Sound Quality Vs. Measurements

[jacco vermeulen]

their 627, 637 and similar op amps were always damn expensive, at least in Europe. And consequently, EXTREMELY hard to find.

Not so.

$30 each in 1991, if ordered at a local electronics store.
$17 each, if ordered at Thel audio in Germany, had an ad with prices and phone number in several magazines for years, e.g. (Holger Barske's) Klang & Ton.
Couple of years later, followed by Mr Rother's Schuro mail order business, still before the www days.

(1991 US$ exchange rate, AP type)

 

[dvv]

I might add a great period for the British film industry. They pushed the envelope. Oliver Reed, Richard Lester, Linsay Anderson, Peter Sellars etc. were involved in so many great films.

Yes, and that, as well. Lindsay Anderson made his "If", Ken Russel made "Tommy", "Women in Love", etc, even my favorite war movie (I'm generally not very fond of war movies) "Battle of Britain" was made then. As was Tony Richardson's "The Charge of the Light Brigade", and so on and on.

It was a time of an intellectual awakening, and we all so naively thought we could change the world for the better.

 

[dvv]

I was thinking mainly of the current noise, MM really wants FET's.

I figured that was what you were getting at, Scott. I certainly won't say you're wrong or unreasonable, but if memory serves, there weren't too many Burr-Brown FET front end op amps in the economy class at the time - remember, that was published in 1990.

And quite obviously, it had to a Burr-Brown op amp, which rules everybody else out.

Personally, I'm glad they didn't. My most hated op amps are from BB, their 134 family; it has come to the point where I can actually pick them out by their sound signature inside some device say 7 or 8 times out of 10. They sound shrill and irritating to me.

Oddly enough, one of my most favorite op amps is also from BB, the venerable 2604 - now, that sounds really sweet, it stands apart quality wise, and I use it gladly.

Go figure.

 

[dvv]

Not so.

$30 each in 1991, if ordered at a local electronics store.
$17 each, if ordered at Thel audio in Germany, had an ad with prices and phone number in several magazines for years, e.g. (Holger Barske's) Klang & Ton.
Couple of years later, followed by Mr Rother's Schuro mail order business, still before the www days.

(1991 US$ exchange rate, AP type)

$30 is cheap to you?

And remember, $1 was worth more then than it is today. And we had less of them then.

Also, I said "hard" to find, not "impossible".

What I was getting at is the fact that when talking to Americans, you need to point out that they pay for these parts at worst 1/2 of what we do, more typically 1/3 or even 1/4.

The same is true of you, for example. At the time, that may have been a viable option for you, but not for many of us - on its way from say Germany to me, its price would go up by 10% customs duties and 18% for VAT, or simply multiply its cost plus shipping by 1.29. So, what cost $30 in Germany would cost me say $35 with shipping, and ultimately $45 at home.

That's DAMN expensive even today, let alone those days. To me, at least.

 

[nigel pearson]

Films

Modest-power tube amps and horns. Check out the history of Western Electric, among others. My refs are buried in storage or I could be more specific.

One problem with all film optical audio tracks, even relatively recent ones: they wear out in a hurry. That's one of the reasons DTS is a continuing success, supplying audio that merely has to be synchronized to the film. And the "flywheel effect" of their servo system is significant, so some fairly large dropouts can be readily accomodated.

Thanks for that . Lee de Forest started working on it about 1923 . Kellogg of Corn Flakes fame also ( Paul Voigt ) . My feeling is all hi fi is dirived from cinema . The man from Star Trek may be related .

 

[nigel pearson]

Booster

I took the liberty of lifting this from basic amp design thread running now item 31

This idea is often in text books . I have tried it with feed forward cap . Anyone ever succeeded with this as a booster ? 51 R is the usual text book feed forward .

 

[dvv]

I took the liberty of lifting this from basic amp design thread running now item 31

This idea is often in text books . I have tried it with feed forward cap . Anyone ever succeeded with this as a booster ? 51 R is the usual text book feed forward .

What liberty? Everything posted here by me at least is public domain.

Question - does it HAVE to be a single line power supply?

 

[gk7]

... remember, that was published in 1990.
...

There was an earlier version with two OPA37s and passive RIAA published
in a now defunct austrian electronics magazine in the 80s. (I know because
that´s what I used at that time). The OPA627, 637, 606 and the likes did not exist yet.
Ronald Reagan was president in the US back then and it was not allowed
to sell such stuff to the eastern block. That´s probably why Burr Brown and
similar companies had representatives in Vienna ;-) (The OPA37s I have are "military grade" because these where the only ones they had in stock...)

 

[nigel pearson]

What liberty? Everything posted here by me at least is public domain.

Question - does it HAVE to be a single line power supply?

Hi D V V . It is split supply but that's his idea not mine . I was thinking use feed forward but not global feedback ( one can use it naturally as belt and braces ) . Here is a good version of feed-forward .

http://www.diyaudio.com/forums/solid-state/74861-single-darlington-line-preamp.html#post856982

 

[john curl]

That circuit is a 'dog' that was first used in the Studer A80 line amp, and measures with LOTS of Xover distortion. The normal R5 is 100 ohms. I made a 'current dumping' version of it that measured 10 times better, and offered it to Studer, in fact Willi, himself, in 1975, but they ignored my measurements, that were backed up with Polaroid pictures of the distortion.
Haven't any of you tried a good, clean, but easy enough to make, discrete design for serious recording?

 

[jacco vermeulen]

$30 is cheap to you? And remember, $1 was worth more then than it is today.

The US$ value in the period after desert storm was at a near identical level as now, actually.

In the early '90s, an LT1028 had a similar price level. Fancier AD opamps then were not exactly cheap by today's standard either.

All things relative, NE5532/4 then cost several times the current rate.
Power parts as Hitachi J50/K135 were ~$9 each, Toshiba A1095/C2565 ringemitters did a dollar less.
Compared to current average price levels, Motorola MJ15024/25 and similar were the real expensive parts in those days.

Import of US goods here now is worse than 20 years ago.
19% sales tax, plus 10% import tax (as countermeasure for the US not swallowing the Brussels trade agreement) , that would make ~31% added.
That's where Dutch customs pull their own hat trick, charging the 31% over the declared value in US dollars, but charging the number in € currency.
Effectively +70% added at the current exchange rate (31% taxes x 1.30 $/€), so order $100 in parts and end up paying $170 and change.
Screw-EU !

In my lean student days i bought $500 average per year in electronic components, couple of years ago it was at a level of 1k a month.
Depending on how to value parts, i have $50k to way over $100K in amp components on the shelves.
The $60 they charged me for two OPA627 in April 1991 were worth the extra cash, compared to a $6 each PMI OP27.
Personally, i find "expensive" hard to define nowadays, a power transistor is expensive if it costs more than a buck fifty, while folks gladly pay several dollars for a JFET that cost me 10 cents for years on end.

 

[SY]

That circuit is a 'dog' that was first used in the Studer A80 line amp, and measures with LOTS of Xover distortion. The normal R5 is 100 ohms. I made a 'current dumping' version of it that measured 10 times better, and offered it to Studer, in fact Willi, himself, in 1975, but they ignored my measurements, that were backed up with Polaroid pictures of the distortion.
Haven't any of you tried a good, clean, but easy enough to make, discrete design for serious recording?

Well, it's not 1975 any more- you can make a MUCH better performing circuit with a monolithic IC.

 

[simon7000]

Granted that lytics can have terrible soakage and forming issues, at equilibrium decent ones are pretty much settled down, with the current piecewise-linearly dependent on applied voltage. At low-enough voltages the I-V curve really does pass through the origin. So what is the worst-case d.c. offset between the emitters of Q1 and Q2? That will determine the current through the capacitor, and piling on the terms to determine the offset, I agree with Scott that you just can't conjure all that many millivolts, nor all that big a transient if VR1 even goes from a short to an open. For the more likely failure of wiper lifting, but track continuity preserved, the effect is even smaller.

I think of the conjectures so far, the notion that R14 sets a minimum gain and trims variations in VR1 out (which means the schematic should have an asterisk beside the value), and possibly that it has an effect on the effective VR1 taper, are the most plausible.

Actually if you make the pot 100K and R14 10K then the circuit makes sense. It gives better feel to the gain trim control. So I suspect this is really an argument about a typo.

 

[simon7000]

Someone from a very large telecom firm insists that the dark current in a photodiode is in the dark at 0V, i.e. the thermodymanically impossibe result. Same is true of a capacitor at equilibrium.

But Scott when I use my Simpson Model 260 V.O.M. it does read zero! Surely that is enough reference for you!

Then there was the follow who asked if I could repair a V.O.M. someone gave him as it didn't work right. I cut the jumper across the meter that was installed for shipping purposes. Hmm wonder what that big red tag that came on the meters face was all about?

Actually one of the interesting effects with older wound electrolytics where the foil extended out the ends for connection was that if the was a temperature differential they could produce voltage!

 

[nigel pearson]

That circuit is a 'dog' that was first used in the Studer A80 line amp, and measures with LOTS of Xover distortion. The normal R5 is 100 ohms. I made a 'current dumping' version of it that measured 10 times better, and offered it to Studer, in fact Willi, himself, in 1975, but they ignored my measurements, that were backed up with Polaroid pictures of the distortion.
Haven't any of you tried a good, clean, but easy enough to make, discrete design for serious recording?

I was always building versions of it . Never much success . Was very good as a motor drive where it seemed very stable .

 

[john curl]

It was designed that way to REDUCE output stage heat. They used them with up to 24 tracks, so a Class AB output stage would add extra heat. As a motor drive, it should be superb!
All you, or SY has to do is to build it and measure it under 600 ohm load, that it was designed in for. I had MORE distortion contribution from the line stage, AND higher order too, at -10Vu, than the tape, itself at that level. I have been looking for a suitable design that I used for Crystal Clear recording and Dave Wilson (Ultramaster) etc. I will post 'something' to get started, until I find a clear copy. Then we can discuss what is possible, and 'successful'. (Beats a 5534!) '-)

 

[SY]

In this application, the chip to beat is THAT1510/1512, not the 5534.

 

[john curl]

Sorry, all I can find at the moment is a 'working' copy from 1977.

 

[john curl]

Thanks SY, does every manufacturer know this?

 

[SY]

They are insanely popular, yes, except in the cheap gear. Same with the very similar SSM2019. I think TI also has something competitive (spent the weekend with their apps guy so I'm embarrassed that I don't know the part number). Really excellent CMR, better than the transformers I usually use. 1nV/rt Hz voltage noise at 80dB gain.