Discrete vs opamps... - Page 6

Hold on a minute. Now we are talking about output stage? That's a completely different circuitry. I thought the discussion about discrete vs. monolithic was purely on opamp itself and has nothing to do with the output stage. (It is true that every monolithic opamp has an output stage but if the language here is audio I think you can pretty much ignore that "incompetent" output stage for lack of a better word) Of course if you buy a monolithic opamp from AD or TI it is understood that the opamp is not designed to have the output stage you want to have in audio applications. You can add additional circuitry to the opamp to make it do class A or whatever but from an engineering point of view this is obvious inferior to designing an opamp that does exactly what you want in the first place. Just pull up any opamp data sheet and you will not see any description on "optimized for audio amplifier only", therefore a discrete opamp will have everything an audiophile wants, including output stage, it is like making an opamp of your own, except you realize it not using silicon fabrication technology in a clean room on a wafer but with individial discrete transistors on a PCB. However, If a manufacturer such as AD or TI is willing to design an opamp for us, such as turning a discrete opamp design in audio amplifier to a monolithic chip then we can compare the two, here you maybe able to see the advantage of having laser trimmed transistors on silicon,
otherwise I don't think the comparison is 100% valid, maybe 97%.

THE OUTPUT STAGE OF THE OPAMP




The part that CAN NOT be made into class-A operation no matter how bad many wish it to be so.Not possible,can not happen get over it and accept reality.No matter what you hang on the output you will still have a Class A/B or class B output stage on the output of the actual chip itself and this WILL have an impact on the sonics *


no you can NOT add circuitry to "make it do class A".You can make the output Class-a TO A POINT then it switches back to AB for the rest of the signal or IT WILL burn up the output stage transistors of the opamp and once they go thermal runaway takes out the rest instantly .
At best you get 5mA to 7mA of class A from an opamp capable of 50 mA output current so that is FULLY 45mA of AB or less than 10% of the output in class A bias.

are you arguing a point just to do it ? It is totally obvious you have no clue even though I broke it down into plain non technical language so you are not understanding the obvious or you are intentionally being obtuse for whatever reason.

Again for the cheap seats.......................

YOU CAN NOT HAVE A CLASS-A OUTPUT I.C. OP-AMP !

there are none and there will BE none made in your lifetime.Deal with it

100% correct,100% valid,100% undiluted Rick

don't come to a gunfight with a gunslinger armed with a stick man.Not a good move

__________________________________________________ _______________
__________________________________________________ _______________

The * from above : see followup post coming right up

By adding circuitry to bias opamp into class A i meant this:

http://tangentsoft.net/audio/opamp-bias.html

Why do you insist the output stage of an opamp? I already said, quote myself that "(It is true that every monolithic opamp has an output stage but if the language here is audio I think you can pretty much ignore that "incompetent" output stage for lack of a better word)"

So I already ruled out the output stage of opamp. Because just like what you said, output stage of opamp cannot be made to do Class A. That is why when I saw the phrase "Class A" in the posts in this thread and everywhere in my work and social life, the thing came to my mind is Class A of the system, it is dumb to think "of the opamp"

As promised what is maybe to me the "coolest" Op-amp in current production,the AD744.

It specs out great and sounds fine "as is" for general audio use and even beterr with a discrete Class-a "current pump" add-on stage but what separates this from all others is the ability to toally bypass the entire output stage and use it purely as a JFET input front end to whatver output stage you can make work !

Check out Fig.28 :

http://www.analog.com/UploadedFiles/...516AD744_c.pdf


Get the TO-99 package,put a TO-5 heat sink on it,add a current source on the output to push this to pure class-a (if it can take it,don't really know off hand) then add on a totally discrete Class-a output and you just saved yourself from designing a front end.discrete is still better but this is better than all I.C. so there ARE choices if you know where to look and once finding them realise what you are looking at.A simple signal path is a good signal path if you are of the "use what you need to do the job and no more" thought group and this is a simple as it gets in the all the areas that count :

good dual fet input matching,a simple second stage (both appear to be class A) and the ability to toally choose what the output stage is to be.Like i said,pretty cool !


(credit for this part used in audio stages goes to Walt Jung formerly of ADI and the author of a damn serious amount of ADI linear app notes for audio specific circuits plus many audio DIY magazine and electronics magazine articles)

In my system design career I have used power opamps that can output tens to a couple of hundreds of watts. Voltage rating from 10 all the way up to 1000V, and current rating up to 30A or so.
So take that 50mA example of yours and change it to 5A.

My post in the "Audio Pet Peeves" thread in the member lounge :

so with that in mind this will be my last post on this topic since it is more than obvious what is going on here

the topic of this thread is

"Discrete vs. Op-amp" and on that which is superior and why.

ONE MORE TIME for those pretending not to understand :

A discrete Op-Amp is superior for audio use to ANY monolithic Op-Amp because with discretes you can have full class-a operation from intput to output and with an I.C. Op-Amp you can not.Class-a sounds better than class AB so the class-a discrete Op-amp will whip the I.C. opamp in SOUND every time hands down.

Simple and basic and nothing to "misunderstand" except by intent


on what basis ? Since when is the output stage not part of the opamp ? do you know something no other humans does not ? What is the point you are trying to make and why ?

I think i should refer you back to the quote at the top of this post.If you have no clue...............figure out the rest

.

So you are implying that biasing monolithic opamp into class A doesn't come close to a discrete opamp in terms of better sound.

It is interesting that you believe bias opamp into Class A is different than designing a discrete opamp in Class A.

said i would post no more but dude you have serious issues that need to be dealt with before you accomplish your goal here which is to confuse an already confusing topic for many members.You are doing it on purpose though why I do not know.Anyway :

SO ?

Pure class B and pure crap ! What does that have to do with audio and this particular discusssion ? Why bring it in if not to confuse the issue (sonics) in the minds of the non technical ?

It is far more interesting that someone that obviously has no clue is so willing to get into a pissing match with someone who does rather than go read damn book and learn something.

PARTIALLY biasing an opamp output into class-a does not make that Op-Amp Class-A and it seems everyone but you gets that so you need to chill or I will have to start editing posts since you objective seems to be to confuse others for no good reason.

This thread was supposed to be about information and understanding not your personal petty whatever it is you are up to so chill it or make some sense but do not post nonsense when ewcverything has already been fully explained to the satisfaction of you hence my beleif you are either trolling or thread crapping.

enough dude or bring something to the table that others can take something from and not this disinformation campaign you seem bent on pushing

rick out

ok, thread was supposed to be about information and understanding, so I am here to understand why every method mentioned in tagentsoft article is PARTIALLY biasing an opamp output, in other words, why is sourcing a constant current from an opamp not the Class A mode achieve in discrete design. Hope someone here can enlighten me on this. =)

It all depends on the ouput loading does it not?

If this is meant to be a serious question then i will answer it but let it be known it was answered MANY TIMES in this threa already and because of this side track we now have questions like this which is exaactly what I was trying to avoid and why your responses were starting to piss me off.

From clarity and full understanding to foggy and no clue is reverse info not info.

Read the Tangentsoft article very carefully and take note of at what mA point the op-amp output stage is biased at.At NO POINT is any Op-Amp either in real life or in the article biased up to the full output potential of the Op-Amp !

the first part of thew article deals with driving a buffer stage and the opamp running at 5ma clas-A by tying the N-CH resistor to the power supply rail through a current source,that means whatever the former output current capability was (35 mA for a OPA2134) it is now limited to 5mA in the class A region.If the output current goes above this point it goes BACK into Class-AB operation for the rest of the 35 mA so fully 30 mA is back to class AB.The only reason it does notr is because the output is not tied to a buufer stage for the output current but this TOO is not pure class A but AB usually up to some that are MOSTLY class A but none full class A.

take the initial OPA2134 and force the output to true class A operation (the full 35mA) and you will have no opamp to worry about.it will burn up.

The closest you can come is an opamp with a class-a bias to 5mA (and anything over back to AB) driving a to-220 buffer stage running 75-80 % class-A (and anything over back to AB) and STILL the all discrete all Class-A opamp will whip its a*s in sonics as was stated my initial post in this thread.this is not conjecture but proven beyond all doubt in every head to head test I have ever performed (plus most others find this to be true also)and why I do not use opamps anywhere in my home system.They are not worthy amd are totally reserved for portable duty where i have no use for class a operation anyway and my opamp of choice is a Class AB MONSTER that puts out gobs of current and still runs cool.no buffer,no jfet bias,no bells and whstles just straight in your face opamp power with sound "good enough" for the end use.

You can not drive a headphone with a class A opamp because not only are there none avbailable it could not even be made at that level of output current without having some means to heat sink the output section so it does not overheat and burn up.Simple laws of physics man and not muy rules just the rules

Hopefuly that was clear because to be honest this topic is boring me and there is no more for e me to post on anyway sooooooo

rick out

When op-amp chips are biased into "class A", they only stay in class A up to a point. Crank the volume up a bit and they drop out of class A into class AB or B operation. Only the first 10% or so of its maximum rated output is class A.

A discrete class A design stays in class A no matter how much you crank the volume. 100% of it's max rated output is class A.

Here's the problem, class A is very inefficient, you're lucky to get 5-10% efficiency, the rest goes up as heat. In other words, for a 1W output, you'll need at least 10W of power, which means large parts with big heatsinks. It doesn't matter where the volume dial is, the amount of power it sucks is proportional to the maximum rated output of the device, not how much power it happens to be putting out at the moment. The 1W amp may only be putting out 0.01W at a given time, but it'll still suck a full 10W.

To use an example, let's look at the OPA627. According to the specsheet, the max output voltage is 12V, and max output current is 45mA. That works out to about 0.5W for max output power. Meaning if you want to run it full class A, it'll need to dissipate at least 5W of power as heat, most likely it'll be closer to 10W. There is no way in hell a chip package the size of your pinky nail is going to dissipate that much heat without going up in smoke faster than you can blink. Look at the size & weight of a 10W resistor, look at the size of the chip. It's not happening.

As an "add on" :

My last few responses on reflection are a bit "snarly" and for that I apologise.No one needs me verbally beating them up in a public forum without just cause and maybe I assumed things not in fact.

In my defense ( ) I am at day four of the most wicked back pain I have ever experienced and just sitting at the keyboard is really kicking the rickmonster behind.no excuse but I just got busted for snapping at my wife and SHE is not taking that so realising I was not good company for someone I actually LIKE it came to me maybe i went a tad overboard with a stranger so for that please accept my himble apology and at this time I am taking my a*s to bed 'cause MY FKN BACK IS KILLING ME and no good will come of me posting in pain considering how I can be when I feel GOOD !

BTW-tomorrow if I remeber I will post a couple of very simple discrete opamp schemos with full commentary (not mine ) to flesh out what I attempted to say


G'night John Boy



Rick out.................................

It sounds like your back is running class AB at best, Rick. Dropping out to B?