On the other hand...
...this....
http://www.amazon.com/exec/obidos/t...f=sr_1_4/104-2384137-2161515?v=glance&s=music
.....is probably one of the world's worst CD transfers ever!!!!!!
...this....
http://www.amazon.com/exec/obidos/t...f=sr_1_4/104-2384137-2161515?v=glance&s=music
.....is probably one of the world's worst CD transfers ever!!!!!!
fdegrove said:
I have the misfortune of having about 80 percent of my CD collection from the '80s when I went on a big CD buying spree. Big mistake. Most of my CDs from that era are completely lifeless sounding. The "DDD" disks I have from back then are mostly on the Denon label, and most of them have hard digital clipping, which ironically sounds like a badly mistracking phono cartridge. For a period of several years, I completely lost interest in listening to music, mainly because of the poor quality of the source material I had bought.
But things are getting better now! I found out about high-res, and bought a Denon DVD-2900 player. Got a couple of Grateful Dead DVD-A's (Workingman's Dead and American Beauty) and was blown away by their quality. Being a big jazz fan, I've been buying a bunch of the Blue Note "RVG" CD reissues, remastered at 24 bits by Rudy Van Gelder, the original engineer. Most, but not all of these sound really good to me. They're not anywhere near what one would consider modern audiophile recordings, but they sound much better than I expected, based on LP listening from the '70s. The bottom line is that only very recently have I gotten the enjoyment from music that I expected long ago from the promise of digital. Better late than never I guess.
Digital mostly sucks, I think that the 'poor digital transfer' is just an excuse. How hard can it be? Negative feedback should be reduced or avoided if practical. Preamps sound better without global negative feedback, at least in my experience.
It is difficult to make a power amplifier without global feedback, unless it is truly class A, which means low power and/or BIG heatsinks. Also, it won't meet THX specs, which is important in the home theatre marketplace.
The problem is that the amp needs to put out almost 28 times more voltage than the preamp. This means that it will make about 750 times more 3rd harmonic distortion (the lowest order distortion not balanced out), so the distortion gets very high, very quickly with increasing power
Now, I might be able to make a preamp with .01% distortion at 2V out. This would produce 400 watts from my power amp. However, my power amp, even class A, and built with exactly the same quality as my preamp, might have approximately 750 times more distortion, so it would have 7.5%, at the very least, probably much more. Do these figures look somewhat familiar?
Still, listening from the first watt to 10W, I would suspect that the open loop design would be superior, as the distortion would be only .02-.2% or so.
I hope that this gives Steve Eddy, some clue as to my design decision to use feedback in power amplifiers, with respect the fact that feedback is not my first choice. This has confused him in the past. ;-)
The problem is that the amp needs to put out almost 28 times more voltage than the preamp. This means that it will make about 750 times more 3rd harmonic distortion (the lowest order distortion not balanced out), so the distortion gets very high, very quickly with increasing power
Now, I might be able to make a preamp with .01% distortion at 2V out. This would produce 400 watts from my power amp. However, my power amp, even class A, and built with exactly the same quality as my preamp, might have approximately 750 times more distortion, so it would have 7.5%, at the very least, probably much more. Do these figures look somewhat familiar?
Still, listening from the first watt to 10W, I would suspect that the open loop design would be superior, as the distortion would be only .02-.2% or so.
I hope that this gives Steve Eddy, some clue as to my design decision to use feedback in power amplifiers, with respect the fact that feedback is not my first choice. This has confused him in the past. ;-)
John,
I can't follow you. Where does that 750 come from? I guess it's 28 squared (about), but why the squaring? And why would the distortion scale with the gain?
If the preamp has .01%, and assume the poweramp is ideal, we still have only .01% at the power amp output, no?
Jan Didden
I can't follow you. Where does that 750 come from? I guess it's 28 squared (about), but why the squaring? And why would the distortion scale with the gain?
If the preamp has .01%, and assume the poweramp is ideal, we still have only .01% at the power amp output, no?
Jan Didden
john curl said:
If this is true there must be a good, rigorous technical reason why............
'Explanations' such as 'it sounds good in my experiance' should be discounted without a micro seconds consideration.....
For the simple reason that there are 100s of subjective reviewers who take the contrary view, because according to them...'it sounds good in my experiance'.
Who then are we to believe??????
john curl said:
Again, WHY???
Here are a couple of subjective reviews of Halcro's DM10 preamp., which uses large amounts of feedback, but is considered excellent by said reviewers.....
http://www.stereophile.com/amplificationreviews/404halcro/
http://www.halcro.com/pdf/Halcro_dm10_Stereo_Sound_Magazine.pdf
Ultimately, it is impossible to take subjective designers seriously, for the simple reason that NONE of them agree on what makes good audio.......beyond the ubiquitous .....'it sounds better in my experiance'.
http://diyaudio.com/forums/showthread.php?postid=476671#post476671
http://diyaudio.com/forums/showthread.php?postid=479509#post479509
One simply cannot hope to design good reliable audio electronics by merely taking 'untestable' and technically unjustified advice about 'design philosophy' from folk who not only cannot give reasons for their positions beyond.....'it sounds better in my experiance', but who moreover, universally disagree on what constitutes good practice..... 🙄
In general terms, the following constitutes sound, verifiable advice with respect to the design of linear audio electronics:
Make your foward path as linear as possible, and with as few singularities as possible, and then apply as much major loop feedback as you can get away with, without neccesitating the use of heavy (slew-rate sapping) compensation.
Jan, 1 volt into an amp (from the preamp) outputs to 28 Volts, so the preamp has an easier job. The square 28 is about 800 and third harmonic distortion will increase this much, all else being equal.
Halcro is an esteemed competitor. They use lots of negative feedback in everything that they design, not just power amps. 'Stereophile' puts both Halcro and the JC-1 power amp of my own design at the class A rating level. Both are global feedback designs, but Halcro takes it much farther than me. However: Martin Colloms in the June, 2004 issue of 'Hi-Fi News' criticizes the Halcro DM38 on several points, but we all agree that their distortion specs are lower than anything else around. Of course, you have to pay 3 times the price, and get 1/2 the power of my design, but the specs look great! So much for the 'cost savings' of a high feedback design. ;-)
The Halcro preamp did not fare much better in Martin Collom's review in 'Hi Fi News', April 2004.
My no global feedback preamp, the CTC Blowtorch, will never be reviewed by 'HFN' or 'Stereophile' because we don't have dealers, and our design cannot be considered for review, but it was reviewed in 'Ultimate Audio' which put it ahead of the pack. This is to be expected.
My 'true' rivals are Nelson Pass and Charles Hansen, both contributors on this website. Why, because they think like I do, in many ways, but there are still interesting differences that make for a 'horse race' in audio design.
Halcro is an esteemed competitor. They use lots of negative feedback in everything that they design, not just power amps. 'Stereophile' puts both Halcro and the JC-1 power amp of my own design at the class A rating level. Both are global feedback designs, but Halcro takes it much farther than me. However: Martin Colloms in the June, 2004 issue of 'Hi-Fi News' criticizes the Halcro DM38 on several points, but we all agree that their distortion specs are lower than anything else around. Of course, you have to pay 3 times the price, and get 1/2 the power of my design, but the specs look great! So much for the 'cost savings' of a high feedback design. ;-)
The Halcro preamp did not fare much better in Martin Collom's review in 'Hi Fi News', April 2004.
My no global feedback preamp, the CTC Blowtorch, will never be reviewed by 'HFN' or 'Stereophile' because we don't have dealers, and our design cannot be considered for review, but it was reviewed in 'Ultimate Audio' which put it ahead of the pack. This is to be expected.
My 'true' rivals are Nelson Pass and Charles Hansen, both contributors on this website. Why, because they think like I do, in many ways, but there are still interesting differences that make for a 'horse race' in audio design.
john curl said:
Yes, you said that before. You just repeat yourself. I asked if you could enlighten me why this is so. I take you seriously, may I ask you the same courtesy?
Jan Didden
I will try again: It can be shown that second harmonic will increase directly with level, so: 10 times level eg 1V to 10V will give you 10 times more distortion.
Third harmonic will rise as the SQUARE of the level, so 10 times level gives you 100 times more distortion. "Thus for small distortion, HD3 increases 2dB for each dB in signal level'' This is from class notes in a course in nonlinear signal analysis taught by Dr. RG Meyer at UC Berkeley about 30 years ago.
Found on p138 of 'Analog Integrated Circuits for Communication' Pederson/Mayaram, KAP
Also found on pp374-376 of 'Analysis and Design of ANALOG INTEGRATED CIRCUITS' by Gray/Meyer.
I hope that these references will make it more clear.
