This might appear very outdated ? Perhaps even better as many op amps today would seem to leave these for dead . The TL081 perhaps is most interesting . It deals with the time when problems were mathematically real . Now it is a mystery why one almost zero distortion op amp sounds very different to another . Not a big mystery no doubt . A mystery on first glance .
I enjoyed reading it ( briefly I admit ) . I must print it and do a proper read . Forgive if too ancient and boring .
http://waltjung.org/PDFs/SID_TIM_TAA77_P2a.pdf
I enjoyed reading it ( briefly I admit ) . I must print it and do a proper read . Forgive if too ancient and boring .
http://waltjung.org/PDFs/SID_TIM_TAA77_P2a.pdf
Is this the one Brad?
http://www.analog.com/library/analogDialogue/cd/vol4n1.pdf
(In any case, thank you for pointing to these publications)
George
YES! That's the baby, Demrow's piece. Also I recall Jim Williams did some work on measurement techniques not long before his unfortunate demise. Thanks for finding that George.
There are two problems with settling time measurement. A tough one is generating sufficiently stable flat-topped test stimuli. The tougher one is suffering a huge overload and instantly recovering to see the highly magnified activity after slewing.
I like that series of Walt's very much, particularly as it is full of real measurements. And he listens a lot too, while thereafter (out of some other articles) being taken to task in the letters column by Lipschitz and Vanderkooy (tough cop and con cop as I said to John once, with a twinkle
).Also his candid discussions of insights and puzzlements are refreshing. In one of them, he missed (at the time) the nature of intermodulation distortion arising from specific types of nonlinearity, as he was looking for a particular byproduct in the spectra and didn't see it sometimes. I don't believe that one was ever discussed in LTEs. This was all well before my time for following audio, for the most part, but that was soon to follow. I caught up with those old TAA issues only recently, and they make for fascinating reading.
Demian's point about the overall settling behavior, not just the slewing, is very well-taken, and not very evident in those days in audio. It took the advent of sampled-data systems for the symbol domain to really drive the understanding and development of fast accurate settling behavior.
What is incredible is that the Analogue Dialogue paper is June 1970 . It talks of things common now which were almost unknown then . This was later on .
http://cds.linear.com/docs/Application Note/an10f.pdf
Does anyone have a rule for slew rates and real music ? Real can included electronic . Music not dangerous to tweeters .
1/ Op amps
2/ 100 Watt amp .
http://cds.linear.com/docs/Application Note/an10f.pdf
Does anyone have a rule for slew rates and real music ? Real can included electronic . Music not dangerous to tweeters .
1/ Op amps
2/ 100 Watt amp .
Last edited:
You will see perfectly square wave, hard to spot any distortions, so no use for such "show". Actually, it is shown in that thread. 500 KHz on full 100V swing was taken on purpose, to examine on "stretched" view what happens on 20 KHz, but of much lesser degree, hard to spot/measure. I use such technique very often, instead of measuring hard to measure things in nominal regimes I push regimes, measure, then extrapolate.
There are so many things that get missed for a long time, or reinvented, because the specific field is somewhat apart from another to which something is pertinent. I recently ran across prior art for the control electrode current recycling technique, from 1957, developed for bipolars by Frank S. Boxall, and a bit later in the bootstrapped cascode version by Aldridge. These two are better known to audio as attributed, respectively, to Larson-Baxandall and Hawksford. But because they were invented/developed for telephony/communications use, apparently no one in audio noticed them or recognized their applicability.
Another example: "memory distortion", which is very old, dating from (at least) oscilloscope vertical amplifier designs, where it is referred to explicitly as thermal distortion. In the 'scope case a lot of the information remained proprietary, but it is still unlikely that someone reading only audio-oriented publications would have noticed it if published.
I was given an old Techtronics storage scope which enhanced the memory of the trace . I think that was common on better scopes ?
I discovered the other day a MOSFET patent by HC Lin for 1969 . He is the most un-talked about person in Audio .
http://www.ece.umd.edu/lin/bio.php
I discovered the other day a MOSFET patent by HC Lin for 1969 . He is the most un-talked about person in Audio .
http://www.ece.umd.edu/lin/bio.php
Last edited:
Tektronix storage scopes were interesting beasts. The scope that really impressed me was their 7104, which, by using a microchannel-plate-intensified CRT, displayed single-shot events of subnanosecond duration in normal ambient light conditions. John Addis, one of the almost-unsung heroes at Tek in those days, writes about the instrument and his championing of it in one of Jim Williams' collections*. John opens the article, Good Engineering and Fast Vertical Amplifiers, by discussing product development and suggesting that "The Time To Market God needs to be shot, and engineers must not adbicate their responsibility to determine when a product is saleable. That is part of their job."
On the other hand, when I mentioned that the photodiode array spectrometer development took four years before the entire system was turned on, my supervisor at the time said "Around here, if you tried to do that, you'd spend three and a half of those years looking for another job."
Brad
*Analog Circuit Design, Art, Science, and Personalities".
I discovered H C Lin after being told that Harold Leak firmly told either Mr Tobey or Mr Dinsdale he would see them in court if they were so foolish . They were indignant that his amplifier resembled an article in Wireless World by them . He then added " I am very rich , who is going to win " ? No doubt none of it true ? However it got me looking at who did invent quasi complimentary amplifiers . As Douglas Self points out Tobey and Dinsdale seem to forget Mr Lin when accepting praise . Mr Lin himself mentions a colleague who inspired it . Lin got the biasing sorted out .
One story I know to be true is Mr Leak was approached over a weekend whilst the staff were at a hi fi show . Apparently he was crying the Monday morning afterwards . It had been bought for the then fortune of £3 000 000 over that weekend . Rank bought it . Apparently one of their staff loved hi fi . They were rich with Xerox money . Soon they abandoned the almost military standard of testing . Then surprise surprise reliability suffered . Next it was given to Roland Electronics / Rotel . The source of this Alan Tisdale who was his right hand man . Leak was a significant company . Out of the ashes was born Rotel as we know it , albeit not directly . Naim copied Leak almost to a tee , the way of testing espeacailly .
I am no lawyer . I can see T&D had no chance . The famous RCA amp with LTP ( Naim , Harman Kardon ) is called modified Lin . No offense to T&D and I accept the story is most likely untrue . The Byran amplifier was unmistakably theirs .
Oral History Lin Index RCA Germanium Transistors Audio
https://www.youtube.com/watch?v=49He8QwlAU0
Bryan Amplifiers - Anyone have any Information? - pink fish media
One story I know to be true is Mr Leak was approached over a weekend whilst the staff were at a hi fi show . Apparently he was crying the Monday morning afterwards . It had been bought for the then fortune of £3 000 000 over that weekend . Rank bought it . Apparently one of their staff loved hi fi . They were rich with Xerox money . Soon they abandoned the almost military standard of testing . Then surprise surprise reliability suffered . Next it was given to Roland Electronics / Rotel . The source of this Alan Tisdale who was his right hand man . Leak was a significant company . Out of the ashes was born Rotel as we know it , albeit not directly . Naim copied Leak almost to a tee , the way of testing espeacailly .
I am no lawyer . I can see T&D had no chance . The famous RCA amp with LTP ( Naim , Harman Kardon ) is called modified Lin . No offense to T&D and I accept the story is most likely untrue . The Byran amplifier was unmistakably theirs .
Oral History Lin Index RCA Germanium Transistors Audio
https://www.youtube.com/watch?v=49He8QwlAU0
Bryan Amplifiers - Anyone have any Information? - pink fish media
It doesn't take much to understand just how important testing is. Especially now, in this "brave" new world.
About 3 years or so ago, Wima, te German quality capacaitor maker, realized that they were going off the board. Their response was immediate - they reduced the size of their caps, they reduced the prices by quite a bit (if memory serves, they slashed off around 45% of the original price), but they also slashed they warranty, down from 2 years to 6 months.
Since I use a lot of their caps in my products, this was a mixed bag for me, and I believe others using their products as well. On the one hand, slashing prices like that is not good, not great, but stupenduous good news, but on the other hand, cutting their warranty is all bad news. I offer a 5 year warranty for my products, and now I beging to doubt one of my key suppliers. You don't just wake up one morning and say it's a great day, let's slash prices today.
So we (me and my associates) discussed it, and frankly, we were a little scared. We decided to extend our 48 hour or 2 day test period to 7 days. Statistically, if it works for 7 days under high load conditions, and comes out with its guaranteed minimum specs all above the norm, chances of its failing in the field are extremely small.
So what we saved on the price of the caps, we paid for extended testing, new test racks, etc. But we all still sleep soundly. In 11 years of operation, only one unit ever failed, and even that was traced back to a power amp which relied on fuses for "protection" - the amp burned, the fuses remained intact, in all fairness, the amp's fuses AND our internal fuses. Frankly, this is something I am very proud of.
We test components when they come in, not 1 in 100 or 1,000, but 1 in 10. We test and measure individual filter boards and we test the finished product. Sure, it costs time, space and money, but at least we are sure we are selling a fully functional product which works as specified. It does cost money, but we all, and especially myself, feel its money well spent. No buts, no ifs, no maybes.
Leak was a company with a rock solid reputation for the longevity of their products, all of them. After Rank took them over, they very quickly went into oblivion, obviously the Rank people had no idea of how to run an audio company. Even more obviously, they went about it by cutting costs first, and the first victim was Leak's testing procedures, which were said to be second to none. The net result was a cheaper product line, which soon became known for its poor reliability. This sunk the company, and if I were Mr Leak, I'd cry too.
They tried to bring them back once or twice, for example by rebadging a Rotel integrated amp (which, in my view, looked way better than the Rotel version of it), but it never caught on.
Oddly enough, I knew all this when I started out in 2000, and Leak was one of the examples I always had in mind when extensive and sometimes cruel testing procedures were introduced. But we are a small operation, so our logistics are relatively simple, and I always justified them by saying we're in this for the long haul, so we have to pay maximum attention at all times. 12 years down the road, and I think we should extend our test procedures to hit the 10 year warranty period.
If you look at an early Leak amplifier nothing about it suggests reliability . Equally some of the hastily made Sony's look perfect and were not . Sony were replacing ranges very quickly in the 1980's , no time to spot long term problems . The point is Leak tested every component and every sub assembly . Then every complete product . Rank gradually took it down to testing 1 in 10 amplifiers . They even sort of blamed the UK production staff for problems . Funny part of the story is that Tony Morpath took over the remnants of the Leak branded Rotel disaster . He sold it as Rotel UK ( Game Path Ltd latter ) . I got involved as the first Flat Earther as we were called ( those who prefer turntables ) . I sold it when no one else would and thought the Stan Curtice design RA820 astonishing ( still do ) . Apparently Tony was told by Rotel to sell more or loose the brand . He calmly said he could sell less of the rubbish , but not more . He then said " shame as it is well made rubbish " . Rather than get angry Rotel said if you are so clever why don't you tell us what to make . Apparently the box , PSU and sockets was predetermined . The PCB was the UK bit . Tenuously Rotel came out out of Leak . Reliability of those Rotel is second to none at any price . Stan played a bass guitar through it as he thought if it coped with that it would cope with real music . We debated capacitor choice the other day . That's how Stan did it I'm told . It saved money . I think he said he was very short of budget and wanted a few special components . He redesigned the PCB many times so as to avoid wiring . That got him the money for the parts . It sold in the UK for £79 inc VAT and 25% dealer margin . That was the same price as the discount houses so I could even give students 10% . Fuses were 95% of the repairs . One output transistor a rare exception . I can still remember the bias as 43 mV through 0R22 .
When Rank took over Leak the prices tumbled . Alas so did the reputation . Some how it never recovered . Quad must have been delighted as Leak did what they did better and cheaper . Leak strongly resembled American Acoustic Research in no nonsense science .
I think if done with care Leak would be a brand worth restoring . In fact via Rotel would have been OK . It might even be possible to look at the point where the brand stopped and pretend a logical path of improvement took place over the years . If it were me I would even keep the all NPN output stages . I would also make the TL12 plus in revised form . A superb amplifier not unlike the Heath designs using EL 84 .
Naim did an exact copy of good Leak practice . Julian Vereker even said it saved money as 5% components would do . The tolerances were laid out in bins and paired . Exactly as prototype and no great cost . He once joked to me that Naim had the lowest profit of all UK hi fi companies . He then said , I do spend more on research than them .
When Rank took over Leak the prices tumbled . Alas so did the reputation . Some how it never recovered . Quad must have been delighted as Leak did what they did better and cheaper . Leak strongly resembled American Acoustic Research in no nonsense science .
I think if done with care Leak would be a brand worth restoring . In fact via Rotel would have been OK . It might even be possible to look at the point where the brand stopped and pretend a logical path of improvement took place over the years . If it were me I would even keep the all NPN output stages . I would also make the TL12 plus in revised form . A superb amplifier not unlike the Heath designs using EL 84 .
Naim did an exact copy of good Leak practice . Julian Vereker even said it saved money as 5% components would do . The tolerances were laid out in bins and paired . Exactly as prototype and no great cost . He once joked to me that Naim had the lowest profit of all UK hi fi companies . He then said , I do spend more on research than them .
All of you who think there has been progress in Audio listen to this . Where is the hiss ? Where is the distortion ( allow a little and flutter ) . This is a direct transfer and is only a fraction as in data compressed form . I never bettered it .
https://www.youtube.com/watch?v=MPSSx4jYEmg
https://www.youtube.com/watch?v=EY7lvuVjjX4
Use headphones if you can , not computer speakers .
https://www.youtube.com/watch?v=MPSSx4jYEmg
https://www.youtube.com/watch?v=EY7lvuVjjX4
Use headphones if you can , not computer speakers .
- Status
- Not open for further replies.