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.

ES9038Q2M Board
ES9038Q2M Board
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 13th June 2018, 09:40 PM   #1461
Mikett is offline Mikett  Canada
diyAudio Member
 
Join Date: Aug 2003
Location: Toronto
I cannot see a trace where I can easily separate L & R channels from the source of AVCC cleanly. So right now I am relegated to using a common feed to both AVCC L and AVCC R. I will accept that.

I guess the question is not whether an SRC is good or not. I would tend to believe it is good. Now all of this stuff is kind of new to me. But I would think that I could accept USB on a external box, upsample as high as I can go and output to SPDIF. As opposed to getting an XMOS and then outputting to I2S.
I guess then is I am not sure what exactly the difference between an SRC and one of those XMOS or Amanero USB boards is? Do the functions cross or are they the same or does one do more than the other? I am kind of confused about this as they both accept USB but some do it in an external box and some as in an add-in card.

I'm in Canada so the proximity is not there. But hey thanks. I wish I was closer.
  Reply With Quote
Old 13th June 2018, 10:36 PM   #1462
Markw4 is offline Markw4  United States
diyAudio Member
 
Join Date: Jul 2016
Location: California
Regarding SPDIF, Amanero boards, etc., SPDIF is only good up to 192kHz. For people wanting to go higher and or who want DoP or DSD, it is necessary to use a different kind of interface which may be referred to as PCM or I2S, or native DSD. All those things require multiple signals operating in parallel sometimes at much higher frequencies than SPDIF will go.

Regarding upsampling, there are pretty good arguments for upsampling everything to 211kHz that would otherwise be lower, and either not support higher sample rate formats or not upsample them. It is possible to upsample with SPDIF, TOSLINK or PCM, I2S. DSD is a whole other thing, and it uses different reconstruction filtering. Upsampler hardware can work with TOSLINK, SPDIF, AES, and or PCM/I2S. The TI SRC4392 is more oriented towards SPDIF, IMHO, although it has PCM/I2S ports. The AK4137 SRC chip is more oriented towards PCM/I2S, although there are SPDIF-to-PCM and PCM-to-SPDIF chips made by ATK, it takes more chips to do some things as opposed to using SRC4392. The AK4137 also supports higher sample rates than SRC4392. OTOH, SRC4392 might be better for jitter reduction than AK4137, which is maybe a bonus feature of some SRC chips, although sometimes jitter reduction can be the primary purpose for using one. In our case we certainly don't want more jitter, but we are mostly thinking about helping to mitigate some reconstruction filter issues.

The difference between PCM and I2S has to do with the timing of the DATA signal bits relative to the LRCLK signal. There is also a BCLK signal which is the highest speed clock. PCM comes in two basic varieties, left and right justified. Right justified has to have word length specified. Shouldn't be necessary for left justified PCM or I2S.

Native DSD uses the same number of signals as PCM but they are different signals. For that reason, the same pins may be reused for PCM, I2S, and DSD.

DoP is a type of DSD that is sent using PCM signals instead of DSD signals.

Amanero boards can provide the highest speed PCM, I2S, DoP, and DSD sample rates, but they cost the most. In theory XMOS boards could do all that too, but drivers for native DSD and multi-channel sound are not free. There are free XMOS drivers for stereo SPDIF, PCM, I2S, and DoP.

Because of the very high frequencies involved, PCM signals can only run short distances. Most of the PCM signal output pins on devices cannot drive low impedance coaxial lines, so short length interconnections are necessary. It can help as I described previously to treat PCM signals like old parallel ATA disk drive signals where every other wire is a ground. That helps isolate signals from cross-talk with each other and helps to make a semi-well defined transmission line of not too low impedance for the chips to drive. Still, it helps to keep the wires short and to connect each ground wire in the cable individually at each board's ground plane. Unfortunately, the pins on most boards are not set up for such cables, so it is something you can mod if you find it necessary.

Last edited by Markw4; 13th June 2018 at 10:55 PM.
  Reply With Quote
Old 13th June 2018, 11:48 PM   #1463
Mikett is offline Mikett  Canada
diyAudio Member
 
Join Date: Aug 2003
Location: Toronto
Thanks for the explanation.
That means that even if one were to stick to FLAC and forget DSD for the time being. SPDIF/TOSLINK would not suffice to really get the best benefits of upsampling. One would need to have a USB to I2S interface to the DAC. This would then put external USB to SPDIF or DOP as not really being able to get you there being limited to 192KHz and the only way is to then have such a USB interface essentially right on the DAC board.

Am I on the right track here?

I will pretty much always use a PC - network- NAS as the source. I have too many PCs lying around. So I am trying to understand what I will need to get it right. Right now I am using foobar on the PC. So if I want to do proper upsampling like over 192KHz, I must output via USB because Toslink and SPDIF will only allow up to 192KHz. I take it then that the driver on the PC will upsample and then send it to the Amanero Board or Xmos and then this will pass it on via I2S to DAC at at the higher upsampled rate.

