• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Review and Measurements of Holo Audio May --- Probably the best discrete R2R DAC

maxxevv

Major Contributor
Joined
Apr 12, 2018
Messages
1,872
Likes
1,964
This is spectacular by any R2R standards. Step ladder or otherwise.

As @WolfX-700 mentioned, those R2R makers who claim that its not about measurements are plain just making lazy lousy excuses for their incompetence.

Thanks for the measurements and showing the wider world how far R2R discrete step ladder can go if proper engineering is done to it!

BUT ... ouch to that pricepoint though. :p
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,498
Location
Monument, CO
The biggest objective detraction I have seen regarding delta-sigma converters is the potential for noise modulation (hand-wavingly, the noise floor changing with the signal). My impression is that issue, like earlier issues with tones and other artifacts, has been solved with modern designs using dither (noise decorrelation), higher-order and multibit loops, and better filters. Don't know for sure, though, but Amir's measurements don't seem to show any problems... I have lost touch with my old teacher and mentor that helped me many years ago (Dr. Gabor Temes) but imagine Scott and others could answer (at length). My delta-sigma experience (ADC and DAC) was at RF and used much simpler designs than used at audio frequencies.
 
OP
WolfX-700

WolfX-700

Addicted to Fun and Learning
Forum Donor
Joined
Jul 26, 2019
Messages
537
Likes
4,508
Location
China
The biggest objective detraction I have seen regarding delta-sigma converters is the potential for noise modulation (hand-wavingly, the noise floor changing with the signal). My impression is that issue, like earlier issues with tones and other artifacts, has been solved with modern designs using dither (noise decorrelation), higher-order and multibit loops, and better filters. Don't know for sure, though, but Amir's measurements don't seem to show any problems... I have lost touch with my old teacher and mentor that helped me many years ago (Dr. Gabor Temes) but imagine Scott and others could answer (at length). My delta-sigma experience (ADC and DAC) was at RF and used much simpler designs than used at audio frequencies.

Modern (audio) delta-sigma converters use upsampling to push modulation noise away from the audio frequency band and use LPF on analog circuits to filter it.

In fact, this has been done from TDA1541 and even earlier DAC chip architectures (though not necessarily a delta-sigma architecture)
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,498
Location
Monument, CO
Modern (audio) delta-sigma converters use upsampling to push modulation noise away from the audio frequency band and use LPF on analog circuits to filter it.

In fact, this has been done from TDA1541 and even earlier DAC chip architectures (though not necessarily a delta-sigma architecture)

Almost all delta-sigma converters I have ever seen use oversampling (*); in my experience upsampling is less common and somewhat different than the noise shaping performed by the D-S loop. The noise modulation issue I was addressing is a different thing and I am less familiar with it. One cause is when the signal frequency and amplitude "aligns" with the sampling frequency, loop stage bandwidths, and digital filters to create interactions that cause the in-band noise floor to vary with the signal. What I have observed (bearing in mind I have not designed one of these in a while) is the ENOB (effective number of bits, based upon SINAD per the IEEE Standard) would vary with signal frequency and amplitude, and spectral analysis (FFTs) showed the noise floor being modulated by the signal (sometimes creating wider skirts around the signal, sometime more broadband).

I worked with some sharp folk to try to figure it out for an ADC design of mine, and we could correlate it to phase lag within the modulator itself and to a lesser extent the digital filters following. All very specific to the particular design. I have some slides going over it for a college lecture I gave many years ago but have no hope of finding them now. Again, others here will have much better and more recent knowledge, both on the cause and effect and whether it is even a thing of concern now.

HTH - Don

(*) The ones that did not used Hadamard sequences and/or interleaving to provide the equivalent of oversampling without changing (raising) the actual sampling rate. That made circuit design easier from the standpoint of needing much less bandwidth, but properly combining a number of interleaved channels (let alone Hadamard multipliers) at GHz speeds was daunting. After a year-long research effort I abandoned a Hadamard design, much to my own and my sponsor's dismay, as looking great on paper but simply not realizable at the rates and resolution desired.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
The biggest objective detraction I have seen regarding delta-sigma converters is the potential for noise modulation (hand-wavingly, the noise floor changing with the signal). My impression is that issue, like earlier issues with tones and other artifacts, has been solved with modern designs using dither (noise decorrelation), higher-order and multibit loops, and better filters. Don't know for sure, though, but Amir's measurements don't seem to show any problems... I have lost touch with my old teacher and mentor that helped me many years ago (Dr. Gabor Temes) but imagine Scott and others could answer (at length). My delta-sigma experience (ADC and DAC) was at RF and used much simpler designs than used at audio frequencies.
The 9018 series did have the noise floor modulation issue. The noise floor is modulated by the signal. This phenomenon is also shown by Okto research hear in ASR. (I can confirm this too). The following generations have other modulation issues which are solved by external design. Such issues have never been found on any chip (that I'm aware of) from AKM, TI or, Cirrus logic.
 

scott wurcer

Major Contributor
Audio Luminary
Technical Expert
Joined
Apr 24, 2019
Messages
1,501
Likes
2,821
but imagine Scott and others could answer (at length)..

Retired now, I only have touch with guys like Bob Adams and Martin Mallinson as old friends, not sure any new light will ever be shed on the R2R vs D-S issue. I was at Rudi's original ISSCC presentation of the TDA1541, a great example of using the technology we have in hand to get a job done.
 

Esotechnik

Member
Joined
Mar 9, 2019
Messages
72
Likes
5
Location
Russia
I want to see comparison Holo vs. Mark Levinson 360s (PCM1704):
https://www.audiosciencereview.com/...ts-of-schiit-yggdrasil-v2-dac.3607/post-86611
Delta-sigma topping dx7 generate strange 2 sidebands.

And someone can make "true-SAR" vinil rip records via AP HF channel (1.25MHz)?
For compare with DS-rip via LF-channel (90 kHz max).

FYI comments about ADC architecture:
"Delta sigma's suppose to be a cheap alternative of the expensive and more advanced SARs, this article is sponsored by Texas Instruments and of course they aren't great at SARs, SAR is the first choice of the experienced professionals, everything else is cheap advertise"
https://www.electronicdesign.com/adc/what-s-difference-between-sar-and-delta-sigma-adcs

R2R DACs and SAR ADCs have a guaranteed settling time, deltasigma - not.
 
Last edited:

maxxevv

Major Contributor
Joined
Apr 12, 2018
Messages
1,872
Likes
1,964
I want to see comparison Holo vs. Mark Levinson 360s (PCM1704):
https://www.audiosciencereview.com/...ts-of-schiit-yggdrasil-v2-dac.3607/post-86611
Delta-sigma topping dx7 generate strange 2 sidebands.

And someone can make "true-SAR" vinil rip records via AP HF channel (1.25MHz)?
For compare with DS-rip via LF-channel (90 kHz max).

FYI comments about ADC architecture:
"Delta sigma's suppose to be a cheap alternative of the expensive and more advanced SARs, this article is sponsored by Texas Instruments and of course they aren't great at SARs, SAR is the first choice of the experienced professionals, everything else is cheap advertise"
https://www.electronicdesign.com/adc/what-s-difference-between-sar-and-delta-sigma-adcs

R2R DACs and SAR ADCs have a guaranteed settling time, deltasigma - not.

That's supposed to be parody right?
 

Esotechnik

Member
Joined
Mar 9, 2019
Messages
72
Likes
5
Location
Russia
Old digital recordings (before 1990-1995) are SAR frequently.

My opinion not first:

"On the recording side, if sound is recorded with crappy sigma-delta ADCs, there is no possibility for hi-fi reproduction. This is the main lacking area. I would like someone to make a Successive-approximation style Analog-to-digital converter which is the highest fidelity architecture possible. If someone can design a circuit around existing SAR ADC chips like the Analog Devices AD7621"

"Linearity is important for audibility (even though audibility is not a valid standard by which to judge DACs or ADCs), because the Shannon-Nyquist sampling theorem states that a bandlimited continuous time signal may be reconstructed by a set of samples (Dirac impulses) in discrete time, and the accuracy of the reconstruction of the signal (as long as it is bandlimited and Fs is twice the bandwidth) depends on ONLY two factors: the accuracy of the amplitude, and the accuracy of the timing of each sample."

https://www.head-fi.org/threads/hi-fi-audio-signal-chain-no-more-sigma-delta.747265/
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036

Literally the first reply:
YdOISpi.png
 

Esotechnik

Member
Joined
Mar 9, 2019
Messages
72
Likes
5
Location
Russia
Literally the first reply:
YdOISpi.png
"It is not a new technology now, and has been well tested over the years."

LTC2380-24 is new ADC chip.
APx555 "high bandwidth mode" (SAR 2.5 Ms/s) is new option.
MSB Select II DAC (3072 kHz/24 bit PCM R2R) is new, have 1000V/us slew rate.
Digital oscilloscopes don't use "lying" delta sigma converters.
I want to compare test noise signals and music fragments recorded through SAR ADC, in 24b/96k mode. Before buy LTC demo board.
 
Last edited:

Killingbeans

Major Contributor
Joined
Oct 23, 2018
Messages
4,089
Likes
7,552
Location
Bjerringbro, Denmark.
From the OP:
This is a commemorative prototype

It's not in production yet.

MSB Select II DAC (3072 kHz/24 bit PCM R2R) is new, have 2000V/us output speed.

Correct me if my math is off, but isn't that like 40MHz for an 8Vpp signal? Why would you ever need that much?
 
Last edited:
Top Bottom