NAP-140 Clone Amp Kit on eBay

• add an inductor in series with the amp output.

Also I’m intrigued that at least one person claims to have immeasurably improved/transformed the sound by replacing the 0.22 ohm output resistor with (I think) a 10ohm one. I might try that as well.

Hi Pattox,

If you added an inductor as less does on his boards, it should be paralleled with a 15R resistor which is suppose to dampen it. Which 0.22R do you mean then? or have you kept the output resistor which the boards were delivered with and just bypassed them with the coils?

Personally I found the coil on my ncc200 amps to make the sound dull as it killed the treble so I bypassed it with a 0.22R ala naim style which mushes the sound a bit but gives much better balance to my ears. Have you listened to the boards without the coil?

cheers
 
Member
Joined 2010
Paid Member
Hi Pattox
The VAS + CCS transistors you refer to on your Ebay NAP140 boards are (on my example) 2SB647 and 2SD667. These are very close cousins of TO126 cased 2SB649 and 2SD669, a great Hitachi favourite over the years for drivers or VAS. It would not seem a bad choice because the ZETEX pair are not fast, modern types either and their dissipation is rather high for the tiny case. I guess you can't be accused of non-original parts but these Renesas/Hitachi types are pretty good choices. I think 10 ohms in series with the output is just not on! An error, surely.

I have to say how peeved I was at seeing the ink was hardly dry on the "pulls" from discarded equipment that the 2SC3858s were claimed to be. They were new, like yesterday, without any sign of former use. I don't need to start another thread with more pictures of the same shabby printing attempt and lack of heat spreader. They are sure fakes and at the very low price, it's all one might expect.

So, I figure that replacing the drivers with the MJE243/253 and going for slower outputs like the MJL21193/4 might be more like the original TO3 types and at not such a great expense, to get closer to original sonics than what I have with this clone now. 'Wouldn't mind comments from others who have been swindled and taken similar measures.
 
Last edited:
prebuilt nap140 clone module

Hi Juhleren:
The 0.22 ohm resistors is the one just before the amp output (not either of the two 0.22 ohm emitter resistors connected to the output transistors). I put the coil in series with this resistor because I wanted to make sure there was no likelihood of rf oscillation which the early Naim amps notorious for doing. I seem to remember a post by Les Wolstenholme suggesting precisely this (ie putting an inductor in series with the output). Btw I did replace the resistor on the board with a better quality one I had. Also I had run some simulations which showed no difference in the major characteristics of the amp with the inductor in series or in parallel. I’m reluctant to change anything on the ncc200 partly because, sadly, unlike so many pristime and neat wiring examples I have seen, mine is a veritable rat’s nest

Re the coil on the ncc200; on my system the ncc200 is a trifle bright anyway so I don’t perceive any problems with the coil. Eventually I want to change to a bi amp or tri amp system with an active crossover which will give me more flexibility in adjusting the sound balance.

Hi Ian:
I decided to replace the 2sd667 and 2sb647 because I couldn’t find them for sale anywhere in the world including Mouser and Farnell and was therefore very suss about using them. Actually at that stage I was tempted to replace all of the other transistors as well but decided to stay with them for the time being. I was never under any illusion about the Sankens being genuine because, as you said, the “kosher” ones would be simply too expensive to include in such a cheap product. Just to finish the exercise I have some mje243/253s and some mjl3281s so I might try them.
Re the 10 ohm mod, its not a mistake; at least one person swears by it.

See p25 of this forum (post by Ruwe)
http://www.diyaudio.com/forums/solid-state/112453-nap-140-clone-amp-kit-ebay-25.html

“12.I found by experimenting that the sound is by far better if you use 12-15 turns 0.75-1.0 mm output coil+10 ohms/3W resistor instead of the 0.22-ohm resistors. Some of the kit versions are with coils, some other are “original” with resistor. I thought the “original” should be better, but it's not. Try for yourself if you wish. For me the difference is night and day (or music - no music)! I wound the coil on a pen and didn't put the resistor inside, but kept them separated. I think we don't need to affect the resistor with some minor magnetic fields. It's so easy to avoid it, so, let's avoid it... put one on top of the other and not into.”​

I too would be interested in hearing about other people’s experiences with these clones and mods.

To wrap up, I will try these mods but I am looking to hear vast improvements in the sound. If that doesn’t happen then I will probably just give up on the naim clones. My thinking is that if these don’t sound great “out of the box” and one has to experiment by trial and error, and virtually try to redesign the whole thing, in the hope of getting a good result then it’s simply not worth the time and effort involved. While this experience satisfies my curiosity, I am really happy and feel vindicated about my ncc200 experience – getting a really excellent amp with minimal mucking around and for a relatively small price premium.
 
pattox,

If I read you correctly, then you have misread the output coil tweak somewhat. According to both Less W and Ruwe the 0.22R is removed and instead the coil with a larger damping resistor is used (in parallel with the coil, Less W puts in inside the coil). Otherwise you´ll get the worst of both worlds; screwed damping factor with added resistance and screwing the treble with the added inductance. (I have actually bypassed the coil and the resistor (15R) with a small link wire and then applied my own resistor (0.22R) between board and the output terminals on the amp)

My ncc200 actually seem stable when i ran it with neither coil or resistor but the sound seemed a bit hot so I went with the resistor as it was more balanced to my ears. Maybe the coil would work if it was less inductive and/or were paralleled with a resistor that is smaller than the 15R Less W uses on his boards. This has actually been suggested on the SKA forum...

cheers,
 
prebuilt nap140 clone module

Hi Juhleren

this is a very old post by lesw
speaker wire and naim [Archive] - pink fish media
06-09-03, 01:27 AM
The limited drive capability of the Naims should have been addressed long ago. I've yet to come across another make of amplifier which locks the owner into this kind of cable nonsense.

Quick DIY Tip: Make a simple inductor from some 0.75mm enamelled copper wire. Wind about 15 turns on a 6mm drill bit and scrape the tails clean. Disconnect the positive lead to the red loudspeaker terminal inside the amp and introduce this inductor into the circuit. (For those of a nervous disposition, it's easily reversed without trace)....... Hey Presto..!! The amp will now drive anything - easy innit..??​

I realize the ncc200 arrangement is different with the inductor in parallel with the output resistor. However I am sceptical that an output inductor, placed in series rather than in parallel would turn a silk purse into a sow's ear.

I assume that the published nap 140 circuit is accurate and therefore I continue to be puzzled about why the finished clone product sounds so ordinary. It certainly doesn't sound like the naims I have heard. Is it the dodgy transistors? Or perhaps the other components such as capacitors? Or the poor quality resistors. Without the protection the circuit is relatively simple and not dissimillar to the ncc200 but it just doesn't sound great. ?But I'm not sure that I will spend mich more time persevering with this.
 
Member
Joined 2010
Paid Member
I decided to replace the 2sd667 and 2sb647 because I couldn’t find them for sale anywhere in the world including Mouser and Farnell and was therefore very suss about using them.............Re the 10 ohm mod, its not a mistake; at least one person swears by it...........I found by experimenting that the sound is by far better if you use 12-15 turns 0.75-1.0 mm output coil+10 ohms/3W resistor instead of the 0.22-ohm resistors. Some of the kit versions are with coils, some other are “original” with resistor. I thought the “original” should be better, but it's not. Try for yourself if you wish. For me the difference is night and day (or music - no music)!
I concur with the difficulty in buying these or any Renesas transistors through the souces you mention. They do exist though, for direct OEM supply or through their agents. I satisfied myself with testing breakdown voltage and hFe of parts used, which were well above the mark.

Juhleren has seen the issue with the damping resistor being in parallel with the output coil. I would agree that you would likely change the sound using an undamped output coil. It's purpose is to prevent ringing with capacitive loads and without the resistor of 5-10 ohms, it is asking for trouble with any linear SS amplifier.

When these amps were still on the market, I was asked to service several where the customers complained of severe RF breakthrough and some speakers sounding bad because users insisted on huge cables with low inductance. Naturally, this was not what Naim intended but their wimpy cable didn't look impressive or cost enough for audiophile appeal so trouble and disputes ensued.

The 0.22 ohm resistor was itself inductive and the combination somewhat effective but poor at keeping out RFI and this can still be a real sound spoiler. Many experimenters have settled for low inductance cable with a damped 0.15 uH coil, like just about any DIY design. I think that if such small coils are so obviously affecting sound, there is a more fundamental problem there.

I would still agree with your comments on the poor sound of these clones out of the box but using newer, higher performance semis and getting lower distortion is not going the right way, IMHO. Realizing I don't have the time or skill to tweak an old design to new tastes, I opted to seek the original sound which so far, seems easier, given what likely matching techniques were used and thankfully having access to an original NAP 140.
 
prebuilt NAP140 clone module

I've been doing some simulations and found some interesting results re the inductor.

Relative to when the inductor is in series and the "protection circuit" is disabled, changing the inductor to parallel increases output distortion by around 25 times (from 12mv to around 300mv)

Again relative to when the inductor is in series, enabling the protection circuit increases distortion by around 20 to 25 times.

In the simulations I used an input voltage of 0.5v swith a 2khz sine wave as the input.

Also the simulation shows that changing the 0.22 ohm output resistor to 10 ohms doesn't change distortion or transient response or ac and phase response. Someone with an oscilloscope might want to check to try to validate these simulation results.
 
pattox,

Could you please elaborate on the specifics of coil, resistors, and the configuration of these in relation to your results for transparency. Are you talking about 0.22R // inductor (what app. value), 0.22R in series with the inductor, 15R // inductor, only 0.22R in series with output, only coil in series with output or? Is the distortion a background noise like the typical naim hiss or is it proportional to the input signal?

Ian, I may be overly sensitive to the negative effect of the small coil on HF performance, but try to add that value to the speaker wire or speaker crossover without any compensating capacitance as what seems to be done in the ncc200; It really changes the balance and kills everything delicate. I guess that other amps that employs output coils either compensates for it in the amplification or by also employing a capacitor to balance the top end...
cheers
 
Member
Joined 2010
Paid Member
Simalong

Interesting result, and the magnitude is nothing I have ever seen reported or heard from solid state using a standard parallel coil up to even 10 uH. Being a very standard Lin quasi amplifier, the Naim would be no exception, even with it's resistor in the output, VI limiter and Baxandall diode.

O/T but what you don't find out is that the 10 ohm resistor just reduces the load from say 8 ohms to 18 ohms, so you would surely expect same or better THD results. Worse though, the loudspeaker is a complex RLC and it is now driven with a damping factor and useful power output shot to pieces. To hope to make sensible sim. predictions of output stages, you need a realistic model of the load, otherwise it's just doodling.

Try varying the total resistive load (including 0.22 ohm resistor) between 0.5 and 50 ohms in the simulation for a reality check. Distortion is but one, usually minor, factor in assessing output stage performance.

Just saw your post Juhleren, I'll come back on the coil but there is so much contrary comment on other forums about this matter, I wouldn't try to follow it all.
 
Member
Joined 2010
Paid Member
Hi Juhleren
Assuming you mean altering the global feedback compensation (Miller) capacitor to maintain stability, if needed. Certainly. you could be right and the VAS, being a large factor in the "sonics" is always sensitive to this value.

The issue here is simple though - NAIM provided or specified a high-inductance cable of three meters and users changed this to short, low inductance as it became fashionable and any heavy cable looks and usually is better than typical light duty twin flex sold as "speaker wire". However, this compromised performance, which was marginal IMO.

It is one thing to say that a stable amplifier with added inductance sounds worse somehow and quite another to see an amplifier overheating with parasitic oscillation because you changed the semis and also have more ringing and RF at the output. No question, that sounds bad too.

I consider that replacing the .22 ohm series resistor with a low value coil/resistor is merely providing the equivalent and perhaps a little more stability with low inductance cable. It stretches the memory, but audiophiles at the time heard no difference on NAP80,140, 250 I think, whilst the cable change alone was discernible to 2/3 guys. Admittedly, these guys often made up their minds how things were going to sound long before listening.

The inductance used was 0.8 uH, still very low in comparison to "unconditional stability" values of 6.8 or compromise 3 uH values of the day. It worked to restore stability and sonics with the real thing then.

How that applies to this latest clone with fakes is still a puzzle, but obviously every clone and its components in the topology is an issue if not the same and as I found, high Ft semis really do change the sound and stability margins. It may seem better at the time of experiment to some folks but then it's not Naim sound. I have also seen a lot of "amazing improvements" trashed after extended listening periods too, so I take most unqualified reports on sound quality with a grain of salt.

I've only heard one Avondale clone and liked it but it was not the same design approach or sound as Naim NAP140 IMHO. It shouldn't need saying but the sound of VI limiting in Naims (even if objectionable) is still part of the signature.
 
Last edited:
prebuilt NAP140 clone module

Firstly: re the inductor, when I rechecked the circuit and reran the simulations I found only very tiny differences in THD between when the inductor (3uh) was in parallel and in series. THD was around 0.012% (0.5v 2khz sin input) in both cases. My apologies. But it also also suggests that it doesn't matter which way the inductor is placed.

I still get bad THD figures (around 0.8%) with the protection circuit in place.

Juhleren, I've attached a jpg of the tail end of the circuit.

Ian, I changed the output resistor to 10, 50, 500, 5000.. and I get an unchanged THD profile. So I'm not sure what that says about the simulation.

Btw I agree with everything you say, even about the doodling (although it is fun). It is possible to simulate real loudspeaker loads but I don't have the time or skill or inclination. I'm not sure though that using a resistor to approximate a speaker invalidates the whole process. I find the simulations both seductive in their accuracy (eg being able to simulate changes in bias current and see them happening in the same way as with the physical circuit). On the other hand it is very easy to change anything then forget to change it back-as I did, end up with weird results, then have to recheck everything.

Juhleren, I've also attached a distortion plot using zero input. The distortion is high (relative to a tiny signal which I guess simulates hiss). I'll leave it for you to interpret.
 

Attachments

  • Clipboard01a.jpg
    Clipboard01a.jpg
    402.9 KB · Views: 1,248
  • Clipboard02a.jpg
    Clipboard02a.jpg
    341.7 KB · Views: 1,182
Last edited:
Do we have some kind of consensus with these amplifiers? ie a sum up of 66 pages, to save trawling thro them and reading the lot?

It would be also nice to have some kind of consensus regarding the "best" transistors for each position, am I right in saying the faster the tran, the better?

which ARE the best trannys to put in?

I have a kit amp, not one of these, a white noise, and it sounds fantastic, its a mosfet circuit with a modern opamp input stage, perhaps kits ARE not a waste of time, but its a testiment to this amplifier design that 1/ its so popular commercially, and 2/ its KIT version is so popular, ie ppl trying to get it on the cheap.

it sounds good imo bcos of its relative simplicity, I believe the more complex an amp becomes to get the meausred performance technically better, the more the sound quality, subjective as opposed to objective, suffers. these being "old" 1970s or earlier designs, have less evolved circuitry and as such suffer technically and objectively, which in turn gives more character to the subjective performance.
 
Member
Joined 2010
Paid Member
Do we have some kind of consensus with these amplifiers? ........It would be also nice to have some kind of consensus regarding the "best" transistors for each position, am I right in saying the faster the tran, the better?....which ARE the best trannys to put in?

Hi lt cdr data
This thread is old now and passed from one group to another, looking at several variations and sources of the kits. It would be difficult for anyone to summarise the content with so many personal views. This recent set of posts looks at a current Gigaworks issue that is distinctly different to the 2009 version at the beginning of the thread.

As a high-end design from 30 years ago, it's simple, standard technology, using premium quality parts of the day. The big missing piece in a cheap clone kit is the components that are significantly different and not selected or matched - nor are the amplifiers typically even wired up as we can see the originals were.

If there is a consensus, it would probably be that the sound is not a patch on the original, which may well be expected for the very low cost approach. The original amplifier price was stratospheric, so expect big compromises if just assembling without attention to selection and type of parts, layout, wiring etc.

Regarding faster transistors being better; that seems to be the general consensus regarding all modern design amplifiers, particularly in the VAS and drivers but maybe this is not what you need for old designs like this one, with a certain desired "sound". ;)
 
nap140 clones

Again I agree with Ian. In the end I think the quality depends on the quality of the components, perhaps also selected components with individual characteristics, and perhaps to some extent on pcb topology.

This whole thread seems to be based around clones of the ncc200 (gigawork) and the Naim Nap 140 (tubeshunter & others) with varying degrees of operational and (untested) sonic success. My recent posts relate to a Naim Nap140 prebuilt clone compared to a real (not clone) ncc200. I don't seen reports of a comparison between a clone ncc200 and a genuine ncc200 though people claim their ncc200 clones sound very good.

Out of the box the NAP 140 clone I bought doesn't even begin to come close to the ncc200 in terms of sound. I suspect that if I rebuild it with better components sound would improve dramatically but until I do so I can't be sure. Curiously the NAP140 (without the protection circuit) is a fairly simple design and it would seem difficult to go wrong. But, out of the box, the sound doesn't begin to stack up, and as Ian said earlier, many of the "improvements" claimed don't stand up to prolonged listening, or comparison with known quality products. In my case I was initially happy just to hear my clone work -this thread is littered with reports of failure -and produce half decent sound before concluding that it wasn't great.

In the end I think you get what you pay for. In my case I bought some ncc200 boards from Avondale with together with a listing of quality parts that Les Wolstenhome had obviously selected after years of experience. And the sound has has stood the test of time and comparison with other products.

I haven't seen the nc200 pcbs on the Avondale but they do sell completed modules which, while around 7 times the price of the Nap140 prebuilt clone, are still a cheap way of building a top quality amplifier.
 
Member
Joined 2010
Paid Member
Partsfest

Hi - I couldn't leave out this for fans of the Avondale NCC 200 kit.... DIY Audio Power Amplifier modules : DIY Audio.
The newbs will cry in anguish to see from the images that the input cap is TAG tantulum, the feedback appears to be wet tantalum and the output TO3s are "industrial" MJ15003 :eek: :rolleyes:

Well, this should tell us about the sense of scattergun parts buying - just religious selection of boutique parts and then assuming these will be best in all similar applications. As I have seen, NAIM did use Tantalums in the audio path, still gets criticised for it, but still earned their superb reputation using them regardless! Got that, guys? There are indeed some nice little foil caps in feedback and HF bypass applications and Zetex VAS semis are also used. Note the very low value O/P coil!

Now, if that array of components was used on the Ebay clones, what would be the audible results? Perhaps it would be even worse without selection but seeing matching done on other competing Ebay products, I think it could be done to great benefit for quite tolerable extra cost, even using genuine parts.

Really. DIYs who are serious about their audio should routinely measure and match critical part parameters, not just the easy DMM stuff. In that long ago period, some top-drawer amplifier kits even were supplied with pre-matched pairs of drivers, LT Pairs and often selected beta types in the VAS. We might consider too, that whilst designers now aim to avoid parts selection, the old designs and parameter spreads are still with us.
 
Hi,

Just to put in my 5 cents:
The ebay Naim clone kits aren't that bad as a topology, I'd say they are all quite good. If you find on Internet original Naim amps board pictures and you'll see that they are quite similar.
Problem are the parts. Generally, I assume that all semiconductors are fake or of lower quality.

So (to ptempel), try to replace them all, and if you wish replace all capacitors and then you would be able to compare ebay kits with NCC200.

Avondale definitely have selected better quality parts. However, I don't like some of the mods that they are using, especially the 100 ohms degeneration resistors, and the diodes and 220 ohms resistors on the supply lines.

I tried degeneration resistors and in my opinion anything above 22 ohms is changing too much the sound from the original "Naim sound". For the supply lines I personally use 22 ohms/1000uF, instead of 220 ohms/100uF, and no diode. Theoretically in these amps, adding degeneration without decreasing the VAS capacitor means that the open loop gain of the amp was decreased significantly, and probably more high freq. distortion is present, but less distortion in the input pair, because of the better input pair balance.

Anyway, I'm not familiar with the sound of the NCC200 kit, but I suspect it's quite good reading all great reviews.

I measured my amp, not in lab conditions, just at home with a scope, and to me it works pretty good. I have a file but don't know how to attach it. It's about 80 MB because of all the scope pictures that I took.

Latest findings about components:
- Input transistors: ZTX692 - huge gain => very low base current => very low input distortion due to source equipment impedance and so on.
- VAS, current sources: XTX753/ZTX653
- Drivers: MJE243/253
- Output transistors: MJ3281 - best possible compromise current/price/bandwidth
- power resistors - metal element, not wire wound. You can try that one (say 0.12-0.15 ohms) instead of a coil/resistor output combination. It works and sounds similar.
- capacitors - styroflex (polystyrene) wherever possible, bipolar electrolytics of good quality for input and feedback. I use Mundorf. Or use tantalum, but watch for polarity.
- don't underestimate the power supply bank capacitors. Some high current computer grade work. Currently, I use Mallory 2x10000uF.
That's it in general. The rest is proper grounding, which means, no loops, no redundant grounds, no thin wires.

I still don't understand the superiority of tantalum capacitors and the secret that Naim had discovered about them. By all books they are fantastic for radio frequencies (I see them, wet or dry type, in almost any old radio that I repair), but have some distortion problems in the audio range. Also very easy to damage with wrong polarity. From troubleshooting and reliability point, they very much like to go short.

Naim tend to claim some aspects of their engineering, not achievable by simple humans, and we usually accept as a given stuff like that, especially for audio, especially for hi-end. I respect Naim very much as a company (one of the not so many doing real business, not some garage operation). So, I wouldn't say "snake oil", but I'd say "it's just good engineering". I don't believe in magic.
Take for example the NAP-500. Look at the pictures online. It looks to me as 2xNAP250/ch. (bridged or parallel, I don't know which) + 2xregulated power supplies. No miracles, except the price range.

Anyone with knowledge could improve any amp, just because the main purpose of all engineers in the world is to meet their project expectations and reduce cost. I'm sure it's valid for Naim operations either. Then, why not using better parts in their circuits? DIY is basically cost-no-object :) If you don't improve anything, you won't ruin anything with higher quality parts, if you know how and what to read in a component datasheet. (I don't even want to start on the nonsense in the "hi-end components" business).

Very sorry for the long post, it was meant for good :)
Cheers
 
I have read a fair bit cloning Naims, and my summary would be that no one really understands what exactly it is that makes the naim sound so good. I don't think any one kit really matches it. If I were going to try I would probably go for Neil Mcbrides version, simply because I it reads well to me and the other option NCC200 sounds overpriced for a kit and probably close to the price of a real used Naim.
 
Member
Joined 2010
Paid Member
Very sorry for the long post, it was meant for good :) Cheers
It's a great post, Ruwe, honest opinion and informative.

I also found with 3 clone pairs built and tweaked, that parts were the issue and though using fast stuff can certainly "improve" performance (lower THD etc) it does not get closer to Naim sound which is surely the signature distortion and general characteristics of the amplifier in a system.

The only fakes I have found with about a dozen Ebay kits have been these M-200 Sankens, which we could guess with the low price. I agree about the quality of the NCC200, it does sound good, but is way overvalued for a simple module and does not, IMO, sound quite like a NAP 140, even swapped in an original amp, as I have been fortunate to test with Linn Saras.

BTW, if you want to avoid fakes, Tubeshunter sells bare PCB pairs for USD 20. These accept T03 or plastic power as output transistors. Fit your own original spec. or as you wish. I'll try this next - after the 20 other projects I hope to do...;)
 
I have read a fair bit cloning Naims, and my summary would be that no one really understands what exactly it is that makes the naim sound so good. I don't think any one kit really matches it. If I were going to try I would probably go for Neil Mcbrides version, simply because I it reads well to me and the other option NCC200 sounds overpriced for a kit and probably close to the price of a real used Naim.

no one does, but it HAS to be in the quasi complementary output stage, its a combination of all factors, toplogy, components, etc. configuration is a massive part.

see my post here....

http://www.diyaudio.com/forums/solid-state/193484-naim-nap-250-135-specs-8.html

Hi,
roubleshooting and reliability point, they very much like to go short.

Naim tend to claim some aspects of their engineering, not achievable by simple humans, and we usually accept as a given stuff like that, especially for audio, especially for hi-end. I respect Naim very much as a company (one of the not so many doing real business, not some garage operation). So, I wouldn't say "snake oil", but I'd say "it's just good engineering". I don't believe in magic.
Take for example the NAP-500. Look at the pictures online. It looks to me as 2xNAP250/ch. (bridged or parallel, I don't know which) + 2xregulated power supplies. No miracles, except the price range.
Cheers

yes you are not meant to question the naim ideology, its like a cult in the strongest sense, anyone disagreeing gets banned from their forum, they certainly won't answer tech questions or disclose schematics.

amazing that they still "invite" discussion and debate to this day, and will probly continue forever, naim controversy. drives some ppl mad


nap 500 afaik IS a bridged 135/250 with regulated psu's.
 
Last edited: