roibm, you don't read too well... similar class... A power supply would cost exactly the same for a clone of 20W vs another s/s class B/AB amp of 20W. And yes, you'd need a much bigger transformer for class A but for a miserly 20W it could probably still be done for $100. Even in Stuttgart.
Now since you ask politely that I should, ahem "cut the crap", why not explain your link to the Chernobyl disaster. Exactly what are you trying to say? I fail to see the relevance. In fact, if you want to save the planet then leave Chernobyl alone, it did more mental harm than actual physical damage. Turn your attention to greenhouse gases. Did you know that besides CO2, methane is the biggest culprit. And did you know that cows are one of the biggest sources. So if we all stop eating beef and drinking milk then there will be fewer cows. And therefore fewer cow farts, resulting in less methane and therefore less global warming.
Cow farts or high-dissipation amps..? Hmmm... lemme see now that's a difficult choice for me. How about you?
Now I know why your gainclones didn't sound all that good.DrG said:
Have you used a solid state like PSU for your gainclones?
good amp is good amp,no matter is it tube or discrete or GC;
but,(big but!) only when you pair your amp with adequate speakers.
in that light-speakers are at least 50% of story,along with room,in which speakers are positioned......
proper GC is good amp,no matter what anyone can say;
I know that ,just because I build ,and repair ,almost any living 😉 type of amp in this world
tube is tube,
GC is GC,
discrete A class is A class.
in fact ,only bad and good amps exist;
one thing I know-I have at least one smaller or bigger system with various types ,and I like them all.
LBNL-and every amp in question is waaay above crappy class of amps,so offten advertized in HiFiFi magazines as best buy,amp of the year,amp of decade etc.
is your mama better than mine?
not for me.
same as my mom to you.
but,(big but!) only when you pair your amp with adequate speakers.
in that light-speakers are at least 50% of story,along with room,in which speakers are positioned......
proper GC is good amp,no matter what anyone can say;
I know that ,just because I build ,and repair ,almost any living 😉 type of amp in this world
tube is tube,
GC is GC,
discrete A class is A class.
in fact ,only bad and good amps exist;
one thing I know-I have at least one smaller or bigger system with various types ,and I like them all.
LBNL-and every amp in question is waaay above crappy class of amps,so offten advertized in HiFiFi magazines as best buy,amp of the year,amp of decade etc.
is your mama better than mine?
not for me.
same as my mom to you.
Of course. You know another kind for this application?
Do tell... Is there something magical about connecting a diode bridge to filter caps and using a decent grounding layout? Some sonic erratum I inadvertently made which destroyed the gainclone's sound?
I believe I understand now why this thread is ignored...
DrG, you have your own world, you master it, you are the CIC.
Enjoy 😀
DrG, you have your own world, you master it, you are the CIC.
Enjoy 😀
DrG has made several valid points here that should not be disregarded. The price for a design of SIMILAR design is not so different. We understand that your op-amps are dirt cheap and that they sound decent. They do however place limits on your design. The largest opamp you guys can use is about 56w and can push 6 amps without blowing up. $2.60 in discrete output devices buys me 260wrms of power handling with the damping and low impedance drive of 25 amperers. If you want a nice 200w subwoofer amp, can you do that with chips? What about something that can drive 2 ohm loads?
From the National's datasheet:
I'm not going admit that I can easily hear 0.1% THD, but many discrete amps distort far less than that. A lot of gainclone guys claim the quality is phenomenal and I'm really having a hard time understanding why. 0.1% THD isn't good, it isn't bad, it's OKAY. As to how 47 Labs can charge thousands for it, I have no idea.
From the National's datasheet:
I'm not going admit that I can easily hear 0.1% THD, but many discrete amps distort far less than that. A lot of gainclone guys claim the quality is phenomenal and I'm really having a hard time understanding why. 0.1% THD isn't good, it isn't bad, it's OKAY. As to how 47 Labs can charge thousands for it, I have no idea.
DrG said:
It is very easy to build a working gainclone. Even if you use the correct topology for grounding etc, this does not necessarily mean it will sound amazing.
My first attempts sounds nothing like the PCB kit that I got from BrianGT. I was quite disheartened by the hard and tiring sound I got with my earlier attempts, but Brian's kit sounds 100 times better.
Maybe the results of your experiments with chip amps are on a par with my first attempts?
The major differences were:
PCB layout instead of P2P.
discrete (MUR860) rectifier diodes instead of bridges
better PS caps - Panasonic FC, 1500uF
use of Caddock & Riken resistors, rather than metal foil.
