discrete opamp/ discrete output stage
Nov 27, 2008 at 4:43 PM Thread Starter Post #1 of 18

athenaesword

500+ Head-Fier
Joined
Aug 22, 2007
Posts
707
Likes
12
can someone briefly explain to me what the above are? I keep reading that discrete output stages eliminate the use of opamps, yet I read about discrete opamps; so what's the difference?

sorry if it seems really basic; but I've really no clue and it's hard to get anything specific with the search function.
 
Nov 27, 2008 at 7:22 PM Post #2 of 18

deltaydeltax

100+ Head-Fier
Joined
Nov 18, 2008
Posts
312
Likes
10
Discrete OP amps are made of transistors, diodes, resistors and other electronic components which are put onto a circuit board.

A monolithic OP amp is one you buy in an SOIC, DIP , or whatever package, such as an OP134.

The discrete ouptput stage is pretty much the same. So, you have transistors and other components instead of one package such as the LM3886 IC.
 
Nov 27, 2008 at 9:39 PM Post #3 of 18

nikongod

DIY-ku
Joined
Jan 24, 2005
Posts
8,882
Likes
124
Quote:

Originally Posted by deltaydeltax /img/forum/go_quote.gif
The discrete ouptput stage is pretty much the same. So, you have transistors and other components instead of one package such as the LM3886 IC.


I disagree with that.

An opamp by definition is a device with (practically) unlimited gain, and infinite input impedance. Its disadvantages come from the fact that you MUST have lots of feedback to select the gain you would like. An opamp can be converted to a nearly-zero input impedance for current-output DACs, but its an effort of still more feedback.

A discrete amplifier stage if well designed can overcome the perils of lots of negative feedback. There are differences to be sure. In the case of current-output dac chips the input of an amplifier stage can be made to be a perfect current input (zero input impedance)

At the end of the day, NO global feedback is probably not the answer, but 75dB of global feedback (around an opamp) is worse. Tastefully applied global/local loop feedback Is FAR better than a single high-feedback global loop.
 
Nov 28, 2008 at 1:33 AM Post #4 of 18

athenaesword

500+ Head-Fier
Joined
Aug 22, 2007
Posts
707
Likes
12
wow I wish I could've understood what you just said :S

I guess I'd better ask in specific examples. for example the Oritek OMZ modded zhaolu's discrete output stage. how is that different from a Zero DAC with modded HDAM OPA Sun V2, Earth and Moon opamps. and which is superior?

Quote:

Originally Posted by nikongod /img/forum/go_quote.gif
I disagree with that.

An opamp by definition is a device with (practically) unlimited gain, and infinite input impedance. Its disadvantages come from the fact that you MUST have lots of feedback to select the gain you would like. An opamp can be converted to a nearly-zero input impedance for current-output DACs, but its an effort of still more feedback.

A discrete amplifier stage if well designed can overcome the perils of lots of negative feedback. There are differences to be sure. In the case of current-output dac chips the input of an amplifier stage can be made to be a perfect current input (zero input impedance)

At the end of the day, NO global feedback is probably not the answer, but 75dB of global feedback (around an opamp) is worse. Tastefully applied global/local loop feedback Is FAR better than a single high-feedback global loop.



 
Nov 28, 2008 at 3:57 AM Post #5 of 18

Sovkiller

Proved that despite its huge size the CD3000 can be shoved down one's throat.
Joined
Jul 31, 2002
Posts
12,885
Likes
22
OPamp by definition...

Usually a discrete output stage are called those that are made of discrete components, like transistors, resistors, caps, whatever you need for the given design. Those give you certain freedom as a designer, as you could design and customize them for the given application or load, but also you have to be careful with tolerances and matching, specially if you want to achieve almost the same results for both channels...in case of stereo, or more in case of multichannels...
 
Nov 28, 2008 at 12:10 PM Post #7 of 18

cotdt

Headphoneus Supremus
Joined
Aug 29, 2005
Posts
3,220
Likes
15
Quote:

Originally Posted by athenaesword /img/forum/go_quote.gif
Do does my omz have opamps at all?


oh no, don't ask that question again. it will spark a huge flame war over the semantics of opamps
 
Nov 28, 2008 at 12:52 PM Post #8 of 18

athenaesword

500+ Head-Fier
Joined
Aug 22, 2007
Posts
707
Likes
12
Quote:

Originally Posted by cotdt /img/forum/go_quote.gif
oh no, don't ask that question again. it will spark a huge flame war over the semantics of opamps


My intention wasn't to spark anything. How can this be anything other than fact? If it has opamps it has opamps. If it doesn't it doesn't. If someone says that a discrete output stage has discrete opamps then that'll settle my confusion from what ppl have been telling me: discrete has no opamps, and yet there's such things as discrete opamps.
 
Nov 28, 2008 at 1:07 PM Post #9 of 18

majkel

Headphoneus Supremus
Joined
Jun 1, 2007
Posts
2,783
Likes
56
Some people don't accept the fact that many low distortion op-amps have the same schematic as some "one voltage stage" discrete amps, output stages, etc. It's a folded cascode architecture, with a diamond buffer or other class AB output stage, and a feedback loop connected to the input differential pair via a resistor divider.
The discrete op-amp resellers compare their designs to lower grade op-amps, then formulate statements that their circuits outperform the LT1028. Sure, especially that the ACN8 series sounds different and better than the lesser CN8 series. Same thing with other chips having A and B (better) series. People say it doesn't matter for audio but my ears tell me something different. Guess why there is not much talk about the OP27 here? First, because it sounds completely different when biased into class A, second, it's rarest versions exceed $100 per device, at least in Europe, while the cheaper GP series from Burr-Brown (these I know might sound great) are impossible to find these days.
 
Nov 28, 2008 at 1:58 PM Post #10 of 18

StanleyB1

Banned - Shares IP addresses with DC Lee Herandu
Joined
Jul 11, 2008
Posts
265
Likes
10
I must say that I am surprised how much a discrete stage output is pushed on headfi, which is then soon followed by suggestions as to which product with a discrete output stage is recommended. One would be easily fooled by such talk that discrete is better than opamp. That's like saying that electricity north of the equator is less noisy than electricity south of the equator.
 
Nov 28, 2008 at 2:52 PM Post #11 of 18

cotdt

Headphoneus Supremus
Joined
Aug 29, 2005
Posts
3,220
Likes
15
Quote:

Originally Posted by StanleyB1 /img/forum/go_quote.gif
I must say that I am surprised how much a discrete stage output is pushed on headfi, which is then soon followed by suggestions as to which product with a discrete output stage is recommended. One would be easily fooled by such talk that discrete is better than opamp. That's like saying that electricity north of the equator is less noisy than electricity south of the equator.


discrete output stages are pushed for a reason. that's because people here like how they sound.

have you ever actually listened to opamps vs. discrete analog stage? they sound completely different. opamps have VERY different properties as compared to discrete even if it uses the same internal circuit, and needs way higher negative feedback and typical audio opamps have much lower slew rate. as a result the sonic differences between integrated and discrete is immediately noticeable. actually the different opamps sound different from each other as well, and same applies for discrete.
 
Nov 28, 2008 at 3:20 PM Post #12 of 18

mrarroyo

Headphoneus Supremus
Joined
Jul 27, 2005
Posts
19,073
Likes
29
I have listened to both and IMO it is not a function of the op-amp vs. discrete. It is a function of the designer/builder.

A good designer/builder can make either sound great. A poor designer/builder will make either sound like $hit. Just my opinion based on side by side comparisons.

Kind of like apple pie! A good cook will make a great apple pie out of good apples. A bad cook will make a lousy pie out good apples. :) I like food analogies!
 
Nov 28, 2008 at 3:38 PM Post #13 of 18

Champ04

100+ Head-Fier
Joined
Jun 24, 2007
Posts
147
Likes
10
Quote:

Originally Posted by StanleyB1 /img/forum/go_quote.gif
That's like saying that electricity north of the equator is less noisy than electricity south of the equator.


I would have to totally disagree. I find that the electricity south of the equator is less noisy.
biggrin.gif
 
Nov 28, 2008 at 4:40 PM Post #14 of 18

StanleyB1

Banned - Shares IP addresses with DC Lee Herandu
Joined
Jul 11, 2008
Posts
265
Likes
10
Quote:

Originally Posted by Champ04 /img/forum/go_quote.gif
I would have to totally disagree. I find that the electricity south of the equator is less noisy.
biggrin.gif



I know that
wink_face.gif
. Let's leave the rest to figure it out...
 
Nov 28, 2008 at 5:51 PM Post #15 of 18

deltaydeltax

100+ Head-Fier
Joined
Nov 18, 2008
Posts
312
Likes
10
Quote:

Originally Posted by nikongod /img/forum/go_quote.gif
I disagree with that.

An opamp by definition is a device with (practically) unlimited gain, and infinite input impedance. Its disadvantages come from the fact that you MUST have lots of feedback to select the gain you would like. An opamp can be converted to a nearly-zero input impedance for current-output DACs, but its an effort of still more feedback.

A discrete amplifier stage if well designed can overcome the perils of lots of negative feedback. There are differences to be sure. In the case of current-output dac chips the input of an amplifier stage can be made to be a perfect current input (zero input impedance)

At the end of the day, NO global feedback is probably not the answer, but 75dB of global feedback (around an opamp) is worse. Tastefully applied global/local loop feedback Is FAR better than a single high-feedback global loop.



True. A lot of that makes sense to some people. The OP I think needed a much simpler answer. One of my favorite acronyms from school is KISS, "keep it simple stupid". I think the only way we may have confused the person more is by getting into holes and majority carriers.
 

Users who are viewing this thread

Top