I would rather be concerned with the amount of distortion than its order. 0.114% is not great, odd or even. Cleaner solid state amps are available, probably at lower cost.
Solid Snake said:
My TDA7294 gainglones measures pretty well, with 0.005% THD (no, it's not a typo)
The measurements have been reapeted several times, and donw with a computer analyser, and a pair of lindos. Always approx 0.005%
"My TDA7294 gainglones measures pretty well, with 0.005% THD (no, it's not a typo)"
That is consistent with the data sheet. This is at 1kHz over a power range of 5W-55W output with an 8-ohm load. Double the figure for 4-ohm loads. Above 55W the distortion figure shoots up pretty fast and it rises in the typical fashon at higher frequencies. This is the best measuring IC amp I've seen figures on. The data sheet calls it a 100W amp and I guess it can be operated that way, but if you operate it within the limits, above, I imagine its a pretty impressive little chuck of silicon.
That is consistent with the data sheet. This is at 1kHz over a power range of 5W-55W output with an 8-ohm load. Double the figure for 4-ohm loads. Above 55W the distortion figure shoots up pretty fast and it rises in the typical fashon at higher frequencies. This is the best measuring IC amp I've seen figures on. The data sheet calls it a 100W amp and I guess it can be operated that way, but if you operate it within the limits, above, I imagine its a pretty impressive little chuck of silicon.
The first time I measured it, I was a little sceptic about the results. The value was so low compared to what you usually see on hifi amps.
That was with the lindos, and I didn't know this stuff well. But since, I did the measurements again on my computer with RMAA, and got the same numbers (and that was with crappy interconnects)
I never thougth that I would have the same numbers as in the datasheet 🙂
That was with the lindos, and I didn't know this stuff well. But since, I did the measurements again on my computer with RMAA, and got the same numbers (and that was with crappy interconnects)
I never thougth that I would have the same numbers as in the datasheet 🙂
I couldn't agree more, analog_sa. Too damn easy.
Joe,
I may be a cynic but I am not wrong. I did not deny the existence of altruism and philanthropy nor whatever advice you may offer freely to whomever. What I maintain is that you are a commercial amp manufacturer and PROFIT is your motive, not the "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 make the JLTi to make other people (with less money) happy and that is not a bad vocation..." drivel you're trying to convince us of. And furthermore, after re-reading your post it smacks even more of a sales pitch.
Having said that I repeat that I'm happy with the profit motive. Hope you sell plenty of JLTi's to happy customers. Truly. But please don't try to acquire a halo of altruism under which to hide salesmanship. It's unnecessary.
pinkmouse, this is not a personal attack but an interpretation of Joe's public post. And an accurate interpretation by all applications of business logic I think.
Joe,
I may be a cynic but I am not wrong. I did not deny the existence of altruism and philanthropy nor whatever advice you may offer freely to whomever. What I maintain is that you are a commercial amp manufacturer and PROFIT is your motive, not the "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 make the JLTi to make other people (with less money) happy and that is not a bad vocation..." drivel you're trying to convince us of. And furthermore, after re-reading your post it smacks even more of a sales pitch.
Having said that I repeat that I'm happy with the profit motive. Hope you sell plenty of JLTi's to happy customers. Truly. But please don't try to acquire a halo of altruism under which to hide salesmanship. It's unnecessary.
Actually I'm a radiologist. And I do a lot of real altruistic work such as teach for free at a university on my afternoon off. Four years now. But I don't try to sell anything in the process and neither do I feel the need to advertise the fact. But this is irrelevant.
Right, I don't know. But my guess is that like any businessman you need to sell to have a successful business and make money. And having enquired about dual JFET's previously (do you possibly remember?) I was met with horrendous pricing. Astronomical!. These are not altruistic or philanthropic traits, Joe. They are the characteristics of business in a market economy. Nothing to be ashamed of but certainly not samaritan motivation either.
My gut instinct is that you are a very competent engineer fortunate to have his hobby as his job but that like the rest of us you require income, derived at least in part from your audio business. And as such you will extoll the virtues of the products in your catalog.
pinkmouse, this is not a personal attack but an interpretation of Joe's public post. And an accurate interpretation by all applications of business logic I think.
DrG said:
OK, but this part of the discussion stops here. Let's all stick to technical matters now please.
analog_sa said:
Actually the JLTi is not that easy to make. There are significant differences between the VIGC and the JLTi that makes the latter quite more complex. The power supply to the Tube Buffer is an order more complex (Class A SuperReg'd) and the buffer itself is more involved. There are no real balance acts in the VIGC, you just put them together and they work. This cannot be said about the JLTi, but there are several elements that have to be carefully balanced. The end result is not that simple but sonically superior.
