I sim my last amp and get THD = 0.000625% at 262W/8Ohm, 20kHz. I use 77VDC PSU. 😎 Bias current only 30mA. Simple amplifier with "sophisticated" compensation.
I think most of audiophile like an effect box. I know a Indonesian famous sound engineer, he like lower distortion of an amplifier.
Another sound engineer suggest me to use different amplifier to different purpose, like editing and mixing in studio or fun listening or serious listening in a room.
I still amaze to how complex human psycho acoustic is....
I have a simple amp with less than 1 ppm distortion at 100w.
But can you have 0.005% from JLH69 topology?
All amps are 'effect box' by my definition. And yes, I don't like what I hear with most JLH.
If you read my previous posts you will understand my views. All amps are effect boxes. Even amps with ultra low thd.
Why people like amps? What characters are liked from amp sound? There are many! As entrepreneur you must understand what is liked by the majority of your market. As an individual you should know what you like.
Those who like 'effect box' imo tend to have better ears. It is just not easy to create amps having the best of both worlds.
Last edited:
Member
Joined 2009
Paid Member
That's why we're all here, addicted to building amplifiers and suffering from 'upgrade-itis' because we never quite get there
At least we know that popular amplifiers have something 'right' which appeals to people so when we clone their approach we have a chance to learn and a good probability that we'll like the end result.
You remind me to a member who has "golden ears". He "disappeared" few month ago.
Maybe you've not read much Naim clone discussion here in recent years - not that it's all worth reading but there is some background to my understanding which can be gleaned there on occasion. This from mjona at #1885, regarding a NAP200 clone in the big thread, could spark an interesting if patchy read, spread over many pages. here.
If you follow Pinkfishmedia forums you'll find more Naim design and DIY discussions, with contributions from some UK manufacturers too. It's a good place to read and check leads out for credibility, if your interest is in real amplifiers as much as an intellectual one.
This confused quotation from JV in 1999 hides some other clues:
"In the very early years, the actual transistor types did vary, but the planar type Motorola pre-drivers and the Ferranti drivers are still the same as those we used in '74. The input pair hasn't changed either, and these have to be very closely matched, because this is where the feedback meets the signal. There are no circuit configuration changes of any sort."
The LTP match though, is not in my experience an equal one - just enough (~10% Hfe difference) to correct for the LTP imbalance because there was one in the several amplifiers I've worked on and checked out in past years and the transistors fitted appeared to be matched for <30mV offset rather than for how one would normally interpret the claim.
What's really required is the harmonic spectrum plot of the IPS, with and without Hfe balance to see the shift in odd/even harmonic dominance. Self's method is not something I can still set up properly and I'm not about to assemble it again from scratch with an original, agency maintained NAP 140 just to make a point.
Member
Joined 2009
Paid Member
From what I've read from many sources (although some may just be a derivative of a few sources) our ears are not that adept at detecting differences in 2nd harmonic distortion - meaning that if you increase or decrease the LTP matching you will not be sensitive to changes in the level of the 2nd harmonic unless it changes by a relatively large amount. However, it seems that the key is to ensure you have enough 2nd harmonic to mask higher order harmonics.
Nelson Pass said he discovered that about 1/3 of people prefer dominant 2nd harmonic distortion, another 1/3 prefer dominant 3rd harmonic and the rest didn't mind if 2nd or 3rd.
The key is that the harmonic profile is dominated by low order distortion products, 2nd and 3rd. I expect that a perfectly balanced LTP compensates the 2nd too much, leaving the 3rd as overly dominant. You want to have 2nd of similar order to the 3rd. An LTP can be operated to acheive that.
These 2nd and 3rd low-order harmonics act to mask the higher orders. SET amps are famous for this. Hiraga popularized this approach. But you want the higher order harmonics to be low and to fall off fast as they are much more objectionable.
Simulations of my clone amplifier show dominant 2nd order, then 3rd and after that it falls off quite a lot.A supposedly ideal scenario. The alternative, if you can't achieve this profile of distortion is to go in all guns blazing and get all distortion products down as low as possible. So far, I've found both approaches produce amplifiers that are nice to listen to, but they do sound different.
Nelson Pass said he discovered that about 1/3 of people prefer dominant 2nd harmonic distortion, another 1/3 prefer dominant 3rd harmonic and the rest didn't mind if 2nd or 3rd.
The key is that the harmonic profile is dominated by low order distortion products, 2nd and 3rd. I expect that a perfectly balanced LTP compensates the 2nd too much, leaving the 3rd as overly dominant. You want to have 2nd of similar order to the 3rd. An LTP can be operated to acheive that.
These 2nd and 3rd low-order harmonics act to mask the higher orders. SET amps are famous for this. Hiraga popularized this approach. But you want the higher order harmonics to be low and to fall off fast as they are much more objectionable.
