Kuei Yang Wang
Sorry if I offended you.
You're taking this way too personally.
Seems like you are married to the 3886. Chill out bud!
This is a forum for open exchange.
You need to learn self-control, and to stop bulying other members.
There are many very bright individuals out there who believe the best discrete amps can outperform the best chip amps. Are you going to bully them in to submission? All of them?
No amount of rhetoric on your part will convince me otherwise.
Thank You
Sorry if I offended you.
You're taking this way too personally.
Seems like you are married to the 3886. Chill out bud!
This is a forum for open exchange.
You need to learn self-control, and to stop bulying other members.
There are many very bright individuals out there who believe the best discrete amps can outperform the best chip amps. Are you going to bully them in to submission? All of them?
No amount of rhetoric on your part will convince me otherwise.
Thank You
I'd love to hear your amps. I especially like the versatility of your amps to handle whatever comes down the pike. Have you noticed any differences in sound quality between the different typologies? I bet the biamping aspect is the best way to go as it allows you to get the "one chip" sound while essentially doubling the power put to the speakers.
I asked myself the same question.
In answer to your question about the quality of chip amps,
amazing, is my answer.
I have had many amps in the last 30 years, including Naim,Rotel Nad, to name a few.
my last setup consisted of,
Naim nac42 pre amp,
snaps power suply,
NAD 3155 as a power amp (links removed)
and Linn Sara speakers.
I was very happy with this sound,
But. as I had a huge toroidal gathering dust,(24v 600va 4x 150va)
I decided to buid one just for the hell of it.
I used the LM3875 chips, folowing the vikash non inv circuit (attached)
the only variation from the diagram were 2 Large, 1500 uf 100v caps 80mm high x 25mm. and KBPC3506 35AMP 600V Bridge Rectifiers x4
after building it into a nice 3u rackmount case, off ebay,
powering 1x 12" woofer chassis for the r chanel and 1x 4.5 " woofer chassis from a wharfdale diamond for the L chanel.
I left it running in in my garage for 8hrs or so.
After listning to my existing setup for an hour, I installed the Gainclone.
I was gobsmacked,!
the sound was stunning, it blew the Nad into the weeds.
I could not believe my ears, bass lines I never heard before were leaping out all over the place, vocal lines were crystal inrstruments became seperate, and distinct.
It was to good to be true, I even asked hi fi freinds to come and listen, they too where amazed.
One freind said" if it had a NAIM badge on it, and i'd paid £800 for that sound, I would be very happy !"
(I have been asked to make another 3)
I dont know if I have been exceptionaly lucky, but the rest of my system just seems to be a perfect match for the "clone"
as I have never had a better sounding system.
Have a go, build one, you won't regret it.
"Make it so"
Neil
In answer to your question about the quality of chip amps,
amazing, is my answer.
I have had many amps in the last 30 years, including Naim,Rotel Nad, to name a few.
my last setup consisted of,
Naim nac42 pre amp,
snaps power suply,
NAD 3155 as a power amp (links removed)
and Linn Sara speakers.
I was very happy with this sound,
But. as I had a huge toroidal gathering dust,(24v 600va 4x 150va)
I decided to buid one just for the hell of it.
I used the LM3875 chips, folowing the vikash non inv circuit (attached)
the only variation from the diagram were 2 Large, 1500 uf 100v caps 80mm high x 25mm. and KBPC3506 35AMP 600V Bridge Rectifiers x4
after building it into a nice 3u rackmount case, off ebay,
powering 1x 12" woofer chassis for the r chanel and 1x 4.5 " woofer chassis from a wharfdale diamond for the L chanel.
I left it running in in my garage for 8hrs or so.
After listning to my existing setup for an hour, I installed the Gainclone.
I was gobsmacked,!
the sound was stunning, it blew the Nad into the weeds.
I could not believe my ears, bass lines I never heard before were leaping out all over the place, vocal lines were crystal inrstruments became seperate, and distinct.
It was to good to be true, I even asked hi fi freinds to come and listen, they too where amazed.
One freind said" if it had a NAIM badge on it, and i'd paid £800 for that sound, I would be very happy !"
