Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › Audio-Gd Master 7 - Discrete Fully Balanced DAC (PCM1704)
New Posts  All Forums:Forum Nav:

Audio-Gd Master 7 - Discrete Fully Balanced DAC (PCM1704) - Page 66

post #976 of 1922
Thanks I own the mytek sterio 192 . As the chip is the same kind I do not know if it is the same chip. And then comes implinmintaion . The mytek is very detailed but the M7 is much sweeter . The pwd mkii is more fowArd , but with headphones especially the he 6. Or hd800 it is just perfect. Thanks for answering. As I see a lot of people have the NFB. And not to many the m7.

Al D
post #977 of 1922

I feel it comes down to what kind of sound you are looking for.

 

Generally:

(Given my experience of a few PCM1704 chipped Dac's/players including the Audio-gd Ref5.32 and hearing a few Sabre32 Dac's/players as well as owning a Sabre32 DAC)

 

 

The Sabre32 gives a more vibrant, exciting upfront and forceful sound. Which majors on speed, detail, clarity, dynamics and bass control.

 

The PCM1704 gives a more laid back refined sound. Which majors more on tone and texture and bass weight with maybe slightly better sound staging.

 

But these are just general observations I have no doubt there are many exceptions.

 

A poor Sabre32 DAC can be bright and thin and a poor 1704 DAC can be too dark and distant.

 

Personally I find most 1704 chipped players sound quite boring and prefer the Sabre32 dacs that I have heard but I listen to rock & metal which, given my previous comments, makes sense.

 

So pick your poison  :wink_face:

post #978 of 1922

To weigh in a third opinion - I found Ref7.1 to have a weightier sound than Sabre - more mid centric and "organic."  I have not heard the Master 7, but there have been several comparisons in this thread.

 

In terms of technicalities I think the Sabre is ahead of Ref 7.1 - it is faster, more extended in both directions and cleaner in it's presentation.  M7 is apparently better than Ref 7.1 in technical aspects so my guess is it closes the gap.

 

For my tastes and priorities I prefer the Sabre sound and don't care much for the particular strengths of PCM1704.  

 

Personally I don't get the "texture" thing.  Ref 7.1 had more weight and a little murkiness but I'm not sure if this counts as texture.  Maybe the DSP in M7 and Ref 7.1 is doing a better job at retrieving micro-detail for high jitter transports?  I know Sabre can tend to sound a little smoothed in the mids when the jitter is too high, while other chips retain midrange texture better with high jitter digital inputs?  From experience I found the Wolfson NFB10SE to have better midrange texture than the Sabre NFB10-ES (strangely the bass texture was better on the Sabre).  WIth low jitter transports like Audiophilleo 2 with Purepower texture is not something I would call the Ref7.1 as being ahead in - again M7 is probably a different beast again and for all I know might present better texture irrespective on transport jitter?


Edited by drez - 11/26/13 at 5:27am
post #979 of 1922
Quote:
Originally Posted by drez View Post
 

To weigh in a third opinion - I found Ref7.1 to have a weightier sound than Sabre - more mid centric and "organic."  I have not heard the Master 7, but there have been several comparisons in this thread.

 

In terms of technicalities I think the Sabre is ahead of Ref 7.1 - it is faster, more extended in both directions and cleaner in it's presentation.  M7 is apparently better than Ref 7.1 in technical aspects so my guess is it closes the gap.

 

For my tastes and priorities I prefer the Sabre sound and don't care much for the particular strengths of PCM1704.  

 

Personally I don't get the "texture" thing.  Ref 7.1 had more weight and a little murkiness but I'm not sure if this counts as texture.  Maybe the DSP in M7 and Ref 7.1 is doing a better job at retrieving micro-detail for high jitter transports?  I know Sabre can tend to sound a little smoothed in the mids when the jitter is too high, while other chips retain midrange texture better with high jitter digital inputs?  From experience I found the Wolfson NFB10SE to have better midrange texture than the Sabre NFB10-ES (strangely the bass texture was better on the Sabre).  WIth low jitter transports like Audiophilleo 2 with Purepower texture is not something I would call the Ref7.1 as being ahead in - again M7 is probably a different beast again and for all I know might present better texture irrespective on transport jitter?

 

I don't either to be fair. I have no issues with 'lack' of texture with the NFB-7 even with a phone like the HD800's, which have often been described as thin and themselves lacking in texture.

Everything just sounds very correct and real to me.

post #980 of 1922
The only sabre i have is the mytek
It sounds a little thin to me
Weather headphones or speakers.

Other davs i have are the m7. And the pwdmkii
Witch is more detailed and very updront
Great for speakers but a little to personnel
With some headphones
post #981 of 1922

I made some oscilloscope measurements that had me scratching my head about what is going on with the Off Ramp 5 and Master 7 I2S interface.  There's still work ahead for me as I need to rethink the cable type and termination values but I have good clue to why I am hearing so much variation with the cables used so far - CAT5, 6, 7 or SATA.

 

Conclusion #1 - The Off Ramp 5 seems to expect 75 ohm termination to ground.  The original thought that 120 ohm termination was used was based upon Off Ramp's spec page stating the termination is 120 ohms output impedance.  Off Ramp supplied a terminator RJ-45 plug to be utilized if the RJ-45 interface is not used and the resistor here measures 75 ohms.  See the last scope capture.

 

First, the scope capture below of Off Ramp's MCLK output with 120 ohm resistor only.   The resistor was crimped directly to an RJ-45 plug so no cable.  Clearly the termination is the wrong impedance at 120 ohms.  The waveform should be square without the leading/trailing edge overshoots.

 

Conclusion #2 - The Master 7 terminates I2S with 220 ohms.   I removed the M7 Ethernet connector board and measured the four surface mount resistors and they are also visually marked 2200.   The waveform below is the Off Ramp 5 MCLK signal with my favorite short CAT7 cable measured at the end of the line (DSP board header pin 5).   Larger overshoots with the increased 220 ohm termination resistance (and cabling).   I also measured the six cable types that I built and they all have this same basic wave shape.  Not good, see directly below!

 

 

And to confirm the Off Ramp likes to drive a 75 ohm cable interface I attached an RJ-45 to a 30-inch piece of coax and soldered 75 ohm termination at the end.   Now we are talking!  The waveform measured at the termination resistor is square and no large reflections even with 30 inch length.  See below...

 

(Edit: Earlier picture replaced with the same 2V/div scale as the others for easier comparison)

 

What to do?   I believe I will first rework the Ethernet board and remove the current 220 ohm resistors (0805 package) and replace them with 75 ohm resistors.   Afterward repeat the cable tests for the six cables made.  So first things first... find some 75 ohm surface mount resistors.


Edited by DACLadder - 12/1/13 at 1:06pm
post #982 of 1922

Wow I'd love to see how this affects the SQ when you have a chance to test it :3

post #983 of 1922
This is interesting in the least. According to Steve coax is 75 ohms as you said and it must me exact , he has post,s on audio circle. But please correct me if I am wrong are you saying that at the spec page for the off ramp 5 it says 75 ohm for coax??

And what is the imp of the rj45. To be ??

I did consult kingwa and Steve so it would conform properly. As I a glad you Re doing this , I am getting a little confused as to why your results are what they are... I am not saying you are reading it wrong either ...wow

Al d
post #984 of 1922
I just emailed kingwa and Steve to ask. I just rec my short block today so now I have everything I need to test , and now I am reading your post and

}{#}^}*{%]#. not cool.
post #985 of 1922
Outputs (all outputs are active)
Isolated S/PDIF coaxial – 75 ohm Canare BNC jack
AES/EBU – XLR male
I2S on RJ-45, Empirical Audio standard (5V, optional 3.3V) output impedance is 120 ohms
I2S on HDMI - LVDS format, pin compatible with PSAudio and Wired for Sound DACs.
post #986 of 1922

No industry-wide I2S standard continues to dog us.  The closest thing we have is the PSAudio/Wyred HDMI standard.  This makes me more happy I decided to go with the Audiophilleo although replacing the resistors on the M7 ethernet board seems like a easy enough fix so I haven't ruled out an Offramp. 

 

Really good stuff DACLadder, much appreciated.

post #987 of 1922
Impedance? Coax is 75 ohms. RJ-45 I2S is whatever cable you use. If you use 50 ohm, then terminate the end device
with 50 ohms. If you use 75 ohms, terminate with 75 ohms. If you use twisted-pairs, then probably 110 ohms.

Ted of off ramp replied. This is interesting.

Ald
post #988 of 1922
Kingwa. Answered.

Dear Albert,
The all our design, coaxial input impedance is 75 ohm.
The I2S input is 150 ohm.
Kingwa

So it appears the coax is fine but the i2s is wrong and the resisters must be changed
s
post #989 of 1922
Quote:

But please correct me if I am wrong are you saying that at the spec page for the off ramp 5 it says 75 ohm for coax??

The spec states the Off Ramp 5 I2S output impedance is 120 ohms.  I haven't read anything about termination impedance.  The 75 ohm theory comes from the supplied Off Ramp RJ-45 terminator plug which has a 75 ohm resistor for MCLK.  And the demonstration with 30-inch coax and 75 ohm terminator seems to indicate 75 ohms is better choice than 120 or 220 ohms.   A different I2S box may require a different termination value.  One fact is for sure that more testing is needed...


Edited by DACLadder - 11/29/13 at 5:06pm
post #990 of 1922
Please explain , it is above my understanding as you stated it. What I think I understand does not make sense to me.

Steve's reply for a twisted pair was about 120 ohms imp. For the i2s. Termination so please explain for me . I only gathered info. From both parties and based on what you just stated I do not understand the perimeter. To follow.

Al D
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Dedicated Source Components
Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › Audio-Gd Master 7 - Discrete Fully Balanced DAC (PCM1704)