(1) The improved DAC output circuitry is used only for the L,C,R channels, not the remaining channels.
In that case, we even have a stronger case in trying to lobby Denon/Marantz to offer a FW upgrade (even if a paid optional one) to make it possible to disconnect the center channel as well for the X3600H, if not for the 2020 models. That is so that many users who only want to use it with a 3 channel ext power amp rated >200 WPC and not having to limit themselves to those with at least 28 dB gain.
How low is too low? How low significantly affects performance? One way to view the question is that output impedance of the RCA's from the AV8805 that have output buffers is about 320 ohms. If we use 10x a figure of merit that makes an input impedance of at least 3.2k ohms a reasonable number. Any of the above scenarios work.
I think that 10X rule of thumb is fine for most but not some die hard audiophiles who probably would be looking at 20X or higher.
Evaluating the 8-channel volume control seems more complex since the output impedance may vary with the attenuation setting and internal opamps in the volume control aren't very robust. The specifications for the NJU72343 are taken at 47k ohms, but lower than that must be acceptable or the control would be unusable in most situations. Looking at the datasheet:
www.njr.com/electronic_device/PDF/NJU72343_E.pdf
On page 14, looking at the graph: Output Voltage vs Load Resistance
With 4.2V in, the output voltage falls as the control is unable to supply enough current into lower frequencies to support 4.2V. Of course 4.2V isn't often required. Pick your spot on the curve and you have your answer.
For another data point: Yamaha uses the Rohm BD34701ks2 or similar, 8-channel volume control in some of its AVR's/AVP's.
https://www.mouser.com/datasheet/2/348/bd34701ks2-e-1807077.pdf
The BD34703ks2 has a somewhat more robust current output. The distortion/noise specifications seem similar. On page 27 of the datasheet the minimum load impedance is given as 10k ohms. There is a graph on the same page showing output vs. load resistance at 1% THD+N. While this may not be a valid way to view all this, if the minimum for the Rohm unit is 10k ohms the the minimum for the NJR unit is likely higher. The actual THD+N driving these minimum impedances at various output levels is not specified.
Well, we are both guessing on this one. My guess is that, the 47 kOhms load resistance is just something NJU picked, perhaps because it is an old existing standard for phono cartridges lol.. If such high load resistance is necessary for the specified typical THD+N, they (and users like Denon) would not have skip a cheap OPA buffer. In an interview, Denon even bragged about this chip that they claimed they developed it with NJU, and they used it in their $4,000 AVR-X8500H.
https://translate.google.com/transl...,15700186,15700191,15700253,15700256,15700259
"Takahashi: The integration level of AV amplifiers has increased year by year, and circuits such as volume and selector have evolved in a single chip.
However, in order to get a better sound, we should go back to the basics and prepare a volume circuit dedicated to that function and optimize the signal path on the board in order to make it more audio-like design. I wanted to design a circuit that prioritizes sound quality.
So I asked JRC (a semiconductor maker) to make a circuit dedicated to the volume as a custom device. This development took about two years.
In addition, a simple and straight signal path is realized by using a dedicated device specialized for each function for the input selector, output selector, etc."
The AVR-X8500H allows individual power amp channel disconnections, sort of like an individual channel preamp mode.
Also, as I linked many time to HTHF's AVR build quality article, in which Dr. Rich discussed the JRC (NJU32743) chip and the Rohm chip in some details:
https://hometheaterhifi.com/technic...vr-audio-video-reciever-build-quality-part-v/
He said:
"One op amp per channel is inside the NJU72343 exactly like the Cirrus. There are eight of them in total. Compare that to the Rohm BD34705KS2 AVR LSI with 24 op amps. With 24 of those op amps packed into one die along with over 80 switches, the size of each op amp has to be smaller. Smaller size translates to more noise and distortion. This is discussed in detail in my article cited above.
Worst case specs show the New Japan Radio volume chip cuts distortion which is reduced by 50% at 1 kHz compared to the best Rohm. Rohm does not supply a 10 kHz THD specification as New Japan Radio does. Worst-case noise is down by 40% between the Rohm BD34705KS2 and the NJU72343."
So he obviously thinks the NJU chip is better than the Rohm chip Yamaha used in their AVRs and AVCs in terms of noise. He also cited that in the worst case scenario, the Rohm chip's THD+N is 2X higher.
The Rohm one is a LSI chip that has 80 pins with much more parts jammed in, including even an ADC, it is hard to imagine it would be more "Robust", with so many more OPAs and switches build in. Just because it specified a much lower test load, is not a strong evidence in its own imo.
Sadly, none of them publish enough details in their spec sheets, leaving us engineers guessing. I actually asked their customer support a few questions about what we have discussed, they acknowledge (automated response obviously) but no response whatsoever, basically a dead end. Dr. Rich, being an insider, appears to have more info based on what he wrote is his HTHF article, but he is hard to reach.
A straight forward way to render this question mute would be to follow the volume control outputs with opamp buffers. These opamps would serve the same function as the HDAM's in the AV8805. Lots of opamps and supporting passives components would add cost and heat to a unit. Really excellent opamps, higher rail voltages, and circuit boards with more layers would add even more cost and heat. Frankly I doubt people want to pay for them, Don't hold your breath. Yamaha does use opamp buffers in some of their higher end AVR's/AVP's for some channels.
Agreed on this in general, but I would counter that for a lot of applications such as one when paired with an external amp such as ATI's where the input impedance is >20,000 ohms, rated only 200-300 WPC, and gain >28 dB, skipping an additional discrete (e.g. HDAM) or chip OPA buffer is actually better because fewer parts = lower noises/distortions so more parts should only be used if the +ve out weight the -ve effects.
While it is interesting to speculate on the effects of these components, and dream about all sorts of enhancements, measurements of the products provide the bottom line.
Also agreed, unfortunately without detailed measurements like ASR's, marketing hypes such as but not limited to "superior low leakage flux toroid transformer, copper shielding, gold plated connectors, etc. etc., would rule. Just look at the $5000 AV8805's measurements, not really any better than the $1,149 is it?
Again, I raised one speculation of my own on one possible thing that could affect bench measurement consistency, that is the wide tolerance in THD+N specs of chips such as the JRC volume control chip and the Rohm chip.
That is, typical 0.0004%, maximum 0.01% (the NUJ72343), and maximum 0.02% (the Rohm BD34705KS2). The gap between the typ and max value is huge! No one has offer a counter argument on that yet so far, I still hope to be completely wrong on this one, that is, no need to worry about the "max.", though I would love to know why, it that is the case and would like to see links to supporting evidence/source info for any reasons/explanations cited.