ECdesigns

[realDHT]

3 minutes ago, realDHT said:

and tell the DACs apart. 

it should read like "tell the sources apart" 

[realDHT]

16 minutes ago, seeteeyou said:

  

 

That's so interesting to compare the stark contrasts among different experiences with the same DAC model.

 

Yes indeed. 

 

16 minutes ago, seeteeyou said:

OTOH, others who didn't report much of a difference between two sources should have connected PowerDAC-R to < $1,000 headphones

 

 

 

 

Not true, my headphones that I used in evaluating are 2000-3000 USD and the headphone amp (equally important IMO) would easily cost more than 10000 USD if it was sold commercially (just the components for building it cost >5000 USD).

 

But, as you said yourself, the price is not a good predictor for resolving power anyway.  

[realDHT]

11 minutes ago, hopkins said:

Anyways, price does not correlate with "transparency".

 

A funny recent experience about this: this summer I visited the best hifi shop in this country and had a long listen to a system with a total cost of more than 400000 EUR. It sounded very good for sure, transparent and 'musical' (BTW it also included the Taiko Extreme 😃). A few days later at Rowli's place, the same recordings sounded even better although his system is much more modest price-wise. Above a certain level, price does not tell much about the performance.    

 

11 minutes ago, hopkins said:

 

In my opinion, the only explanation that makes sense is that differences are either "psychological" (or due to some non audio related motivations) or due to other factors (ground loops, etc...). 

 

 

 

Or that the data from the sources differ in some way, i.e that at least one of the sources is not bitperfect.

[realDHT]

12 minutes ago, bodiebill said:

I think @matthias is misunderstood. He started with a legit question:

Does the 30k music server sound like a laptop (1) or the laptop like the 30k music server (2) or do they meet somewhere in the middle? (3)

 

Those who think that the PD is perfect would answer (2), those who think it is terrible (actually nobody I know of) would answer (1), and others (realists?) would answer (3).

 

I am confused by both Mathias initial question and now also your follow up? Am I the only one that think your posts does not make sense at all? What do you mean?

 

It has nothing to do about if the sound/performance of the powerdac is good or not, the initial discussion was about source independence. If the PD is source-independent it means the Taiko and the laptop will "sound" exactly the same ie they have no "sound signature" at all. This does NOT mean that the PD "hides" differences, but that the PD will not be degraded by RF transmission from the source, as long as the digital data is the same. Me and @hopkinsclaims virtually perfect source independence. Others might find they disagree on the basis of their own testing. However, some people seem to disagree just for the sake of it, without any experience at all.  

 

 

 

[realDHT]

9 hours ago, Norton said:

Exactly what I found with the MOS16/UPL.  Sounded great with  a plastic QED cable from my spares box, so I decided to see what an “upgrade” to a fancy glass cable would bring, turned out this made things sound bright and harsh - what I would attribute to increased  jitter.  ECD outlined that they concurred with this from their testing. 

 

I made a text search on the long ECD DIYAudio thread. It seems on the contrary that John found the glass toslink cables to sound better with the MOS16, although they increased the random jitter. His explanation was that random jitter can mask the effects of deterministic jitter.

 

https://www.diyaudio.com/forums/digital-line-level/79452-building-ultimate-nos-dac-using-tda1541a-post5600610.html

 

Quote:

Hi John, What is the connecter required for the toslink cable? Do you still recommend glass toslink? Best

 

On the new MOS DACs I use standard rectangular Toslink inputs. You can see this on the rear view of the MOS16 DAC picture I posted.

I compared real glass fibre (hundred's of very thin glass fibers bundled to a single optical conductor) vs a cheap $1 single plastic fibre optical interlink.

I measured greatly increased phase noise on the Toslink optical receiver output compared to a single plastic fibre.

It turns out that it is impossible to make the hundred's of thin glass fibres exactly the same length (to the nanometer).

So what you get is different light paths, all with a different propagation delay, resulting in timing "noise" as all signals arrive at a slightly different time.