Just a thought.
Steve
Solid Snake said:
You can build a 200W amp with chips, but they really weren't designed for that kind of work. You could also build a 50W amplifier out of discrete jellybean transistors if you wanted to. Horses for courses.
Steve
Not to sound like a broken record, but the measured distortion performance is MUCH better than their THD figure seems to indicate. THD figures are meaningless on their own. You should be able to easily hear 0.1% THD if it's due to crossover distortion.Solid Snake said:
It all depends at what power you measure the distortion. If you measure it at 5 watts into a 16 ohm load, it better be less than 0.1% THD.
There are certain undeniable truths here, whichever side one opts for. The most basic is that amplifier topology and choice of active devices has by far the greatest influence on sonic signature. Far more so than fiddling with caps, diodes and feedback resistors. And herein lies the Achilles heel of chip amps. Along with further issues such as quasi-complementary output topology and thermal proximity to the VAS and input stages.
And there is nothing companies like Linn, Jeff Rowland or 47Labs can do about it. They make choices for fiscal reasons, adhering to minimum sonic requirements and premium pricing. The latter bordering on the insane in the case of 47Labs. And spawning hordes of copycat cloners in the process who have great fun I think, getting into diy-fi. Which is great!
However, a fringe of these people persist in extolling imaginary (I believe) virtues of these little devices. Exchanging all manner of input, feedback, PSU and buffering variations and possibly leading some readers to believe that the holy audio grail has been discovered at National Semiconductor. Which it has not. For all the real reasons I mentioned in paragraph one.
Sonically I (and others) think they're average. Which is what they're designed to be. Measured performance is average too, for adherents of that approach. They hold only two apparent advantages: cost and convenience. And considering the similar expenses of PSU's and casework for a discrete design, the cost issue is not as significant as one might think. Even more so once the hardcore devotees have thrown Riken resistors, Black Gate caps, tube buffers and milled aluminium casework into the equation. Leaving only one real advantage viz convenience, the fortuitous culmination of various design compromises, as I said at the beginning.
And there is nothing companies like Linn, Jeff Rowland or 47Labs can do about it. They make choices for fiscal reasons, adhering to minimum sonic requirements and premium pricing. The latter bordering on the insane in the case of 47Labs. And spawning hordes of copycat cloners in the process who have great fun I think, getting into diy-fi. Which is great!
However, a fringe of these people persist in extolling imaginary (I believe) virtues of these little devices. Exchanging all manner of input, feedback, PSU and buffering variations and possibly leading some readers to believe that the holy audio grail has been discovered at National Semiconductor. Which it has not. For all the real reasons I mentioned in paragraph one.
Sonically I (and others) think they're average. Which is what they're designed to be. Measured performance is average too, for adherents of that approach. They hold only two apparent advantages: cost and convenience. And considering the similar expenses of PSU's and casework for a discrete design, the cost issue is not as significant as one might think. Even more so once the hardcore devotees have thrown Riken resistors, Black Gate caps, tube buffers and milled aluminium casework into the equation. Leaving only one real advantage viz convenience, the fortuitous culmination of various design compromises, as I said at the beginning.
I.ve never built anything before.
Chip Amps are an attractive proposition to beginners ( not too many bits ).
This forum and others has given me the confidence to tackle other more complicated projects should I wish to.
Not only have I built my first amp, which actually works.
I find that it equals and possibly outperforms my £800 solid state amp and my kt88 tube amp.
My chip amp cost less than a tenth of this and is a remarkable achievement. which I can play about with and hopefully improve even further.
So you see I don't care whether it can be called High End, or if a £3000 Krell sounds better.
By the way would someone like to define Hi-end audio.
Could it mean....... Big expense, Small upgrade?
Albo
Broken record again...
0.001%. Measured performance is NOT average. It beats many high end amp designs, I'd go so far as to say "nearly every" amp that's not zero NFB / heavily class-A / single ended. (That would be if it wasn't for the fact that a lot of modern consumer amps are nipping at the heels of so called "perfection".)
Believe what you want to about the sound, I am not in a position to argue about that, but the measurements are verifiable.
0.001%. Measured performance is NOT average. It beats many high end amp designs, I'd go so far as to say "nearly every" amp that's not zero NFB / heavily class-A / single ended. (That would be if it wasn't for the fact that a lot of modern consumer amps are nipping at the heels of so called "perfection".)