In any case, WHEN the preamp is at 1V, the power amp is putting out 28V. It is difficult, if not impossible to make a power amp without global negative feedback to have very low distortion (below .01%) at 28V. What about 56V? (the rated output of the JC-1)
Third harmonic will rise as the SQUARE of the level, so 10 times level gives you 100 times more distortion. "Thus for small distortion, HD3 increases 2dB for each dB in signal level'' This is from class notes in a course in nonlinear signal analysis taught by Dr. RG Meyer at UC Berkeley about 30 years ago.
Found on p138 of 'Analog Integrated Circuits for Communication' Pederson/Mayaram, KAP
Also found on pp374-376 of 'Analysis and Design of ANALOG INTEGRATED CIRCUITS' by Gray/Meyer.
I hope that these references will make it more clear.
In any case, WHEN the preamp is at 1V, the power amp is putting out 28V. It is difficult, if not impossible to make a power amp without global negative feedback to have very low distortion (below .01%) at 28V. What about 56V? (the rated output of the JC-1)
Thanks John,
This is new for me, I'll research it.
Jan Didden
Edit1: Yes, I KNOW that 1 times 28 is 28. No need to overdo it.
Edit2: Gray/Meyer is a 155$ paperback! Jeez.
This is new for me, I'll research it.
Jan Didden
Edit1: Yes, I KNOW that 1 times 28 is 28. No need to overdo it.
Edit2: Gray/Meyer is a 155$ paperback! Jeez.
janneman said:
Nope....:
http://www.amazon.co.uk/exec/obidos/ASIN/0471321680/qid=1096145031/202-0774040-8971823
Yes, found it also on amazon, new, 2nd hand for 64$. But that's the version without the Spice disk.
But it made me wonder. If you talk in terms of IC desing, I can imagine that there is something like the law John quotes for distortion, looking at isolated single FET or BJT stages. But that is different for a 28 gain power amp system. Should I shell out $ 65 just to find out that the discussion is not applicable? After all, I have done a lot of reading, and I have never encountered this in audio amps. Anybody can shed some light on this and save me 65$?😀
Jan Didden
Edit: Now playing: Mark Curry, "Let the wretched come home". Somehow it seems to fit...😉
But it made me wonder. If you talk in terms of IC desing, I can imagine that there is something like the law John quotes for distortion, looking at isolated single FET or BJT stages. But that is different for a 28 gain power amp system. Should I shell out $ 65 just to find out that the discussion is not applicable? After all, I have done a lot of reading, and I have never encountered this in audio amps. Anybody can shed some light on this and save me 65$?😀
Jan Didden
Edit: Now playing: Mark Curry, "Let the wretched come home". Somehow it seems to fit...😉
I have hundreds of tech books at my side. How can you hope to understand, if you won't invest in a textbook, once in a while? I pulled two texbooks that were within arms reach with the info. Unfortunately,'The Art of Electronics' did NOT have any info. However, you can also go on the internet and get the same info, I'm pretty sure. I can't write the equations here. I am just not equipped.
It doesn't make sense. If you have an amp with a gain of 28, and you have an input signal of 1 V with .1% 3rd harmonic, that means there is an input signal of 1V say 1kHz and 1mV 3kHz.
Amplify that 28 times in the power amp gives 28V 1kHz and 28mV 3rd harmonic, 3kHz.
Now if that power amp has ITSELF .1% 3rd harmonic, you also find at the output another 28mV 3kHz and, let me see, 0.28mV 9kHz. The two 3kHz components can cancel, add or something in between so you get anything between 0mV 3kHz and 56mV 3kHz.
Anyway, where is that 750 (or 800) times come in? I'm beginning to think that it only refers to some isolated single stage non-feedback BJT or FET but has nothing whatsoever to do with audio power amps.
Jan Didden
Amplify that 28 times in the power amp gives 28V 1kHz and 28mV 3rd harmonic, 3kHz.
Now if that power amp has ITSELF .1% 3rd harmonic, you also find at the output another 28mV 3kHz and, let me see, 0.28mV 9kHz. The two 3kHz components can cancel, add or something in between so you get anything between 0mV 3kHz and 56mV 3kHz.
Anyway, where is that 750 (or 800) times come in? I'm beginning to think that it only refers to some isolated single stage non-feedback BJT or FET but has nothing whatsoever to do with audio power amps.
Jan Didden
- Status
- Not open for further replies.