I really like E-I. Toroids are very different. It would be very interesting to get a NAP250 and use a 1000VA E-I. Without me wanting to say why I would imagine slam would be less whilst low level detail greater.
I agree!
However my amps are used for >100Hz in a semi-active 3-way system.
My bass speakers are flat to >>20Hz with a 1st order rolloff. They only "slam" if it's on the recording.
Actually despite my willy waving above, the character of the bass seems to be largely determined above 100Hz.
Anyway, I've exorcised toroids out of my entire system and I'm happy with the result.
BTW, Naim claim that that the benefit of big transfomers is the short conduction angle. They seem to say that the less time the amp is connected to the mains supply the better the sound. My approach is either brute force filtering (v difficult in practice with power amps) or side-step it by connecting it to the supply with a lower bandwidth transformer = EI core etc.
However my amps are used for >100Hz in a semi-active 3-way system.
My bass speakers are flat to >>20Hz with a 1st order rolloff. They only "slam" if it's on the recording.
Actually despite my willy waving above, the character of the bass seems to be largely determined above 100Hz.
Anyway, I've exorcised toroids out of my entire system and I'm happy with the result.
BTW, Naim claim that that the benefit of big transfomers is the short conduction angle. They seem to say that the less time the amp is connected to the mains supply the better the sound. My approach is either brute force filtering (v difficult in practice with power amps) or side-step it by connecting it to the supply with a lower bandwidth transformer = EI core etc.
Huge power supplies...can't help thinking it helps justify the product price. Music at normal household levels in normal speakers, even when unbearably loud, hardly consumes much power. A well-designed regulator with suitable RFI isolation will obviate the need for a gross transformer, IMO.
It is not your imagination. The room effect is so strong. But now you know that it is not from the speaker. If recorded this way it is good enough, in reality itu should be better.
You know this is a worthwhile discussion. Noboby needs to be right or wrong ( a rare thing at DIY Audio ). All we need say is the Naim circuit never was the bigger question. It was how Julian though the power supply was the true sound people were hearing. I remember in print Julian approximately saying " People say my amplifiers are 80% power supply, when the NAP250 that's not far from the truth". It's not important that it is true, it is what he thought. I strongly doubt that making money was all he wanted. He told me the Nait was his best product as it had more failiures as prototypes sonically. To quote him " I don't know how to make cheap stuff ". I asked if that was his dream" Oh yes, that's where the money is ". The best bit was Julian telling me in front of some of his management how he was surrounded by " yes men ". He said he tested them by saying " So you won't mind having it at home for a month" . Their faces dropped and that's how Julian got them to raise their game. The problem with sales staff is they tend to be "yes men". I always liked Doug.
BTW Baffles help with room problems, even no carpets although that's not recomended.
http://www.sakurasystems.com/reference/gaincard.html
http://www.sibatech.co.jp/fal/products.html
This was an amplifier I thought better than Naim in their style. It has a built in loudness circuit which I found to be OK. I used one with Isobariks and loved every minute. It went loader than NAP250. I suspect Isobariks were a happy accident with it. Just the right sensetivity so as not to make the loudness obvious.I think it was a feed forward type?
https://www.hifiengine.com/manual_library/rational-audio/z1.shtml
http://www.sakurasystems.com/reference/gaincard.html
http://www.sibatech.co.jp/fal/products.html
This was an amplifier I thought better than Naim in their style. It has a built in loudness circuit which I found to be OK. I used one with Isobariks and loved every minute. It went loader than NAP250. I suspect Isobariks were a happy accident with it. Just the right sensetivity so as not to make the loudness obvious.I think it was a feed forward type?
https://www.hifiengine.com/manual_library/rational-audio/z1.shtml
Last edited:
Thanks for elaborating. I agree: it is possible and necessary to correlate sonic differences to circuit characteristics although tricky sometimes as different circuit problems can give rise to similar sound problems. But this process is the only way I have found to prioritize. There is always a circuit issue behind every sonic issue. 😎
THD is more sophisticated a measure than, say, frequency response but it's still crude and hence the unsatisfactory correlation with sound once it is reasonably low.
Yes about transistors but not sure why compensation caps get such a bad rap. Well I do know. 😉
Last edited:
Member
Joined 2009
Paid Member
I've attached a screen shot of my parts list. It was dumped into Excel from Eagle pcb layout and then I went through and identified digikey part numbers against most of the items. I need to go through and check it and maybe update some of the items before I place my order.
The idea is to use modern transistors for the power outputs and the drivers. The front-end will be more faithful to the original Naim (NAP 160 ?) and I will make it a priority to use the same Zetex part for the VAS. I will use tantalum caps where Naim used tantalum. I will use C0G/NP0 ceramic caps instead of polystyrene caps. The aluminium power rail capacitors will be modern state-of-the-art parts, mostly by Nichicon.
I won't be trying to source NOS 1980's parts, they will be newly minted parts.
Perhaps calling this project a 'clone' is not strictly accurate, as per the thread title it is more 'based on' but I am using the Naim design as a key factor in decisions about parts choices, topology and operating points. So I won't be claiming that this amplifier will produce the 'Naim' sound. However, the spirit of the original pervades it, the original inspires the design and the discussion we have been enjoying in this thread. It remains my homage to the NAP legend.
I'm not going to try and reproduce some specific specification for the power transformer. It'll be an off-the-shelf toroid with 2 x 30Vac secondary windings which will give me around 42Vdc which is a bit higher than desirable but within the realms of close enough.
In hindsight I would have preferred to pay more for double copper thickness on the pcb but that horse has left the barn.
The idea is to use modern transistors for the power outputs and the drivers. The front-end will be more faithful to the original Naim (NAP 160 ?) and I will make it a priority to use the same Zetex part for the VAS. I will use tantalum caps where Naim used tantalum. I will use C0G/NP0 ceramic caps instead of polystyrene caps. The aluminium power rail capacitors will be modern state-of-the-art parts, mostly by Nichicon.
I won't be trying to source NOS 1980's parts, they will be newly minted parts.
Perhaps calling this project a 'clone' is not strictly accurate, as per the thread title it is more 'based on' but I am using the Naim design as a key factor in decisions about parts choices, topology and operating points. So I won't be claiming that this amplifier will produce the 'Naim' sound. However, the spirit of the original pervades it, the original inspires the design and the discussion we have been enjoying in this thread. It remains my homage to the NAP legend.
I'm not going to try and reproduce some specific specification for the power transformer. It'll be an off-the-shelf toroid with 2 x 30Vac secondary windings which will give me around 42Vdc which is a bit higher than desirable but within the realms of close enough.
In hindsight I would have preferred to pay more for double copper thickness on the pcb but that horse has left the barn.
Its confusing. You dont want Naim sound but you use tantalum.
I listened to Naim thru youtube and I heard how it differed from the accurate sound. It is seductive but it is wrong. Distortion is very high. If JV wanted low distortion he would have used fast transistor like C5200 in your list. But to get to the Naim sound he didnt do that.
The sound is to exposed the midrange to highlight musicality. Some kind of "bandwidth limiting". Overrated supply is necessary in ANY amplifier. Here a lot of things had been done not to improve the Naim sound but imo to mask the Naim weakness.
May be someday I will try to make a clone (im not sure because I dont like it), but first I will check the speed of the Naim transistor. May be it is 8MHz. The vas transistor is a match to this output transistor.
Member
Joined 2009
Paid Member
No, it's simple. I am not avoiding the Naim sound, rather I am accepting the limitations of my approach which does not guarantee the Naim sound. I hope to get some sense of it but realize it isn't going to be perfect.
there's no need to be a fan of Naim here but I'm curious about what aspects of this hobby do you enjoy ?
I like to have the best amp for several categories (power size, topology, etc.). The mission is to have all the goodies (musicality, distortion, etc.) in one amp, which is kind of impossible, so each goodies is given a design threshold.
In the process of doing the above, I enjoy finding the "what causing what" aspect.
Listening to the Naim, I think JV understands that most troubles with amp sound is caused by impurities in HF (excess). "Theoretically" (re PSRR etc) we think that it will not mess up with the amp but somehow it does. From what I heard from youtube, the HF from the Naim sound (with Harbert monitor) is sooo clean (may be the best that I have heard).
I think this quasi topology have something. You either produce a bad sounding amp or one that seems to be seductive. I think what I don't like mostly is the high distortion. From looking at your schematic I saw low distortion, which is uncommon. The question is, is fast transistor like C5200 common in this topology? Or you are the only one who can make it stable with such a (I think) difficult topology? I don't know but now I'm eager to see the result of your choice of the output transistor.
I'm afraid you are being deceived by listening to more recent Naim products via the internet with reprocessed, compressed dynamic range, low resolution audio quality.
Newer Naim electronic products also have considerably less euphonic distortion, which was the hallmark of all early NAP model sound quality up to around the passing of JV in 2000. The difference in sound quality between current models and other high-end amplifiers has now reduced to small levels and is quite subtle. Most importantly, listen to the the real thing before jumping to conclusions with the expectation that others should agree with your impressions gained via the 'net with your undisclosed sound system.
Whilst I can scarcely distinguish the sound quality of recent models to those of high-end competitors, my "chrome bumper" NAP 140 remains clearly different, like other models of the 80's period. That is what the topic is about and what Big'n is looking to find in his clone, I believe. For relevance, it makes sense to stay with the 140 model or at least one of the same era and of course, using the same speakers, source etc. in each comparison.
No, that is NOT how I listen to sound. I mentioned that the way amateurs listen to sound is by looking at how smooth the treble, how low the bass, how good the depth or soundstage. That is why they think listening through youtube will give you nothing. And you posted a link from Naim company where sounds are polished. I never listened to such polished sound. I don't know how to assess such sound. I found that more information can be found from listening to the rough condition, posted by individuals.
Yes, this is a newer Naim, Nait XS-2 (I didn't pay attention because anyway I'm not familiar with the line but I think Nigel just posted that Nait is the newer generation):
https://www.youtube.com/watch?v=mxfhD0TVh7k
But from listening other videos (didn't pay attention to the model because there are too few video available) they have the same house sound: musical and distortion.
I just listened another video of the same series as previous youtube channel. This time it is Eva Cassidy's "Imagine" and I must say I'm impressed. Eva Cassidy is very hard to produce soulfully. If the details are not there the music will be flat and boring.
I don't wan't to divert the thread but you just admitted to the obvious problem of standard Youtube audio quality. Anyway, Naim used high definition MP3 quality and there should be optional screen settings for playback quality, video and audio. Try them and see what the differences are in your system, if your data rate and audio card are capable.
Something no one has said. In the 1980's " Super compatablity " was said of lets say LP12 NAC32 Hi CAP NAP140 and perhaps Linn Kans. Naim were very insistant that a dealer should know these ideal combinations so as to best demonstate the true intentions of the company priest so to speak. In Julians words anything else was to cheat the customer. Mostly I go along with that. An example was the Mission 770 speaker which was all the rage at the time. A really nice design which was about 70% as good as the reviews said which isn't bad. For me the equally quirky Spendor BC1 was a better choice, the SP1 variant being a bit like the Mission. The Naim 770 combo was not really giving the best of either. The Meridian 101/105 was another story. Now if ever an amp should be cloned it should be the 105. With LS3/5A it is a dream set up, care needed as it goes very loud and 3/5A's can't. Gale 401 works with both amps. Superficially the two amps are alike.
I feel very sorry for people who never heard this sound. It's a bit like owning a special car and not learning it's qualities. These Naim clone amps should be close enough to hear a bit of it. Like an AC Cobra replica it should be like the original, more so than a Ford Focus.
Linn Sara Isobarik Main / Stereo Speakers | eBay
At the pub last night we had a conversation about the sound of amplifers. We concluded the VAS cap is the problem and is the sound. It's a horrible idea.
Never underestimate the PSU. I recently helped a friend design Hypex Modules into a Nait box. It sounded poor. Boomy and limp. We tried the Hypex switch mode PSU which was far better but sounded metalic ( measurements easilly said why ). Having had a hint at the possibilities I designed a special transformer which for fun we made look like the Naim. We made it as large as possible. I had to use the most compact capacitors I could so only what would fit the case rather than special types. The result was a very Naim like sound and 90 watts per channel, by mistake I had it runniong all day at 90 watts and no problems! I can't tell you how to make this PSU. Make it big and you will be close. The major difference Naim to Hypex is Hypex is better. It is bright in a pleasent way and has PRaT. If the PSU is wrong it's just another very boring amplifer. I was very surprised as class D is something I didn't want to like. Julian was said to have said " an amplifier is a signal modulated power supply ". You can actually use a LM317 as one!
I feel very sorry for people who never heard this sound. It's a bit like owning a special car and not learning it's qualities. These Naim clone amps should be close enough to hear a bit of it. Like an AC Cobra replica it should be like the original, more so than a Ford Focus.
Linn Sara Isobarik Main / Stereo Speakers | eBay
At the pub last night we had a conversation about the sound of amplifers. We concluded the VAS cap is the problem and is the sound. It's a horrible idea.
Never underestimate the PSU. I recently helped a friend design Hypex Modules into a Nait box. It sounded poor. Boomy and limp. We tried the Hypex switch mode PSU which was far better but sounded metalic ( measurements easilly said why ). Having had a hint at the possibilities I designed a special transformer which for fun we made look like the Naim. We made it as large as possible. I had to use the most compact capacitors I could so only what would fit the case rather than special types. The result was a very Naim like sound and 90 watts per channel, by mistake I had it runniong all day at 90 watts and no problems! I can't tell you how to make this PSU. Make it big and you will be close. The major difference Naim to Hypex is Hypex is better. It is bright in a pleasent way and has PRaT. If the PSU is wrong it's just another very boring amplifer. I was very surprised as class D is something I didn't want to like. Julian was said to have said " an amplifier is a signal modulated power supply ". You can actually use a LM317 as one!
Last edited:
The datasheet of those BDY56's shown in the pics at the very beginning of this thread reads a transition frequency of 10 MHz. So your assumption is nearly right.
Best regards!
The very early NAC22 NAP160 was a beautiful sound. Warmer. The Naim transistors made for Naim were not the same. Neither product worked well in other Naim combo's. I found this out when I paired the two. I thought the NAC22 very very poor until then.
- Status
- Not open for further replies.