As you might expect if the transistor pairs were selected on test. It means that in production, when there are many pairs to assemble, you would preselect them as pointed out in one of the more recent Naim factory tour videos. In your case, the differentials are quite consistent at 437/455 = 96% or 688/711 = 96.7%. The actual hFE numbers are unimportant, as long as there is enough for the particular amplifier design.
To change the subject a little and return for a moment to a previous discussion about the VAS transistors and sound quality, I thought some of you might be interested in what goes on in newer Naim models. I've been interested in the Nait 5i because it looks like it would suit my humble needs but not everyone likes its sound quality so I was pleased to find its schematic already linked/posted. Here it is again.
.
Well, the schematic design is still much the same in SMT as it was in original Nait, thru-hole parts days. Get this though: The VAS transistors are still the same beefy types but now in SOT-223 size, on steroids and cascoded, it seems. That's all good so far but the Cob of FZT853/953 is double, at 65pF! That may be acceptable for a driver transistor but look at any other DIY or commercial design here on DIY audio and all the design experts are using as tiny as possible capacitance in voltage amplifier stage. The KSC3503/KSA1381 transistors and others often specified for that application have only a few pF Cob.
Take a look at your own solid-state linear amplifiers, kits or maybe your next dream project, What type of transistor(s) is there in the VAS?
. Well, the schematic design is still much the same in SMT as it was in original Nait, thru-hole parts days. Get this though: The VAS transistors are still the same beefy types but now in SOT-223 size, on steroids and cascoded, it seems. That's all good so far but the Cob of FZT853/953 is double, at 65pF! That may be acceptable for a driver transistor but look at any other DIY or commercial design here on DIY audio and all the design experts are using as tiny as possible capacitance in voltage amplifier stage. The KSC3503/KSA1381 transistors and others often specified for that application have only a few pF Cob.
Take a look at your own solid-state linear amplifiers, kits or maybe your next dream project, What type of transistor(s) is there in the VAS?
Last edited:
In Nait 5, caps are bigger than nap's. Do you know the type of caps (C301, C321,C313 and perhaps c305,c306)? I'm asking that, yesterday I tried to change my nap's input cap (tantal) to polyester high voltage cap, result interesting. I don't wait that much change (better result).
my transistors BC550C/BC239C, ZTX753, ZTX108,ZTX653, MJE15030, MJ15003. What they said/wrote
my transistors BC550C/BC239C, ZTX753, ZTX108,ZTX653, MJE15030, MJ15003. What they said/wrote
Last edited:
Your guess on the smaller capacitor types is as good as mine. However, there is a good, high resolution pic. of Nait 5i-2 here: https://www.diyaudio.com/forums/class-d/340304-hacking-naim-nait-5i-2-hypex-nc-module.htmls
It is not an identical amp to the 5i but you still should be able to identify what caps do what they are supposed to in the power amplifier section at least. The types are common electrolytics in SMD and through-hole types. Tantalums are the yellow chip style components .
There is a mixture types according to capacitance value and size, more than any sound effects we may imagine they contribute. The large electrolytics will likely be Rubycon types.
It is not an identical amp to the 5i but you still should be able to identify what caps do what they are supposed to in the power amplifier section at least. The types are common electrolytics in SMD and through-hole types. Tantalums are the yellow chip style components .
There is a mixture types according to capacitance value and size, more than any sound effects we may imagine they contribute. The large electrolytics will likely be Rubycon types.
Hmm...404 error on the link to our own pages??. I'll try again: Hacking a Naim Nait 5i-2 with Hypex NC module?
It raises some questions. I'm sure JV would not approve of this design. It breaks a lot of the design axioms of the earlier amps. I personally would not buy this because I can see many weaknesses in it. It looks almost as if Naim hired a college graduate to design it...or maybe Douglas Self
. I doubt they lost the formula, but if not they have deliberately decided to compromise the sound quality for some reason. Could be to ensure it fits in the price/performance curve or maybe its cheaper to manufacture. It has more parts so it would only be cheaper if there was less, manual select on test.
The accuracy of the input differential amplifier, or long tailed pair, is crucially important in a NFB design. Any errors will be unfixable. There are plenty of mathematical explanations of LTPs around to explain what conditions, like transistor gain and Ic and so on make for the smallest error.
the thing about "name and business buyers" is that they rarely think in the right ways.
it would be me, I would have taken the exact diagram of the first naims, stuffed all this with SMD in a pretty new old fashion designer shoebox with bluetooth and everything and everything and neither seen nor known I confuse you
today what matters is the form, not the content, the image, not the content of the image.
almost all people don't give a damn about JV as their first diaper, what they want is a "slightly exotic" brand in a pretty box.
Misere, I'm going to go to sleep ...
it would be me, I would have taken the exact diagram of the first naims, stuffed all this with SMD in a pretty new old fashion designer shoebox with bluetooth and everything and everything and neither seen nor known I confuse you
today what matters is the form, not the content, the image, not the content of the image.
almost all people don't give a damn about JV as their first diaper, what they want is a "slightly exotic" brand in a pretty box.