Believe what you want to about the sound, I am not in a position to argue about that, but the measurements are verifiable.
Albo said:
Ahah!
From somebody who has built amplifiers for more than a quarter of a century and also someone with a long time bias (all persons are biased, otherwise a person would have no opinions) towards tubes, I can say that a VIGC (as it has become known) constructed with care (get the wiring right) and high quality components, does indeed sonically outperform most tube amps I know off. I may not like it and there is mot much I can do about. Is anyone asking that I be dishonest with myself? OK, I admit I am talking about the JLTi and I make them - but that is why I make them, and yes, they are better than the VIGC posted here on this forum.
I am a co-developer of the VSE DPA-300B power amps that were given a Best Of 2003 Award by www.enjoythemusic.com and also State-Of-The-Art mic preamps used in recording studios, including doing SACD recordings, also the A$25.000 VSE RTP Phono Preamp. These are true tube SOA products and indeed push the very boundaries of audio. Why do I mention this? Because the JLTi was developed with those benchmarks in mind, I know what the very best equipment on the planet sounds like. The JLTi and its slightly lesser VIGC cousins are no slouches and they really do make music.
Beyond this I am not willing to go back and fro discussing endlessly the attitudes of sceptics. Their loss and not mine. Oh BTW, there are actually good reasons why chip power amp chips should have this potential (but sadly rarely do) and indeed be superior to discrete circuitry. Besides I haven't found a discrete SS power amp that I have wanted to live with, and that doesn't apply to my JLTi (which is hybrid). It is one of those rare amps that gets both the beginning and the end of a note right. Sorry, but discrete SS has never done that for me. Even some highly regarded ones sound like a trailing grain on every tone and the JLTi is not totally free of that and I have only ever heard one that is and that is tube, and heaps more expensive. I make the JLTi to make other people (with less money) happy and that is not a bad vocation.
Also the JLTi has one of the best spectral distribution of harmonic distortion I have ever seen (yes, I can measure that) with higher order odd types completely missing and the lower order cascading the way the best zero feedback tube amps do. This was identified in a paper by Jean Hiraga decades ago as good thing and was re-printed be a recent copy of AudioXpress mag (March 2004). Amps that have this characteristics invariably sound good... I speak from experience and that of others. All amp makers should include this measurement/graph with their products, I DO!!!
Joe R.
PS: For 2003 Awards, see www.enjoythemusic.com/magazine/equipment/0104/bestof2003.htm and for VSE DPA-300B review www.enjoythemusic.com/magazine/viewpoint/0104/aachapter52.htm
No, Joe that is not why you make them. If it was you'd be giving them away to your friends. You make them for exactly the same commercial reason 47Labs and any other brand does. Money. Which is fine but at least be honest about it...
There is no big-hearted altruism or philanthropy at work here and I doubt anyone here doesn't see through it. Or through your veneer-thin ad campaign for the JLTi. Whatever it's sonic accomplishments might be.
I'm curious, exactly what are the forum rules regarding thread-jacking for commercial gain...?
DrG said:
Well, you are a true cynic and also PLAIN WRONG!
If there are no such thing as altruism or philantropy, then why do we have those words in the English language. I suspect you are an accountant, at least you should be. Either that or you have a mean spirit which you like to show off publicly. Why? I just don't understand you, and just as well I don't.
There are indeed people who truly have 'head and heart' - but you don't seem to know about that. A wiser man once said that there 'is more happiness in giving than receiving' and the only way to discover that is to try it.
Some of the world's greatest artists did not do it just for money. They had greater motivations than greasy notes could ever supply. One of the greatest saticfaction one can have is achievement, but far too many count achievement and success in dollars. How sad! Much has been lost because of greed. Much is to be gained by having an 'expansive' outlook and one that has genuine interest in the other person. They enjoy things that money could never buy!!!
BTW, have you heard about impugning bad motives, don't you realise this is wrong? What do you even know about me? How many JLTi's have I sold? How much have I made from it? Maybe, and you have no way of contradicting this, I have made an actual loss? Maybe I am independantly wealthy and I do this for love? Don't discount it because you don't know!
As for forum rules... maybe you should read them before judging adversely others and assigning bad motives to anyone. As for the VIGC, have you seen the Tube with Power IC Output Stage... thread here. Have you seen the effort that went into posting the "VIGC" there? Have you checked the link below for the DIY website set up specifically for it? That actually has cost me money. Do I expect those doing the DIY version to buy one from me? It hasn't happened that way and I didn't expect it. Why did I do it? For money? How would you know? But what does your gut instinct say? Maybe you're wrong? Huh?
Joe R.
Now onto something more constructive (please note I am charging nothing for the time - hundred hours plus by now- I've put into this project DrG).
Better post this follow-up:
IF this seems a bit high @ 0.114% then please consider this is the JLTi with closed loop gain of 33dB - a fair bit higher than NS recommendation - but it means little or near no feedback above 200KHz where the 90 degree phase shift (caused by the internal com cap) deteriates significantly (feedback starts to go pisitive). It sounds better that way.
Note the following: The Phase intersect the 100 degree @ 200KHz. Next the open loop gain is 20dB @ the same 200KHz. Since we have chosen closed loop gain to be 33dB there is insuffiicient gain at 200KHz to generate feedback. Yes, we get higher measured distortion and a shift upwards in the noise floor (which likely contributes to increased measured THD. In fact this noise floor is quite close to tube gear.
The extra 10 degrees beyond the critical 90 degrees do no damage here as the amp is open loop now at all times. There is to me an audible reduction is slew rate induced distortion. Note lack of high order stuff, the lack of crossover artifacts. Hiraga was able to demonstrate the adding feedback reduced overall THD but these were only lower order and the higher order got pushed up. So we decrease inocuous distortion and increase the horrible stuff, how logical?
This chip also has something else, the internal Class A driver (we know because it has a current mirror) is highly capable of sucking base storage effects of the output Darlingtons and amazingly good at controlling crossover switching artifacts. The use of Darlingtons is also significant here as the Class A driver sees only the first tranny lower storage and thsi more effectively controls the final high current tranny. The packaging and thermal coupling here must be critical to make this work. Can you see the logic?
Time to goto bed, bona notte.
Joe R.
Better post this follow-up:
An externally hosted image should be here but it was not working when we last tested it.
IF this seems a bit high @ 0.114% then please consider this is the JLTi with closed loop gain of 33dB - a fair bit higher than NS recommendation - but it means little or near no feedback above 200KHz where the 90 degree phase shift (caused by the internal com cap) deteriates significantly (feedback starts to go pisitive). It sounds better that way.
An externally hosted image should be here but it was not working when we last tested it.
Note the following: The Phase intersect the 100 degree @ 200KHz. Next the open loop gain is 20dB @ the same 200KHz. Since we have chosen closed loop gain to be 33dB there is insuffiicient gain at 200KHz to generate feedback. Yes, we get higher measured distortion and a shift upwards in the noise floor (which likely contributes to increased measured THD. In fact this noise floor is quite close to tube gear.
The extra 10 degrees beyond the critical 90 degrees do no damage here as the amp is open loop now at all times. There is to me an audible reduction is slew rate induced distortion. Note lack of high order stuff, the lack of crossover artifacts. Hiraga was able to demonstrate the adding feedback reduced overall THD but these were only lower order and the higher order got pushed up. So we decrease inocuous distortion and increase the horrible stuff, how logical?
This chip also has something else, the internal Class A driver (we know because it has a current mirror) is highly capable of sucking base storage effects of the output Darlingtons and amazingly good at controlling crossover switching artifacts. The use of Darlingtons is also significant here as the Class A driver sees only the first tranny lower storage and thsi more effectively controls the final high current tranny. The packaging and thermal coupling here must be critical to make this work. Can you see the logic?
Time to goto bed, bona notte.
Joe R.
Phew, this thread got hot quickly!
DrG, we have no strict rules as yet on commercial postings, as we tend to judge each poster on a number of aspects, the most important being how much thay have put into the forum. On this count, Joe is well ahead of the game, and his assistance in all aspects of GC construction, and especially the advantages of buffering, have lead to many good quality discussions.
However, we do have strict rules about personal attacks, and your last post was very borderline, so please restrict your critique to technical matters please.
And now, back to your scheduled programming...😉
DrG, we have no strict rules as yet on commercial postings, as we tend to judge each poster on a number of aspects, the most important being how much thay have put into the forum. On this count, Joe is well ahead of the game, and his assistance in all aspects of GC construction, and especially the advantages of buffering, have lead to many good quality discussions.
However, we do have strict rules about personal attacks, and your last post was very borderline, so please restrict your critique to technical matters please.
And now, back to your scheduled programming...😉
- Status
- Not open for further replies.