Last edited by Mikett; 13th June 2018 at 11:55 PM.
  Reply With Quote
Old 14th June 2018, 12:05 AM   #1464
Markw4 is offline Markw4  United States
diyAudio Member
 
Join Date: Jul 2016
Location: California
Actually, SRC4392 could do SPDIF up to 216kHz, and the DAC should probably work with that, have to try it though. Turns out 211kHz is what people use since a 27MHz clock will get you there, although I haven't seen a low jitter one at that frequency.

Even with low jitter clocks that work up to 192kHz, that is fine for most music which is on CD, DVD, or MP3. It certainly covers the audio frequencies we can hear and a lot more, and leaves plenty of unused frequency space for reconstruction filter slow transitioning.

Remember, the reason for all the high sample rate formats is to try get better sound quality for somewhere in the frequency range we can hear. The argument from people like Benchmark and Crane Song is we can get excellent sound quality without going over 192kHz, and there is no advantage from going higher with really good data converters. In fact, some things start getting worse as sample rate goes up. For example, the lowest jitter clocks are below 50MHz, and the lower the better from that perspective.

Last edited by Markw4; 14th June 2018 at 12:10 AM.
  Reply With Quote
Old 14th June 2018, 02:15 PM   #1465
Mikett is offline Mikett  Canada
diyAudio Member
 
Join Date: Aug 2003
Location: Toronto
Got a new Asus X470 MB with a ALC S1220 Codec. Actually it sounds better than an older external Assemblage DAC2 with the factory upgrade . Dynamics, pace, clarity. It is all over the older DAC. Never expected a built in sound chip on the MB to be superior....easily discernible. Tells me that this digital stuff have to considered as short term stuff. Keeps jumping in performance and older stuff needs to be put away. I was thinking I would take the old stuff and stick it in the office and it would be half decent. Oh Well.

Last edited by Mikett; 14th June 2018 at 02:19 PM.
  Reply With Quote
Old 14th June 2018, 02:18 PM   #1466
freezebox is offline freezebox  Germany
diyAudio Member
 
Join Date: Jun 2010
Default new clock on board

Hello,


Yesterday I installed my new NDK SDA 49,152 MHz clock on my board (v.1.04). As the new clock is very tiny (2,5x2mm) I had to build an adaptor board with same size as the orig. clock. I used a 2-line print board and cut the 2 lines so that I ended up with 4 lands to solder the clock on. For soldering the clock I first put 4 stripes of copper film on the outer land parts to make the connection from the lands to the solder points on main-board on bottom side of the adaptor board. After placing this copper stripes I added some smd solder paste on each land covering the copper stripes and the lands completely. Than I placed the clock in the Center of the "cross" so that the clock contacts touched the Corners of the land fields and sticked by the solder paste. Then I heated up my new solder hot air gun on lowest Speed with around 275 C and soldered the clock - worked perfectly. Then I installed the small board onto the main-board by soldering the copper stripes to the orig. solder points of the old clock. Before that I had to de-solder the old clock and used some modified solder iron. I removed the orig. tip and replaced it with some thick copper wire parts that I bended so that I could heat up all 4 solder Points at the same time - it worked.


How does it Sound? I would say the difference is not big and some people say the memory of sound lasts no longer than 1,5 sec. So a comparisson before / after is very difficult. Anyway I had the impression that it further cleaned up the heights (more silk like) and the stage is a little wider.
The deep bass got some more power towards kick-bass which gives a "rounder" impression with more substance and more natural feeling. As I changed the AVCC caps the same time from my "normal" 47uF aluminum to 47uF polymers 47uF (K50 with lower ESR) this might also have influed the bass performance. All in all I am very happy now with the sound allthough I already was before this clock change.


As a ranking of most efficient mods I did to my board I would judge it like this:


1) Change from voltage to current modus by i/v board -> improves dynamic and makes the sound very natural by getting rid of hiss in heights and less fatiguing , less edgy sound.
2) apply AVCC opamp PSU (inkl. own reg. for Vref) -> gives a tight and controlled bass "in-time" with the higher f-range. Pulses are very controled now and come with kick- and deep bass at same time
3) clock change and clock PSU with seperate reg. -> gives finer heights and removes some harshness. Soundstage increases a little in depth and wide.


