Fractal Speaker DAC Specialities Masturbatory Conjecture Talk

What DACs are you talking about exactly?
Are you simplfying "power resistor network" to mean resistor I/V conversion, if so then yes that is not new either.
The Wadia article doesnt mention anything about resistors, PWM (1bit) powerDACs are not very special, that's true, but this ECDesigns DAC is multibit.
I dont understand how/if this DAC is will work in practice but assuming the "power resistor network" is describing I/V conversion resistors then in order to get an output impedace of 1.2ohm at 18Vpp you would need an output current of 15 amps.
ECDesigns mentioned the terrible efficiency (20% or something) of this "powerDAC" to achieve this power output.

It's a multibit NOS DAC, there would be no noise shaping.
Do your NOS DACs all have full reconstruction filters? if not then you've already been connecting your headphones to the equivalent of this DAC's output.

 

That's great! High wattage resistors are known to be so accurate.

 

here is a multibit power dac based on PCM53. this particular one is old as 2014. they used to sell the DAC board separately but I cannot find it anymore http://audioalt.ru/product.php?cat=69&product=2359

so, this idea is not new. by the way, the removal of all active components and stages is like throwing out the baby with the bathwater, but the listener can judge for themselves

you can see the working scheme here:

 

Yeah, that looks like the same idea.
I have tried a really shitty version of this kind of 'power DAC' driving headphones from passive resistor I/V on the DAC output,
If you ignored the fatal lack of macrodetail there were some appealing traits about the sound of this 'passive' output.

Hearing a proper example of a 'power DAC' would be interesting, of course the engineering feat isnt just building a DAC with those kind of output specs, it's building a good one since the benefits of having no active components seem too small to listen to an awful DAC just for sake of it being 'passive'.

 

Hello all,

I am the man who said that the Fractal DAC with UPL96ETL (usb stick player) was far better than the Denafrips Terminator with Sotm SMS200 Ultra Neo and all the Tweaks of very high level. I suspect SOTM streamer to be so bad that it could have ruined the sound by itself. Also : USB is not a great way to tackle with musical data... (super complex systems and sound quality loss - for the moment).

Here is more information about the PowerDAC that sounds unique from what I understand. For more information : please consider asking to John BROWN directly as they are open to discussions (sometimes they work so hard that you have to wait 1 week before getting an answer but this is not a problem according to me - especially regarding the high level of answer we get each time).

"The conventional analogue audio signal path is highly problematic. It was already problematic with analogue sources. Similar to USB we have all kinds of degrading here that kills resolution, adds distortion and mixes noise spectra together, creating new frequencies that do not belong to the audio spectrum. If one uses one wrong component, volume control, pre-amp, amplifier, speaker, interlink, the music is killed instantly. We had an example with one single capacitor in a speaker crossover (part of an impedance equaliser circuit that one can find in many speakers) when it is installed, the music is gone. So chances are that somewhere in the analogue signal path, some component messes up the music, there are more than enough components to cause this. Semiconductor input can only work in a narrow range (basis of a transistor or FET usually works in the linear region and this translates to a maximum signal swing of approx. 200mV). The signal on the collector can be much higher of course. But strictly speaking, every transistor compresses the signal into 200mV, again and again at every transistor (and there are many in an OP-amp or comparator). What happens is that the lower bits are pushed deep into the (thermal) noise floor, are amplified again, pushed into the noise floor, amplified, pushed into the noise floor. This can happen say 20 times before the signal leaves the Op-amp or amplifier. By that time most of the input resolution is lost (converted to white noise). Semiconductor amps are most destructive for resolution.

Tube amplifiers can have larger voltage swing on the grid and operate with much higher voltages. Usually there are only 3 or 4 tubes in the signal path, so here it is likely that more resolution survives. The biggest degrading in sound happened when sound studios started using mixing desks based on semiconductors (OP-amps). Then most of the resolution was already lost at the studio before the signal was put on record or tape.

Class D (digital) power amps are even worse as resolution is lost in the comparator (that is used for PWM generation), jitter, crosstalk between DAC frequencies and class D switching frequencies and noise spectrum of the SMPS inside the class D amp, ripple voltage on the switched output signal and degrading introduced by the required low-pass filter on the output of the class D amp. In my humble opinion class D amps are -totally- unsuitable for audio.

The tweaking that has become common practice with audiophiles is basically a result of incorrectly designed circuits and or the given limitations of components.


The PowerDAC skips the -entire- analogue signal path and -all- related problems. Tweaks that are required for highly critical analogue systems no longer apply to the PowerDAC as signal generation is fundamentally different. The PowerDAC has no transistors, tubes or other components that could distort a signal or could mask / reduce resolution. PowerDAC output = Speaker input = no degrading at all. The only ways to degrade a PowerDAC are to connect a flawed speaker or to use a recording with poor recording quality. Speaker interlinks don't matter much either, the PowerDAC has no issues with capacitive loading (passive circuit) and slightly increasing the value of the golden ratio matching resistor (speaker interlink DC resistance) is a non issue either. I use plain equipment wire to connect my PowerDAC o my speakers.

The PowerDAC only contains 32 digital switches that switch between -2 ... -9V and +2 ... +9V. Each switch can only have two positions, 1 / 0, true / false. There is no in between voltage like with analogue circuits, it is either fully positive or fully negative. The switches introduce almost a dead short (20 milli Ohms when switched on) so voltage losses are very low.

These 32 logic outputs drive very high precision (0.01%) 10W thin film resistor clusters. This way the digital fractal patterns that consist of a one's and zero's are translated to a highly accurate output voltage that corresponds to the sample value in the digital audio data. So the data is instantly translated into a voltage with almost zero losses / degrading. The most significant pattern resistors determine output impedance, the lower the value, the lower the output impedance. This way we can directly drive speakers with resistors and switches only, but still generate highly accurate analogue output voltages.

The downside is that lower resistor values and higher supply voltage will increase power consumption (also if no music is being played). This explains the low efficiency of "passive" PowerDACs of 15%. Efficiency cannot be improved, this also applies to class A and class AB topologies. I can improve efficiency in the voltage regulator and I already did by using low drop-out voltage regulators.

I can also vary the supply voltage (low when low volume is selected, high when high volume is selected). This way power consumption stays low when listening at lower volume setting.

The PowerDAC volume can only be varied by varying the supply voltage (gain) and or by shunting the output signal with selectable shunt resistors. This will be handled by a small micro controller in the PowerDAC that controls the supply voltage and the relays that switch the shunt resistors. Relays and shunt resistors no longer cause degrading here because these are in parallel with the speaker, the voice coil also sits in a strong magnetic field, so the very weak magnetic field of the single energised relay for the single shunt resistor is a non issue in this application.


Toroidal transformers offer very low magnetic field, offer high efficiency and are cheap. Peak currents are a bit higher as these transformers are more efficient. The only problem with conventional analogue systems is coupling capacitance (input-output). The PowerDAC is isolated by ElectroTos and it is not connected to any other mains powered devices, so no ground loops can be formed and the transformer coupling capacitance is a nom issue.

The PowerDAC works with very low impedance and the signal to noise ratio is huge. Its power supplies are fully stabilised (precision, programmable LDO regulators). So supply ripple and noise are non-issues here. Analogue power amps do not run on stabilised supply voltage, so all noise enters the amplifier output stage and feedback loops. Here the transformer and rectifier noise are a big problem and every different transformer causes a different noise spectrum and a different sound. Another reason to abandon analogue audio components. Power amplifiers in general sound better when powered by a stabilised (lab) power supply.

Forget about switch mode power supplies and class-D , in my humble opinion these are totally incompatible with audio reproduction systems. Produced sound (impossible to produce music) is totally flat, highly analytical and resolution is close to zero. Causes are very powerful switching noise spectrum, inter-modulation, jitter (PWM signal), ripple (PWM signal) and practical limitations of passive low-pass filters on the class D amp output.

Other problem is switching speed, the PWM switching frequency would need to be much higher than the sample rate and there are limits to the switching speed of power circuits. Higher switching speed = more losses = more heat dissipation. In a (Power) DAC we also want to minimise switching frequency to lowest possible value (higher frequencies = more degrading through crosstalk). The highest frequency present in the PowerDAC equals the sample rate (44.1, 48, 88.2 or 96 KHz).

The PowerDAC output impedance is pure resistive and acts like a series resistor, so there is no need for an external resistor. The PowerDAC has no local / global feedback loops that could become unstable due to speaker back emf.

My Open baffle speakers dip to 2.66 Ohms (bass) so I calculated the golden ratio matching resistance: 2.66 / 1.6 = 1.66 Ohm output impedance. T use 13.6 Ohm power resistors in the most significant pattern of the PowerDAC. This value can be divided by 8 (I use patterns with 8 sub-bits to all resistors are in parallel for ac). So 13.6 / 8 = 1.7 Ohm. This output impedance of 1.7 Ohm (resistive) acts as series resistor for the speaker.

When using shunt resistors for volume control, the speaker back emf is attenuated according to the volume setting and causes no problems.


The PowerDAC concept is the logical consequence of systematically eliminating imperfect components from the audio signal path. It now serves as reference for comparing different digital audio receivers and speakers."

Anyway, I am just here to share informations, not answering technical questions. I have just never heard digital music systems (and analogue music systems too) sounding this way like the the UPL96 and Fractal DAC is able to do : it is a very relaxed and transparent sound and the PowerDAC will go further without a doubt (as analogue paths is always problematic (coloration and/or lass transparency) from my experience).
They expect to launch the PowerDAC late 2021 and at this occasion they will go further than the UPL96ETL that, in my opinion, is very good but not excellent. John is a master in what he does and you will have to listen to their pieces before jusging the quality of their products.

Could you, please, change the title of this trhead that looks funny but not really representative of John BROWN attitude (very pragmatic, highly skilled, serious and passionated)?
Best,
Benoît

 

thanks for the info Benoit. no offense to you, the thread title is accurate until someone hears this thing (fractalDAC or powerDAC or whatever) and can provide a detailed report.

i would be happy to try this DAC/transport and compare it to my own digital system which is constructed with a very different audio philosophy in mind.

there is no such thing as true transparency, there is the aesthetic of the recording and the aesthetic of the playback system, and the constant balancing of the two. you may find that the 'elimination' of an analogue signal path and all of its 'problems' upsets this balance, to where you no longer want to listen to your favorite music, and then the system has failed.

 

I think true transparency does exist (at 99%) and I understand why you can't even imagine that is possible. I was like you before I heard the UPL96ETL and Fractal DAC combo - which was really good but not perfect but gave me the feeling that EC Designs is doing something in the right direction. It brought so much "truthness" in the sound that we were able to listen to poor recorded music very pleasantly. That is a proof (to me) of the true transparency of one system.

What do you mean by "detailed report"? A professionnal ad payed by EC Designs and that will never be accurate? A feedback from somebody who listened to it and will speak like an "idiophile" (who I am) - meaning that is only able to point out relatively good point of a system, because the sound is subject to colorations, etc. ?
I think both methods are wrong as only your ears (you and your friends) can tell you the truth. You should, as you said, listen to it. However, it seems difficult for the moment for EC Designs to send the product abroad - further than in Europe. Perhaps would you ask them how to proceed if you want to get a try to they electronics?

Anyway, it would be nice to discover what is your setup! Could you describe it and what experience you get from it?

Until the time somebody will share his experience with ECD combo or PowerDAC and next gen sources, perhaps would you mind changing the pejorative title of this thread that I find unfair.
Best,

Ps : True transparency is an ongoing process that could be achieved only by true gifted people and I think John is this kind of man (with his brother). This is pure projections I make here (as I have not heard from their PowerDAC and next gen receiver), but I choose this posture based on my experience so far. Listening tests will be critical.

 

I would quickly add that the Fractal DAC is really the true gem here. An excellent DAC at a very low price.
The Denafrips Terminator was not even as great in my opinion but we will have to make more tests in the future to figure it out.

For your information : John BROWN worked on very tough industrial projects and designed/assembled over his life (he is 60 now) tens of thousands of PCBs. He also made DACs over the last decades, from different technologies (USB, Optical, ElectroTOS etc.) It seems that he is going to reach the goal he had in mind since he started his quest (creating the company in 1990 and working on it as a hobby).
If you want more information about the past of John, I can find more if you want.

 

Unfortunately (or fortunately) life experience for itself doesn't always come with the ability to produce great products, whatever the subject at hand is.

I think it's pretty normal to be skeptic about new products, especially the ones with big claims/promisses.

You have to understand this type of situation has happened a lot before and most came out as "meh" or worse. Let's wait until people can actually try it and the title will change accordingly.

 

So, until then EC Designs is treated like masturbatory' people?
Have you already treated people like wanker before listening to their products? Please show a little bit more maturity to your argumentation.

Anyway, I agree that the major "revolutionnary" products sounds bad. This is not the case for EC Designs.

 

No, no. I ain't talking about agreeing with the choice of words, since they are not something i would particularly use, as i said, people hypeing products that end up as being pretty bad is pretty common around these parts. As far as i can tell the word was not towards the product itself but about the attitude (that comes with that hype).

Ironic to talk about argumentation maturity when your interpretation denotes lack of it.

 

To my uninformed glance two questions immediately pop out:
1. It looks like the DAC requires very clean and stable DC power. How is that achieved without active components? If it works with active components what about the discussed imperfections that they bring?
2. How the noise from switches is suppressed? It looks likes it will be pretty noisy.

I'm sorry, but for me, without engineering knowledge in the field (what most potential customers are), your description looks like a sale pitch for some kind of "perpetuum mobile".

 

Looks like a lot of posers and claimed professionals here shutting something down without merit to something they know nothing about and the masturbatory is in the presenter of the title .

 

It was more than three years ago.