traderbam,
You would not believe how many people there are who think that nothing but the most expensive is good enough! People without real knowledge and understanding but with deep pockets. I have contact with them almost every day. Two of them have invested two thousand euros in Kubala Sosna loudspeaker cables only. The sound they got from it is uninspiring. Wasted money.
You would not believe how many people there are who think that nothing but the most expensive is good enough! People without real knowledge and understanding but with deep pockets. I have contact with them almost every day. Two of them have invested two thousand euros in Kubala Sosna loudspeaker cables only. The sound they got from it is uninspiring. Wasted money.
Generally, that's right but close matching is questionable for a single pair of transistors in a quasi-complementary output stage like Naim's. Having transistors from the same manufacturing sequence (as in the same tube from the supplier) is a good start to consistent performance and sound quality though.
If you have only random parts and you are particular about sound quality , stereo imaging etc, a bad mismatch may cause problems. At least try to find same grade parts and/or get a match within 20% for Hfe.
None of this is essential just to get an idea of what Naim sound is about but as I suggest, if you are fussy about audio, then go the extra and do as much as you think you can afford without spending silly amounts of money on an untried DIY project.
a few years ago, we went (the complete team) to put a pair of grand utopia in a mansion in a rich neighborhood at the exit of brussels.
at that time, we were about 80000 € a pair, which is not nothing, and something like 200 or 250 kilos for each speaker.
after having put them in place, we ask the customer where is the stereo system (which was invisible) and he designates us a ridiculous little piece of furniture in "empire" type wood, my boss opens it and finds an akai AM U01 ...
Member
Joined 2009
Paid Member
At least the very expensive speakers ensure some of that money flows back into circulation!
I suspect wealth is an exponential distribution, wherever you sit on an exponential curve it looks the same (it's self-similar) so you feel poor when you look 'up' and rich when you look 'down'. This is why enough is never enough.
I suspect wealth is an exponential distribution, wherever you sit on an exponential curve it looks the same (it's self-similar) so you feel poor when you look 'up' and rich when you look 'down'. This is why enough is never enough.
With all these poorly mixed, over-compressed and over-processed recordings I see less and less reason to buying super expensive audio. But people do what they have to do. Sometimes making a statement is more self-gratifying than common or engineering sense.
Looking back, I realize that I spent probably the money for a brand new Mercedes E-class on DIY audio projects... I don't think that building, re-building, upgrading, updating, re-working, buying piles of test equipment and so on, sounds normal to most people either. By all means DIY audio is money spent without return of your "investment".
To me, though, it's all about the good music, personal satisfaction, learning something by doing it, and the pleasure of this hobby. DIY audio brought my interest in electronics and put me in my professional path. I play with better toys than I have at home and I'm well paid to do so. Could it be better if I just had spent these 60-70k on a stack of brand name audio and after that being ridiculed by 1% of the people that I have an average system and sound, and by the other 99% that I'm an idiot? 🙂
Looking back, I realize that I spent probably the money for a brand new Mercedes E-class on DIY audio projects... I don't think that building, re-building, upgrading, updating, re-working, buying piles of test equipment and so on, sounds normal to most people either. By all means DIY audio is money spent without return of your "investment".
To me, though, it's all about the good music, personal satisfaction, learning something by doing it, and the pleasure of this hobby. DIY audio brought my interest in electronics and put me in my professional path. I play with better toys than I have at home and I'm well paid to do so. Could it be better if I just had spent these 60-70k on a stack of brand name audio and after that being ridiculed by 1% of the people that I have an average system and sound, and by the other 99% that I'm an idiot? 🙂
Just a quick update on my DIY kit that started with a bare pcb board from ebay. I received most of the required parts and have completely assembled one channel (the second channel is waiting on a couple of caps).
The board does indeed have a ground plane. It would be much better to not have a ground plane and keep the 2 ground connections isolated on the board and only connect at the star point. This probably increases noise pickup.... It is hard to tell right now with all the flying leads and open chassis.
I had difficulty getting the boards biased correctly. One of the resistors is marked 2.2K but should be 1.2K.
The quiescent bias point does not seem to be critical. Using a spectrum analyzer to monitor distortion, I was able to set it over a wide range from about 12ma up with similar THD.
I plan to run it with a +/-35VDC power supply. In my initial experiments, the board works equally well at 35VDC and 40VDC (in terms of distortion, not total power). The critical input and VAS biases are set via current sources that are not referenced to the power supply voltage. I may need to adjust some of the component values in the SOR limit circuit.
Between the ground plane, the mislabeled resistor value and lack of documentation, I can not recommend this board.
The board does indeed have a ground plane. It would be much better to not have a ground plane and keep the 2 ground connections isolated on the board and only connect at the star point. This probably increases noise pickup.... It is hard to tell right now with all the flying leads and open chassis.
I had difficulty getting the boards biased correctly. One of the resistors is marked 2.2K but should be 1.2K.
The quiescent bias point does not seem to be critical. Using a spectrum analyzer to monitor distortion, I was able to set it over a wide range from about 12ma up with similar THD.
I plan to run it with a +/-35VDC power supply. In my initial experiments, the board works equally well at 35VDC and 40VDC (in terms of distortion, not total power). The critical input and VAS biases are set via current sources that are not referenced to the power supply voltage. I may need to adjust some of the component values in the SOR limit circuit.
Between the ground plane, the mislabeled resistor value and lack of documentation, I can not recommend this board.
If that's the Zerozone kit, the bias pot value is probably also wrong at 1k. I'm also unsure whether the kit designer figured that it was OK to confuse the design values between the standard NAP250 design that you find used for all models on the net, with the real NAP140 design or not. It probably depends on whether you use the right supply voltage for either model (+/-34V for NAP140, regulated +/-40V for NAP250). Goodness knows what a mess is likely when the kit information is either wrong or doesn't even exist.
That's because it's quasi. With quasi outputs upper half (Darlington) and lower half (Sziklai) requires different bias. Quasi can never have optimal bias. Therefore you have similar spectrum profile with different bias settings. In the older times it was usual to bias it to 50-60mA so that lower half is overbiased and upper half is underbiased.
Last edited:
In quasi stages, like the one Naim used, Darlington and Sziklai supertransistor have very different behavior. In the late sixties Baxandall tried to make them more similar with his "Baxandall diode". One way to bias them for proper operation is to use two bias circuits, one for each half of output. I do not know that anybody used that method. So, whatever bias you choose it's bound to be improper. There is not proper value. But I guess that's what makes quasies special. That's quasi sound, the sound of it's characteristic distortion spectrum. In fact it is better to have low bias to get more of that sound, if you like it in the first place. If you bias it deeper into class A, you get the sound of class A, but if you bias it low you get quasi sound.
A cursory look at any conventional, single BJT pair AB output stage, shows that you can't separate the DC bias currents from the AC signal and a similar DC current has to flow through both output transistors, rail to rail. Even with massive high-pass filtering, I doubt it's possible to design a practical output stage where the fixed bias currents of the output transistors are independent. This is essentially the same for quasi-complementary designs but the likely cost and complexity of separating the bias make it a pointless exercise where complementary devices are cheaply available.
Quasi-complementary distortion characteristics are a fundamental part of 20th century Naim sound and it doesn't seem to matter what type of power devices you fit as long as the switching speeds are fast enough - much as JV stated in interviews. I'm not suggesting Sanken's premium LAPT transistors are wasted on Naim amplifiers because they also represent the most linear high speed types available and IME, that changes some characteristics for the better.
Quasi-complementary distortion characteristics are a fundamental part of 20th century Naim sound and it doesn't seem to matter what type of power devices you fit as long as the switching speeds are fast enough - much as JV stated in interviews. I'm not suggesting Sanken's premium LAPT transistors are wasted on Naim amplifiers because they also represent the most linear high speed types available and IME, that changes some characteristics for the better.
I don't see any significant difference in bias current between top and bottom followers. The bottom one will additionally sink the base current of the top driver but this is some 3 orders of magnitude smaller than the total bias current.
That would be ironic because the reason JV said he used a quasi complementary output is because fully complementary darlingtons behave very differently.
Last edited:
The way I see it is that the only reasons for using quasi is to get it's distortion profile or if you have big stock of N transistors or for reasons of economy of scale (if it's company). The question is not how much current pass but what distortion profile has Darlington vs. Sziklai. If quasi was more linear than fully complementary Darlington there won't be necessary to produce P outputs at all. Quasi was used out of necessity when there was no P complements.
Yes, exactly. If I didn't allow for the existence of such small currents though, it would still draw comment.
Happy Halloween (or as appropriate) to all
Naim have complementary output stage amplifiers now, starting with NAP100, I believe. As you are in the high-end audio industry, what do you think of the sound quality?ivanlukic;5589173...... said:
Last edited:
I think that JV has done as we could when we come across an interresting circuit, that is to say to try to get the best at the time (at the time) where we do it.
he found this interesting RCA circuit and developed it until he got the result he wanted.
I do not think that it has asked more questions than it has about complementary or quasi complementary
he found this interesting RCA circuit and developed it until he got the result he wanted.
I do not think that it has asked more questions than it has about complementary or quasi complementary
I think that what we like about old Naim is the sound of specific distortions. After Naim embraced high linearity engineering principles the sound changed for worse. Now it's sterile, boring and tasteless. Like any other product based on the same high linearity principles. traderbam thinks that I am bashing quasi stages but I am not. I only want to say that quasi has some character that some people like in spite of it's non-linearity.
I am in the high-end audio industry only in a way that I have rich friends that change expensive systems every season.
I am in the high-end audio industry only in a way that I have rich friends that change expensive systems every season.