Yes, this nonsense goes on year after year.
I remember in the BT thread, that I rebuild his 741 test circuit some 5 years ago,
measured it, got about the same spectra and proofed that what he claimed
to be FM distortion was ordinary IMD of order (9, 5) by stepping the signal
generators 10 Hz to the side and seeing the spurs move 90/50 Hz.
And getting the exact frequencies from the spectrum analyzer showed that
the math was OK.
At this time it is my opinion that this is outright fraud.
Gerhard
Last edited:
Gerhard,
Nice to see you have a sense of humor. I think several of the posts here were the first time I saw it.
On the topic another thing wrong with opamps is the marking. Laser etched gray on black in type sized to fit on an already too small package.
Now in opto isolaters there are some that come in white packages! So why can't we start getting opamps in color? Not only easier to identify the boards will look prettier!
Really would help to be able to tell some of those other parts are not really just grains of salt.
Nice to see you have a sense of humor. I think several of the posts here were the first time I saw it.
On the topic another thing wrong with opamps is the marking. Laser etched gray on black in type sized to fit on an already too small package.
Now in opto isolaters there are some that come in white packages! So why can't we start getting opamps in color? Not only easier to identify the boards will look prettier!
Really would help to be able to tell some of those other parts are not really just grains of salt.
Well, I'm surely confused by your explicit anger.
It sure seemed to me that you were saying something about picking one of your opamps! Of course, since I wasn't there at the show in 1979 I have no way to know what it is you were saying exactly. SURE SOUNDED like someone used one of "your" opamps to retrofit a board.
Guess not, guess you were saying that the board already USED "your" opamps. So, given that this was the case, WTF import does that have??
Who cares? Lot's of opamps are and were used in boards that all sorts of people "like" and use. Even today, right now.
And, a large proportion of these live shows sound horrific. Horrific for a variety of reasons, most of which probably have not much to do with opamps. They'd have to get more things "right" before anyone might notice the effects (assuming they actually exist) of opamps. SO, the idea that Fleetwood Mac used opamps in a board in 1979 is virtually meaningless.
If you had managed to make this clear, I would have asked (if I posted at all) what it was about this story that made sense WRT the topic.
So...
_-_-
EDIT: Curl quotes the year 1979, but it's rather irrelevant what year Fleetwood Mac used a board with opamps...
Last edited:
Ed: you could paint them. That way you could have your own colour scheme and also test the different paints to add to you list😀
Ok, so back to the bottom line... whenever you folks get tired of bickering like a bunch of old women (apologies to any old women who happen to be reading)?
SY, who is AWOL, and even I hope is ok, wherever he is... in a published article states that he could hear a string of 6, not 5 unnamed buffers.
Well, wouldn't it be interesting to do another experiment to see if this is a "magic" threshold, or unique to the particular part that SY picked?
I think it would be.
So, the idea would be to compare buffer X and buffer Y where buffer Y is swapped out in sequential tests with other buffers. Thereby gaining a group of data points.
The same thing could be done (as I have suggested a few times now) with opamps.
Wouldn't it be interesting to see if perhaps with one part that the number of parts in the string for being "indistinguishable" was substantially higher?? And how much they varied? Maybe a few were at "2" or "3" in series??
Forget what Curl's conjecture is, right or wrong, the question is IF there are audible differences - of course with proper implementations.
None of you "heavy hitters" here seem to really want to do much more than talk EE minutia, rather than to get to the pragmatic real-world.
_-_-
SY, who is AWOL, and even I hope is ok, wherever he is... in a published article states that he could hear a string of 6, not 5 unnamed buffers.
Well, wouldn't it be interesting to do another experiment to see if this is a "magic" threshold, or unique to the particular part that SY picked?
I think it would be.
So, the idea would be to compare buffer X and buffer Y where buffer Y is swapped out in sequential tests with other buffers. Thereby gaining a group of data points.
The same thing could be done (as I have suggested a few times now) with opamps.
Wouldn't it be interesting to see if perhaps with one part that the number of parts in the string for being "indistinguishable" was substantially higher?? And how much they varied? Maybe a few were at "2" or "3" in series??
Forget what Curl's conjecture is, right or wrong, the question is IF there are audible differences - of course with proper implementations.
None of you "heavy hitters" here seem to really want to do much more than talk EE minutia, rather than to get to the pragmatic real-world.
_-_-
Yes it would be interesting to do a properly conducted test, and as you know I have thrown a few together over the last few months (and years).
It needs someone else to pick up the mantle and produce some files. I also find that interest is often very limited (particularly in relation to the work involved), at least up until the point all is revealed. Also, those whose input could in many ways be considered the most valuable... can they tell or tell a difference, or can they not... never take part. That's a shame although I can understand the reasons why they abstain.
For those interested PMA (Pavel) has an immaculately presented comparison of two files over in this thread:
http://www.diyaudio.com/forums/everything-else/299168-abx-listening-test-link-stages.html
I've had four attempts and could not discern a difference. Anyone willing to give them a try ?
It needs someone else to pick up the mantle and produce some files. I also find that interest is often very limited (particularly in relation to the work involved), at least up until the point all is revealed. Also, those whose input could in many ways be considered the most valuable... can they tell or tell a difference, or can they not... never take part. That's a shame although I can understand the reasons why they abstain.
For those interested PMA (Pavel) has an immaculately presented comparison of two files over in this thread:
http://www.diyaudio.com/forums/everything-else/299168-abx-listening-test-link-stages.html
I've had four attempts and could not discern a difference. Anyone willing to give them a try ?
Bit of a shame, actually. I like parts you can tweak like that.
I guess it's not very popular any more because people are no longer expected to know what's going on inside the part, quite the contrary. We're living in the era of mystery blackbox devices, or actually have been since about the '80s when scant datasheets became all the rage for some reason, and you could only hope to learn about internals from some literature if you were lucky. Were manufacturers suddenly afraid that their stuff would be too easy to clone? Silly. Or was it simply considered too much work? Thankfully things have gotten more verbose again, though somehow some interesting key parameters still rarely or never make an appearance (I'm thinking common-mode / input impedance distortion, CMRR(f), and transistor datasheets never seem to spec Rbb').
Reminds me, still not seen Jay around...
I know they are big effort, but I really do appreciate all the tests you have done, even though they sometimes say more about people on this forum than on the test (as you noted).
The almost invisible (for us) marking is probably easy to see by a pick-and-place with IR camera. But the colors is a good idea. Most ICs can be identified by 4 digits, and a resistor-like color band or dot marking would be great.
Quick - call the patent office ! 🙂
Jan
Bear, you can't play chess with pigeons - at some point they will knock the pieces over, crap on the board, and fly back to their flock to claim victory. SilentlY or, in other cases, preparing their exit with an "you are an uneducated idiot that I will subsequently ignore" comment.
hands up those who aquired mil spec DIL parts (white ceramic with gold plated legs) for homers in the past.🙂
Changing the molding compound used in production isn't easy. A substantial amount of engineering work goes into using compounds which are robust over a wide range of temperatures and humidity levels (while also being RoHS compliant). Most products go through temperature cycling tests to try to expose delamination issues between the mold compound and the leadframe before they are qualified for mass production.
This gets even more difficult for high precision products where package stresses produced by the mold compound curing process can introduce errors. Or in ultra-low bias current amplifiers where leakages through the molding compound itself can degrade the amplifier's performance.
It's not just any old black epoxy...
It appears that he also interpreted the numbers in a certain way. John's contention is with the interpretation not the numbers.
I know you can't state it in public, but if we ever meet I'd love to hear what you really think about this thing "ROHS" 😉
John I appreciate that, but depositing a few colored dots would not jeopardize that, would it? It's extra cost sure but hopefully disappear behind the comma. I can dream, can't I 🙂
Jan
I think the point was mentioned once long ago and forgotten. There seems to be an opinion that the phase transition in the open-loop transfer function, the word quadrature was used (improperly of course), matters. That is for IM AM and FM sidebands are at the same frequencies but have a different phase relationship. A little math will show that that is not happening here.
The waters were muddied by claims, from trying to read a xerox of an old chart recording, that the distortion was actually in-harmonic. Which of course is impossible.
- Status
- Not open for further replies.