Please excuse my delay in addressing Peter and his minion Mani's criticism of my design choices and philosophy with respect to the Yggy. At the time they initially went off, I was in a pre hospitalization mode where I could barely walk. I then went in for some fairly major neck rebuild surgery and just now can function well – much better! I am still wearing a medieval torture device neck collar which prevents me from looking down, and for that reason may leave some typos in this missive. So now that I am healthy, please allow this retort:
1. Audio vs. Industrial DACs. Forty years ago, I designed and built the first no feedback, passive RIAA equalized, and 6DJ8 exclusively tubed audio preamplifier. At the time, the audio tubes in use were the 12A_7 family which indeed were designed to be for audio use. They were proper and inexpensive enough to be designed into the cheapest of phonographs. Although embarrassingly cheap, they were used in the finest preamps of the late first tube era, MacIntosh and Marantz. The 6DJ8 was a lower noise, far more inherently linear, and much wider audio through RF bandwidth device. Those capable of qualitative differentiation realized that the 12A_7 was a subset of the 6DJ8 – the latter featured much lower noise and much better distortion as a bonus. A total freebie was the much wider bandwidth which enabled much higher slewing rates. At the time, I was savaged by an engineering troglodyte of the day who called me out for not using audio parts and high amounts of feedback. Well, here we are 40 years later as the second tube era enters maturity. It seems the 6DJ8/6922/etc tube family has been well established – screw all of you Luddites.
Fast forward to today. We have audio DACs (largely used by engineering insophisticates or troglodytes). We also have industrial DACs which are faster, and far more accurate. Accurate you say? You bet. There are two specs that inexpensively designed audio DACs leave unspecified. DNL and INL. I don't want to turn this into a tech wiki, but those specs are less than 1LSB on both the AD5791/81. DACs unspecified thusly are prone to errors much higher than -110 db, possibly as high as -6db according to which code is problematical. This is why cheaply designed audio DACs are totally unsuited for weapons (missile can hit the orphanage instead of the real target) or medical imaging (Doc is going to think cancer is metastasized when it isn't or in the wrong place). Industrial DACs are just like 6DJ8s – you get more good stuff. THERE IS NO INHERENT ADVANTAGE IN USING AN AUDIO DAC FOR AUDIO EXCEPT COST – PERIOD. Again, I am savaged by trogs and Luddies on using industrial DACs for all of the wrong reasons – see immediately below.
2. The alleged Yggy AD5791 glitch problem. The shill Mani has pointed out this “massive” glitch with the approval of his tech guru Peter. A number of posters have become fully erect and piled on to the incorrect notion that the AD5791 has a glitching problem. This incorrect magnitude characterization (massive) is so widespread, it has even been used by the forum moderator. It is unfortunate that those accusations are not only based in bullschiit, but the accusers are so lame they are not even aware they are wrong.
The glitch referred to on the AD5791 data sheet occurs once per conversion – that's once every 2.5 to 2.8 microseconds or so. The glitch displayed graphically by Peter/Mani occurs every 500 microseconds. It seems these geniuses have not enough fingers and toes to count. Either that or they really cannot read sophisticated DAC data sheets.
The glitch pointed out (and agreed to by me as glitch before my surgery) is DAC glitch proper to multibit DACs driven by 2s compliment math at zero crossings, already pointed out by one poster, but ignored by many on this thread in their haste to condemn the DAC. This glitch worsens with lowering decoded output on any DAC.
The rest of the world deals with this zero crossing phenomenon is by adding dither, which is random noise just above the level of the glitch. This can either be done on purpose digitally or accidentally with an overly noisy analog section. Mani's claim that he can hear the glitch but that he does not hear the stuck toilet noise or dither at nearly the same level visible in the plot on his reference DAC tells me he is tossing himself off.
It is ordinary to employ dither. For the record, there is no dither employed in the Yggy. Then again, the Yggy in not an ordinary DAC. As of yet, I am not convinced that dither is the best to do. That is why the zero crossing is so obvious. If this changes, it is a trivial software upgrade.
I'm done. Perhaps my accusers may study 2s compliment math, dither, and R2R zero crossing glitches, and sophisticated DACs in particular before they jump on me again.