ECdesigns

[murphythecat87]

3 hours ago, hopkins said:

Concerning the open baffles, they really require careful placement to optimize the sound. 

 

My living room is long, but the ceiling is not very high (2.5 meters, standard for new constructions). I felt like I was experiencing too much reverberations, and it was muddiyng the sound (when not listening very close to the speakers). 

 

I had the idea, of testing acoustic panels on the ceiling.

 

 

The sound quality improvement is significant. Everything is more defined. Now, my main problem is to stop listening and not remain glued to my chair all the time :)

 

I have decided to invest in proper ceiling panels to cover my listening area in the living room (roughly 1/3 of the ceiling). 

 

The open baffle speakers are définitely hear to stay!

good for you. My cloud was one of the biggest improvment in my system!

[Qhwoeprktiyns]

Error

[Qhwoeprktiyns]

I opened a topic some time ago asking whether audio reproduction was a matter of taste: 

 

https://audiophilestyle.com/forums/topic/63915-audio-reproduction-is-a-matter-of-taste/

 

 

The discussion quickly got very heated, but some interesting points were made, for example here by @Jud

 

 

The presentation refered to is this one:  https://www.youtube.com/watch?v=2V6YN-mshmY

 

 

The idea here, which I think is really fundamental to understanding a lot of aspects of our hobby, is that distortion is unavoidable, and that we choose equipement that has a suitable "disortion profile" according to our tastes. As a result, the low quality of audio equipement (regardless of cost) could explain why some claim that  most equipment (at least DACs and amplifiers) sound the same. But we are getting sidetracked ?

 

The architecture of the powerDAC-S minimizes distortion, according to ECD. I asked John Brown for some more specifics and got this answer:

 

"With the PowerDAC-S the D/A conversion error (distortion) equals the distortion on the speaker that is connected directly to the D/A converter output. In other words, the typical cumulative distortion found in all conventional audio sets is now reduced to D/A converter bit error only. I measured the D/A converter bit error using suitable test / reference signals (data) and that is better than 0.005%.

The human auditory system is not only sensitive to harmonics distortion (purity of tones) but also to temporal distortion (timing errors) as our 3D perception of sound origin relies heavily on accurate (L/R) timing information.

Timing also relates to the signal strength of the reflected signals and so on. So simply low distortion at only one 1 KHz test tone says very little about the realism of the reproduced audio signal.

The PowerDAC-S offers very low temporal distortion because the signal path is unusually short and almost completely free of non-linear components.

Power source -> on/off switches -> resistors -> speakers.

The small non-linearity is caused by the power supply. But given the massive decoupling caps and the Wheatstone bridge output configuration, non linearity errors remain exceptionally low.

Wheatstone bridge can be compared with a balanced system where common mode noise like power supply ripple and non-linearities are present on both speaker terminals. This way the power supply related noise and imperfections on the speaker output between the output terminals) are almost completely cancelled."
 

In addition, there is no filtering of any kind in the powerDAC-S (so the associated trade-off mentioned by Jud is not applicable).

 

So, assuming that the powerDAC-S + open baffle offer levels of distortion unheard of and unattainable by "conventional" systems (DAC + preamp + amplifier + speakers with crossovers and box coloration...) regardless of cost, can we assume that this system is "the most accurate" (to the recording) ? This is very obvious when listening to the system, but that is my personal opinion, so it can be discarded.

 

I would love to see some research done on this topic to determine whether we are able to recognize "distortion-less" sound (relative to the recording) when we hear it. After all, we can always tell that a voice or instrument is being played from a recording and is not heard live, even if we have never heard that particular voice or instrument before. So there must be something there...

 

I would also love to see some speaker manufacturers using such a system (powerDAC-S) in order to design and offer speakers that are truly capable of preserving the sound quality of the amplifier.

 

Does this all mean that audiophiles would prefer this system to others offering higher distortion levels ? That remains to be seen.

[Jud]

Thanks for the mention.  I talked about compromises between some different types of distortion in the post you referenced.  Let me talk about a couple of others, since you raised the topic of "distortion-less" sound.  I would say there isn't such a thing because of unavoidable compromises, and that what we may want to look for is a set of compromises that minimizes the distortions that bother each of us most.  These compromises (and hence the equipment) will almost certainly not be the same for everyone.

 

My Vandersteen speakers have baffles for the tweeter and midrange that extend almost no further than the edges of the drivers.  This will minimize diffraction compared to the speakers you have.  On the other hand, the wider baffles on your speakers may do a better job of minimizing back wave cancellation.  So, two different sets of compromises, each unavoidably emphasizing one type of distortion while minimizing another.

 

The Vandersteens also have linear phase crossovers.  This necessitates a small frequency hump in the midrange.  So depending on whether you are sensitive to temporal (phase) response and the imaging qualities that come with linear phase, or whether you want something as close to a flat frequency response across the band as possible, you will either like the Vandersteens very much, or you will vastly prefer speakers that achieve a closer to flat response.

 

You mentioned that you would be discussing filters.  If indeed there is actually no filtering in the power DAC you have, then this would mean no ultrasonics would be filtered out and you would have huge intermodulation distortion (like MQA filters but even worse).  I'll wait to read more on this topic rather than speculate.

[Qhwoeprktiyns]

8 minutes ago, Jud said:

Thanks for the mention.  I talked about compromises between some different types of distortion in the post you referenced.  Let me talk about a couple of others, since you raised the topic of "distortion-less" sound.  I would say there isn't such a thing because of unavoidable compromises, and that what we may want to look for is a set of compromises that minimizes the distortions that bother each of us most.  These compromises (and hence the equipment) will almost certainly not be the same for everyone.

 

My Vandersteen speakers have baffles for the tweeter and midrange that extend almost no further than the edges of the drivers.  This will minimize diffraction compared to the speakers you have.  On the other hand, the wider baffles on your speakers may do a better job of minimizing back wave cancellation.  So, two different sets of compromises, each unavoidably emphasizing one type of distortion while minimizing another.

 

The Vandersteens also have linear phase crossovers.  This necessitates a small frequency hump in the midrange.  So depending on whether you are sensitive to temporal (phase) response and the imaging qualities that come with linear phase, or whether you want something as close to a flat frequency response across the band as possible, you will either like the Vandersteens very much, or you will vastly prefer speakers that achieve a closer to flat response.

 

You mentioned that you would be discussing filters.  If indeed there is actually no filtering in the power DAC you have, then this would mean no ultrasonics would be filtered out and you would have huge intermodulation distortion (like MQA filters but even worse).  I'll wait to read more on this topic rather than speculate.

 

I appreciate your reply! Yes, "distortion-less" is an ideal no system can achieve. It is probably impossible to quantify how close we get to this ideal. At some point, we have to listen for ourselves, it is hard to convey listening impressions in writing. The explanations I am providing here only serve to point out the originality of the system. 

 

The types of issues you mention with speakers may be unavoidable - it remains to be seen how significant they are with respect to all the other aspects discussed. As I mentioned, I am very curious myself to try the powerDAC-S with multiple speakers. 

 

I'll provide more info on the lack of filtering. 

[fas42]

There is actually a good, subjective, "measure" of distortion-less: this is when any disturbing artifact is so low in level that the listening ear can't pinpoint the location of the sound generator. This is not going to work with headphones of course, but is easy to test with loudspeakers; one can get right next to a working speaker, literally inches away from one channel, and still be 'blind' as to where the sound is actually coming from; the illusion of a sound element originating from where it is placed in the recording space still dominates.

 

Along with this behaviour is the sense of, well, realness - once heard, never forgotten. All the hallmarks of live sound are in place; and it would be quite easy to fool someone, even if they knew the characteristics of some instrument well.

[Qhwoeprktiyns]

ECD send some explanations on filtering, which are technical, and highlight the limitations of all filtering approaches as you pointed out. The answer can be summarized in the following points:

 

- the powerDAC generates a sine wave analog signal using what is called "zero order hold" -  the sample value (voltage) is kept constant until the next sample. Sine wave signal now looks like a stepped sine wave signal. The consequence of zero order hold is slight trebles roll-off (approx. -3dB). Trebles are not boosted, so there can be slight trebles roll-off depending on speaker, room modes and listening position. This trebles roll-off can be easily corrected by the speaker. The Tangband W8-2145 offers such trebles boost (see datasheet), this way trebles are reproduced correctly. Most multi-way speakers already have an attenuator circuit on the tweeter because the SPL of a tweeter is usually higher compared to the SPL of midwoofer and woofer. So this L-pad attenuator can be easily adapted for 3dB trebles boost. Speakers can also be corrected externally using a simple filter (resistor in parallel with a capacitor, in series with the speaker).

 

- PowerDAC-S can output 4W rms max. The ultrasonic content at max. power is much lower. There is no (pre) amplifier that could become unstable due to ultrasonics. So in practice, the speakers cannot be damaged by ultrasonics as the power levels are way too low, and because there are no (pre) amps that could become unstable and oscillate. Because the signal goes direcly into the speaker, the speaker mechanical properties block the high frequencies and act as mechanical reconstruction filter.

[opus101]

19 hours ago, hopkins said:

 

- the powerDAC generates a sine wave analog signal using what is called "zero order hold" -  the sample value (voltage) is kept constant until the next sample. Sine wave signal now looks like a stepped sine wave signal. The consequence of zero order hold is slight trebles roll-off (approx. -3dB). Trebles are not boosted, so there can be slight trebles roll-off depending on speaker, room modes and listening position. This trebles roll-off can be easily corrected by the speaker. The Tangband W8-2145 offers such trebles boost (see datasheet), this way trebles are reproduced correctly. Most multi-way speakers already have an attenuator circuit on the tweeter because the SPL of a tweeter is usually higher compared to the SPL of midwoofer and woofer. So this L-pad attenuator can be easily adapted for 3dB trebles boost. Speakers can also be corrected externally using a simple filter (resistor in parallel with a capacitor, in series with the speaker).

 

As a designer of NOS DACs for many years, I take issue with the claim that NOS droop is 'easily corrected'. It can be easily corrected only in a very approximate way, to correct it more accurately isn't easy. The shape of the boost curve isn't one found in first order slopes, it takes at least a 2nd order system and then the curve only fits over a part of the boost range. 3dB boost only applies at a frequency of 20kHz, below that the boost is lower, falling to 0.1dB around 5kHz.

[Qhwoeprktiyns]

2 hours ago, opus101 said:

 

As a designer of NOS DACs for many years, I take issue with the claim that NOS droop is 'easily corrected'. It can be easily corrected only in a very approximate way, to correct it more accurately isn't easy. The shape of the boost curve isn't one found in first order slopes, it takes at least a 2nd order system and then the curve only fits over a part of the boost range. 3dB boost only applies at a frequency of 20kHz, below that the boost is lower, falling to 0.1dB around 5kHz.

 

Even if you did correct it perfectly, assuming you could do that without any other compromises, you would never get a flat frequency response listening to music with speakers in a room... I think the benefits of the solution provides by ECD far outweigh these small shortcomings. 

[opus101]

Yes I agree, the frequency response deviations of speakers in rooms are generally of a greater magnitude than for electronics. I'm not taking issue with the overall solution, merely with the particular claim I addressed.

[ZzZzZzZ]

Has anyone tried powerdac-s with hard to drive planars like Hifiman Susvara ?

Can it drive Susvara directly without the need of an amplifier and how is the overall sound quality of this combination?

[Perri]

On 4/17/2022 at 2:50 AM, hopkins said:

Concerning the open baffles, they really require careful placement to optimize the sound. 

 

My living room is long, but the ceiling is not very high (2.5 meters, standard for new constructions). I felt like I was experiencing too much reverberations, and it was muddiyng the sound (when not listening very close to the speakers). 

 

I had the idea, of testing acoustic panels on the ceiling.

 

 

The sound quality improvement is significant. Everything is more defined. Now, my main problem is to stop listening and not remain glued to my chair all the time :)

 

I have decided to invest in proper ceiling panels to cover my listening area in the living room (roughly 1/3 of the ceiling). 

 

The open baffle speakers are définitely hear to stay!

These look of decent quality where did you acquire them ?

[Qhwoeprktiyns]

On 5/28/2022 at 4:19 PM, Perri said:

These look of decent quality where did you acquire them ?

 

If you are referring to the speakers, they are not sold as-is, you need to build them yourself using ECD's specifications.

If you are referring to the acoustic panels, I got them here: https://www.panasorb.eu/lng/fr/mousses-a-relief-pyramidal/pyramidenschaumstoff-aus-basotect/   I purchased a few panels to test their use in various configurations, but will consider proper ceiling panels at a later date.

 

[Qhwoeprktiyns]

My short review of the powerDAC-S and open baffle speakers: https://paulstephane.github.io/powerDAC-S

[fas42]

Again you point out the importance of accuracy ... something that many audio enthusiasts pay lip service to, but rarely, IME, achieve. Key are the statements, "(They) have shown me aspects of sound reproduction that a conventional audio system ... may never be able to offer, regardless of cost. In this case, less IS more. For me, there is simply no looking back."

 

Yes. Once you appreciate what a well sorted setup can do, in regard to revealing the contents of recordings in a convincing and immersive manner, then everything else is an unacceptable compromise ...

[Qhwoeprktiyns]

12 hours ago, fas42 said:

Again you point out the importance of accuracy ... something that many audio enthusiasts pay lip service to, but rarely, IME, achieve. Key are the statements, "(They) have shown me aspects of sound reproduction that a conventional audio system ... may never be able to offer, regardless of cost. In this case, less IS more. For me, there is simply no looking back."

 

Yes. Once you appreciate what a well sorted setup can do, in regard to revealing the contents of recordings in a convincing and immersive manner, then everything else is an unacceptable compromise ...

 

To temper your enthusiasm, I did try to point out the drawbacks of the speakers on some aspects, which can be open to discussion (as to the degree of their significance) but cannot be completly dismissed.  I am trying to keep a low profile here because I do not want to get involved in lengthy discussions any longer, but of course I'll answer questions if there are any.

[Perri]

Are there any firmware updates to the power DAC ? or has anyone modified it ?

[Qhwoeprktiyns]

3 hours ago, Perri said:

Are there any firmware updates to the power DAC ? or has anyone modified it ?

 

There is a procedure for updating the firmware, which involves installing the chip manufacturer's software (nothing terribly complicated). I did update my powerDAC-R and S but this was required as I had an early test version. I believe that the versions sold since all have the lastest firmware - but I suggest you confirm that with ecdesigns directly.

[Qhwoeprktiyns]

ECD is now offering a USB interface that handles up to 192 kHz sample rates: https://www.ecdesigns.nl/en/shop/ut192

 

The interface uses an XMOS XU208 processor with asynchronous feedback and low phase noise audio clocks.

 

It has a dual output: Toslink and ElectroTos. The ElectroTos output was added as ECD (and others) found that Toslink cables do not always support 192 kHz - knowing which Toslink cable does can only be ascertained by trial and error.

 

The ElectroTOS connection was discussed here previously. It is compatible with the Toslink input of any DAC. The use of an optical signal offers in both cases perfect galvanic isolation from the source. As can be seen in the picture below, the ElectroTos cable has the LED placed at the end of the cable, so the light signal is directly placed at the DAC's input. The benefits of this interlink, in addition to guaranteed 192kHz compatibility, is lower jitter due to the absence of signal degradation associated with standard Toslink cables (reflections...).

 

As a result, the ElectroTos connection may be of interest to both users of ECD DACs (to guarantee 192kHz compatibility), and to users of other DACs which may not display the same "source immunity" as the ECD DACs, at least on their Toslink input (refer to previous discussions on the topic). 

 

 

The other side of the interface has a single USB B input to connect to your source. The interface is powered by the source. ElectroTos cables are available in lengths from 1, 2 or 3 meters. Check with ECD if longer cable lengths are required.

 

I have been using the UT192 for a few days and it works flawlessly, up to 192 kHz as advertised.

 

[Huubster]

8 minutes ago, hopkins said:

ECD is now offering a USB interface that handles up to 192 kHz sample rates: https://www.ecdesigns.nl/en/shop/ut192

 

The interface uses an XMOS XU208 processor with asynchronous feedback and low phase noise audio clocks.

 

It has a dual output: Toslink and ElectroTos. The ElectroTos output was added as ECD (and others) found that Toslink cables do not always support 192 kHz - knowing which Toslink cable does can only be ascertained by trial and error.

 

The ElectroTOS connection was discussed here previously. It is compatible with the Toslink input of any DAC. The benefits of ElectroTOS over Toslink, aside for guaranteed 192kHz compatibility, are lower jitter due to the absence of signal degradation associated with standard Toslink cables (reflections...).

 

As a result, the ElectroTos connection may be of interest to both users of ECD DACs (to guarantee 192kHz compatibility), and to users of other DACs which may not display the same "source immunity" as the ECD DACs, at least on their Toslink input (refer to previous discussions on the topic).

 

 

The other side of the interface has a single USB B input to connect to your source. The interface is powered by the source. ElectroTos cables are available in lengths from 1, 2 or 3 meters. Check with ECD if longer cable lengths are required.

 

I have been using the UT192 for a few days and it works flawlessly, up to 192 kHz as advertised.

 

 

Hi Stephane, I recently came by this thing on the ECD website. Is the idea that this would also be beneficial to PowerDAC users? I now use Toslink directly, but the word goes that the USB output of my streamer is of better quality.. So it might be worth a try for me, although I'm a bit skeptical whether it might improve things here. 

 

 

[Qhwoeprktiyns]

1 minute ago, Huubster said:

 

Hi Stephane, I recently came by this thing on the ECD website. Is the idea that this would also be beneficial to PowerDAC users? I now use Toslink directly, but the word goes that the USB output of my streamer is of better quality.. So it might be worth a try for me, although I'm a bit skeptical whether it might improve things here. 

 

 

 

Hi Huub - I do not expect there would be much benefits for PowerDAC users, aside for 192 kHz compatibility, but if you try let us know what you find. Sorry I cannot be of more help :)

[Huubster]

On 6/3/2022 at 1:03 PM, Perri said:

Are there any firmware updates to the power DAC ? or has anyone modified it ?

Hi Perri, I have first hand experience in modifying the DAC for the better. Most of the internals is SMD and I'm not touching that, but John uses 4 capacitors which can be replaced pretty easily, 

 

I replaced 3 of them with capacitors which are widely known and appreciated in the DIY community,

- The 2 big ones on the side of the 2 boards I replaced with Japanese version of the Elna Silmic II's, with the gold stripe, This was a subtle improvement over the original ones used by John, certainly not a major one.  

 

 

- Later on I replaced one of the smaller Nichicon caps (6,3v 330uF) used in (if i'm not mistaken) the power supply part on the board. These are sandwiched between the 2 DAC boards. I used a (absurdly expensive nowadays) BlackGate n-type (6,3v ; 220uF). This was a hole-in-one upgrade. This made the soundstage very big and more transparent, without losing the magic and refinement of the PD. I'm now thinking about replacing the other bigger cap (6,3v ; 1.000uF) there as well, but that value of 1.000uF is hard to come by in that small form factor. I have 2 other BlackGate n-types (220uF) which I could use in parallel, but then I'm still far of 1.000uF. The thing is that a technically skilled friend of mine thinks that the 1000uF is way oversized there and it could work fine with less. 

 

 

[Norton]

58 minutes ago, hopkins said:

I have been using the UT192 for a few days and it works flawlessly, up to 192 kHz as advertised.

I noticed this recent addition to the ECD site.  Is there any benefit is using this. vs. the previous U192ETL?  Have  you compared the two?  
 

This thread has been relatively quiet recently, would be good to hear from PowerDAC  R and (especially) S owners and their evaluation vs other DACs, inc previous ECD where possible.

[Qhwoeprktiyns]

1 hour ago, Norton said:

I noticed this recent addition to the ECD site.  Is there any benefit is using this. vs. the previous U192ETL?  Have  you compared the two?

 

I sold the U192ETL several months ago, so can no longer compare. If you are using the U192ETL with their previous DAC then you have no other option (since their previous DAC uses a special protocol, which the new UT192 does not support). 

[Qhwoeprktiyns]

1 hour ago, Norton said:

This thread has been relatively quiet recently, would be good to hear from PowerDAC  R and (especially) S owners and their evaluation vs other DACs, inc previous ECD where possible.

 

It is difficult to compare the PowerDAC-S to another DAC on its own.

 

You could compare it to a combination of DAC+preamp+amp. You could also  use the powerDAC-S simply as a DAC with an external amplifier (ECD offer adaptées that plug into the speaker outputs and have two RCA connectors). In both cases, the results are really dependant on the amplifier. This does not mean you won't get good results, but it will be a little different. It's all difficult to quantify...