NAP-140 Clone Amp Kit on eBay

[rensli]

Hey,

But, when it tells me that all the hFEs are 116 or 137 or 201 for 100 devices, I don't believe it !
Same for Vbe and lots of other parameter readouts.

I know, this HFE measurement is done for ppl who has exactly the same device as i do..i mean noob's.

In overall, i Ended up:

1. Rewired the amplifier PSU and ground(still not twisting). Ian was right, that the Oscillation in Audio range was caused by wiring!!, but... i still had "silent" oscillations left with TIP41/42CG (high gain) drivers. I cured the bias oscillations by replacing the 1.8K feedback resistor with 1.2k, it is the resistor which is connected in series with 39pF capacitor in feedback network. It was stable even with Chinese Philips BC550C LTP transistors.
​

2. Removed stupid Inductor's from PSU DC path.
​

3. Sound is still way more pleasant with BD139/140 as drivers and BC550C as LTP.​

4. The Driver base filter networks are adjusted. NB! every time i change drivers or input pair transistors... driver network filters need different resistor values to give best sound.​

5. Removed soft-start/overheat control's and they'r PSU-s from the case.
​

I have reed here that someone was not sure about rhythm on naim clones... Probably it was DaveS . That happens when your grounding is wrong AND, OR input LTP are messed up.

For short: I have better dynamics compared to H-140 naim clones, same deep bass, alot better separation in high's AND it is Sooo rhythmic. ALLIN1, i prefer less powerful(500VA H-140 clone vs ~150VA Very close NAIM ??), but more interesting amplfier for now.

Will post detailed pictures and schematics in a month, as i have run out of DIY time
Also, i will search for BC550 guys of a lower gain, about ~200-380. All the BC550C's and B's i have are ranging from 400-to-600 (.

For the reference i want to say, that the BIGGEST changes in Sound are coming from transistors/resistor values used in the NAIM Circuit, Speakers/Room/building and YOUR own position in the room. Not the PSU, or inductors or whatever else!

I could even hear a MOTOR, that was powered 2 houses away from my place . I heard it from my both speaker trough naim, it was quiet but hear-able... still, the sound of this amplifier is good!!

 

[rensli]

Hello,

I was able to make the amplifier quiet without inputs connected, by replacing overheat/control systems and they'r PSU with much simpler solution:
Capacitive reactance based PSU's straight from AC power line. If everything is done correctly, there will be minimal heat and noise.

So i continued the journey for better sound that, i have a feeling, will not be reached

Basically, i do not know how original NAIM sounds like, but at my first guess, they definitely sound DIFFERENT to each other because of the PSU's(regulated and voltage ratings) and Different component values.
If amplifier is so good @ TIMING then it should produce very attractive instrument separation. Comparing it to naim.
This effect should bring out aggressivnes/explosive sound in some musical conditions.
Separation means less masking of the sound(more tones will fit between the notes) which results in better background perfomance, everything will sound more musical and involving.​

This was experimented with (NAP 140/160/180)that is seen in pictures below.

Final results were never reached, cuz there are always some downsides.
At least Tidu34 pre-amp is doing good job, transparency at its best

Few facts:

*I have not yet tried BC239C, but second candidate would be BC550 series, they are much more linear to my ears in naim LTP then any other transistor i have tried here.
Third candidate would be BC546B, they affect bass a little, some may like it. Also, drum instruments have more power with BC546B's inside.

*Every output decoupling capacitor had negative effect for me, nothing positive for sure.
*10nF capacitor in parallel with rectifier diodes will affect high frequency perfomance in negative way. Removed, will be removed on my first amplifier too.
*Removing inductor from to the DC PSU and placing it back again after a week or so, i was not pleased by its effect. It was removed shortly after few hours of listening.
*New wireing vs old one do not sound different...
*Using Tantalum-Polymer SMT 50V 10uF capacitor on naim LTP input will have negative effect on the sound(becomes metallic).
*Trying different grounding techniques do not fix problems.
*H-140 clone shown in the last picture is better @ any kind of piano music.
*Original naim circuit is a killer, if configured correctly, it is very hard to make it enjoyable in every aspect with separate(OUTPUT, Front-END) PSU's.
*If one tries to compare amplifiers with low sensitivity and/or tiny or not properly made speakers, it could be very hard to hear any difference at all.
​

Tought's:

* I think, that LEFT and RIGHT channel NAP140 are fighting each other when used with 1 PSU,
I always had good results powering 1 channel only, from laboratory PSU(regulated).
*620R CCS emitter sounds worse @ 33V than 560R.
*In my experiments, using front end PSU on an amplifier that was not designed to use it in a first place will have negative effect on high volumes.
The effect is simple, one will get everything amplified as volume rises, even those frequency's(sounds) that should stay more or less attenuated. Simply put, it will suck the life out of the music.


Is there a point to search for BC239C transistors ? Maybe someone here can suggest correct mods/adopt the circuit to the Front-end/Output separate psu ?

 

[erikovsky]

Hi,

Since it's been a long thread, I started a new thread here:
http://www.diyaudio.com/forums/solid-state/294864-naim-amp-2016-build-thread.html#post4799613

And I think this will be a little bit more friendly for new comers, as after I read hundreds of page, I think I've learnt a lot but also exhausting

Regards,

Eric

 

[rensli]

1st picture is now what i am using for about 1.5months
2nd. is the previous configuration that was used over a year.

Yes, new setup wich actually includes component removal is better.
Highs frequencys got cleared up, more damping factor = less but more
controlled bass.. like punchy. Still plays pretty low.

I noticed, that sound got more interesting again... a little bit front-
sounding on certain freq. spectrum.

OK, went over to my colleague house, we had an audio night. 4 guys, 3 amps
and a pair of speaker. No place for women this time

First was the CLASS A amp, pure tube. Yep, it has that vocal that goes

deep and to the end.. very good with instumental, vocal, slow chill
music.

Second was my naim clone...Highs are good, but there was no Drive
power...atleast not enaugh.

Speakers that we used were dome tweeter + 5inch some high-end mid-bass
in tall Hornlike very low frequency tuned wooden case.

We then praised, if only the tube amp had the same Class AB or B high
frequency playback profile.

Third amp, Class D philips chip, external filtering, linear power-
supply.
This one suprised me... it CAN DRIVE just about everything And also, was not so bad... just solid amp.

After, collectively deciding to see if naim can put out some bass into 5"
speakers with low freq. tuned case.
1R on the output was enaugh but highs got messed up, at the end nobody liked it.
Now follows the feeback capacitor change... Previosly it was 47uF ultra
low ESR TANT(0.2R), we changed it to 47uF simple SMD electro(0,7R).

1 guy said it made some changes in lower register for the better... i
was very skeptical about it... no profile was changed for me in sound.

So i got home, back to my setup and noticed that with 47uF electro in feedback, front-sounding spectrum moved a little bit lower..

I then realized how non-sensitive speaker system we had back there
The same colleague is now after F1 with, 85€ on laterals, dual mono + 8" full freq. spectrum speakers with correctly tuned case and filters.
Actually, we all kinda agreed inside, that hes speakers sucked... but nobody was telling it out loud to him Speakers were designed correctly, using simulators and math.

I had an idea, to try 68uF 16V (0.55R ESR) CSS type tant in feedback and 47uF (2.5R ESR) very lossy High ESR tant at bias bypass position.

And idea was right, i get more front-sounding profile on much higher freq spectrum where bass is still punchy, dynamics there, everything is cool on reasonable volume, all this in my home with my 8" inch sensitive bookshelf speaker setup.

Next thing is Font-End supply removal only, to listen the effects... it should be positive as i described long ago, cuz it is not ment to be in the design, when looking at naim amps ) ...

 

[Ian Finch]

Hi rensli, interesting experiments, even if they are as much about different speakers as the various amplifier designs. Were you using the original spec. VAS semis or other parts, since they have a marked affect on compensation and sound, as discussed previously?

Front end decoupling measures are important because the gain applied to noise and distortion there has the largest affect on the input signal. A common issue is that a simple power supply derived straight from the output stage rails with little decoupling, will always contain artefacts of signal from the output stage and loudspeaker back-emf. These modulate the supply to the LTP and VAS to a small degree but enough to colour the sound noticeably. In some experiments of my own, many years ago, I found that regulating or otherwise filtering the front end supply, kills the effects which can actually be what give the desired sound quality to simple but respected amplifiers.

A modern amplifier that contains RC filtering, current sources, mirrors, extensive supply decoupling and various other circuit tricks to achieve very low levels of noise and distortion, is a different matter. There, you start with a design that primarily has ultra-low distortion and neutral sound quality as its aim. Early Naim amplifiers and many other old favourites from the 1970s may have begun with that intention but now, are just not in that class. If you read about the early years in Julian Vereker's interview docs, you'll see what I mean.

I think though, that this is beyond the scope of Ebay clone build discussion and deserves a new thread about tweaking simple amplifiers for audiophile effects. This is a huge thread and due to so many off-topic posts in recent years, it has lost direction and the interest of readers, since it takes days to read and track the OP thread. It began when these kits were new and had issues. The few kits now promoted either use the LJM PCB or they are Caowei's proper 140 clone kits, both showing Sanken M200 output devices . These are as close to the late original product appearance and design as possible but if that's not your interest, I think it's time to give the topic another thread with another "Naim"

 

[rensli]

I think though, that this is beyond the scope of Ebay clone build discussion......

You are propably right

Hi rensli, interesting experiments, even if ..........

1.

No, original naim spec'd amplifier is in another case, also demonstrated lately in this thread. Running on +/- 27.5V DC with separate front-end supply which is +/- 33V DC just rectified and filtered.
This is the amplifier, where i played with Driver base network-filters. The design is the closest to naim NAP 140, expect the rail voltages, too low i guess....​

2.

My main amplifier and the same experimental one is the Ebay-clone H-140, where 2SC2705 and 2SA1145, both Y- type(gain) are used in VAS.
BC550B PH-03 LPT and
MPSA06 CSS
2SC5200 on outputs as always with MJE243/253 drivers.​

Front-END is regulated here. But as i promised, will be removed.

I actually think, that moving from Stereo to dual-mono will get me alot answers here, it has something to do with powersupply pumping where both channels driven from the same supply are interfering each another.

Another option is to use 1 opamp for each AMP board, inverting to 1 channel and non to the other... this should elliminate all the bad effect when running 1 psu for stereo.

 

[Ian Finch]

It has been discussed many times that substituting low Cob transistors for the VAS affects the frequency compensation. ZTX parts have a huge 30pF Cob and original 2SC2705/A1145 have a very small 1.8 pF. They are not comparable and with only 39pF for Cdom, there will be instability unless there is further frequency compensation somewhere, such as using 47-68pF for Cdom or adding some lead-lag shunt compensation across the LTP. However, if you do play with compensation in any amplifier, you need to have reality checks to be certain you haven't made a time-bomb instead. For that you could use simulation but a soundcard oscilloscope and audio spectrum analysis software provides real, direct and simpler results. Otherwise, I can assure you that adjusting compensation by ear is unsafe because it does not show what is happening at radio frequencies where any peaking occurs and this may burst out spontaneously or build slowly - perhaps unnnoticed. Take care and check regularly for heating of the output stage.

By using significantly lower power supply voltages like 33V, you are, in fact, using the correct supply for original NAP140. The kits actually contain parts and supply recommendations for a NAP250 and these use different resistors to regulate bias in the LTP and VAS stages. So what you have is wrongly "Naimed" but as long as you don't confuse the designs, supply voltages and parts values, it shouldn't matter.

Dual mono designs are a good idea but any "pumping" effects with conventional stereo amplifiers only occur if the supply is too small. Since most listening is done at very low levels where the supply doesn't significantly drop, this becomes a imaginary. The feature of dual mono designs is the isolation of the grounding and supplies in an attempt to eliminate all crosstalk, which has a broader meaning. The odd part about this, is that I believe it over-emphasizes the production stereo mixing that is modelled on a standard type of 2 channel amplifier. You tend to get a stereo image that is unrealistically spacious, but I guess many folk like to hear such effects which appear to give them more of everything, definition that is holographic etc .

 

[rensli]

Hi Ian.

I can assure you that adjusting compensation by ear is unsafe because it does not show what is happening at radio frequencies where any peaking occurs and this may burst out spontaneously or build slowly - perhaps unnnoticed

Yes, i had this kind of instability's, were @ low volume bias acts normal and any higher input AC(music) would make amplifier unstable and overheat, bias goes crazy instantly and NEVER goes down...

All of this has been experimented and there is more to come.

56pF VAS comp. is used on chinese clones. I have never moved that value up or down... there was no need for it.

The feature of dual mono designs is the isolation of the grounding and supplies in an attempt to eliminate all crosstalk

Yes, this is true and its what i belive. but channel interaction is also a part of 1 PSU for stereo@ higher volume(where bias current would go crazy and non-linear and start affecting each-other ?).

 

[Ian Finch]

......but channel interaction is also a part of 1 PSU for stereo@ higher volume(where bias current would go crazy and non-linear and start affecting each-other ?).

It would have to be an awful design in which the bias current of one amplifier was so affected by the load of the other channel. As the greatest power requirement is for bass, which is usually common to both channels, it is probably not the issue many imagine or preach about.

For Naim amplifiers, which used single supplies for all their early domestic models, there has never been an issue as they are fairly insensitive to bias levels. At higher power, such as in the excellent NAP200, separate windings are used in that dual mono design but still on the same toroid core and subject to similar core loading.

In reality, I think its just simpler for many DIYs to understand and source parts for 2 amps using 2 of the same transformers as you would use for a commercial stereo build. That way you get the traditional, satisfying overkill off power supply ratings and parts that audiophiles love to see

 

[penmarker]

***snip***

Chinese boxes are often ≥ 3mm aluminium which is fine as this isn't a huge amplifier; just 2*75W. This case might suit you as it has 4mm sides and 3mm top and bottom in the style and finish of your preamp. The price isn't scary and offsets any freight cost, though I think the seller also has a European distribution warehouse which should save some cash: Aluminum Chassis DIY Case Enclosure Cabinet FOR Preamp DAC 430 95 330 MM | eBay

I'm sorry if this may sound ignorant or dumb, but I don't know how to attach standoffs on chassis.

When I look for them, a lof of the standoffs are labeled as self tapping. So if I get those, does that mean I only need to drill correctly positioned holes in the chassis floor?

And will hand drills be sufficient - provided that I make some pilot hole marks prior to drilling? Or do I still need a drill press for better accuracy?

There are also stick on standoffs but they're all made of plastic which worries me.

It is very obvious that I am a beginner. In fact, the NAP200 clone is my first build.

Thank you very much for any input.

 

[Ian Finch]

Hi penmarker
"self tapping" means that the standoffs have plain through holes and likely need to be used with specially hardened screws that have a coarse, tapering thread like those you find used to hold the cases of most modern electronics gizmos together. They are forced to cut their own thread as you screw them into softer materials like plastic, aluminium or brass but take care by using a to and fro action, not to snap them off in the process. The correct size pilot hole and some experience with these screws is useful there because you may not be able to extract broken screws without making a mess of the materials. Self tapping fasteners are not really a good idea for repeated fitting, disassembly and repairs though, unless the original fit is perfect.

PCBs usually are mounted with about an 8mm length of standoff tube but the length is dictated here, by the power transistors being mounted to the case bottom with an aluminium heat spreader bar between them. If the standoffs are say, >12mm with your particular assembly, you could use short 3mm dia. bolts+shakeproof washers at both ends and of course, bolts would be neatest, both underneath the case and inside. Here's an example of threaded standoff , similar to what I would use with short bolts+washers.

There are lots of variations like a threaded through hole one end, threaded tails either end or plain hole for use with self tapping screws but there you need to get the drilled pilot hole size just right if it is long, or you'll never get it in without jamming half way.

With short standoffs, you may not be able to fit standard length short bolts or screws both ends and a plain tube standoff with a single long through bolt nut and washer is easiest but looks untidy inside if you want to make a showpiece of your work. YMMV

Re drill types: It's fine to use a hand drill if you can clamp the materials you are drilling properly and use a centrepunch to ensure holes don't wander, particularly with blunt drill bits, since both hands are occupied and small drills bend and are very fragile. I did all my chassis work as a kid with a hand drill and though it started rough, I learned so that by the time I was teenager, You couldn't see what sort of drill was used - up to about 1/4" dia. at least

 

[penmarker]

Hi Ian, thank you very much for your thorough but concise reply. Yes using female threaded standoffs with short bolts and washers does make a lot of sense. This way we could avoid breaking a tap inside the hole. I'll try drilling out the holes once I got the parts and measure them out.

But for everything else, I'm good to go. At first I got worried about the uncommon 28v-0-28v x2 transformer thinking it would cost significantly more than the more common 30v-0 ones. But then it's not too expensive, I'm so relieved.

**A little off topic
What speakers pair well with this NAP200 clone?

 

[AndrewT]

..................A little off topic
What speakers pair well with this NAP200 clone?

8ohms speakers.

 

[Ian Finch]

A leading question but yes, I have to agree, AndrewT, 8Ω would suit nicely
How many opinions and disagreements would you like here? I've Googled plenty elsewhere for Naim products and the differences are vast - many views read like promotional blurb and others are just the experience of happy customers. It really is a question of budget if you consider the range of choices made by Naim owners in the UK, Europe and North America. You are talking of €10k prices recommended for original Naim gear. DIY is of course, a budget market and wallets are not as not fat so if you buy finished speakers new or used, the price could be relatively high for even average to good quality full range speakers.

If your needs are modest, the field opens up with stand mounted speakers and so many more quality models to choose from - too many to suggest really and you will need to try with your amplifier before buying, whatever the size, format, condition or the seller's location. My suggestion is to look for extended bass response. Quality, well integrated midrange and treble is fine but speakers still need to deliver fundamental bass convincingly to blow you away.

 

[rensli]

8ohms speakers.

++ try to get "big" speakers...atleast 8" for mid-range or bass

 

[penmarker]

I'm not sure whether this is a good or a bad news, but I managed to get my PSX as CD player working again.
The remote and controller ports stopped working, and found out that the 3.3v line fuse to the socket blown due to a cheap controller. Now its working again, I might finally build a linear PSU for it. So that means I'll have to hold out the NAP200 for now.

Or should I build the NAP200 first?.
Hmm... decisions..

 

[penmarker]

The board had arrived. I opted for the assembled and tested board instead of the unassembled.

The quality of the board I must say is pretty good. I have underestimated the size of it. The power filter caps are bigger than my big toes! And I noticed a star at the left side of the board - I guess Naim took the "star grounding" a bit too literally

Also got myself new beefier speaker binding posts on the left. The right ones were free from PayPal credits I received. Too small and no insulation washers. I can fabricate new washers, but then they won't look too good.

Just look at the size difference between the two.

The case was bought from Taobao, but it must've been dropped some time during shipping and damaged one of the legs. I've reported and they're sending a replacement for free. Gotta love good sellers.

The screw snapped from inside the threaded hole

 

[penmarker]

Wired the inputs and outputs and turned it on today. Managed to snag a pair of old satellite speakers from my hometown and hooked them up.
***edit
I know the connections look like bare copper wires. They're actually polyurethane insulated wires, like magnet wires, but at 1.12mm width.

An externally hosted image should be here but it was not working when we last tested it.

They're so small and draw so little power, the amp stayed cool. Its been half an hour or so.
I'm getting the sound signature though. Quite nice.
One question, my amp thumps when turning on, but when I googled around Naim forums said they thump when they're turned off and were designed to stay turned on.
Do you guys have the same thump too?

 

[AndrewT]

LOOP AREA !
where are the signal pairs?

 

[penmarker]

By signal pairs you meant RCA input? It is in the middle between the two speaker terminals. Here is a better picture of it and the RCA connection to the R+ L+ Gnd.