Misere, I'm going to go to sleep ...
schematic graphics
I think people put a resistor there for "symmetry" without considering that the other side is actually pulled up by the VAS BE junction, 0.65V from the rail. But the higher Z on that collector creates an unnecessary "Miller" capacitance. You could argue that it reduces the power dissipation of the transistor; not really an issue. Remember this is a trial and error design so if it didn't cause any problems then it stayed.
Something similar is the diode in the emitter of the PNP driver, although that could actually be useful for cross coupling if they only used it that way, but rarely do.
As I said before, there is no accounting for the "magic" sound. And DIYA is full of beautiful schematics that have serious problems, like the symmetrical IPS amps.
I think people put a resistor there for "symmetry" without considering that the other side is actually pulled up by the VAS BE junction, 0.65V from the rail. But the higher Z on that collector creates an unnecessary "Miller" capacitance. You could argue that it reduces the power dissipation of the transistor; not really an issue. Remember this is a trial and error design so if it didn't cause any problems then it stayed.
Something similar is the diode in the emitter of the PNP driver, although that could actually be useful for cross coupling if they only used it that way, but rarely do.
As I said before, there is no accounting for the "magic" sound. And DIYA is full of beautiful schematics that have serious problems, like the symmetrical IPS amps.
Naim or Blame it all on Steve Sells. He's the guy in a current Naim Audio video, who sits in front of the big screen, wearing a white dust coat and telling you "this where we make the design of the Naim amplifier". You never know, he may even have permission to defend his work outside the confines of the official website/forum.
That forum is a claustrophobic, micro-managed and oh-so-formal place anyway so it might be refreshing to get out and about in the rough and tumble of DIY type discussion here. OK OK, that would be a miracle but I believe in Santa Claus too
Steve Sells – the man behind Naim’s electronics | What Hi-Fi?
That forum is a claustrophobic, micro-managed and oh-so-formal place anyway so it might be refreshing to get out and about in the rough and tumble of DIY type discussion here. OK OK, that would be a miracle but I believe in Santa Claus too
Steve Sells – the man behind Naim’s electronics | What Hi-Fi?
This diode/resistor combination has a specific role in any quasi-complementary amplifier. It's often referred to as a "Baxandall diode" after P.J.Baxandall who is usually credited with defining the application - same guy as the one who developed the Baxandall tone control system that is still used in analog audio circuits. See post #7 here: Help in design: adding a Baxandall diode across emitter in Quasi Complimentary Stage
I just read his interview and to be honest I don't know what to think.Naim or Blame it all on Steve Sells. He's the guy in a current Naim Audio video, who sits in front of the big screen, wearing a white dust coat and telling you "this where we make the design of the Naim amplifier". You never know, he may even have permission to defend his work outside the confines of the official website/forum.
That forum is a claustrophobic, micro-managed and oh-so-formal place anyway so it might be refreshing to get out and about in the rough and tumble of DIY type discussion here. OK OK, that would be a miracle but I believe in Santa Claus too
I do not know the man (in real terms) and it is easy to form a negative or positive image of a stranger, nevertheless I retained a passage:
"As an engineer it’s really hard to sacrifice specs, but we do it all the time. Generally, we’ll design on the bench from experience. We know what sounded good in the past so we’ll try those things again. Then when we’ve got a couple of different prototypes we’ll engineer those till they measure well. Then we go to the listening room and we play around with components to fine-tune the sound. Invariably, when you go back and measure things they’ve got worse".
39p/1.8k is for additional stability. Because NCC circuit has degeneration LTP resistors the open loop gain of the amp is much lower, which makes the amp more stable with its usual closed loop gain. So, they omitted that RC circuit.
Stability also depends on PCB topology and wiring. Cloning a circuit doesn't mean you'll make it same as the original.
Stability also depends on PCB topology and wiring. Cloning a circuit doesn't mean you'll make it same as the original.
NCC200 is not promoted as Naim clone anyway. Avondale believe theirs is an upgraded design but their schematic is popular for DIY experimenters because it gets updated from time to time so that mostly common and easily found parts are specified. Still be careful with the issue date or you may find yourself with a different schematic to the one you began with.
The NAP250 schematic we find on the net, is unchanged from the original, issued more than 45 years ago, as I recall. This is what most Chinese kits are based on but as Ruwe mentions, they may not be exact copies either - many will omit the phase correcting networks referred to. If you wanted to buy the PCBs or a kit of parts for the NCC200 though, the price would be many times the price of Chinese kits. So you will have to work through what changes should or should not be made if you are using particular PCBs designed for a different clone version.
The NAP250 schematic we find on the net, is unchanged from the original, issued more than 45 years ago, as I recall. This is what most Chinese kits are based on but as Ruwe mentions, they may not be exact copies either - many will omit the phase correcting networks referred to. If you wanted to buy the PCBs or a kit of parts for the NCC200 though, the price would be many times the price of Chinese kits. So you will have to work through what changes should or should not be made if you are using particular PCBs designed for a different clone version.
It's called "phase lead compensation" if you want to look it up.In Naim's schematic in feedback path there are 27k // 39pf & 1k8. NCC schematic for 39pf & 1k8, wrotes "not necessary". what happened or changed on the schematic, how it happend not necessary? I'm not an engineer. If it's silly question, forgive me
I find his statement quite surprising. I thought they had a more objective process than this. Perhaps some of the original recipe got lost. Or perhaps JV didn't have a complete recipe.