Matti Otala made up his mind about TIM in the late '60's, years before Japan came seriously on the scene. He also worked with 'Silver' of Japan for years. IF Japanese amps had flaws, it was because they ignored what Otala had found and went entirely for 'specs'. It was their own fault.
At the end of the 60s there was no fast output devices and eventualy he could had been right, but ten years later OSs had one order of magnitude better Ft.
That said i believe that overdesigned amps like the JVC AX series where out of reach for the rest of the world, hence the criticism.
JVC A-X9 Manual - Stereo Integrated Amplifier - HiFi Engine
Cascoded paralleled fet differential + differential + CM loaded differential folded cascode + diamond buffer , that s for the RIAA preamp...
About the same for the power amp + non switching + error correction + TEF.
How much knowledge is required to design successfully such a beast given that it must be reliable as it s a commercial design..?.
It s likely that these are these kind of amps that prompted the audiophile gurus to go the opposite direction as an escape route, not counting the psychogical factor as i remember that there was a lot of ressentment against those japanese firms that wrecked havoc above the other countries firms as in a few years the European audio industry, among others, was litteraly crushed and disappeared.
Fast output transistors did not come in the 70's to the USA. However, Matti Otala et al, and I, as well, designed amps that did 100V/us in the early 70's. Few others did. Of course, we use the faster output devices NOW!
I remember them well. came out a year after the 800A, 5 years before Hawksford, 9 years before Cordell.
😎
Attachments
Mr.Curl
Long time ago I have read one text that claim that actually TIM theory originator was Mr.Tapio Koykka , very close friend and countryman of Mr.Otala(RIP) , here is that text :
-
BTW Matti's surname is written Otala with one T only.. He is living not too far from my home.
But I really want to tell you about another man than Dr. Matti Otala:
In Finland we had one certain genuine world class amp designer, the grand old man. Somewhat sensationally, Dr. Otala has publicly given praise for the TIM ideas to this man as the originator. His name was Tapio M. Köykkä. This man invented (I cannot translate this word correctly!) the counterphaseamplifier (please correct!) in 1953 which was later applied throughout the world. Electro-Voice was the first asking for a licence and Westinghouse invited the man to work with them in U.S. with no success. He stayed in Finland.
Mr. Köykkä also used a specific amp design curing transient distortion from 1955 (no feedback over many amp stages) and he wrote an article about the crucial transient distortion in 1969 taking heavy critics from scientific engineers as the idea was against the accepted dogms. A bit later Mr. Matti Otala grabbed bravely this controversal idea and continued to study it and showed that Mr. Köykkä was right - the muddy sound of amps was caused by transient distortion.
If I have understood this correctly, it seems that Mr. Otala took the international fame for curing the transient distortion and made his career using it. But later at least in Finland, Mr. Otala admitted openly, that Mr. Köykkä really invented the whole idea. Mr. Köykkä was sitting on the first row in the 40th anniversary seminar of Finnish Assoc. of Acoustics when Mr. Otala started his keynote speech by nodding to Mr. Köykkä saying out loud that it was really him who originally invented the TIM phenomenon. Mr. Köykkä said later, that he had blushed then.
Mr. Köykkä also created so-called Orthoperspecta system (pat. appl. 1962 - being the mother of quadrophonic stereo) which employed a big whole range mid speaker and small high freq. side speakers working with turned phase. For instance Bang & Olufsen had to pay to Mr. Köykkä by violation of his patent in Finnish markets. Mostly Mr. Köykkä created tube amps called VOIMA. The most important model being the receiver VOIMA OP3. People keep telling it could give quite astonishing concert phenomenon with it's 3 spkrs. Later, transistor design from 1976 was called WATTRAM and finally rivalled the tubes by the creator.
We could also say that PRaT was first realised by this man. He kept talking about the leading edge at least from the 60's. He geniously showed how important the leading edge is by playing a tape wrong way. Exactly the same information looses all it's intensity, when it is played backwards. And still there is everything. Only when genuine transient start a chord, the music has it's original power (!).
He also said that measuring an amp by a sinewave is like testing a camera by shooting a white screen. And squarewave is the worst measuring signal, he said, by masking the problems. He insisted testing with information instead and appreciated measuring IM distortion.
Mr. Köykkä was a real character being in his own spheres when not too many understood hifi too well. He gave demos using classical music from vinyls in the backroom of his workshop in Helsinki. His company was called Voimaradio. And he was telling people in 60's and 70's that speakers don't spoil music but amps do. Thus the speakers he did use were truly simple and primitive.
This man invented many things. Also the multispark idea for car engines for lower fuel consumption. So the Alfa Romeo's Twinspark etc. is originally from his brains.
Here's Tapio M. Köykkä with his VOIMA OP3 receiver (chip tuner, transistor pre, tube power amps)-
I think you are right, banat, Matti learned from this man, and then took up TIM theory, adding lots of math and measurements. Matti got plenty of criticism too! (as we well know)
It was surely feasible using BDY56-58 transistors, wich had 30MHz Ft, but the amp had to be a quasi complementary, and this OS topology is apparently not as stable as a true complementary OS, perhaps that you could provided us a few details about this latter issue.
As for fast complementary devices there was none available in France in the late 70s, excepted a fews that were sold by L' Audiophile (Jean Hiraga s magazine) for the DiY designs published in their pages.
Thank you for the pics, that s fond memories..🙂
Undoubtly it was an amp that was a summary of enginering excellency and boldness, not that i like specially complexe designs, but as your point it all the sub circuits that nowadays are considered as instrumental for a high perfs amp are concentrated in this design.
This amplifier was in a group of 50 tested by Stan Curtis in the Hi-Fi Buyers Guide to Amplifiers published in 1982. I kept this for Curtis advice on what to listen for in testing/auditioning equipment. The JVC A-X9 cost 599 pounds and did not really undercut the European market.
Apart from the fact the A-X9 received no endorsement - on Curtis authority better performance could be had more cheaply from a Technics model such as the SU-V3 or SU-V7 at 119 pounds or 199 pounds respectively or from a Denon PMA-510 at 139 pounds.
Last edited:
In '74 we made do, as did John a couple years later, with 4 Mhz complementary devices from Motorola, 2N58XX.
😎
The Sansui AU-D9 was tested in Curtis 50 amplifier line up. It boasted technical innovation too - at a cost of 360 pounds. Money well spent according to him.
Actually, I used the 2N5884-6 in amps starting in 1970. I used them for the JC-3 power amp from 1974-1976 as well. Mark later went to the 5684-6 for the ML-1 just to get that added peak current and thermal dissipation, but this was a bad idea. I originally used the 2N5684-6 at Ampex research in 1969 in a very large 4Q amplifier for a motor drive. They were powerful, but had lousy safe area. I presume you, Nelson, are talking about the amp I designed for Noel Lee in 1977 or so, that again used the 2N5884-6 devices.
SUV7 has also a TEF but it s stabilised by diodes in serial with the signal path between pre drivers an drivers, it didnt worth the money, neither did the SUV3 wich used RC4559...
These were cheap and mediocre designs given they appeared years after the AX9...
I did find a lot of thoses in the local scrap yard when getting my trash there , but i never saw a JVC AX, not even a low power one..
Anyway he would had been of bad advice, I m 100% sure that he would say completely otherwise if asked nowadays..
Nice diamond differential but all is simplified, RIAA is mediocre as they seems to have anticipated that the CD would dominate, i guess that it was recommended due to its power equaling the AX9..
Thanks as always for your comments. Should your last sentence be read to mean that a series of two devices, one real and non-linear and the other ideal and perfectly linear, will have a different harmonic distribution than the real device by itself? That implies difficulties in making real-world measurements, at the least. Or have I misread?
As a separate question, would two series'd non-linear devices necessarily have the same harmonic distribution under the conditions in the case 1) of minimal "local" feedback and larger "global" feedback that they would have in the case 2) of more "local" feedback and less "global" feedback, both tests having the same total gain, etc. ?
If it were a simple linear case, I could imagine both cases to be identical, ignoring issues of internal signal levels and such that confound the real world. But with non-linear stages, I'm not certain that the cases are necessarily equal.
Thanks very much for any comments and corrections,
Chris
I m not sure myself but i guess that if the two components caracteristics are a straight line and a parabole the result will be an hyperbolic carateristic.
And in many real world situations, a "SET" amplifier, if well designed and operated very conservatively, will be the lower distortion device. It will have a hard time being anything other than monotonic, something that modern efficient amplifiers have a great difficulty approaching.
No Free Lunch rears its head at low signal levels, where most of the action really is, for any efficient linear amplifier. The "SET" amplifier school, including the N. Pass FET improvements, is commonly said to be non-linear, but because it's inherently monotonic, is instead inherently linear at the smaller signal levels of real music.
Thanks,
Chris
Can anyone really hear distortion of any kind below -90 or -100 dB?
That is the question.
I think humans just like 2nd and 3rd harmonics, maybe with a bit of 4th
It's clear many prefer it to the pristine sound of zero distortion.
Nothing wrong with that.
That is the question.
I think humans just like 2nd and 3rd harmonics, maybe with a bit of 4th
It's clear many prefer it to the pristine sound of zero distortion.
Nothing wrong with that.
dB is a dimensionless ratio. I like to start thinking about signal levels from the "VU" standpoint of an old school mastering studio (before "mastering" became a clipping plug-in). A standard volume level, usually a *very* loud 85dB SPL averaged, is set and headroom of 20dB is allowed for. Then speaker sensitivity is cranked in and required amplifier power to handle +20dB peaks is calculated.
For home use, a VU more like 75dB SPL is more my speed, and with nearfield listening to typical speakers of say 88dB SPL/0dBW/1M and 3dB gain from two non-correlated speakers, amplifier outputs at 0VU average to -16dBW each channel. If we could agree that signal at say -40VU should be preserved cleanly for acoustic music, then we need clean reproduction down to at least -56dBW, or .0000025 Watts average, and at lower listening volumes proportionately less. And music with greater than average dynamic range even less.
I contend that the performance of amplifiers in the "first watt" and even the first milliwatt is way more important than our continual discussion of their full power output performance. That applied in the 1950's but not now. We're looking for our missing car keys under the streetlight.
Thanks for any comments or criticisms.
Chris
I think much of the problem is the near-dichotomy between listening to music and listening to equipment, especially amplifiers, and especially when the latter is done with friends (or even enemies) to "show off the system". Almost inevitably the levels are pushed near or into clipping for system demo. Most of the time the selections are the "sonic spectaculars" to which most have become inured but as well sick of. And as soon as a bit is heard it's off to something else. Tedious.
This may to some extent explain why many are content to listen to relatively low-fi equipment when listening to music per se.
Let see where the amplifiers harmonics are located musically speaking, starting with the fundamental, frequency is divided by a power of 2 to fit in the first octave as this make the comparison easier :
H1 1.00
H17 1.0625
H19 1.1875
H9 1.125
H5 1.25
H11 1.375
H3 1.5
H13 1.625
H7 1.75
H15 1.875
H2 2.00
A musical piece played in a tempered scale will contain those fequencies ratios, and their even multiples, starting from the fundamental, i add the closest harmonic provided by the amp :
1.0000000
1.0594630
1.1224620 H9 1.125
1.1892071 H19 1.1875
1.2599210 H5 1.25
1.3348398 H11 1.375
1.4142135 H11 1.375
1.4983070 H3 1.5
1.5874010 H17 1.0625
1.6817928 H13 1.625
1.7817974 H7 1.75
1.8877486 H15 1.875
2.0000000 H2 2.00
We can see that set apart for H2 an amplifier will produce only false notes with its hamonics.
H3 has quite little error but wich has an importance for the sound as it s the equivalent of the effect provided by a non sweeping flanger but with a single note of the piece, this enhance the sound.
H5 error is more pronounced but still low enough as to get the flanging effect enhancing sound as well..
H7 is too far and will degrade the musicality.
H9 is of the same suspicous wood as H7.
H11 is a disaster as it s litteraly a quarter tone.
H13 is no good as well.
H15 is not as awfull as the two previous ones, so the saying that the higher the hamonics the more the awfullness is wrong as it s not systematic.
H17 is no good.
H19 is not as awfull as H17 and H11, this validate the previous observation about H15.
The conclusion is that high THD amps like the Cary are no more than non sweeping flangers and hence sound enhancers, and it will be the case of all amps whose (high) distorsion is located up to H6 as this latter is the same note musicaly as H3, likewise with H5 and H10 or H7 and H14.
This is for musics made with a tempered scale, with non tempered scales some of those harmonics will sound worse and other not as disastrous, but in all cases a good amp must be accurate with all musics, wich is not possible with high THD amps.
H1 1.00
H17 1.0625
H19 1.1875
H9 1.125
H5 1.25
H11 1.375
H3 1.5
H13 1.625
H7 1.75
H15 1.875
H2 2.00
A musical piece played in a tempered scale will contain those fequencies ratios, and their even multiples, starting from the fundamental, i add the closest harmonic provided by the amp :
1.0000000
1.0594630
1.1224620 H9 1.125
1.1892071 H19 1.1875
1.2599210 H5 1.25
1.3348398 H11 1.375
1.4142135 H11 1.375
1.4983070 H3 1.5
1.5874010 H17 1.0625
1.6817928 H13 1.625
1.7817974 H7 1.75
1.8877486 H15 1.875
2.0000000 H2 2.00
We can see that set apart for H2 an amplifier will produce only false notes with its hamonics.
H3 has quite little error but wich has an importance for the sound as it s the equivalent of the effect provided by a non sweeping flanger but with a single note of the piece, this enhance the sound.
H5 error is more pronounced but still low enough as to get the flanging effect enhancing sound as well..
H7 is too far and will degrade the musicality.
H9 is of the same suspicous wood as H7.
H11 is a disaster as it s litteraly a quarter tone.
H13 is no good as well.
H15 is not as awfull as the two previous ones, so the saying that the higher the hamonics the more the awfullness is wrong as it s not systematic.
H17 is no good.
H19 is not as awfull as H17 and H11, this validate the previous observation about H15.
The conclusion is that high THD amps like the Cary are no more than non sweeping flangers and hence sound enhancers, and it will be the case of all amps whose (high) distorsion is located up to H6 as this latter is the same note musicaly as H3, likewise with H5 and H10 or H7 and H14.
This is for musics made with a tempered scale, with non tempered scales some of those harmonics will sound worse and other not as disastrous, but in all cases a good amp must be accurate with all musics, wich is not possible with high THD amps.
Last edited: