A little short for an answer.
...My question (under the question) was: "why".
Last edited:
What can we expect tweeter's side ? Their inertia provide a good HF filter. So, what ? Increase of temperature in the coil due to the presence of HF signals ?
I dont know..... some tweeters are elctrostatic other planar and others cone etc . Maybe compression from heating? Who knows for sure?
The ML-9600 has 10-20 mV of HF going to power amp (assuming no additional preamp gain before PA). Whats the gain of a power amp? 30dB? What does that do to the PA distortion in-band and what does it do to some tweeters. Does anyone actually know? That was my curiosity and want to find out. Modern, wideband opamps should do well ... I am talking only of PA.
I would like as clean a signal as we can do. Is this the best we can do with CD? It just seems lacking or sloppy to me. And the G-D and the 10KHz? I'm going to 24/96+ if for no other reason that it is a better solution on paper and in T&M. Its just a better solution IMO. And, it sounds more real. IMO.
Meanwhile, I like the results better with the 50KHz filter on this CD analog output; Aesthetically and sound-wise.
The OPPO 105D just now arrived.............
THx-RNMarsh
The ML-9600 has 10-20 mV of HF going to power amp (assuming no additional preamp gain before PA). Whats the gain of a power amp? 30dB? What does that do to the PA distortion in-band and what does it do to some tweeters. Does anyone actually know? That was my curiosity and want to find out. Modern, wideband opamps should do well ... I am talking only of PA.
I would like as clean a signal as we can do. Is this the best we can do with CD? It just seems lacking or sloppy to me. And the G-D and the 10KHz? I'm going to 24/96+ if for no other reason that it is a better solution on paper and in T&M. Its just a better solution IMO. And, it sounds more real. IMO.
Meanwhile, I like the results better with the 50KHz filter on this CD analog output; Aesthetically and sound-wise.
The OPPO 105D just now arrived.............
THx-RNMarsh
Last edited:
And a good opportunity to see if your filter improve your listening experience as well on it. A way to know if it is correlated with HF removal, or only some Phase /frequency response change ?No HF when playing CD in the OPPO !!!
It is a good beginning.)
Just did some tests for THD and THD+N. It is also very good. All the way to 20KHz, the THD+N is shown below at or just below.01% (100K load, .88v rms out.
And, No analyzer filters were used.
The test was done via a test CD made on the ML-9600 recorder. I had recorded very low distortion sine waves at various freqs in CD Redbook format.
THx-RNMarsh
And, No analyzer filters were used.
The test was done via a test CD made on the ML-9600 recorder. I had recorded very low distortion sine waves at various freqs in CD Redbook format.
THx-RNMarsh
Last edited:
OPPO Tr: The sq wave was also the same CD made by and used for tests by ML-9600.
The Tr going into CD system is faster than system so we still get Gibbs ringing. But Tr is about 20uSec. (no passive filter used, of course.)
All-in-All, very nice. Appears there Have been improvements in some units since the -9600 was designed
But the DVD Blue Ray/3D player and others are not this good and also not this expensive.
Tomorrow I will listen and evaluate.
THx-RNMarsh
The Tr going into CD system is faster than system so we still get Gibbs ringing. But Tr is about 20uSec. (no passive filter used, of course.)
All-in-All, very nice. Appears there Have been improvements in some units since the -9600 was designed
But the DVD Blue Ray/3D player and others are not this good and also not this expensive.
Tomorrow I will listen and evaluate.
THx-RNMarsh
Last edited:
Do you have any amps with the typical rising thd at the high freq. at 20KHz and test one of those for distortion with 50-100KHz HF.
The question I had for myself to answer is the typical GNFB amplifier with its falling nfb and increasing distortion, cause more distortion in audio range when HF is present?
I might not have to do that test myself
THx-RNMarsh
Well, I tried something like that. Please see attached THD measurement (almost pure 2nd harmonic). The amp is quite fast though, slew rate is >80V/us. The HF CCIF signal described yesterday was used, with frequencies about 900+910kHz. Test signal amplitude is 45mV peak. Two measurements, one without additional input filter, just amplifiers input RC onboard, and second with additional input 2RC filter 15kHz/-3dB. We can see that this amplifier produces HF intermodulation distortion at the difference tone, which disappears with input 15kHz/-3dB filter. 0dB(rel) in the spectrum plots is 6Vrms. Please forgive some mains lines because of complex setup.
P.S.: we can see that high slew rate, JFET input and low OLG are not a complete cure against HF intermodulation.
BTW, the Oppo square you are showing is typical for all nowadays reasonably engineered CD players.
Attachments
Last edited:
Your PA under test is much, much better (.5% at 900KHz) than most average/good amps!! Most are no where near this good at even lower freqs.... if they would even go to almost 1MHz. But, you get what my idea was about.
You get worse inband distortion as the out of band distortion increases in presence of HF/RF.
Maybe you can find a more typical VFA amplifier/receiver to test, sometime?
THx-RNMarsh
You get worse inband distortion as the out of band distortion increases in presence of HF/RF.
Maybe you can find a more typical VFA amplifier/receiver to test, sometime?
THx-RNMarsh
Last edited:
I think it is a forgone conclusion that with worse amp THD above 20KHz will do poorly inband in the presence of HF/RF. And, like the HF from an unfiltered HF DAC the frequencies are many and wide band... there would be many more IM products produced than a two tone test at HF.
I just culled some typical well known amp performance from Stereophile that people buy which fits the bill for higher audio distortion with HF.
How bad things can get depends on design, of course. And, it has been another variable which is uncontrolled from listener to listener environment and connected equipment. Out of band signals coming in on the audio coax might be helped for a lot of gear if it had a filter on its input (50KHz maybe).
If you see a review with THD heading for the moon (removing the 30KHz analyzer filter) then you may be a candidate for such an input filter.
Today, I am listening to the OPPO.
THx-RNMarsh
I just culled some typical well known amp performance from Stereophile that people buy which fits the bill for higher audio distortion with HF.
How bad things can get depends on design, of course. And, it has been another variable which is uncontrolled from listener to listener environment and connected equipment. Out of band signals coming in on the audio coax might be helped for a lot of gear if it had a filter on its input (50KHz maybe).
If you see a review with THD heading for the moon (removing the 30KHz analyzer filter) then you may be a candidate for such an input filter.
Today, I am listening to the OPPO.
THx-RNMarsh
Last edited:
I have measured a THD of the power amp under test at 900kHz and at test signal amplitude. THD is about 0.5% at 900kHz, so the amp is not any "crap". The HF intermodulation measured in audio band just reflects this non-linearity. Now I (finally) see what Richard was pointing to
Is it that with input LP filter connected?
Yes, amplifier's input LP filter is always connected. But it is only 1k + 220pF (723kHz/-3dB).
FYI, during the test, HF CCIF signal at input was 89.3mV peak-peak and amplifier output voltage was 1.255V peak-peak. Gain 14.05 (22.96dB) at 900kHz .....
The first one is simply horrible. We also need to know slew rate of the amps, I know an amp (no names) with textbook THD for small signals even at 20kHz but with horrible SR, and this will dynamically saturate with input HF interference.
Maybe. But in my experience, twin-tone testing method is more sensitive. With multitones I usually get nothing (just a nice plot), as the amplitudes of spectral lines in multitone or noise are much lower than those in the twin-tone. IMO the twin-tone tuned over certain frequency range is a good, reliable and repeatable method.
- Status
- Not open for further replies.