As for DrG, I did say I was not going back and fro. So I shall refrain from responding to him as it is obvious that this guy doesn't know me at all and makes no attempt to. BTW, I don't see 47 Labs here or helping the DIY fraternity. They are the ones making the money, not me. Where is my dictionary (turning my head around), wonder what definition it gives to enthusiast.
Re complexity, indeed the VIGC is more complex than the standard IGC, but hey, it makes a great first tube project. There are so many out there capable of putting an IGC or NIGC together (yes, it is simple) , but then with a little stretch they can actually end up making a tube amp. alas a hybrid, but still tube. How about that!!! No lethal high voltages, as easy to make as SS. This to me is a real buzz, I mean really.
So go VIGC builders of the world!!! Hey, maybe I will acccused of spreading the word of tubes by sly underhanded means?
Joe R.
Solid Snake said:
The actual figure is not that important. This is the THD figure where the "T" stands for "Total". What is important is not the Total Harmonic Distortion but its distribution of those harmonics.
Believe me, there are solid state amps that do measure 0.1% and even a bit more. They eschew feedback and suddenly the distortion magically appears, well not really, it was just masked by the f/b. What f/b does is re-distribute harmonic distortion. We really need someone to come up with a weighting system where THD is mathematically arrived at by shifting mathematical emphasis to high order distortions, especially the odd ones. So 0.1% 2nd order is equavalent to 1% (or more) 5th or 7th order. If this could be implemented with a degree of accuracy we would find that by increasing feedback, the weighting system would show we are not decreasing distortion. In fact we may be increasing and that is what we are hearing, again harking back to Hiraga since his data and observations confirms this.
There is another thing to look at re 0.114% figure: This was taken @ 2.83V into 8 Ohm, thus 1W. It was not taken at nearer full power 35W. The noise floor now sits higher. Also I use much less feedback than NS does in their data sheets. Their data sheet is probably very accurate, in fact thet are. I use 13dB less f/b than NS. Indeed, similarly 47 Labs use 11 dB less f/b, so theirs would measure 'worse' as well. Add to that, I use a Tube Buffer with zero loop f/b (but lots of local f/b). This all add up.
1) Noise floor different - affects THD
2) Less f/b - higher gain
3) Tube Buffer.
With my measurements on the JLTi, the THD goes down slightly with increased power, the noise floor is getting suppressed. The reverse happens when we go below 1W. At 10mW (not a typo) it is 0.177%. I think you can see the picture better if you look at graphs on this link:
http://members.ozemail.com.au/~lisaras/gainclonesound.htm
I wish I could scan Hiraga's article and post it, but that contravenes copyright. It makes for required reading, even if the original article was from the 70's. You say 'clean' - yet amps with higher THD can sound cleaner. I keep coming back to this point, Hiraga shows it's the spectral distribution that is important. If I may say so, my own data utterly confirms this.
Joe R.
"The actual figure is not that important."
I can't logically agree with either this or the opposte position when stated as categorical absolutes. Both the Total and the distribution are important sometimes and sometimes not.
A if the total is sufficiently low as to be inaudible, the distriburtion is irrelavant.
B. If the total is high enough, this distribution only effects the degree intolerability even if all there is is second harmonics.
C. If the evaluation of distortion is bandwidth limited, the area of disput is reduced because as you look at increasing higher frequencies high order harmonics become irrelavant. For example at 10kHz the only audiophiles who will be concerned with 3rd and higher harmonics are those with 4 legs, bad breath and who answer to the name or rover.
D. There IS a range of THD figures where the distribution maters. This may be a matter of disput, but I'll guess it lies between 5% and .005%. Outside of that range one of the above conditions dominates.
E. Although this is only a generality, it seems to me that in the real of all discrete amps the distribution of harmonics falls in to a fairly regular pattern of diminishing with order number. Among IC amps there seems to more variety to the patterns.
I can't logically agree with either this or the opposte position when stated as categorical absolutes. Both the Total and the distribution are important sometimes and sometimes not.
A if the total is sufficiently low as to be inaudible, the distriburtion is irrelavant.
B. If the total is high enough, this distribution only effects the degree intolerability even if all there is is second harmonics.
C. If the evaluation of distortion is bandwidth limited, the area of disput is reduced because as you look at increasing higher frequencies high order harmonics become irrelavant. For example at 10kHz the only audiophiles who will be concerned with 3rd and higher harmonics are those with 4 legs, bad breath and who answer to the name or rover.
D. There IS a range of THD figures where the distribution maters. This may be a matter of disput, but I'll guess it lies between 5% and .005%. Outside of that range one of the above conditions dominates.
E. Although this is only a generality, it seems to me that in the real of all discrete amps the distribution of harmonics falls in to a fairly regular pattern of diminishing with order number. Among IC amps there seems to more variety to the patterns.
Hi Joe,
Absolutely.
I recall Jean Hiraga articles on this matter very well. I think out of the 40+ issues of L'Audiophile I'm only missing a few.
Which reminds me, the articles I have are "bien entendu" written in French and I assume this is the case of your copies as well but... As the magazine is now long defunct and I've seen other extracts taken from these issues translated and published on the web already, wouldn't it be legally O.K. to put them on one of your servers?
I wouldn't even suggest this if it weren't so important.
Oddly enough, or probably not, this seems to have sunk in much better with the tubeheads (I'm guilty in that department myself) than with the sandmen.😉
Another, translated from English into French I suppose, very interesting topic that was published by them, was a study by Matti Ottala on IMD and the importance of phase behaviour.
Just to let you know that:
a) I don't mind doing some translation work here.
b) I'm always glad to find a kindred soul in the audio department.
c) The way you help out these people deserves a warm round of applause instead of undeserved accusations.
Cheers,😉
Absolutely.
I recall Jean Hiraga articles on this matter very well. I think out of the 40+ issues of L'Audiophile I'm only missing a few.
Which reminds me, the articles I have are "bien entendu" written in French and I assume this is the case of your copies as well but... As the magazine is now long defunct and I've seen other extracts taken from these issues translated and published on the web already, wouldn't it be legally O.K. to put them on one of your servers?
I wouldn't even suggest this if it weren't so important.
Oddly enough, or probably not, this seems to have sunk in much better with the tubeheads (I'm guilty in that department myself) than with the sandmen.😉
Another, translated from English into French I suppose, very interesting topic that was published by them, was a study by Matti Ottala on IMD and the importance of phase behaviour.
Just to let you know that:
a) I don't mind doing some translation work here.
b) I'm always glad to find a kindred soul in the audio department.
c) The way you help out these people deserves a warm round of applause instead of undeserved accusations.
Cheers,😉
sam9 said:
It is possible for an amp to have low THD and sound bad and yet amps with high THD can sound good and indeed 'cleaner' - like less audible distortion. Then you examanine their spectral distribution a la Hiraga, and find after listening and measuring many amps that the conclusion is inescapable as the pattern is there for all to see.
So what we are realling doing here is listening first (as a priority) and measuring second (as a lesser priority) and trying to find a bridge between the two. Isn't that laudable? I don't know about you, but I much prefer listening to an amp than measuring and rightfully (whew) find myself doing more of the latter. Sorry to be coy. 😉
As to higher order distortions above 10KHz, I think you miss the point, surely. What you are seeing are only the symptons, these are measurements. That is all measurements are really capable fo doing. I tell you this truthfully, it is possible to hear slew rate problems measured at hundreds of kilohertz. Can you see the connection between slew rate and higher order distortions?
A doctor can diagnose your pain (objectively) but cannot experience it. These type of measuremnts only allows us a diagnostiscal insight and we ought to be grateful when they do correlate, since often they don't. It seems that science of audio is even less perfect than medicine.
Joe R.
Joe R.
fdegrove said:
Hi Frank
I don't have the originals but it has been translated by various others and I had some but they've gotten lost. But what they communicated was never lost and in time found myself doing them as indeed so does Lynn Olson on tube amps. He has a number of examples on his web site.
But Audio Xpress mag published it just last month, March 2004. But this is a reprint of a translation into English done for Hi-Fi News & RR in March 1997. Since that is all I've got I cannot scan and post it, and Audio Xpress did the right thing and was given permission to publish by HFN & RR.
Hey guys, go out and get a copy of Audio Xpress March 2004
!!!! This is landmark stuff.
Joe R.
Hi,
I was totally unaware of that. Interesting....
For those reading a little French here's a list of the articles that appeared over the years:
L'AUDIOPHILE ON CD-ROM.
Unfortunately the list doesn't link what appeared in which issue which often means I have to dig through a bunch of issue before I find the one I need.
Some of the members here might sneer at the publishing dates but from what I've seen here at the forum it's all still as valid reading as it was way back than.
At least one other member, my solid state brother and cartoon poster, Jam is well aware of the value of these articles and diy projects...
