NAP-140 Clone Amp Kit on eBay

Well, I replaced all transistors that didn't pass DMM test and checked and rechecked all the leads but still no change in bias voltage and -34vdc at output. Unfortunately, this board I purchased isn't the same as others. I haven't gotten to light bulb test because I maxed out my allotted time given by my girlfriend. I'm done (for at least 6 months). :sigh:

Thank you Ian and Ruwe.
 
Slowly I rebuild my Naim clone.
Actually the schematic based on the NCC200 kit. It sound so much better in any way compare to the Ebay Naim kit not even can mentioned in the same league.
I had two Ebay clone, the first I purchased many years back from Tubes hunter that time he sold his kit expensive and with original transistors (not counterfeit)
At first I did like the amp sound I did listen the amp over 6-8 months.
I ordered other kit for the PC board so I can start to test different transistors and capacitors in each place.
Later I found the NCC200 clone so I took the McBride PC board layout and I modded after the NCC200 circuit.
I built the new amp with the same component like my first Ebay kit.
Sound wise Heaven and earth between the amplifier to favour of the NCC200.
After I spent several (4-5) moths testing each place different transistors and capacitors until I get a really great amp.
Now I did not had chance to compare my clone with the orig NCC200 since that way to expensive just to get it for comparison.
All I can say I did listen my clone few mouths and I went back to the Ebay kit to compare it again after burn in.
I could not listening the Ebay kit any more the difference was so huge.
So I sold both Ebay kit..
Now after all those test I have to rebuild my clone because I soldered and re-soldered that PC board over 4-5 month it became a junk..
Now the new PC board was designed so I can test the Greg Ball modification to.
I hope soon I ca report back the result.
I kept one side the orig and another side with the mode.
I do believe it is important to burn in the components so when I test a amplifier I give min. 50hours to the new component to burn in.
Also important to match the components to each-other. Not enough to pick the best components and build.
My experience that often do not gave the best result sound wise!

Greetings Gabor
 

Attachments

  • Picture 026.jpg
    Picture 026.jpg
    65 KB · Views: 1,227
  • Picture 027.jpg
    Picture 027.jpg
    75.9 KB · Views: 1,219
Member
Joined 2010
Paid Member
...... My hands have a genuine NAIM 3R, 0.22 resistor bias is 2.8MV, now I'm the DIY NAP140 is transferred 2.8MV, sounds okay, but I would like to further optimize the use of the power transistor MJ15003.
The bias current for NAP 140 and most similar NAIM designs is already discussed several times in this thread. The current is in a range of 25-38mV maximum before instability becomes likely in the original Naim design with its small case etc. It will vary for individual DIY ideas.

Remember, the time for bias to stabilise is dependent on the size and type of case because there is no thermal connection to the output stage other than by air transfer or direct radiation. In some installations it may never stabilise and this is part of Naim amplifiers' typical long, >20 min. "warm-up" time.

The voltage is measured across one or preferably both 0.22R "emitter" resistors to determine the bias current. 11mV across 0.44R obviously means 25 mA current. This is a good starting point after first testing with zero bias when building. I hope this answers your question but don't forget to check out other posts discussing the bias setting. Other guys have different ideas and some may be OK if the current is still below the advised maximum.
 
Last edited:
Member
Joined 2010
Paid Member
....... Now the new PC board was designed so I can test the Greg Ball modification to.
I hope soon I ca report back the result.....
No result? Perhaps you found a very different sound with Greg's engineering approach having both LTP collector resistors equal. That usually kills the even harmonic distortion, and for this reason, NAP and NCC200 both have intentional gross LTP imbalance.

When heard a quickly converted trial version of the modification, I have to say I was really unimpressed. Perhaps if the clone sounded typically poor already, it might have been an improvement but this improved clone originally did sound close to genuine Naim. :(
 
The bias current for NAP 140 and most similar NAIM designs is already discussed several times in this thread. The current is in a range of 25-38mV maximum before instability becomes likely in the original Naim design with its small case etc. It will vary for individual DIY ideas.

Remember, the time for bias to stabilise is dependent on the size and type of case because there is no thermal connection to the output stage other than by air transfer or direct radiation. In some installations it may never stabilise and this is part of Naim amplifiers' typical long, >20 min. "warm-up" time.

The voltage is measured across one or preferably both 0.22R "emitter" resistors to determine the bias current. 11mV across 0.44R obviously means 25 mA current. This is a good starting point after first testing with zero bias when building. I hope this answers your question but don't forget to check out other posts discussing the bias setting. Other guys have different ideas and some may be OK if the current is still below the advised maximum.

Thank you for your reply, I try late.
 
Member
Joined 2010
Paid Member
Caowei, at such low bias, odd harmonic distortion begins to increase until you have large amounts of crossover distortion. You will certainly percieve this as clear sound initially, but perhaps after some hours you may not find it very smooth any more.

My questions to you are; does that current setting remain stable in a case or are you using this amplifier open, like a breadboard assembly? As I said, there will be little bias control without an enclosure fitted. Have you tried say, 35 mA bias?

If the amplifier becomes noisy as you increase bias, you have a serious problem with the power supply or amplifier. I'm not sure what, but no amplifier should perform like that. Perhaps you mean to simply say it is not as nice to listen to?
 
It goes away when I moved my hand over the board so I was thinking that it's picking up RFI from somewhere.

Funny thing is DNM make amps with hardly any RF shielding . They insist it sounds better . Getting the shielding just right works well . Amps which are slightly prone to RF intrusion ( interference word intended , time of day makes it worse ) often sound better . Richard Hay of Nytech was asked by a customer about this . Richard could cure the problem , it would make the sound less open .

The usual fix is to have a resistor and capacitor to the amp input . On my amp it is series 2k2 ( input ) and from that 47 pF NPO ceramic between input transistor base and ground . Ferrite rings on mains lead help . Sometimes earthing the chassis of the speaker units helps .

The radio you hear is the rectified radio signal across the junctions of the input transistor ( usually ) . In the cheapest valve radio's slope detection as it was called was common . It reduced the valve count by one as it combined the function of the next stage ( a triode replacing a diode , it also amplified , if I am right with a distortion penalty , using enhancement I think which is a no-no with valves as a rule ? ) . I was told years ago it is slope detection we hear in transistor amps .

Aluminium and stainless steel ( non magnetic grade ) are favoured over steel . Plywood box with a small mu metal shield would be considered excellent . The heat sink should help offer some shielding . The ply wood offering damping . Stopping a PCB from vibrating helps . Sony found out in the late 1970's that there is a small measured effect from PCB's moving in the magnetic field of the power supply . Cork ( floor tiles ) is a good damping material .

My advice if using an input capacitor is adjust it to almost be gone at the worst time of the day ( 20 . 00 hrs ? ) . Ear to the speaker says it is still there is fine .

Twisted input cables might help .
 
I suspect parasitic capacitances may be responsible for performance changes in and out of a conductive case/chassis.

I have reported before (more than once) that I have never managed to build an amplifier that is quieter when inside it's conductive chassis/case. This is particularly so when the case is substantially steel.
 
Member
Joined 2010
Paid Member
Over many years, I have repaired a few Naim original amplifiers and some of their balanced line preamps too. I often heard from the owners about their RFI breakthrough problems. One guy who owned a record shop and used an early Naim amp for his sound system, had huge problems because immediately outside were traffic control lights on a busy road intersection. In the days that taxis, couriers, police etc. used VHF, HF and CB frequencies, the experience in daytime was frightening - Inside the shop was quiet and you would be listening to a soft passage of classical music when suddenly, a rasping loud voice and piercing switching tone would startle you like an electric shock! :eek:

The problem with Naim designs was and is, simply inadequate filtering on the output in comparison to other designs. They need an output coil. A 0R22 resistor just doesn't do it. Others fitted it, as did Avondale on the same design and though I haven't tested one properly, I believe they are much quieter and have a wide group of happy customers. It's easy to make and fit one as illustrated in this thread or simply go to Avondale Audio's website and download the image file for the NCC200 amplifier schematic.

This doesn't stop all interference and one of the worst problems is long speaker speaker leads with the wrong reactance (L and C). The right type of cable, and it's not some exotic overpriced cable - snake oil issue, is detailed many times on the Naim audio site and forum plus the UK forums like PFM that recycle the topic ad infinitum.

For interest, I largely solved the problem with passing the leads 6 times through 32 dia. toroidal ferrite cores 0f 2-10 MHz range and changing the cable to a cheap, close spaced one, moving the amplifier to an elevated shelf between the speakers and removing the excess cable to about 2.5m lengths. Naturally, the owner was mortified to learn that ferrite had been brought anywhere near his precious Linn/Naim audio system but I didn't admit that until he confessed it sounded great and customers were no longer being frightened out their wits. When he later changed to a "Chrome Bumper" NAP200 and the agent refitted long cables, the problem returned.

This idea was not mine - it was recommended for similar locations in our local EA magazine, back in the 1970s. In typical suburban or remote locations, you won't need anything quite as drastic.

I don't suggest that these measures will be a universal solution but the NAP140 design is simple and old, so old that many RFI problems we now have weren't a big issue when it was developed. An amplifier though, is usually more than trafo, handful of caps, rectifier and cheap PCB/kit of electronic parts. Most commercial products now do have the additional filters suggested by Nigel, above. They have to, often to meet EMC requirements. The amplifier though, should be optimised for operation within a metal case and this will almost certainly be down to the specifics of the grounding arrangement and also the power connections for the whole audio system. This is not a simple topic and having read Douglas Self's chapters on grounding, layout and input section, I'm slowly getting a deeper understanding of the issues and how to sort them out for a specific build.
 
