find the LME49720 in the DAC1 PRE. So that's pretty good enough for me. Besides the Musical Fidelity uses el-cheapo JRC5532 and Stereophile both raved about it and it measured well. In my case, I am limited solely by the need for a dual op-amp for I/V. I just couldn't find anything better than runs off +/-12V supplies. I stuck the OPA627 as the final diff-SE driver because I wanted very low output offset. Here, I can take any opamp and I have the OPA627 and LME49990 in stock from previous projects.


Hi Ken,

Unfortunately, I had to hand solder my board. Hence, one of the design goals being no BGA or leadless QFP parts. Even if I sent my board to one of the many proto shops in the Bay Area, it'll still cost at least $1K and I can't afford that (might as well buy a finished product). The 2 TQFP parts (TAS1020B, SRC4392) were a pain to solder. The 2 fancy +/-12V regulators are in a special package that has a thermal pad. My PCB has a hole so you can get to the thermal pad from the bottom.

Last edited by Arius; 8th March 2012 at 03:40 PM. Reason: typo
  Reply With Quote
Old 8th March 2012, 07:03 PM   #5
diyAudio Member
 
Join Date: Apr 2003
Location: Eastern Pennsylvania
Quote:
Originally Posted by Arius View Post
Hi Ken,

Unfortunately, I had to hand solder my board. Hence, one of the design goals being no BGA or leadless QFP parts. Even if I sent my board to one of the many proto shops in the Bay Area, it'll still cost at least $1K and I can't afford that (might as well buy a finished product). The 2 TQFP parts (TAS1020B, SRC4392) were a pain to solder. The 2 fancy +/-12V regulators are in a special package that has a thermal pad. My PCB has a hole so you can get to the thermal pad from the bottom.
Yeah, fine pitch SMD, BGA and other leadless packages are the bane of DIY electronics. I really miss the days when any IC was available in a DIP package. One would think a proto-shop could robotically populate a small PCB for much less than $1k.
__________________
Ken
  Reply With Quote
Old 9th March 2012, 10:43 AM   #6
diyAudio Member
 
Join Date: Aug 2008
Arius

Any chance turning this usb dac into a is2 source by eliminating the dac and analog section ?
  Reply With Quote
Old 9th March 2012, 12:03 PM   #7
zinsula is offline zinsula  Switzerland
diyAudio Member
 
zinsula's Avatar
 
Join Date: Mar 2004
Quote:
Originally Posted by Arius View Post
[....]I am aware of the better jitter filtering spec of the AD1896 ASRC compared with the SRC4192 which is pin compatible.[...]
I really wondere where this fairytale comes from. One guy actually bothered and measured both:

Any feedback on new CS8421 high-res ASRC?

Read the post in the link and some five posts down.
And i would not doubt his words.....
__________________
If you can't trust your ears, then CLICK HERE
  Reply With Quote
Old 9th March 2012, 12:28 PM   #8
Dxvideo is offline Dxvideo  Turkey
diyAudio Member
 
Dxvideo's Avatar
 
Join Date: Nov 2003
Location: Turkey
Send a message via Yahoo to Dxvideo
Dear Arius,

Will you share your PCB files and/or sell some PCBs(and one programmed PSOC of course) ? If so, I am interested about that project..
__________________
Best regards,
Ozgur

Last edited by Dxvideo; 9th March 2012 at 12:31 PM.
  Reply With Quote
Old 9th March 2012, 03:58 PM   #9
Arius is offline Arius  United States
diyAudio Member
 
Join Date: Mar 2004
Location: California
Quote:
Originally Posted by zinsula View Post
I really wondere where this fairytale comes from. One guy actually bothered and measured both:

Any feedback on new CS8421 high-res ASRC?

Read the post in the link and some five posts down.
And i would not doubt his words.....
Err Zinsula,
Perhaps you need to read the references you cited a little better. It was a well done spec comparison, no measurements were made and I agree that the CS8421 sucks. Search this forum and you'll find more info comparing AD1896 vs SRC4192. Even TI came back and admitted that that the filtering on the SRC4192 wasn't so good. That is why they made the SRC4392. Bel Canto used the CS8421 and then moved to SRC4392 for their newer DAC's. Benchmark and Centrance used the AD1896 for the DAC1 and DACmini.

Last edited by Arius; 9th March 2012 at 04:00 PM. Reason: typo
  Reply With Quote
Old 9th March 2012, 06:04 PM   #10
diyAudio Member
 
Join Date: Apr 2003
Location: Eastern Pennsylvania
Quote:
Originally Posted by zinsula View Post
I really wondere where this fairytale comes from. One guy actually bothered and measured both:

Any feedback on new CS8421 high-res ASRC?

Read the post in the link and some five posts down.
And i would not doubt his words.....
1. Here's one reason: http://www.benchmarkmedia.com/discus...RC-Devices.pdf

2. Another reason is that the SRC4192/4193 datasheet is absent a jitter transfer mask diagram. Which typically means a part does not perform well in the omitted parameter. The AD1896 datasheet on the other hand provides just such a diagram.

3. I don't see where Bruno P. provided any measurement of jitter rejection of the SRC4192 versus the AD1896 in the thread you linked. He merely expresses his unqualified preference for the one over the other.

4. Please share what exactly is your evidence for asserting that the jitter suppression superiority of the AD1896 is a fairytale?
__________________
Ken
  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


Similar Threads
Thread Thread Starter Forum Replies Last Post
Maybe have a CD transport source USB to drive a USB DAC? wa2ise Digital Source 0 6th February 2007 12:22 AM
USB powered NOS USB DAC with built-in headphone amp dsavitsk Digital Source 15 10th August 2006 06:11 AM
ASRC, DAC, Digital Filter for Sale Coulomb Swap Meet 4 30th March 2006 01:01 PM
-100dB 16bit data through ASRC 24/96 dac -pics A 8 Digital Source 16 29th April 2004 08:08 PM
AD1853 ASRC/DAC boards FS? tiroth Digital Source 41 1st January 2003 04:06 PM


New To Site? Need Help?

All times are GMT. The time now is 09:12 PM.


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