For all using I2S Input I highly recommend some shielded wires. I installed some UF.L sockets on the bottom side to connect them. As I use SW upsampling to 96kHz incl. room correction both done by my Raspi I am not planning to apply a HW upsampler. Next step will be changing diodes of my analog PSU to HEXFREDS and care for installation into a housing such as a better heat-management. I checked temp. of my opamps and PSU transistors. opamps are around 50 C and PSU transistors are at around 60 C - maybe I will connect them to the housing bottom made of 3mm aluminum plate to improve cooling. Opamps get bigger cooling devices and board will be installed upside-down for better ventilation of opamps on bottom side. Output traces are cut after orig. DC Offset cap solder points and wired from there to new output connectors in the housing.
Attached Images
File Type: jpg clock adaptor card.jpg (207.7 KB, 283 views)
File Type: jpg solder iron mod..jpg (405.9 KB, 266 views)
File Type: jpg de-soldered clock.jpg (379.8 KB, 263 views)
File Type: jpg new clock on board.jpg (351.7 KB, 239 views)
File Type: jpg coax connexions.jpg (380.1 KB, 234 views)

Last edited by freezebox; 14th June 2018 at 02:31 PM.
  Reply With Quote
Old 14th June 2018, 06:54 PM   #1467
Markw4 is offline Markw4  United States
diyAudio Member
 
Join Date: Jul 2016
Location: California
Freezebox, Thank you for the detailed update on all the mods. Sorry you did not get more improvement from the clock. In addition using NDK SDA 49,152 means some of the higher bitrate modes will not be available. But, if the clock is working properly jitter should be much lower for the modes it does work with. There have be some reports of NDK quality control on their clocks (or ones without the A suffix) to be variable. Some people will buy a few and see which ones are the best and use those, or so some of the reports on the internet seem to be saying. Maybe someone else could clarify about the history on that.

Anyway, if the clock is indeed lower jitter then you should get used to whatever improvement it offers over time. Then, you might notice more change if the old old clock was put back.

Also, hard to compare with the equivalent mod done here since the clock I used is a Crystek 575.

Sure wish we had more more build reports to get a better idea of how mods turn out on average. Your contribution for that is very useful to many others who are still thinking about what to do, so thanks again for reporting your results.
  Reply With Quote
Old 14th June 2018, 07:12 PM   #1468
TioFrancotirador is offline TioFrancotirador  Poland
diyAudio Member
 
Join Date: May 2014
Location: Poland
I put CCHD-950-25 on my board with dedicated LT3042 PCB.
CCHD-950-25.jpg - Google Drive

Last edited by TioFrancotirador; 14th June 2018 at 07:16 PM.
  Reply With Quote
Old 14th June 2018, 07:23 PM   #1469
Markw4 is offline Markw4  United States
diyAudio Member
 
Join Date: Jul 2016
Location: California
Quote:
Originally Posted by TioFrancotirador View Post
I put CCHD-950-25 on my board with dedicated LT3042 PCB.
CCHD-950-25.jpg - Google Drive
Thank you for the pic and info. That looks like a pretty tight fit for that big clock. May I ask how did you get it in there?

Also, how do you like the sound?

EDIT: I just noticed it looks like you are using an AK4137 SRC board to drive the DAC digital audio inputs?

Last edited by Markw4; 14th June 2018 at 07:40 PM.
  Reply With Quote
Old 14th June 2018, 07:58 PM   #1470
TioFrancotirador is offline TioFrancotirador  Poland
diyAudio Member
 
Join Date: May 2014
Location: Poland
Quote:
Originally Posted by Markw4 View Post
Thank you for the pic and info. That looks like a pretty tight fit for that big clock. May I ask how did you get it in there?

Also, how do you like the sound?

EDIT: I just noticed it looks like you are using an AK4137 SRC board to drive the DAC digital audio inputs?
To install this clock you just need bend some wires, solder them to the pads and then bend them to adjust to the CCHD-950 footpring.

The sound is much better than with regular clock.

Yes I use ak4137 SRC. I oversample and convert all to DSD 112. The white pcb between the boards is fifo reclock board from diyinhk in such case onboard clocks could be used for reclocking (j03 shorte, j02 opened).

Last edited by TioFrancotirador; 14th June 2018 at 08:07 PM.
  Reply With Quote

Reply


ES9038Q2M BoardHide 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

Similar Threads
Thread Thread Starter Forum Replies Last Post
My Cambridge 500SE, needs an output board (to replace internal DAC Board) slbender Digital Source 1 11th July 2018 03:43 AM
JLSounds XMOS USB board + AK4396 board + Direct DSD PCB ravid Swap Meet 1 29th December 2016 12:36 PM
Unknown Board DC12V TPA3116 amplifier board DIY L163-9 AUS Ben Class D 0 8th December 2016 05:05 AM
FS: Ian FIFO i2s-to-PCM board, Single XO clock board trung224 Swap Meet 3 26th November 2016 03:18 PM
Lost 1 Zv4 ckt board..looking to buy 1 circuit board Blues Pass Labs 1 22nd March 2004 11:09 PM


New To Site? Need Help?

All times are GMT. The time now is 11:01 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki