My evil suspicion.
I have built an LM3886 amp.
I have built a few discrete amps mostly of the "blameless" variety.
At the present state of things, I believe that the discrete approach can (I say *can* not always *does*) exceed the performance of any intergrated chip currently available. However, to be as objective as possible I doubt that under normal circumstances (2- or 3- way nominally 8-ohm boxed speakers and pre-amp or source with normal output impedance in an ordinary sized room) I could really detect the difference. Maybe with very closely conducted ABX testing I could, if boredom didn't set in.
Why then do I wast time and frustration making my own PCBs and fighting oscillations? Probably because lower distortion is "cool" even if pointless and because the process IS harder and requires me to learn stuff that doesn't come naturally. Maybe as a complete klutz and ingoramus when it comes to metal work it diverts my attention from the fabulous enclosues some folks seem to turn out with such ease.
I have built an LM3886 amp.
I have built a few discrete amps mostly of the "blameless" variety.
At the present state of things, I believe that the discrete approach can (I say *can* not always *does*) exceed the performance of any intergrated chip currently available. However, to be as objective as possible I doubt that under normal circumstances (2- or 3- way nominally 8-ohm boxed speakers and pre-amp or source with normal output impedance in an ordinary sized room) I could really detect the difference. Maybe with very closely conducted ABX testing I could, if boredom didn't set in.
Why then do I wast time and frustration making my own PCBs and fighting oscillations? Probably because lower distortion is "cool" even if pointless and because the process IS harder and requires me to learn stuff that doesn't come naturally. Maybe as a complete klutz and ingoramus when it comes to metal work it diverts my attention from the fabulous enclosues some folks seem to turn out with such ease.
Hey JCoffey, Without the wife you can now make huge loudspeakers!!! Better get going on it.... 😉
Well if yall perfer
discreet to IC based amps , take a look at the APEX PA01-PA02 series.Fanstastic specs but a little costly, made of a buncha teeny tiny discreet components (thats why the cost).Its gonna be be the base fer my next amp. I like the LM3875 but hay if there is something better out there ya gotta try it.Maybe the flack will lessen if the chip cost more?
ron
discreet to IC based amps , take a look at the APEX PA01-PA02 series.Fanstastic specs but a little costly, made of a buncha teeny tiny discreet components (thats why the cost).Its gonna be be the base fer my next amp. I like the LM3875 but hay if there is something better out there ya gotta try it.Maybe the flack will lessen if the chip cost more?
ron
Re: Re: At the risk of offending everyone...
I have to agree with peranders.
Discrete design is VERY satisfying. However, I don't believe that the design is compromised in any way. Just as there are good chip amp designs, there are also good discrete designs. The advantages of discretes are that components can vary produciton process and vary in tolerances in the same design. Whereas in a chip design there are compromises in the desing inherent in the process used to produce the chip. On the other hand, the chip design has advantages as well. Thermal stability due to all components seeing for the most part the same temperature could be use to increase thermal stability (If designed correctly).
Personally, I believe that each topology has advantages that the other does not. Myself, with a preference for high rail voltage find chips do not easily accomidate my needs. Thus I design discrete. However, when I need a smilple low voltage amp, I use a chip op amp every time. I'd never even think of trying to design one. (Ok, I've thought about it... ...maybe some day 🙂 )
I think that you've offended no one. Rather, you've only clued us into your bias in terms of amplifier designs. (Lack of openmindedness ???)
-Dan
peranders said:
I have to agree with peranders.
Discrete design is VERY satisfying. However, I don't believe that the design is compromised in any way. Just as there are good chip amp designs, there are also good discrete designs. The advantages of discretes are that components can vary produciton process and vary in tolerances in the same design. Whereas in a chip design there are compromises in the desing inherent in the process used to produce the chip. On the other hand, the chip design has advantages as well. Thermal stability due to all components seeing for the most part the same temperature could be use to increase thermal stability (If designed correctly).
Personally, I believe that each topology has advantages that the other does not. Myself, with a preference for high rail voltage find chips do not easily accomidate my needs. Thus I design discrete. However, when I need a smilple low voltage amp, I use a chip op amp every time. I'd never even think of trying to design one. (Ok, I've thought about it... ...maybe some day 🙂 )
I think that you've offended no one. Rather, you've only clued us into your bias in terms of amplifier designs. (Lack of openmindedness ???)
-Dan
Variac said:
Hehe, Ya, I was actually planning on it, but now I have a gf, and um, as much as she loves toys like that, she still doesnt let me spend any money on them. LOLWell, there is a downside on the Gainclone that you can't deny, the lack of power and ability to drive 4 ohm speakers with low dB/W
When driving my Dynaudio Contour 3.3 with my gainclones, well you just mis everything wat is just so good about these speakers.
I am not saying that when playing at "very" low levels the amp doesn't do its job, it does it perfektly, but knowing that the speakers only come alive with amps >200W makes the gainclone useless.
Still i find it a great project with lots of potential, but it just doesn't fit every speaker that there is on the market, and this is probebly the reason why manufactures don't use this chipamps.
Just my 2 cent
Greetz
When driving my Dynaudio Contour 3.3 with my gainclones, well you just mis everything wat is just so good about these speakers.
I am not saying that when playing at "very" low levels the amp doesn't do its job, it does it perfektly, but knowing that the speakers only come alive with amps >200W makes the gainclone useless.
Still i find it a great project with lots of potential, but it just doesn't fit every speaker that there is on the market, and this is probebly the reason why manufactures don't use this chipamps.
Just my 2 cent
Greetz
moe29 said:So dispel the doubts, have a go. And then be as amazed as the rest of us. These little blighters actually play music. Then you will no doubt be joining us. And we shall welcome you with open arms.
sound like a Cult!
That what some might say when come across what they don't understand or don't want to understand. Accusations like that through history have nearly always been the results of ignorance. Even early christianity was labelled a Jewish sect and cult. What about Single-Ended Triodes? A cult? What about those who lampoon us for listening to tubes? A cult? What about those who knock CDs? A cult? What about those strange people who still listen to vinyl? A cult?
Just because they don't agree or understand, that is no good reason. I am reminded once: I read that one way of boasting was by putting other people down or even humiliating them. Because, in a relative way it elevates one person above the other, it is boasting or bragging.
That is quite right!
And then beyond the simple, there is actually much that can be done to further extract several more levels out of them.
Joe R.
Joe, I see you mentioned a regulated PSU in your amp, well I'm about to implement the exact same thing. I know of only one other member that has implement one and documented it on his site (pedja I do believe).
I was wondering, do you have a thread on his anywhere? If not, care to elaborate on your regulator? I'm just looking for ideas before I get to work on mine...
I was wondering, do you have a thread on his anywhere? If not, care to elaborate on your regulator? I'm just looking for ideas before I get to work on mine...
reading between the posts..
Well now, quite a little hornet's nest got stirred up. Observing the posted replies there are clearly two "teams": the highly defensives and the quietly concurring.
So let's put it this way: anyone wanting a guaranteed, functional and uncomplicated amp to play around with: go ahead and "chip" away at the IC solutions. But besides PSU, multiple forms of buffering, silver wire, P2P connections etc, there simply isn't that much you can fiddle with. The chip designers have deliberately made such experimentation impossible. But that's no reason to get narrow-minded and defensive because somebody points out these real and insurmountable limitations of chips - vide infra.
Those who wish to try tantalum or Holco signal resistors, matched (not to mention truly complementary) transistors, increased O/P stage bias, Elna caps, Black gates or even more fundamental circuit surgery such as NFB adjustment: you know who you are and how successful the results can be. Absolutely NONE of this is possible with a chip! Except the PSU as mentioned.
So what's the problem then... two happy camps ostensibly. Well, given the constraints of working with chips which are undeniable, please don't create the impression to the uninitiated that they represent the holy grail of diy-fi. They can't come close. They suck into low-Z loads. The treble is harsh and edgy. Bridging doesn't help. They are as advertised: well-designed and ecomomical but inherently, unavoidably and INTENTIONALLY compromised building blocks aimed primarily at the lower end of the consumer audio market. No more. No less. No question. No magic.
And as far as pedantic analogies with discrete Pentium processors go: hardly comparable, but you may be disappointed to learn that computers evolved as discrete designs... using tubes no less!
To analog_sa I would say the following, offered as 100% subjective opinion on my part: there are no high-end amps using chips. Let me re-phrase: I do not believe a chip-derived power amp can legitimately be labelled high-end, no matter who whacks it together. And yes, I am aware that some manufacturers do so, usually those on the fringes of "high-end" or those extending the lower end of their model line-up, particularly with surround-sound amps. And always the overriding motive is... you guessed it... cost!
Well now, quite a little hornet's nest got stirred up. Observing the posted replies there are clearly two "teams": the highly defensives and the quietly concurring.
So let's put it this way: anyone wanting a guaranteed, functional and uncomplicated amp to play around with: go ahead and "chip" away at the IC solutions. But besides PSU, multiple forms of buffering, silver wire, P2P connections etc, there simply isn't that much you can fiddle with. The chip designers have deliberately made such experimentation impossible. But that's no reason to get narrow-minded and defensive because somebody points out these real and insurmountable limitations of chips - vide infra.
Those who wish to try tantalum or Holco signal resistors, matched (not to mention truly complementary) transistors, increased O/P stage bias, Elna caps, Black gates or even more fundamental circuit surgery such as NFB adjustment: you know who you are and how successful the results can be. Absolutely NONE of this is possible with a chip! Except the PSU as mentioned.
So what's the problem then... two happy camps ostensibly. Well, given the constraints of working with chips which are undeniable, please don't create the impression to the uninitiated that they represent the holy grail of diy-fi. They can't come close. They suck into low-Z loads. The treble is harsh and edgy. Bridging doesn't help. They are as advertised: well-designed and ecomomical but inherently, unavoidably and INTENTIONALLY compromised building blocks aimed primarily at the lower end of the consumer audio market. No more. No less. No question. No magic.
And as far as pedantic analogies with discrete Pentium processors go: hardly comparable, but you may be disappointed to learn that computers evolved as discrete designs... using tubes no less!
To analog_sa I would say the following, offered as 100% subjective opinion on my part: there are no high-end amps using chips. Let me re-phrase: I do not believe a chip-derived power amp can legitimately be labelled high-end, no matter who whacks it together. And yes, I am aware that some manufacturers do so, usually those on the fringes of "high-end" or those extending the lower end of their model line-up, particularly with surround-sound amps. And always the overriding motive is... you guessed it... cost!
While I love the creative part of diy-audio...it is not about the creative part... It is about building that ultimate system that will provide you with beautiful music...that does it for me
If the gainclone provides that..then I might just investigate... but not until a zillion other projects are finished though...
If the gainclone provides that..then I might just investigate... but not until a zillion other projects are finished though...
Re: reading between the posts..
I would be tempted to speculate that using tiny IC package for the whole amp's circuitry has more advantage than possible use of even best audiophile parts in a discreet circuit. I'm also beginning to discover that so called "audiophile" parts are not really the ultimate solution to the best sound. There are probably not so many good sounding chip amps around, but the one being most popular, doesn't really require any tweaking with bias or output stage substitution. Whatever the chip offers is good enough for careful, subsequent application.
Since you built 2 of those GC circuit, which was your other amp that easily outperfomed them in your system?
DrG said:
I would be tempted to speculate that using tiny IC package for the whole amp's circuitry has more advantage than possible use of even best audiophile parts in a discreet circuit. I'm also beginning to discover that so called "audiophile" parts are not really the ultimate solution to the best sound. There are probably not so many good sounding chip amps around, but the one being most popular, doesn't really require any tweaking with bias or output stage substitution. Whatever the chip offers is good enough for careful, subsequent application.
Since you built 2 of those GC circuit, which was your other amp that easily outperfomed them in your system?
DrG said:
I don't think they missing anything. They just " find chip amps not as DIY as they like, but then that is just a personal taste - after all they don't make their own transistors, for example" (If I can quote one of them, not listed here)😉
Dear DrG
No amp is the Holy Grail, the Holy Grail only exist in stories. I really don’t see your point. All amps are compromises, whether it is intentional or not. It is the art to make the right compromise(s) and if that compromise is a cheap but good chip amp? So what? I have heard various gainclones and original gaincards and they sound good in the right place but don’t drive a BW Matrix or other power hungry LS with it. But that is an open door.
It seems to me that you can’t stand (young) people learning and gaining knowledge and have fun with their hobby and more important are proud of what they actually got working. Please be a bit more merciful to those who know not all. If you feel not comfortable with the “Gainclone club” why are you trying to stir them up? Let’s guess, you have a mission............
Regards
Feeling left out...
DrG: I'm almost insulted that you've completely bypassed my point about studio monitors and chip amps. But I can guess why you've chosen to ignore it.
Both the well respected Genelecs and the highly acclaimed Mackies use ICs in their studio monitors. These chip amped speakers are not "toys" as you would have us believe. Most consumers, including many so-called golden ears don't even know of them, much less know that a good chunk of material is monitored on them. (My apologies to those that do.)
The question is: If ICs are so limited, then why do the engineers that make the sounds you listen to think they are plenty good enough for 16 hour mix and master sessions? Or put another way, are all these professionals somehow wrong?
🙂ensen
With a pair of monolithic chips on my shoulder... heh!
DrG: I'm almost insulted that you've completely bypassed my point about studio monitors and chip amps. But I can guess why you've chosen to ignore it.
Both the well respected Genelecs and the highly acclaimed Mackies use ICs in their studio monitors. These chip amped speakers are not "toys" as you would have us believe. Most consumers, including many so-called golden ears don't even know of them, much less know that a good chunk of material is monitored on them. (My apologies to those that do.)
The question is: If ICs are so limited, then why do the engineers that make the sounds you listen to think they are plenty good enough for 16 hour mix and master sessions? Or put another way, are all these professionals somehow wrong?
🙂ensen
With a pair of monolithic chips on my shoulder... heh!
Dare I say it?
It's easy to make a bad dicrete amp, but much more difficult to make a bad chip amp.
I use chip amps for work. IMO they outperform a great number of discrete amps.
Purplepeople. I agree about the Genelecs. They are so clear and pure in studio, but they as such they're unforgiving, and I wouldn't want them at home.
Perhaps the argument is not what's best, but what you prefer in a given situation.
Cheers,
It's easy to make a bad dicrete amp, but much more difficult to make a bad chip amp.
I use chip amps for work. IMO they outperform a great number of discrete amps.
Purplepeople. I agree about the Genelecs. They are so clear and pure in studio, but they as such they're unforgiving, and I wouldn't want them at home.
Perhaps the argument is not what's best, but what you prefer in a given situation.
Cheers,
Re: Feeling left out...
Hehehe.
Personally, I don't understand why anyone would care how someone ELSE goes about their pursuit of the enjoyment of reproduced music. This sort of elitist absolutism, that if you're not enjoying things the way *I* enjoy them, then there must be something wrong with YOU, displays little more than intolerance of the individualism and diversity that makes this hobby what it is.
People should be ENCOURAGED to try different things for themselves and make up their own minds in the end as to what gives them the greatest satisfaction. And whatever that choice may be, we should all say nothing more than more power to 'em!
Vive la difference! 🙂
And on that note, I'm off to the state fair to partake in some beer drunk from piture size plastic cups and get myself a jumbo serving of deep fried screaming 2 year olds on a stick. Have a safe trip back home, SY!
se
purplepeople said:
Hehehe.
Personally, I don't understand why anyone would care how someone ELSE goes about their pursuit of the enjoyment of reproduced music. This sort of elitist absolutism, that if you're not enjoying things the way *I* enjoy them, then there must be something wrong with YOU, displays little more than intolerance of the individualism and diversity that makes this hobby what it is.
People should be ENCOURAGED to try different things for themselves and make up their own minds in the end as to what gives them the greatest satisfaction. And whatever that choice may be, we should all say nothing more than more power to 'em!
Vive la difference! 🙂
And on that note, I'm off to the state fair to partake in some beer drunk from piture size plastic cups and get myself a jumbo serving of deep fried screaming 2 year olds on a stick. Have a safe trip back home, SY!
se
Some just can't bring themselves to believe that solid state can sound good and that tubes are infinitely better no matter how they aplied, so I'm sure there will be some that will never believe that an IC will sound as good or better than a discrete circuit.
Instead of arguing, build one yourself, if you like it, great! Tell us you do, if not, oh well, be happy you like your previous amp more and don't have to go build another.
Instead of arguing, build one yourself, if you like it, great! Tell us you do, if not, oh well, be happy you like your previous amp more and don't have to go build another.
OK, enough... I give up
To purplepeople: Lest you should feel left out: whereas I've little doubt use is indeed made of the studio monitors to which you refer, I also know that many respected studios use monitors by ATC, Duntech, Rogers and B&W. None of which are "chipped"... Perhaps it has to do with the kind of music being recorded.
Without wishing to get any other factions riled, I believe most amp or transducer subtleties will be drowned into oblivion by the likes of The Dead Kennedy's or Linkin Park, whereas Dianna Krall or Loreena McKennit might be more revealing of circuit variations. And I have some music of all of the above.
And, O purple one, may I add that there is a revival in tube recording desks 😱 ... and those big stage amp stacks... mostly tube, 100% chipless
To Pjotr: I've no "mission", other than my centre-speaker. But I don't think it would be a bad thing to toss around a few more unique discrete ideas and a few less re-hashed gainclones. This is just my opinion, clearly not shared by all. Wouldn't life be boring if it were?
And finally to the inimitable Mr Peter Daniel: I have a selection of "other" amps. How about my "bargain" discrete design which cost me around $150: a tube-mosfet hybrid, EF86>EL86(TCF)>J50/K135(SF). I also use an EAR 518 and two surround amps (AV system). In the pipeline: hybrid EL34PP (cathode-driven by FETs), 300B PSE. All amps are my own design.
Here's a positive suggestion for all amp-on-a-chip fans: heard of an LM12? 5-lead (incl case) power op-amp. Heard a good-sounding amp using these, built by a friend about 10 years ago. And it's even unity-gain stable, unlike the LM's and TDA's favoured here.
Hereendeththelesson
To purplepeople: Lest you should feel left out: whereas I've little doubt use is indeed made of the studio monitors to which you refer, I also know that many respected studios use monitors by ATC, Duntech, Rogers and B&W. None of which are "chipped"... Perhaps it has to do with the kind of music being recorded.
Without wishing to get any other factions riled, I believe most amp or transducer subtleties will be drowned into oblivion by the likes of The Dead Kennedy's or Linkin Park, whereas Dianna Krall or Loreena McKennit might be more revealing of circuit variations. And I have some music of all of the above.
And, O purple one, may I add that there is a revival in tube recording desks 😱 ... and those big stage amp stacks... mostly tube, 100% chipless
To Pjotr: I've no "mission", other than my centre-speaker. But I don't think it would be a bad thing to toss around a few more unique discrete ideas and a few less re-hashed gainclones. This is just my opinion, clearly not shared by all. Wouldn't life be boring if it were?
And finally to the inimitable Mr Peter Daniel: I have a selection of "other" amps. How about my "bargain" discrete design which cost me around $150: a tube-mosfet hybrid, EF86>EL86(TCF)>J50/K135(SF). I also use an EAR 518 and two surround amps (AV system). In the pipeline: hybrid EL34PP (cathode-driven by FETs), 300B PSE. All amps are my own design.
Here's a positive suggestion for all amp-on-a-chip fans: heard of an LM12? 5-lead (incl case) power op-amp. Heard a good-sounding amp using these, built by a friend about 10 years ago. And it's even unity-gain stable, unlike the LM's and TDA's favoured here.
Hereendeththelesson
- Status
- Not open for further replies.