Member
Joined 2010
Paid Member
Clone substitute transistors

I've read sour comments about poor sound quality of Ebay Naim clones here. The comparisons with expensive originals and clones like NCC200 are absurd in cost and content terms but if you believed some of the posts, it could seem that even minor changes would improve them 100%.

I had a bit of time recently to swap transistors on a few damaged boards belonging to a newbie friend. I did this to find out if there was any truth to the comments and despite a great number of differences over the time that these LJM and similar kits have been around, some of the transistors prove to be fairly flexible in the substitutes possible.

TIP41/42 drivers seem unlikely with their stone age performance but somehow, they work with little difference to MJE243/253 at typical domestic sound levels. Output transistors have varied from 2SC5200/A1943 to the suspect MT200 Sankens now supplied but here I used my own supply of 2SC5200/A1943(O), which I am certain are genuine Toshiba.

I could go through all the substitutes for the small transistors but you may note that is essential to fit the right MPSA06/56 limiter types to get correct action and the kits I have seen do retain these. Input stage types don't seem to be too fussy but if you think there should be a benefit in using matched pairs, guess again. The input stage is intentionally thrown way out of balance to ensure that even order harmonics are not cancelled as the LTP is normally inclined to. This is largely what permits a desirable SQ and note that the NCC200 clone is just the same.

Problems at the input stage are usually with confused pinouts of any substitutes. Be careful and check the manufacturers datasheet, if they can be identified.

Now, the big differences I found were with VAS transistors. These should always be the original ZTX 653/753 from Diodes Inc. Why? because they have high Cob (30pF) and speed, an unusual combination in a tiny 2W package. Being expensive at ~$1 ea, they were among the first parts to be ditched from the clone kits when the competition heated up a couple of years ago but rest assured that the low Cob types like 2SD667/B647 (12-20pF)and BC639/40 (On-Semi 6-9pF!) (variable according to manufacturer) begin to sound awful. Why the kit suppliers didn't adjust the 39 pF miller cap. to suit, I don't know but that is what is needed when the VAS transistor is changed since Cdom (the miller capacitance) is composed of both the internal and external values. Ignore this at the peril of SQ and stability when messing with the VAS.

This was a pleasing discovery, considering the other possibilities and I don't suggest this is all you need to get really nice sound but it is a large factor that should be attended to as a basic and necessary start. You can be sure that top quality sound will not come from a jumble of cheapest source bits. The parts price or overall quality is not the issue but the specific properties that affect the design performance are - very much so here, it seemed to me.
I hope some of you are also tempted to drag out those weak sounding clone boards and try again, with attitude this time :)
 
Member
Joined 2010
Paid Member
You're absolutely right there, Nigel. There is little chance of just changing a couple of semis and coming up with an original product replete with Naim sound quality, if that's what you wanted.

You can arrive at a much more enjoyable sound quality than you get straight from the package, though, which makes the basic investment of about $40 for 2 stereo board kits a good one. Otherwise, you get something that will only be used a few times before being binned as a bad idea. Kitsellers are simply reselling a lowest cost selection of parts that fit the basic description.

I look at it this way; If Avondale can supply a simple PCB module version and a significant number up DIYs can arbitrarily assemble satisfactory stereo amplifiers from them that some claim are better sounding than typical Naim product, then they have exploded the myths which keep the original brand alive and got themselves a nice sounding amplifier.

Given the right guidelines (and that's the real issue), there is no reason a canny DIY cannot do likewise with his own careful choice of suitable genuine parts. It just might take a bit of time to get there. :eek:
 
Hi Ian and everyone . Julian Vereker told me that got so fed up with copying that he ended up giving guided tours of the factory . He told them truthfully anything they wanted to know . His mother was a psychiatrist I seem to remember . Julian said he often even showed other companies around ( if you knew him I suspect it was his greatest pleasure to do that ) . He then asked a dealer to supply a sample of that makers product to see if any ideas had been copied . None was the answer . Julian had rightly worked out that no one believes you to tell your secrets .

One secret was rejecting many parts . These were given back to the suppliers by arrangement , none were defective ! If anyone doubts this story maybe history repeats itself .

Making wine in the same . Can be done .
 
I've just built another one of these clones but its not working right....

Power supply configuration is:

300VA transformer into 15A/1000v rectifier bridge into 40000uf smoothing bank with star earth, output at this point is 37.5VDC.

It powers up fine, I set the bias to 7.5mv across R29 on each board, DC offset is below 50mv. Problem is that bias will drop steadily over 5 minutes, when it reaches 3mv it jumps to 0.4mv & DC offset jumps up to 4VDC ! At this point the bias adjustment pots have no effect... Turning the power off & then back on 5 minutes later it starts the cycle again.

Very confusing :confused: Any ideas ?


DSC01426 by j.buckmaster, on Flickr