Simulations of my clone amplifier show dominant 2nd order, then 3rd and after that it falls off quite a lot.A supposedly ideal scenario. The alternative, if you can't achieve this profile of distortion is to go in all guns blazing and get all distortion products down as low as possible. So far, I've found both approaches produce amplifiers that are nice to listen to, but they do sound different.
Last edited:
True.
Thanks, I'll wade through. Yes I build real amps.
The question is whether this collector resistor is there for adding distortion or whether it is in there for an entirely different purpose and any distortion it adds to the LTP is a trade-off. In either case you will get a different spectral content at the output but may be none the wiser.
Bimo,
0.000265% is remarkable.... congratulations! Do you consider it a better amp than earlier amps? Does it 'engage'?
DD,
I read all your posts, and do you the credit of my respect for your comments.
But I sit somewhere with Bimo, Ian, and Bigun. I agree with NP, I think he knows much more than most of us and I am very aware that when we do not fully understand we hypothesise, that is, we set up belief systems which we tend to defend in an amusing but sad way. Religion is not just about theology, after all. I believe that what we measure is missing something that the ear does pick up. I suspect that a good amp is one that does nothing wrong at the limits; many amps behave badly at clip, or close to clip, or in heavy intermodulated signals. I believe that components do have some influence, and I am convinced topology is something to do with this. I do ask questions; I do not throw stones. What we all have in common is a relatively open mind.......
HD
0.000265% is remarkable.... congratulations! Do you consider it a better amp than earlier amps? Does it 'engage'?
DD,
I read all your posts, and do you the credit of my respect for your comments.
But I sit somewhere with Bimo, Ian, and Bigun. I agree with NP, I think he knows much more than most of us and I am very aware that when we do not fully understand we hypothesise, that is, we set up belief systems which we tend to defend in an amusing but sad way. Religion is not just about theology, after all. I believe that what we measure is missing something that the ear does pick up. I suspect that a good amp is one that does nothing wrong at the limits; many amps behave badly at clip, or close to clip, or in heavy intermodulated signals. I believe that components do have some influence, and I am convinced topology is something to do with this. I do ask questions; I do not throw stones. What we all have in common is a relatively open mind.......
HD
Last edited:
I was also disappointed when I built a JLH. Only my Aleph 30 has proved to be a greater disappointment.
Wrt the Naim clones. They can sound remarkably different depending on the PS, boards and components used. Naim's own amps can also sound quite different, so there's no standard to compare to. However if it's possible to DIY something that sounds as good as a NAP 300 then I'm interested!
There are some interesting, engineering based comments by Greg Ball (ex-amplifierguru here) who is well qualified to comment on audio amplifiers from both technical and SQ perspectives. Whether we agree with his priorities is not really relevant.
This thread, on re-working the NAP along technically based lines, also reveals a better understanding of what goes on in the original:
http://www.ska-audio.com/Forum/YaBB.pl?num=1264945548/10
Reply#10 kicks off some helpful comments and observations within the scope of the original design and suggested mods. I suggest others interested in Naim's design techniques read the thread too and glean what is on offer there, at least in the earlier pages.
Last edited:
Bimo wrote 0.000625%. That's less than 1 ppm. Simplest circuit will be achieved by CFA topology.
Many low THD amps do not 'engage' because they sacrifice 'damping' and other parameters for meaningless distortion.
Thanx HD but I don't understand if you think there is disagreement between us.
Yes, he knows much more than most of us, but I don't understand which part that you are 'in agreement' with him. About the distortion profile? I guess so.
I don't understand this comment either. Regarding THD and non-THD I'm aware that there are these 2 groups. If these are religions, do you really believe that there is no better alternative group?
The question is, what have we measured?
Several posts earlier I mentioned a condition where a high THD amp can sound better than a low THD amp when the change in voltage (at listening SPL) doesn't followed by THD increase but the opposite. This is still related to the clipping behaviour. But if the amps are high power and our listening SPL is far from the clipping, I found it okay to have bad behaviour at clipping.
Why bad behaviour at clipping should be aimed in a design? Often there is benefit for having a high gain amp such that clipping behaviour is not nice. I have seen people design an amp where the max output is at the full POT opening and the clipping is so nice. This is to mho is not right (unless we listen to SPL close to clipping).
It is not 0.000265% but 0.000625%.
I do not build it yet. A friend now is making PCB design for it.
I just want to design an simple amplifier with lowest possible THD and still have good harmonic profile, at least in simulation. Definitely, it is not CFA topology although it have high slew rate (122V/uS).
I designed many amplifiers with many different topology and of course many design philosophies. I am an engineer, and technology is not my religion 😀 😎
And a good engineer, Bimo....... hats off......
DD,
Certainly not. We are discussing audio, but I do not wish disagreement. Gotong royong!
