Theoretically it is easy just to do a good thermal tracking of an amplifier. But to do this without changing the known optimal operation of the circuit is the issue. It is obvious (to me at least) that amplifiers of this bandwidth and topology have narrow window for good sounding operating points.
Member
Joined 2009
Paid Member
Ian - I don't think the transistor type has that much influence on the speed of the thermal feedback control - just to nit pick - the real issue in my opinion is the one you mention - dangling the Vbe device in thin air inside the enclosure is the real issue in the original design. It's simply poor engineering. I fixed that in my clone with the benefit of Spice simulations to allow a first-time-right solution.
Just observe that the negative part of a quasi is a CFP, then you will remember.
Design is about making compromises. That's especially true in amplifiers with NAP's bandwidth and topology. You can make small change to the circuit (eg to make the thermal tracking better) but will the sound be similar? Try it with the original/commercial NAP!
I don't think that the speed to reach optimal bias is not part of the compromises to make. Ime, it is part of the compromise. An out of the box solution is to design a dedicated cct that will sense the cold heatsink then QUICKLY heat up the heatsink and stop when a certain temperature (or time) is reached. Poor engineering, huh?
Member
Joined 2009
Paid Member
I'm at a loss to identify any compromise that could explain the choice of thermal management of the original design other than not having the cost associated with mounting a Vbe multiplier device onto a solid surface where it can properly 'measure' the temperature of the output stage and also, no longer being able to tell customers that they need their amp to warm up for half an hour before it's stable just like fancy Class A amplifiers.
The 'speed' of the thermal control loop need not be super fast because the heatsink acts like a low pass filter for temperature changes.
Still, it's also no big deal. It's part of the charm and character of the original design. I decided to 'fix' it because I wasn't trying to clone the amplifier exactly - for starters I used a different pcb layout and I used modern components that I was able to buy new and these for sure are different from the original production units. I accept the clone is not 100% but my experience of the sound is an uncanny match to published reviews for the Naim and that's good enough for me to enjoy myself making it. If I was to insist on an exact copy I'd just go out and buy a real Naim amp on the used market - they keep their value well so it would not cost me anything that I'd worry about. I have thought that as my interest in firing up the soldering iron wanes over the years I might switch to collecting the originals but I wouldn't learn much that way.
The 'speed' of the thermal control loop need not be super fast because the heatsink acts like a low pass filter for temperature changes.
Still, it's also no big deal. It's part of the charm and character of the original design. I decided to 'fix' it because I wasn't trying to clone the amplifier exactly - for starters I used a different pcb layout and I used modern components that I was able to buy new and these for sure are different from the original production units. I accept the clone is not 100% but my experience of the sound is an uncanny match to published reviews for the Naim and that's good enough for me to enjoy myself making it. If I was to insist on an exact copy I'd just go out and buy a real Naim amp on the used market - they keep their value well so it would not cost me anything that I'd worry about. I have thought that as my interest in firing up the soldering iron wanes over the years I might switch to collecting the originals but I wouldn't learn much that way.
Last edited:
I thought that was exactly the point I was making. The NAP 200 and probably other models in that series and maybe some in the olive series that preceded it, were finally reducing the thermal delay problem to a tolerable one by closer proximity to the heatsink.
It was a stubborn acceptance of the long delay and lame excuses by both the manufacturer and defensive owners that turned me off Naim products long ago. DIY experiments and fun here in recent times have been a great education in what could have easily been done 30 or more years ago, just by examining what the competition was already achieving at little extra cost.
Member
Joined 2009
Paid Member
Let's talk about the real NAP design. I don't know about the amps but if you really think that mounting the Vbe transistor to an aluminum plate that is connected to either the driver or the output will solve the problem (without changing anything else), then you are right. Do you think any amp manufacturer might need decades to think about such solution?
I can say tho there are cases where you cannot just do like the above. Look at the driver-output pair of the negative part (CFP). The bias window is so narrow. There is one short-range optimum bias where below this the sound is horrible and above this the sound is not so good either. In other words the bias is sensitive. Then combine that situation with another where you are not allowed to speed up the bias by increasing the Vbe of the driver...
I don't believe that quasi-complementary designs can utilise either driver as a temperature reference. QC is not, thermally speaking, a 50/50 hybrid of CFP and EF output stages and the better reference is an attenuated version of output transistor heatsink temperature. I think Bigun already demonstrated that works rather well with his TGM10 project.
You may not have read yet, that Naim Audio was founded by an amateur, a self-taught but clever guy who modified an existing textbook design to sound the way he liked it. There were no design engineers originally and the "designer" was also the CEO for many years so you have the likelihood of some unusual ideas and innovation that only JV would have dreamed up over the period where he exercised full control and made the decisions. That's where the magic sound and unique audio systems began, though.
