Don't worry, there is going to be an eclipse of the sun and we will be saved.
Anyway, and even compared to that of Bruno P. your communication is 99% poetry free.
Last edited:
So they, including a veteran electronics designer, resort to peeking.
What do you suggest the listening setup should be so that they can form their own opinions? You've been suggesting the listening for a while but still withhold the setup method. It's time you disclose it.
Aren't you the one who's been forcing your own DACs and cables beliefs on others here?
Scott, I must say I have no clue what so ever what you opamp designer people are stuffing into dill critters etc., so I wonder for instance how large are the transistors in terms of physical size and/or power handling around the input/feedback node stage in opamp's comparing to discrete's like say perhaps BC550/560 or other small signal bjt's, or fet's, next I wonder do you opamp designer guys consider the Lavardin stuff when looking at optimizing the input/feedback stage, am considering now mainly parts oriented towards audio use.
Am I being dim here? I was not aware that low drop out was a requirement that walt used when designing his superregs? There may be some Venn diagram overlap but I would never assume it.
Let me guess no measurements?
Plural of anecdote is still not data.
It's certainly not a high profit application, in the opinion of semiconductor manufacturers.
I was talking specs. I should have added noise to the other two parameters I mentioned.
I was based in Asia for 10 years with a large semiconductor house. You’d be amazed at the prices. And as for margins . . .
The term superreg seems to be one of those that has become (sometimes) overloaded, at least I have occasionally seen it used to refer to 'super' LDOs by some people. I more often call them 'high performance LDOs' but not everyone uses that terminology either.
Last edited:
Well, modern regulators are designed for 0 ESR ceramic caps and they probably should have X7Rs all over the place on the 3V3 rail. Obviously, more care can be required than with ancient regulators that don't care and have low bandwidth.
Did you not read my reply to your post several pages back? Almost all of those TI audio op-amps are rebranded or binned versions of general purpose op-amps.
Yes, he is guilty of exactly what he complained about.
Last edited:
May-I suggest-you as most of the things we 'believe' in, even in the domain of science ,has not been verified by ourselves, they are "anecdotal" ?
So why are-we using-it with such a confidence ? As use to say J.C. "It works !"
Here's the Linear Technology LT1115 whose datasheet may be accessed through google. Click on the image to remove foreshortening-distortion and to see it full size.
The little square boxes around the perimeter of the chip, with black circles inside, are "bondpads". They are connections between the die and the outside world. Each bondpad is approx. 5 mils by 5 mils (125 microns by 125 microns).
The set of four comb structures on the left, are the input transistors. I would say they appear to be 2 bondpads x 3 bondpads in size, i.e., about 150 microns by 375 microns. That is a VERY LARGE transistor for sub-100mA analog circuits. But the input transistors need to be large, for good matching and for good noise performance.
The little square boxes around the perimeter of the chip, with black circles inside, are "bondpads". They are connections between the die and the outside world. Each bondpad is approx. 5 mils by 5 mils (125 microns by 125 microns).
The set of four comb structures on the left, are the input transistors. I would say they appear to be 2 bondpads x 3 bondpads in size, i.e., about 150 microns by 375 microns. That is a VERY LARGE transistor for sub-100mA analog circuits. But the input transistors need to be large, for good matching and for good noise performance.
Obviously not the LM parts and not the ones listed by TI as audio parts. They have different numbers because they have differences! They did mention they have a new secret bit to make better parts of some apparently not yet introduced chips.
Parts tweaked for audio use can of course be used elsewhere.
Perhaps the only parts squarely aimed at audio only are the Muse series and the THAT ones, but Bill did exclude the Muse ones from his finger count.
Bill, I think you got our communication exactly right.
Mark,
Although the LT1115 is currently marketed as an Audio chip I suspect it was not purpose built as one. The do market it as general purpose industrial.
When National was designing their audio chips they actually set up a listening room and some were a bit surprised that they could actually hear differences.
I suspect a good determination of audio intent design would be actually having designers listen to the parts behaviors.
ES.
Although the LT1115 is currently marketed as an Audio chip I suspect it was not purpose built as one. The do market it as general purpose industrial.
When National was designing their audio chips they actually set up a listening room and some were a bit surprised that they could actually hear differences.
I suspect a good determination of audio intent design would be actually having designers listen to the parts behaviors.
ES.
Not all of those TI parts have differences. Some have different part numbers because they simply don't meet DC specs. I don't understand why you don't get the point of this discussion. The point is, and what Bill said, was that there are few op-amps designed specifically for audio. Just because TI can tweak some other op-amp very slightly and give it a SoundPlus datasheet does not make it designed for audio.
That implies paying for changes due to these listening tests where cause and effect are blurred or as they say "unmeasurable". These changes are $100's thousands each and weeks of delay.
Ed read the number in the corner of the LT1115 it's 1028, the same goes on everywhere.
Why don't you just look at the datasheet and compare for OPA211 and OPA1611/OPA1612 as one example. Or THS6012 and TPA6120. I'm sure there are more but I don't care to go dig through tables to find matching specs.
There are some unique ones, I will not dispute that. OPA1622 for a recent example. I don't think this changes the general point. Claiming you could count them all on-one-hand is an exaggeration.
I do. He's currently full of gin on a plane somewhere, but I will ask. But like Scott says, unless there is a market for a minimum of 10^7 of a part or a $100 price tag no one is going to do a mask spin for audio. I might be wrong but I think very little is done on 6" wafters these days and an 8" mask set is scary expensive.
- Status
- Not open for further replies.