DD,
Certainly not. We are discussing audio, but I do not wish disagreement. Gotong royong!
I think, an amplifier should not change the "character" with different volume setting before it clip.
But I agree, when it clip it should be "nice". No sticky and no oscillation.
Member
Joined 2009
Paid Member
Greg Ball released his SKA 150W design into the public domain. I liked it. I built a clone, with a couple of improvements to the thermal control. I designed a SOTA pcb with SMD parts that was very compact, good to r.f. with very nice mechanical layout. I shared my progress with others in his forum. I had oscillation problems on first attempt and asked for his input which I received. I then posted a final version, working perfectly. Greg posted questions about my design and asked how I achieved it. I answered his questions, it was great to have the attention of the guru that developed that amplifier. Very quickly after showcasing my results he deleted all my posts back to the point where my design was not working - removing all the successful results and explanations about how I achieved them. He then closed the thread. Others who were participating in his forum contacted me off-line and they couldn't understand his behaviour either. I no longer have any interest in reading information on his forum.
Gareth,
Many years ago I was approached by Greg to buy his design.
I declined, simply because I felt I had good circuits of my own. He was miffed and then began to trash my skills and products. I resisted fighting back for a few months, then relented mildly. He went beserk, and subsequently was thrown out of the forum (and another, actually). The situation would have been ridiculous if it had not been so commercially serious. He did sell a lot of his amps, which were quite good, but I never liked using output devices in collector drive.
He had some tragedy in his family; his wife became seriously ill, life can be cruel. I heard of him living in Jakarta for a time, railing about the Indonesian internet access. He seemed to blame all his problems on others, a telling sign. Perhaps we should be circumspect with a clever guy who lost his way.
HD
Many years ago I was approached by Greg to buy his design.
I declined, simply because I felt I had good circuits of my own. He was miffed and then began to trash my skills and products. I resisted fighting back for a few months, then relented mildly. He went beserk, and subsequently was thrown out of the forum (and another, actually). The situation would have been ridiculous if it had not been so commercially serious. He did sell a lot of his amps, which were quite good, but I never liked using output devices in collector drive.
He had some tragedy in his family; his wife became seriously ill, life can be cruel. I heard of him living in Jakarta for a time, railing about the Indonesian internet access. He seemed to blame all his problems on others, a telling sign. Perhaps we should be circumspect with a clever guy who lost his way.
HD
Well, as I hope you can see, I've tried to avoid mixing personality issues into the topic. OK, so I triggered some unhappy memories there but I hope you won't let that override any merit of the info.
I recall the events and I imagine that having Gareth's SKA clone shown on both forums got up Greg's nose since he didn't exactly leave this forum on good terms and this is where you "reside" if I can put it that way.
Greg has been involved with design for a few high profile audio companies, also had papers published and recognized - even referenced by one of the greats of the industry, so not an audio hack but a local star. Here's an old bio: introducing Greg Ball (aka SKA) - New Member Introductions - StereoNET However, as the history of many pros here illustrates, contracts don't last forever and designers can also be quite ...err...sensitive about their IP, where it is seen and what it's associated with. Their professional image and keeping the right sort of company, acknowledging peers and never being seen to be out of touch with developments is important to them. I call their world a technocracy - a society ordered by technical prowess that doesn't seem to have equivalents in many other scientific fields.
From our point of view, there's no excuse for what looks like prima donna behaviour but frankly, I'm more interested in the facts and I shouldn't care who made the analyses. I may no longer like the attitudes now assumed by some other guys here either; those who have reached some level of recognition here for their designs, simulations and builds etc. and now find us unworthy of joining in discussion. Still, as long as they come up with the goods on occasion, they should get my vote too.
I recall the events and I imagine that having Gareth's SKA clone shown on both forums got up Greg's nose since he didn't exactly leave this forum on good terms and this is where you "reside" if I can put it that way.
Greg has been involved with design for a few high profile audio companies, also had papers published and recognized - even referenced by one of the greats of the industry, so not an audio hack but a local star. Here's an old bio: introducing Greg Ball (aka SKA) - New Member Introductions - StereoNET However, as the history of many pros here illustrates, contracts don't last forever and designers can also be quite ...err...sensitive about their IP, where it is seen and what it's associated with. Their professional image and keeping the right sort of company, acknowledging peers and never being seen to be out of touch with developments is important to them. I call their world a technocracy - a society ordered by technical prowess that doesn't seem to have equivalents in many other scientific fields.
From our point of view, there's no excuse for what looks like prima donna behaviour but frankly, I'm more interested in the facts and I shouldn't care who made the analyses. I may no longer like the attitudes now assumed by some other guys here either; those who have reached some level of recognition here for their designs, simulations and builds etc. and now find us unworthy of joining in discussion. Still, as long as they come up with the goods on occasion, they should get my vote too.
- Status
- Not open for further replies.