Waly,
I didn't say all Japanese equipment was bad, I said there was a whole trove of consumer amplifiers that showed next to no THD values that sounded terrible, it had nothing to do with the fact they were from Japan. Nikko or something like that and many names I can't remember. I will say that some of the Yamaha stuff I have had from the 80's that was consumer grade like cd and tape players I hated the sound. But I did like my Sony of the same vintage. So go figure, it is not an all or nothing thing.
I didn't say all Japanese equipment was bad, I said there was a whole trove of consumer amplifiers that showed next to no THD values that sounded terrible, it had nothing to do with the fact they were from Japan. Nikko or something like that and many names I can't remember. I will say that some of the Yamaha stuff I have had from the 80's that was consumer grade like cd and tape players I hated the sound. But I did like my Sony of the same vintage. So go figure, it is not an all or nothing thing.
It's still a shocker that the clean, low THD amps, shortly after the tube era were perceived as dull, cold, bad, whatever. People never really had a chance to listen to unmodified/undistored source material and when they finally did, they said it sucked, instead of embracing it as their new benchmark, realizing they've been liking distortion all along.
A Japanese guy I met said they made equipment like that because we refused to listen . Instead we bought on specification . He said equipment of that sort wasn't sold on the Japaneses market . To be blunt he thought we were a bit stupid .
Marantz 9 was uncolored . I doubt I ever heard a better amp . It would have had miniscule distortion up to 5 watts ( > - 80 dB ? ) . Dynaco ST70 also . A touch of colour with that one . Mostly accurate . Bad transistor amps sound like tissue paper . And that's the almost OK ones if talking 1970 . The majority of amps that followed tubes were worse . Perhaps 1980 was the turning point ? I use a Quad 33 303 out of pride . 1967 and nearly knocking on the door of the Marantz . Many do not use 303 's properly . If so they will sound pedestrian . They are decent and support the theory well that transistors are every bit as good as tubes . 1967 marks the date it happened . The 303 has excellent measurements . Crown , Bryston , Crimson seem all to have similar circuits ( triples ) .
Marantz 9 was uncolored . I doubt I ever heard a better amp . It would have had miniscule distortion up to 5 watts ( > - 80 dB ? ) . Dynaco ST70 also . A touch of colour with that one . Mostly accurate . Bad transistor amps sound like tissue paper . And that's the almost OK ones if talking 1970 . The majority of amps that followed tubes were worse . Perhaps 1980 was the turning point ? I use a Quad 33 303 out of pride . 1967 and nearly knocking on the door of the Marantz . Many do not use 303 's properly . If so they will sound pedestrian . They are decent and support the theory well that transistors are every bit as good as tubes . 1967 marks the date it happened . The 303 has excellent measurements . Crown , Bryston , Crimson seem all to have similar circuits ( triples ) .
Last edited:
All this seems to be pure legend, far too often repeated.
Can you give the names of these early Japanese solid state equipment wich measured almost no distortion ?
It is quite puzzling that some guys of a younger generation hearing these early japanese amplifiers today found them very good.
At that time, circa 1970-75, numbers for minimal harmonic distorsions for these amps were about 0.03%. Due to slower output devices than those available today, better numbers were unobtainable because the amount of negative feedback was then quite limited. Today, the harmonic distorsions of the best designs which do not need to be extravagantly sophisticated approach 0.0003% 10 kHz., thanks to a high grasp of the feedback behaviour
If an amplifier measures good but sounds awful, it means that the measuring procedure was incomplete. There are no audible defaults which are not detectable by standard tests.
Please, take in consideration that sound engineers are, at least, supposed to make the difference between nicely colored tools (mikes, pre-amps, desks and effects) they use to "CREATE" the sound of an instrument, with machines, as transparent as possible, they use to record and reproduce those sounds they had precisely created.
You are right: Lot of audiophile prefer the poetically blurred sound of vinyls, so far from the original tape, to good digital copy, that no one in the studio can discriminate from the original. (Hundred time verified). Stark reality.
Yes, you are right: Everybody prefer the Beatles vinyls, with all the hidden blunders to the naive remastered CDs, certainly closer to the original tapes, Stark reality.
But, no, good tube amplifiers, at this time, like Macintosh ones were not colored and were more transparent than the first solid state amplifiers i was talking about.
We used Crown DC300 and DC150 power amps in the PA system i managed few years after : they were powerful ... and awful as hifi amplifiers.
On the first part, i agree, and the amp we designed in my R&D department (Mach 50) in 1970 was 0.025%. But it was an exception at this time.
On the second point, I'd love to agree, but I can not. HD is a little part of the story. IM and TIM an other.
There is a lot of things we do not measure with our sin, square etc constant level sources. Dynamic thermal effects etc
Just listen how two different electrolytic capacitances with very close measurements can sound so different.
Or how loudspeakers, with high measured distortion levels, can sound so nice, and some amps, with 1/10 of those distortions, so bad !
Last edited:
Please provide any authoritative references of Japanese amplifiers from the 70's and 80's that measured perfect and sounded horrible. Urban legends do not count.
True.
Not true , the turning point was somewhere in the second half
of the 70s , with very low noise small signal transistors and
fast output devices , as of 1979 state of the art was possible
and our current amps do no better , often quite the contrary
due to obsolesence of said exceptional devices.
JVC A-X9 | Owners Manual, Service Manual, Schematics, Free Download | HiFi Engine
I plan to run the LTP pretty heavy, perhaps 5mA per transistor, and keep Re low, between 47 and 22 probably. Then use MIC to keep the slew rate up. Around 1nV/rt Hz looks achievable with complementary inputs and Toshiba low Rbb transistors. That will do.
Yes. The one sided view of some people provokes an opposite sided reaction.
People should find "what to not like" in both
I did think and do a bit of calculation before I posted. The above should have excellent noise and still fine THD. Current mirror noise can be made below concern. Source needs to be low, of course.
Yes, more current and duplication works, and has some other benefits, in the LTP too, so that's why I plan to use it. But there are limits on how much the current can be usefully increased. What sort of results can the "CFB" achieve?
Best wishes
David
When you achieve the excellent noise with LTP you pay with loss of linearity especially at higher frequencies as I already stated. In the case of CFB current can be increased to very high levels if you really deem noise of prime importance, in this case youll also benefit with increased slewrate as well as higher bandwith with no detriment in stabiity.
Im sorry to say but a noise figure of 1nV/rt Hz is unobtainable with LTP in a power amp in reality. I never managed even close to that, not even using super low rbb transistors like 2sb737. Remember this is power amp we are talking about not MC stages or opamps. I wont design for such low noise in detriment of THD, its totally unnecessary. Speak to any designers in the industry with 30 plus years of experience and theyll confirm this view. Anything at and below 5 nV/rt Hz is deemed excellent and achievable with either LTP of CFB. As for THD a well designed CFB will always outperform conventional LTP designs at frequencies above 1khz. Youd need to make use of modern compensation scheme like TMC in LTP based designs to match the performance and even so I seriously doubt it can. It loses at low frequencies but many like myself deem accross the board low THD better than just at low frequencies. Bruno Putzey mentions this in his feedback paper if Im not mistaken.
I cover this point in my sx-Amp article manso, where I targeted the very wide loop gain bandwidth of CFA. IIRC the loop gain is flat out to about 40 KHz. On my e-Amp, I also can flatten the gain with VAS resistive loading ( but its a VFA). TPC on VFA can also get you to very wide loop bandwidths and improved feedback at HF.
Can't remember the names and I hate to say it but I haven't seen any of those awful amps in the past 20+ years that anybody kept. You want to name the great ones and forget about the early 70's junk that had totally fictitious distortion numbers and power figures. Yes two output devices that were supposed to put out hundreds of watts of power with no heat sinks to speak of, get real.
As I said I am not arguing that there weren't some very good Japanese amplifiers of the time, just not what was selling as most of the consumer junk of the time. It was the exception to the rule, not the norm.
As I said I am not arguing that there weren't some very good Japanese amplifiers of the time, just not what was selling as most of the consumer junk of the time. It was the exception to the rule, not the norm.
Bonsai Ive read the papers on your amps, they are nice clean simple designs, perfect for for home system. BTW you never answered my question in your thread. You could say I tweak the typical CFB to achieve measured .001 THD20. This tweak is simply to use beta enhanced vas. I use resistive loading in my CFB s as well to get 20 khz flat loop gain. This way THD20 performance is near identical to 200 Hz THD. My design has ULGF of near 6 Mhz, not easily achievable with VFB and that at phase margins above 80 degrees.
Im not prejudiced aginst VFB but I prefer CFB designs, I find them easier to obtain allround good performance.
That is exacly my point of view for everything at line level.
May i add that such an analog noise level is anecdotal compared with the ones yet present in near all analog sources (studio background noises, mike pre-amps noises, tape hisses, mixing desk+ artificial reverberations+ sound effects processors etc.) and even beneficial with digital ones ? (kinda dithering, better space perception).
It is a little like in digital photography: a little parasitic background light help to get more details in low lights parts of an image.
Last edited:
Hi Wahab,
I think Yamaha M4 (late -78) represents a well designed late seventies power amplifier. 2 x 100W @ 0.005% 20-30kHz. I've listen to this amp (restoring one right now) and it really sounds great (clean).
Technics SE-A3 is another fine example (at least specs) from the late seventies. 0.002% @ 200W 20 - 20KHz (0.001% half power). Approx 0.005% @ 200W 100kHz. The power supply and high current section of this amp is a fine example in how to design with minimal current loops, interleaved output P N devices and so on.
Datasheet output transistor Yamaha M4 attached, fT 70 Mhz!
I am simming a CFB design now that uses a beta enhancer and it achieves 4ppm at 20KHz at 350 watts into 8 ohms. At 250W it's about 600ppb. So, CFB is as you note, very capable of outstanding performance. Because the loop gain is flat, I also have the same distortion at LF
I could not find your question BTW - can you point me in the right direction
this is not at all obvious - maybe if you add the restriction to only compare "low feedback" designs
but VFA diff pair "added Q in the signal path" are operating a speeds that don't impact allowable loop gain around the output Q at audio - in fact audio power VFA are often compensated to allow even more negative feedback loop gain than CFA at audio frequencies
When does the increased current become counter-productive in a CFB?
Typically there is an optimum value to run an LTP, dependent on source impedance and frequency. How does a CFB behave?
1.0 nV/rt Hz looks like it would require heroic effort, that last few dB become harder and harder.
But near 1 looks achievable. Any number with a 1 in front will be fine.
I would be happy to learn an even better solution, so what are your results?
Some approximate numbers if not details.
What noise value? With what transistors and at what current and THD?
Best wishes
David
Last edited:
why would anyone design, build, listen to amps with audible distortion today
and I don't see the technical or psychoacoustics support for "even distortion with frequency is preferred to rising distortion" claim
music is typically shown to have ~3-5 kHz power bandwidth - it isn't flat to 20 kHz - we really don't like sound/noises with content flat to 20 kHz
the falling amplitude with frequency of music gives a rapid fall off of distortion, IM products – more rapid than a 1st order feedback roll-off for any nonlinearity above 2nd order
the top octave from 10-20 kHz is only 4 critical bands and most here have already lost the top one or two out of the 24 total in critical band theory that appears to explain a lot of human hearing features/capabilities
and today for competent technologists having at least the knowledge available in Cordell, Self's books building power amps with few parts per million to a few 10s ppm is the practical level of achievable distortion over the full audio bandwidth
if nonlinear distortion, harmonics and IMD are inaudible at all frequencies why does it matter if the residual errors deep in the noise floor of any live music has any particular profile way below thresholds of hearing
and I don't see the technical or psychoacoustics support for "even distortion with frequency is preferred to rising distortion" claim
music is typically shown to have ~3-5 kHz power bandwidth - it isn't flat to 20 kHz - we really don't like sound/noises with content flat to 20 kHz
the falling amplitude with frequency of music gives a rapid fall off of distortion, IM products – more rapid than a 1st order feedback roll-off for any nonlinearity above 2nd order
the top octave from 10-20 kHz is only 4 critical bands and most here have already lost the top one or two out of the 24 total in critical band theory that appears to explain a lot of human hearing features/capabilities
and today for competent technologists having at least the knowledge available in Cordell, Self's books building power amps with few parts per million to a few 10s ppm is the practical level of achievable distortion over the full audio bandwidth
if nonlinear distortion, harmonics and IMD are inaudible at all frequencies why does it matter if the residual errors deep in the noise floor of any live music has any particular profile way below thresholds of hearing
Last edited:
Yes, this would be TPC or TMC for example. Certainly in my experiments, TPC can make a huge difference in the 20k THD, while my TMC efforts have also yielded good results.
But, I think the point the pro CFB people are making is that this is a very linear topology capable of performance that easily equals VFA - and you get the wide bandwidths and high slew rates by the by.
Let me also state for the record I am not anti either approach. But, I do recognize and appreciate the unique properties of CFA and that's why I am also designing CFA amplifiers now as well.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.