This will change the jitter spectrum by adding random noise. So the audibility of deterministic (data related) jitter spectrum interference is reduced.

So the signal that looks pretty bad on the scope can actually lead to improved sound as deterministic jitter is masked by random jitter.    

[realDHT]

But then averyone should realize that the powerdac is a totally different beast than all the previous ECD dacs. The former generations implemented clocking BEFORE the interface/cable, therefore inherently affected by added jitter from source/interface/cable. The powerdac is clocked AFTER the source/interface/cable so that it makes no sense that the toslink cable would be important.   

[realDHT]

15 hours ago, Superdad said:

If that is the case (and it seems to be confirmed by at least two PowerDAC owners here), that would seem to blow up the claim of “source immunity.”   No problem if it does—the ECD design topology is intriguing in many other ways. 

To blow up the claim of source immunity, blind tests would be necessary.

 

I do not question the honesty of the powerdac reports here, that claim audible differences between sources and toslink cables.

But I doubt the perceptions would have been be the same if the tests were blind. 

[realDHT]

1 hour ago, Huubster said:

All I know is that the sound coming from my system, with the PD used as DAC, is definitely not insensitive to source material. Batterypacks as power supply for the PD yield different results, as Toslink cables do. Even power cables to feed my LPS which feeds my router and switches is effecting the sound of music played from my local HD.. 

 

The batterypack experience does not surprise me. It would be surprising to me if it did not make an audible difference. 

 

1 hour ago, Huubster said:

 

These kind of changes are my way of finetuning the sound to my liking. You can call me crazy, or say these things can't possibly ever make a difference... But I know what I hear, that's all that counts for me 🙂

I know what you mean, I used to do a lot of source tweaking (and heard differences) with all DACs I had before.  

 

I guess we can just agree to disagree about the source dependence/immunity of the PDAC.  

[realDHT]

Did anyone try 24/176 or 24/192 music with the powerdac-r, and got this working? It should be supported sample rates according to the webpage.

 

I have tried it today without getting it to work (I use a toslink converter that should be compatible with 24/192).

I suspect in my case it can be because I have mod-messed with my toslink module mounting to the board.

 

I would be nice if someone could confirm that 176 and 192 works on their PD (but it will require a converter that can do 192, I think ECDs converter only do up to 96).

[realDHT]

1 hour ago, realDHT said:

Did anyone try 24/176 or 24/192 music with the powerdac-r, and got this working? It should be supported sample rates according to the webpage.

 

I have tried it today without getting it to work (I use a toslink converter that should be compatible with 24/192).

I suspect in my case it can be because I have mod-messed with my toslink module mounting to the board.

 

I would be nice if someone could confirm that 176 and 192 works on their PD (but it will require a converter that can do 192, I think ECDs converter only do up to 96).

I got answer from John already. He confirmed that the powerdac is compatible (and always tested) with all sample rates up to 192 kHz.

 

But he says toslink working at the highest sample rates are sensitive to many things, cable lenght/material, dirt/dust in the connections, connectors not put firmly in place etc. Will try another cable first, as the connectors on the one I use now don't fit that great.  

  

[realDHT]

55 minutes ago, bodiebill said:

 

I was able to play up to 192 with some toslink cables, even my 15m long one. With others the max was 96. I see no pattern, just trial and error.

I tried my other toslink cable and now all sample rates work fine! 

 

So it seems I have to agree with you guys then, toslink cables make a whole lot of difference with the powerdac  😁

 

 

[realDHT]

1 minute ago, hopkins said:

 

To clarify, does that relate to the "source immunity" debate in your opinion ? Seems this is a different aspect (cable quality for 192), and that last sentence may be misinterpreted... 

Hehe I was just joking Hopkins 😁

But you are right, it may be misinterpreted.

 

My position is the same as before, the cable should not affect the sound (as long as the transmission works).

Although I did not test this for SQ with the PD. 