Sorry for the off-topic, 😉
I was totally unaware of that. Interesting....
For those reading a little French here's a list of the articles that appeared over the years:
L'AUDIOPHILE ON CD-ROM.
Unfortunately the list doesn't link what appeared in which issue which often means I have to dig through a bunch of issue before I find the one I need.
Some of the members here might sneer at the publishing dates but from what I've seen here at the forum it's all still as valid reading as it was way back than.
At least one other member, my solid state brother and cartoon poster, Jam is well aware of the value of these articles and diy projects...
Sorry for the off-topic, 😉
"It is possible for an amp to have low THD and sound bad and yet amps with high THD can sound good and indeed 'cleaner' - like less audible distortion. Then you examanine their spectral distribution a la Hiraga, and find after listening and measuring many amps that the conclusion is inescapable as the pattern is there for all to see."
I'm afraid I didn't make my point clear. I didn't mean to say that if distortion is sufficiently less audible, the spectral distribution doesn't matter. I mean to say it does not matter if there is no audible distortion - at all. Possible counter positions might be that "any distortion no matter how low is audible even if an arbitrarily long string of zeroes preceeed the first non-zero digit". Or- "even if you could build amp with literally NO distortion (not "low" or "hyper low" but "NO") there exists some spectral distribution of audible distortion that sounds better. A third one would be "the current state of the art is such that no amp can be built without audible distortion".
Of the three, I would say the third is the most pausible and may be demonstrably true, but I no of no one who has explicitly tested that proposition.
The second is somewhat related to the first but easily degeneraes in to agrument ad absurdum.
The second is most likely a mater of taste and would let both of us walk away feeling righteous. One might, of course, assert that, the implied taste preference is universal or virtually so. That could most convincingly be demonstared by appropriately conducted blind listening tests. Since it is a universal proposition not just limited to the audiophile elite a double blind ABX type of test could be conducted without controversy as a true random sample of the human population would include only a very few of the golden-eared-ones.
I'm afraid I didn't make my point clear. I didn't mean to say that if distortion is sufficiently less audible, the spectral distribution doesn't matter. I mean to say it does not matter if there is no audible distortion - at all. Possible counter positions might be that "any distortion no matter how low is audible even if an arbitrarily long string of zeroes preceeed the first non-zero digit". Or- "even if you could build amp with literally NO distortion (not "low" or "hyper low" but "NO") there exists some spectral distribution of audible distortion that sounds better. A third one would be "the current state of the art is such that no amp can be built without audible distortion".
Of the three, I would say the third is the most pausible and may be demonstrably true, but I no of no one who has explicitly tested that proposition.
The second is somewhat related to the first but easily degeneraes in to agrument ad absurdum.
The second is most likely a mater of taste and would let both of us walk away feeling righteous. One might, of course, assert that, the implied taste preference is universal or virtually so. That could most convincingly be demonstared by appropriately conducted blind listening tests. Since it is a universal proposition not just limited to the audiophile elite a double blind ABX type of test could be conducted without controversy as a true random sample of the human population would include only a very few of the golden-eared-ones.
"I tell you this truthfully, it is possible to hear slew rate problems measured at hundreds of kilohertz."
How many hundreds of kilohertz? Ten, that would be a Megaherz? Can I hear Megahertz? Gigahertz?
Oh my!
How many hundreds of kilohertz? Ten, that would be a Megaherz? Can I hear Megahertz? Gigahertz?
Oh my!
I've been listening to my cheap gainclone amplifier all day. It's been very enjoyable, I might add.
Then I skimmed through much of this thread.
Upon returning to the music, it still sounds great!
Then I skimmed through much of this thread.
Upon returning to the music, it still sounds great!
Sam9,
Seems this thread has taken a familiar direction into the murky depths of THD, IMD and the like. My own non-verifiable experience favours the sound of tubes and hybrids (yes Joe, I'm no stranger to those...) over solid-state, as a general rule. The only measurable differences which SEEM to explain this are the THD but more specifically, the spectral distribution of THD.
As to hearing IMD events in the "hundreds of kilohertz", well I think that's probably bull, for a similarly non-verifiable subjective reason. Namely that the tube amps which I have enjoyed mostly had crummy IMD's and very average slew rates (as I recall). That is, the few whose figures could be found. None came close to the cleanest solid-state examples anyway. Maybe that ABX suggestion of yours could shed some light.
Getting back to the thread topic though, I must say that this endless cloning furore is getting tedious. Yes, they're cheap and cheerful entry-level diy solutions. No they don't sound offensive. But neither do I think they have advanced the state of the art anywhere nor moved any sonic benchmarks.
In fact their greatest achievement is not sonic although it certainly is "audible", reverberating through the pages and threads of this forum. But my own cynical interpretation is that their ease of implementation has simply opened the window of discussion and diy to those who might be intimidated by the geater complexity of discrete efforts. I just hope those folk don't see clones as ends in themselves, but rather a convenient stepping stone into the bigger and more challenging world of diy audio.
Unfortunately their ease of use has also opened a pandora's box of commercial opportunists eager to make a buck for spending a cent. And I can cite numerous examples. But instead of seeing this for what it is, the cloners seem to feel "vindicated" in their choices. Well, feel as you will but believe me, it's about the buck... Only that. Not the sound.
At some point when enough parallel clones, buffering solutions, expensive passive parts and aerospace casework have been tossed around I think reality will resurge, much as tubes have regained credence and pride of place after the grungy solid-state wars of the 70's and 80's. And for me that reality is to control all aspects of discrete design and to optimize each stage as best I can. Or at least to have something which can be enjoyed and tweaked by subtle modification here and there. Clones don't afford this luxury.
I've said enough.
Seems this thread has taken a familiar direction into the murky depths of THD, IMD and the like. My own non-verifiable experience favours the sound of tubes and hybrids (yes Joe, I'm no stranger to those...) over solid-state, as a general rule. The only measurable differences which SEEM to explain this are the THD but more specifically, the spectral distribution of THD.
As to hearing IMD events in the "hundreds of kilohertz", well I think that's probably bull, for a similarly non-verifiable subjective reason. Namely that the tube amps which I have enjoyed mostly had crummy IMD's and very average slew rates (as I recall). That is, the few whose figures could be found. None came close to the cleanest solid-state examples anyway. Maybe that ABX suggestion of yours could shed some light.
Getting back to the thread topic though, I must say that this endless cloning furore is getting tedious. Yes, they're cheap and cheerful entry-level diy solutions. No they don't sound offensive. But neither do I think they have advanced the state of the art anywhere nor moved any sonic benchmarks.
In fact their greatest achievement is not sonic although it certainly is "audible", reverberating through the pages and threads of this forum. But my own cynical interpretation is that their ease of implementation has simply opened the window of discussion and diy to those who might be intimidated by the geater complexity of discrete efforts. I just hope those folk don't see clones as ends in themselves, but rather a convenient stepping stone into the bigger and more challenging world of diy audio.
Unfortunately their ease of use has also opened a pandora's box of commercial opportunists eager to make a buck for spending a cent. And I can cite numerous examples. But instead of seeing this for what it is, the cloners seem to feel "vindicated" in their choices. Well, feel as you will but believe me, it's about the buck... Only that. Not the sound.
At some point when enough parallel clones, buffering solutions, expensive passive parts and aerospace casework have been tossed around I think reality will resurge, much as tubes have regained credence and pride of place after the grungy solid-state wars of the 70's and 80's. And for me that reality is to control all aspects of discrete design and to optimize each stage as best I can. Or at least to have something which can be enjoyed and tweaked by subtle modification here and there. Clones don't afford this luxury.
I've said enough.
DrG said:
OK.
But I DO like the sound of a tubed gainclone, I don't need measurements to tell me that and endless discussions doesn't change anything. My observations are mine, I have never imposed them on anyone else. It just happens that there are significant numbers that agree with li'l old humble me. Does that please me? I think you know. You cannot take that away from them or me and you won't. This thread has become 'sceptics corner' and let that be. But maybe you will allow this ol' tube guy to choose his own foibles. You should hear my real full-on tube amps and then your blues mght be banished.... 🙂
But there is room for gainclones out there and especially the VIGC, and maybe other versions not yet dreamed off. Are they high-end? Absolutely!!! I know of tubes that sounds less involving and yet still regarded high-end. Are they perfect? No! Are they better than most stuff out the? The VIGC (you will notice I'm not even saying JLTi) is.
Is the reason why certain individuals scorn them the same reason why some are condescending of anything that isn't Single-Ended? I have friends who are bit like that (yes, they are still friends). But a little patience and a demo here and there and a reluctant nod... maybe it just takes a little time and a mind ready to absorb other realities.
In the meantime I will enjoy bith my A$35,000 reference amps and then switch to my version of the VIGC, and often in front of friends or visitors, point to the 20:1 difference with a smile. Ah... such is life.
Joe R.
- Status
- Not open for further replies.