So, don't expect the rigorous analyses and measures in modern electronics to have applied to early Naim gear. However, there were plenty of misleading and imaginative claims, silly analogies, wrong assumptions, over-expensive add-ons to systems that required you to buy stacks of accessories, more black boxes, special cables and connectors at highly inflated prices....all this just to hear an audio playback under recommended conditions.
This and flat-earth audio speak were all part of the Naim charm which thankfully has been given a rest and we can get back to humdrum audio speak and clones which are much cheaper and don't need a stack of special accessories.
You may not have read yet, that Naim Audio was founded by an amateur, a self-taught but clever guy who modified an existing textbook design to sound the way he liked it. There were no design engineers originally and the "designer" was also the CEO for many years so you have the likelihood of some unusual ideas and innovation that only JV would have dreamed up over the period where he exercised full control and made the decisions. That's where the magic sound and unique audio systems began, though.
So, don't expect the rigorous analyses and measures in modern electronics to have applied to early Naim gear. However, there were plenty of misleading and imaginative claims, silly analogies, wrong assumptions, over-expensive add-ons to systems that required you to buy stacks of accessories, more black boxes, special cables and connectors at highly inflated prices....all this just to hear an audio playback under recommended conditions.
This and flat-earth audio speak were all part of the Naim charm which thankfully has been given a rest and we can get back to humdrum audio speak and clones which are much cheaper and don't need a stack of special accessories.
Last edited:
But you can't ignore the fact that some people like Ultima Legione and Huggygood really liked their NAIMs
Can you give examples of these pseudo-science?
You know, I don't really like talking about sound quality because I know exactly that people have different hearing and tastes too. For example, I know a cheap headphone that performs better than many headphones 10x the price. I recommended the product to some people and I asked "Can you hear the difference?" The answer was "No". Even if they can hear the difference they still don't hear what I hear.
I know something crucial in audio design, but then what is the point? Even if I demo the proper one along with the wrong one, people can still prefer the wrong one. But I'm happy because my audio system (the music) makes me happy.
"Pseudoscience consists of statements, beliefs, or practices that are claimed to be both scientific and factual, but are incompatible with the scientific method." - Wikipedia
When people explain their audio experiences as better or worse than some previous experience, they may believe their impressions to be factual when it can be shown that people are easily biased by suggestion, false information etc. along with a desire to agree with our peers. That's why we use double-blind testing, often associated with ABX testing, if we want realistic results which will show a credible spread of results for audio listening tests, if the sample is large enough. This usually winds up with smaller differences between options which and may not please anyone but is at least an honest representation.
Why Do Audiophiles Fear ABX Testing?
i agree with you and I still checked it not long ago with a lousy amp and a very nice amp.
two visual tests followed by two blind tests and I saw that the visual was largely part of the test because although the lousy amp was really better, when it was visible is the beautiful amp that won the test, while blind, it was ugly that won 100% .
for my part, knowing that the auditory memory is the least long, I use for nearly twenty years the same source, the two same cd ("flamingo" Grappelli / Pretrucciani and "uni vers l'uni" Michel Jonasz) and the same pair of speakers to do my test
two visual tests followed by two blind tests and I saw that the visual was largely part of the test because although the lousy amp was really better, when it was visible is the beautiful amp that won the test, while blind, it was ugly that won 100% .
for my part, knowing that the auditory memory is the least long, I use for nearly twenty years the same source, the two same cd ("flamingo" Grappelli / Pretrucciani and "uni vers l'uni" Michel Jonasz) and the same pair of speakers to do my test
Ah, so you were referring to that. We are all humans, with unique combinations of strengths and weaknesses. One group is not superior than the other and i know that for sure. If one group feels superior than the other, i call that bias.
Johnego, you are my hidden brother
I stopped looking for the "absolute truth" a long time ago and I keep only the devices I like to listen to and / or watch.
And it is true that I particularly like the airy and relaxed side of my naim clone.
Hi there brother
I don't even understand what 'absolute truth' is
But yes, i think you should keep the speakers if it never let you down.
Pseudo-science is an interesting topic.
You often get this conflict in audio between those who think they are applying science properly and those who think they are listening properly. Both make mistakes.
I think the mis-application of science is the more pernicious because advocates use reference to authority to validate their assertions instead of doing the hard work.
You often get this conflict in audio between those who think they are applying science properly and those who think they are listening properly. Both make mistakes.
I think the mis-application of science is the more pernicious because advocates use reference to authority to validate their assertions instead of doing the hard work.