Hello Gene,
I am new on this forum and actually in the market for a new receiver I read your review of the rx-a2010 with enthusiasm. Being an engineer, the section of most interest to me, was of course the measurements section. I have a few questions and comments and suggestions if you will, some of which will be from a purely academic point of view.
1) In the preamp section, while driving with HDMI you are using -20dBFS signal, I personally would have liked to see a 0 dBFS signal. Using a 0 dB signal will not only allow us to verify DAC/preamps ability to handle a full scale signal without any clipping, distortion etc., but also allow us to see how far down the noise floor is compared to the theoretical limits of the DAC. With a -20dBFS signal the noise floor can be be at the most @124dB from the peak.
Secondly I notice that you are calculating a THD + N using the second harmonic, I don't know about audio industry measurement nomenclature, but normally for a THD + N number I would expect to see all the harmonics and noise over the entire measured bandwidth included in the noise calculation instead of just the second harmonic. Am I missing something here ? It seems like the SNR term is where you likely include the total noise ??
Looking at the FFT plot of the rx-a1010 & comparing it with the plot from rx-a3000 it appears the 3rd harmonic has been cleaned up (either that or is masked by the higher noise floor by @6dB versus the 3000). The noise floor on the 3000 seems to be around -120 dB from 200 Hz onwards to 20KHz with a slight rise of 3dB from there on with a peak at 60 Hz, whereas on the rx-a1010 it seems like a gradual increase in noise from 20 KHz to 20 Hz with a noise floor that is anywhere from 3 to 10 dB higher.Interestingly your calculations show a higher SNR for the 1010 ?? Maybe I am seeing sampling artifacts from a relatively small number of FFT points ?
This suggests a few things to me: 1) The effective number of bits in the DAC/preamp circuitry in the rx-a1010 are less than 24. The noise floor is maybe 1-2 bits higher.
Given that there doesn't appear to be a bump on the 1010 at 60 Hz compared to the 3000 might suggest some additional power supply capacitance/filtering in 'XX10 series versus XX00 series. I'd appreciate your thoughts here ?
From your crosstalk charts any thoughts on why the crosstalk level is much higher for the HDMI inputs versus analog inputs ? Leakage noise in the DAC section ? Is this typical of most receivers/processors ?
Also it seems the crosstalk measurements for rx-a1010 are improved vis -a- vis the rx-a3000 . Does this mean there are likely improvements in the xx10 series versus xx00 ?
I would be curious to see a set of measurements on the preamp outs with a moderate (not max power, maybe 6-10 dB down from max rating) level of signal driven through all channels (4-8 Ohm reactive load ). This might give a better idea of preamp out signal quality with stresses on the power supply.
A lot of people seem to use preamp outs along with some speakers being driven by the receiver.
Going to the power amp measurements: The Power amp FFT plot is taken at max rated power which is typically past the knee of the curve where the distortion increases exponentially. I would like to see a FFT plot on the power/ distortion curve before it hits the knee of the curve where the THD is still low. The same with the SNR measurements. !
I am curious if the power supply 60 Hz and its harmonics still remain in the mix as strongly at lower power ? This might be an area for receiver improvement, except drive it hard enough close to its limits and any amp will start to show these harmonics in.
Of course its a different story whether these harmonics will actually be audible at these levels given that they are @80-100 db below the peak. It is funny how we nitpick the THD numbers of the amplifier/receivers whereas subwoofer/speaker distortion numbers are orders of magnitude higher !
Again I also notice higher/same power amp SNR numbers for the 1010 versus 3000, however the noise floor /harmonic spectra look cleaner to me on the 3000 ? What am I missing ?
Any thoughts on a set of step response tests on preamp & amplifier sections (actually mostly amp section). The thought process here is simulate a scenario similar to what might be experienced in a classical music recording with high dynamic range for e.g a sudden bass drum thwack etc.
Also in your receiver reviews, In the preamp/HDMI section how about a comparison/reference point with the results from a high quality player (such as Oppo 93/95) analog outs versus feeding the signal via HDMI with receiver DACs ? This will give people an idea whether they are gaining or losing something going on way or another. Of course whether the differences are audible is another story
I understand this is all a shitload more tests, but of course its only more time and money
I think that is enough rambling for the night. Thank you for review. Respectfully.