 

WIth the older DACs it for sure makes sense that toslink cables are important for SQ.

 

 

[realDHT]

11 minutes ago, Huubster said:

 

Since you went all balistic on your PD, I have a question for you if I may so 🙂

 

A friend of mine has the PD too and he did a fairly simple mod by exchanging those two huge browns caps on the above picture with well known versions from Elna Silmic ii Gold (Gold = Japanese making).

 

He's quitte shocked by how much the sound changed, actually improved to his opinion. You went through all the trouble of completely modding the PD, have you ever contemplated about doing this type of mod yourself?  

 

I have a few of those Japanese Silmic ii on order coming from China and will try it for myself as well, but it will take some time before I get them.

 

By the way, the 16v 10.000uf of Silmic versions are way too big to fit, my friend replaced them with 1.000uf Silmic versions, which actually have the same size as the brown 10.000uf versions. 

Yep I already changed those too, you can see the new ones at some of the modpics I posted. I used a Nippon Chemcon, a little longer but fits well anyway. There is also a good Nichicon that will fit, but that one was out of stock when I ordered. I could check the particular models when I have more time. 

 

Intresting to hear about someone who did the same. I cannot really say anything about how this particular caps changed sound since I made 1000 mods at the same time :)  

 

  

[realDHT]

6 minutes ago, yogibear said:

Very interesting but isn’t lower capacitance not recommended according to the design ?

 

This is unsettling for me as I never wanted to mod mine …..

Yep I agree it is a bit chancetaking to change value that much, since John choses his values with great care. 

 

I had some idea to parallel those caps with a much smaller highquality type, and see what effect that might give.  

  

[realDHT]

@hubsand

Ah I read more in detail now, so do you mean that it was the 1000uF change your friend was comparing and was happy about?

 

In that case I might try something like that too.   

[realDHT]

2 minutes ago, Huubster said:

 

I did check a few of your pictures on this,  but aparently not the latest ones then :) 

 

I checked and actually all pics show the new ones. For example

[realDHT]

Its max 5V over those, but then 16V are probably chosen for good reasons anyway, lika a lower ESR or other reasons.

 

Just be aware that sometimes strange unexpected things might happen with such changes..since only John knows the details of the design.

 

For example John mentioned that my clockmod might be casing my problems with the highest samplerate (because of frequency drift)

[realDHT]

5 hours ago, Norton said:

a. is source immunity apparent to the listener?

 

b. if it is apparent, is that a good or bad thing, achieved by liberating or limiting the source? Some really poor DACs are likely also immune to source for the wrong reasons.  Another manufacturer might well promote  their DAC, with some credibility, on the opposite characteristic: warts and all transparency to source 

I think point   b.    summarizes some of the confusion I have seen in the thread about source immunity.

 

My take on this:

 

It is important to separate the

1. source file/recording

2. the source hardware (ie computer)

 

A great DAC should be very sensitive to 1.  If a DAC that is not sensitive to 1. it is for sure a bad DAC.

A DAC that is sensitive to 2. cannot possibly be a good thing, it is always a weakness. Because source hardware can never improve a DAC, only make it worse (by transmitting noice and jitter to the DAC). The Taiko is regarded as a very good source HW because it degrades a source-sensitive DAC to less extent, than a bad source would degrade the same DAC.

 

Yes there are some very good sounding DACs that is also sensitive to 2. but they would be even better sounding of they were immune to 2.   

 

The PD is the only DAC  I have heard so far that is not sensitive to 2. 

 

It seems to me that many people regard "source" to be  1. and 2. combined, this leads to confusion and the misunderstanding that source immunity can be a bad thing. 

 

 

[realDHT]

On the same topic, I think the PD is REALLY sensitive to point 1. (highly resolving)

 

I can recommend you download and compare original 44.1 files vs PGGB upsampled 88.2 and 176.4

 

I did and I heard quite big differences with the powerdac.

And the best thing is that the upsampling improved the sound IMO.

I will probably get myself a licence of the software.

 

https://www.diyaudio.com/forums/digital-line-level/371931-makes-nos-sound-post6763055.html 

[realDHT]

4 hours ago, bodiebill said:

 

Thanks for clarifying, I totally agree. My problem is that I am not always able to differentiate between (f.i.) stridency as a result of bad recording vs stridency as a result of jitter. So in practice the assessment of source independence may not always be clear cut: is the DAC reducing fatigue by covering stridency type 1 ('bad') or by reducing stridency type 2 ('good')?

While this is true that we cannot separate Source HW from file in the final sound, it is not relevant to the concept of source immunity, that is my point.  If source immune, the DAC will neither reduce nor increase any stridency type 1, but will always reduce stridency type 2 to undetecable levels.

 

The only thing we need to do to check the source immunity, is to compare several sources, using exacly the same files (bitperfect) and if no-one can detect a difference with statistical significance in a blind test, it is likely immune. If someone can detect the difference, it is not source immune. 

 

 

 

[realDHT]

55 minutes ago, hopkins said:

 

I compared the three version of the "Bach Pastorale" track, in 44, 88 and 176 (24bits) kHz. They pretty much sounded the same to me (with headphones or speakers). I'll make the comparison with some other years to double check, but nothing really stood out. 

The easiest way for me to hear the difference on the pastorale is on the cymbals between 1:25-1:35, the 44.1 is a little subdued and mechanical..with the upsampled versions it sounds more "alive" and resolved. Also around 3:40 there is some loud dynamic piano notes that sound a little piercing and "clonky" on the 44.1, but this sharp clonkyness is gone on the 176.4 version. 

[realDHT]

I actually made a little blind testing on these files tonight before I saw your post.

 

In my JRiver media player I tagged each track (12 tracks total, all 4 songs with 44, 88, 176) with a random character sequence as the name, not to remember the name. Then I shut down the display and use my smartphone as remote, it sorts the track list according to the random names and I cannot see the file info, so I cannot see which song is which resolution.

 

Then I compared the versions for each song and wrote down the random names and my corresponding guess for each file.

The switched on the computer screen and had a check in the player, which random name correspponds to each resolution.

 

I got them all right, except for the "Day0", for that song I mixed up the 88.2 with the 44.1.

I have not tried on speakers, I usually hear this kind of things easier with my headphone system.  

 

[realDHT]

10 hours ago, hopkins said:

Yes, I get confused with all the different ways of stating output power, but John did clearly state 60 watt power consumption, 14 Vpp output, 2 ohm output. Maybe we are not talking about the same watts? 

 

14 Vpp output would be consistent with 3 watts output power into 8 ohms speakers. 14Vpp=5Vrms, and power = U^2 / Z = 25/8 = 3

The power consumption is always higher than the output power, so indeed not the same watts.

 

The powerdac will not be compatible with my current system, but still evokes my interest and surely looking forward to to listening impressions 😃  

[realDHT]

I noticed that John has posted more info and some pictures of PowerDAC-S prototype on DIYaudio. 

 

https://www.diyaudio.com/community/threads/building-the-ultimate-nos-dac-using-tda1541a.79452/post-6878199

[realDHT]

46 minutes ago, hopkins said:

Thanks for pointing out this post which is a good summary. The final product looks the same - at least on the outside (logo is added). 

 

I fully agree with the conclusion: " for me personally there is no way back to analogue circuits as I am very spoiled now." :)

 

@realDHT I'm curious to know what you think of the design and "lack of amplification" ?

 

It is surely a very novel and interesting thing. I would love to hear it. Sadly it would not work for any of my use cases which are 

Stax headphones (requires special high voltage amplification)

Linkwitz Speakers (requires 6 channels and active analog filter/signal processing between DAC and amplifiers) 

 

I think what John refers to with "no amplification" is that there is no traditional amplification, no active amplification devices like transistors or tubes. I would call it amplification in the general sense though.