(I have been asked to make another 3)
I dont know if I have been exceptionaly lucky, but the rest of my system just seems to be a perfect match for the "clone"
as I have never had a better sounding system.
Have a go, build one, you won't regret it.
"Make it so"
Neil
I'm using ribbon tweeters. As I said: Many ribbons
are of lower impedance. The LM3875 does not
sound as good at 4ohm as it does at 8ohm.
Therefore parallelling improves it. Also a
buffer is essential. I'm uing NI configurations.
I find the sound to be realistic (analytical if
you will). However I think it's very close
to real instruments.
What you like best is a personal choice. I think
everybody should compare and make up their
own opninion. I did find that everything affects
the sound though. Thus one implementation
will sound different then another.
For example: I do not use Carbon composite resistors
as they hide detail. In my experience Holco and
Vishay are the best sounding. For a strong bass
large currents are required. Parallelling chips may
be the answer. My bass is definitely very clear
tight and deep.
Harry
are of lower impedance. The LM3875 does not
sound as good at 4ohm as it does at 8ohm.
Therefore parallelling improves it. Also a
buffer is essential. I'm uing NI configurations.
I find the sound to be realistic (analytical if
you will). However I think it's very close
to real instruments.
What you like best is a personal choice. I think
everybody should compare and make up their
own opninion. I did find that everything affects
the sound though. Thus one implementation
will sound different then another.
For example: I do not use Carbon composite resistors
as they hide detail. In my experience Holco and
Vishay are the best sounding. For a strong bass
large currents are required. Parallelling chips may
be the answer. My bass is definitely very clear
tight and deep.
Harry
Konnichiwa,
You did not offend me.
I do not take this (or anything else) personal and I am not married to any LM.
I did take exception to a number of specific claims you made as being factual. I actually merely pointed out that they where rather leaving much to be desired in terms of factual accuracy.
I am surprised that you feel bullied if someone points out that you made inaccurate or prejudiced claims.
I feel equally surprised that you recommend selfcontrol to me, yet you are the one to call others childish, by inference, merely for pointing out inaccuracies in your statements.
First, I am not bullying ANYONE. Secondly, you may be surprised to note that I am completely in agreement that "the best discrete amps can outperform the best chip amps.", the emphasis being on CAN! I am equally convinced that "the best chip amps can outperform the best discrete amps.", the emphasis being again on CAN!
And one amplifier may better the other in one area, but be secondbest in another.
What I object to is the blanket dismissal that claims "IC's can never be as good as discrete circuitry!".
Sayonara
funberry said:
You did not offend me.
funberry said:
I do not take this (or anything else) personal and I am not married to any LM.
I did take exception to a number of specific claims you made as being factual. I actually merely pointed out that they where rather leaving much to be desired in terms of factual accuracy.
funberry said:
I am surprised that you feel bullied if someone points out that you made inaccurate or prejudiced claims.
I feel equally surprised that you recommend selfcontrol to me, yet you are the one to call others childish, by inference, merely for pointing out inaccuracies in your statements.
funberry said:
First, I am not bullying ANYONE. Secondly, you may be surprised to note that I am completely in agreement that "the best discrete amps can outperform the best chip amps.", the emphasis being on CAN! I am equally convinced that "the best chip amps can outperform the best discrete amps.", the emphasis being again on CAN!
And one amplifier may better the other in one area, but be secondbest in another.
What I object to is the blanket dismissal that claims "IC's can never be as good as discrete circuitry!".
Sayonara
About the topology:
Currently with my 4 ohm speakers. I do not hear a difference
between 2 and 4 chips parallelled. I havn't serously
listened to the bridged topologies. This is because I'm
still building the second chassis. The sound if so good
that I wonder if I need it. I have most of the parts and will
build it anyway.
The first chassis is all p2p. This is way too much work.
I spend about 100 hours on this one chassis. I'm developing them
with a friend we both experiment and discus design.
For the second chassis we are going to manufactor PCBs.
This should simplify construction. We are still going to keep
most of the signal path in wires.
I got into the whole chip thing after I've heard the Wilson
Grand Slam X-1 driven by a Jeff Rowland. I thought it
was one of the nicests sounds I've ever heard. I build a
grat set of speakers but the sound was always less
then this system. Now I'm running my BPA design and
the magic from the Jeff Rowland now comes through my
speakers. Th difference is the most pronounced when
listening to complex music: Chorus, Symphony or
just many sounds at the same time. My SET was very
good at a single noite but messes up complex music.
I did build a gainclone with high quality components
but without a buffer and with low capacitance. This
amp was nice in single tones but did not perform well
with complex music. The bass was somewhat floppy.
There was somewhat of a glare in the
sound as well. I thought of it as a competitor of the SET
but I still prefered the SET.
My new design outperforms both by a lot.
Harry
Currently with my 4 ohm speakers. I do not hear a difference
between 2 and 4 chips parallelled. I havn't serously
listened to the bridged topologies. This is because I'm
still building the second chassis. The sound if so good
that I wonder if I need it. I have most of the parts and will
build it anyway.
The first chassis is all p2p. This is way too much work.
I spend about 100 hours on this one chassis. I'm developing them
with a friend we both experiment and discus design.
For the second chassis we are going to manufactor PCBs.
This should simplify construction. We are still going to keep
most of the signal path in wires.
I got into the whole chip thing after I've heard the Wilson
Grand Slam X-1 driven by a Jeff Rowland. I thought it
was one of the nicests sounds I've ever heard. I build a
grat set of speakers but the sound was always less
then this system. Now I'm running my BPA design and
the magic from the Jeff Rowland now comes through my
speakers. Th difference is the most pronounced when
listening to complex music: Chorus, Symphony or
just many sounds at the same time. My SET was very
good at a single noite but messes up complex music.
I did build a gainclone with high quality components
but without a buffer and with low capacitance. This
amp was nice in single tones but did not perform well
with complex music. The bass was somewhat floppy.
There was somewhat of a glare in the
sound as well. I thought of it as a competitor of the SET
but I still prefered the SET.
My new design outperforms both by a lot.
Harry
If you have a crystal ball you can say that but how about this:Kuei Yang Wang said:
"10 x 64 k memory is more than enough what a person need" (Bill Gates)
About such claims above you must add "this is valid in this very moment but not tomorrow.
I remember when I was taught that blue LED's are impossible to make and this wasn't so long time ago.
Remember also that the laser was pretty useless when it was invented.
funberry's claims is made without any contact with the history, am I wrong?
One more thing: 5 computers in the whole world are enough for the mankind said some IBM person in the 50's.
Today somebody said: "IC's can never be as good as discrete circuitry!"
Correction:
I did not call a person "childish".
I called the act of nitpicking at a logarithmic graph "childish".
If I do it, it is childish. So I won't do it . I'll go to the tabular data, because it's more robust.
If you quote, you must quote correctly, otherwise you risk losing credibility.
You can count 10 unlabelled lines, but sometimes the first line past the 10 can be a 15, not a 20, and the last before the 100 can be an 80, not a 90. Since there are no labels, you cannot know FOR SURE.
Thanks
I did not call a person "childish".
I called the act of nitpicking at a logarithmic graph "childish".
If I do it, it is childish. So I won't do it . I'll go to the tabular data, because it's more robust.
If you quote, you must quote correctly, otherwise you risk losing credibility.
You can count 10 unlabelled lines, but sometimes the first line past the 10 can be a 15, not a 20, and the last before the 100 can be an 80, not a 90. Since there are no labels, you cannot know FOR SURE.
Thanks
This is by no means the best there is out there.
McIntosh MC602 Power Amplifier
But it looks like it outperforms a chip I know.
Here and now. Not history. Not future.
McIntosh MC602 Power Amplifier
But it looks like it outperforms a chip I know.
Here and now. Not history. Not future.
funberry said:
So what is the problem? For you this McIntosh outperforms a chip amp, that's fine.
Let it go, man. Who do you want to convince otherwise?
You come here, where people like to build amps based on chips, with DIY love, and you are trying to tell them that they are making an inferior thing? Because in the end it's that what you are doing.
Kuei Yang Wang was extremely nice with you, trying to show you an angle to look at things that was different from what you brought in.
If YOU think chip amps are inferior, OK, that's your point of view. But all others don't see it like that. Jeff Rowland proved they are not.
Can't you go pick a discussion somewhere else? This is not a constructive one, and those are the only ones to care about.
Carlos E. Martínez
I think its time to move along.....
It does no good for people to repeat themselves.
Or to argue over petty semantic details.
There should be no more discussion of whether certain behavior is childish or not OR other similar discussions. You KNOW what I mean, and I will take action if it continues...
Yeah, let's go back to listening impressions and read the 4th part of a chip amp review, that just came out: http://www.6moons.com/audioreviews/audiosector/patek_4.html
I will only quote the best part of it, and I hope I will be forgiven (that shameful plug) 😉:
"the Patek SE is my favorite solid-state amp yet (closely followed by the FirstWatt F2 which, naturally, doesn't work on my Gallo Reference 3s). To be blunt, the Patek puts the screws on lesser (but still more expensive) valve amps that sound slow, thick and indistinct by comparison and offer far less power. Sure, I expect an Art Audio, Audiopax, Lamm or Wyetech Labs SET to redress that statement. But that's a different financial league. Moral of my story? The search for the "perfect" sand amp that fits my current system and sensibilities like a designer glove -- and which I can afford to wear -- is over. Halleluja!
To those readers who will ask why no Blue Moon award - it'd feel like so much empty self-congratulatory back slapping. Remember, this wasn't a formal review assignment at all. It was an a priori personal acquisition. I know how much I like these amps. I know why I didn't take their designer up on his unusual return privilege for a custom order. What more needs to be said? If you're in the market for a $1,200 to $1,800 50wpc stereo amplifier, you still gotta find out whether you like the Patek. I couldn't see how not but in audio, there's no accounting for personal taste and system synergy. All this review intends to accomplish is to report on the existence of AudioSector; give credit to Peter Daniel who, after Kimura-San of 47Labs, deserves the spotlight for pushing this amplifier concept to the limit; and share my exciting discovery especially with those readers who sympathize with my tube leanings but, for whatever reason, must themselves own a solid-state amp (or can't afford the kind of tube amp that would compete with the Patek on sound if not power). This is one you should then put on your books in bright bold letters..."
I will only quote the best part of it, and I hope I will be forgiven (that shameful plug) 😉:
"the Patek SE is my favorite solid-state amp yet (closely followed by the FirstWatt F2 which, naturally, doesn't work on my Gallo Reference 3s). To be blunt, the Patek puts the screws on lesser (but still more expensive) valve amps that sound slow, thick and indistinct by comparison and offer far less power. Sure, I expect an Art Audio, Audiopax, Lamm or Wyetech Labs SET to redress that statement. But that's a different financial league. Moral of my story? The search for the "perfect" sand amp that fits my current system and sensibilities like a designer glove -- and which I can afford to wear -- is over. Halleluja!
To those readers who will ask why no Blue Moon award - it'd feel like so much empty self-congratulatory back slapping. Remember, this wasn't a formal review assignment at all. It was an a priori personal acquisition. I know how much I like these amps. I know why I didn't take their designer up on his unusual return privilege for a custom order. What more needs to be said? If you're in the market for a $1,200 to $1,800 50wpc stereo amplifier, you still gotta find out whether you like the Patek. I couldn't see how not but in audio, there's no accounting for personal taste and system synergy. All this review intends to accomplish is to report on the existence of AudioSector; give credit to Peter Daniel who, after Kimura-San of 47Labs, deserves the spotlight for pushing this amplifier concept to the limit; and share my exciting discovery especially with those readers who sympathize with my tube leanings but, for whatever reason, must themselves own a solid-state amp (or can't afford the kind of tube amp that would compete with the Patek on sound if not power). This is one you should then put on your books in bright bold letters..."
Konnichiwa Daniels San,
Congratulations again on your achievements....
From small acorns.
It is amusing to see what Kimura San kicked off, I am sure he
occasionally has a good chuckle. Even my own standby amp is a gainclone, though I'm afraid I still prefer inverted... ;-)
Still, bloody good batting average, if I may say so....
Sayonara
Peter Daniel said:
Congratulations again on your achievements....
From small acorns.
It is amusing to see what Kimura San kicked off, I am sure he
occasionally has a good chuckle. Even my own standby amp is a gainclone, though I'm afraid I still prefer inverted... ;-)
Still, bloody good batting average, if I may say so....
Sayonara
Thanks.
Not to mention your involvement and inspiration in the chip amp circles over the years. If it wasn't for that motivation, I might not be here today, still building those 'girly' amplifiers...
Regarding certain topologies, as it's been said, there's no accounting for personal taste and system synergy in audio. 😉
Not to mention your involvement and inspiration in the chip amp circles over the years. If it wasn't for that motivation, I might not be here today, still building those 'girly' amplifiers...
Regarding certain topologies, as it's been said, there's no accounting for personal taste and system synergy in audio. 😉
Kuei Yang Wang said:
Yeah. He went completely against most all of the "conventional wisdom" of the status quo at the time. That really took some balls. I've always admired him for that. And I also admire Segoshi San for having the balls to introduce such a product here in the US.
se
I haven't done a lot with chip amps, but to the extent I have done something I've reached a few generic opinions.
- The chips that are good are good within their limitations.
For example many such as LM3886 exhibit significantly higher distortion with lower loads (4 ohm vs 8 ohm) which I assume is the chip verision of BJT beta droop. The droop seems to be grwater than you get in the A1302/C3281 and related transistors. The expediant of usung parrallel devices is probably not available.
The info on datasheets must be read carefully. One chip claims 100W and THD+N of .005%. Both are true but not at the same time. The distorthion at 100W is 1-10%. The .005% figure is at ~40-50W. So if that is all you need it's pretty good.
The LM3886 datasheet claims THD+N of .01% or grater but I measured a Marchand version at .005% with RightMark and that includes the soundcard noise floor.
- For applications requiring no more than 50W and nominal 8 ohm loads, I can't think of a practical reason to use discrete except for the pleasure of doing so or the desire for extreme low distortion (such as Class A).
- For application requiring over 50W and/or low impedance (nominal 4 ohm) loads I think discrete will outperform the chips. A personal example is driving Magnepans.
- From a DIY perspective, I find chip amps to easy unless I'm trying to solve a problem requiring compactness. An example I didn't do myself but suggested to someone else was five LM3886's in a single compact enclosure with a PS that would provide 30Wx5. The intent was a small HT system using either Gallo Micro's or single driver DIY TangBands.
- The chips that are good are good within their limitations.
For example many such as LM3886 exhibit significantly higher distortion with lower loads (4 ohm vs 8 ohm) which I assume is the chip verision of BJT beta droop. The droop seems to be grwater than you get in the A1302/C3281 and related transistors. The expediant of usung parrallel devices is probably not available.
The info on datasheets must be read carefully. One chip claims 100W and THD+N of .005%. Both are true but not at the same time. The distorthion at 100W is 1-10%. The .005% figure is at ~40-50W. So if that is all you need it's pretty good.
The LM3886 datasheet claims THD+N of .01% or grater but I measured a Marchand version at .005% with RightMark and that includes the soundcard noise floor.
- For applications requiring no more than 50W and nominal 8 ohm loads, I can't think of a practical reason to use discrete except for the pleasure of doing so or the desire for extreme low distortion (such as Class A).
- For application requiring over 50W and/or low impedance (nominal 4 ohm) loads I think discrete will outperform the chips. A personal example is driving Magnepans.
- From a DIY perspective, I find chip amps to easy unless I'm trying to solve a problem requiring compactness. An example I didn't do myself but suggested to someone else was five LM3886's in a single compact enclosure with a PS that would provide 30Wx5. The intent was a small HT system using either Gallo Micro's or single driver DIY TangBands.
sam9 said:
Actually if you have very high sensitivity speakers (100dB+) you can bias the output stage into single-ended class A.
se
I know you can force OpAmps into Class A by connecting the output to one of the power pins via a resistor (or more complex arrangement), but I wasn't aware you could do that with chip amps. How would you do it with an LM3886 for example? It doesn't have a lot of contact area on the back to dissapate heat regardless of the size of the heatsink and since it is going to be dissapating power regardless of whether a signal is present or not I imagine that would put some sort of limit on the maximum power rails you could use. What expeiences have folks had in this regard?
Obviously this is academic as long as I keep the Magnepans.😀
- Status
- Not open for further replies.