NAP-140 Clone Amp Kit on eBay

thanh1973 said:
Thanks
I will do as you suggest unless someone says otherwise.
So I can assume the B1 buffer should be fine with this amp.


You can use it with any buffer-stage og preamplifier intended for general use. It is a very ordinary power amplifier when it comes to amplification and input impedance.
BTW, in principle you can bypass the input capacitor if the buffer has an output coupling capacitor. (I do it). Good luck.


To the general discussion:The very slight harshness I experienced in the beginnning, very soon disappeared completely (whatever the reason). The amp may still give a slightly dry treble, but there is nothing unpleasant about it, to my ears. Since the amplifier cirquit originates in an RCA design from a pre-otala age (about 1970), it is amazing how well it performs in this department, can one say that?.

I have been lazy lately, but next I will try it with a tube preamp. I have little respect for the Naim 'ideology', which I suspect originated in the marketing department, not among engineers. As I understand it they want you (actually they force you with their peculiar connectors!) to combine equipment from the same source. (Pure nonsens, IMO).
 
Hi,

The only reason I see in DIN connectors is that only one wire transfers the ground between devices. With RCAs each channel has its own connection to ground, which more or less creates a ground loop - common output point, common input point, but two connections between them.

Otherwise, I agree with you, Sletol, about the marketing tricks. But I assume that the 'system' approach of Naim has some grounds - it helps reduce the possible mismatch between components. Note that most audiophiles, or regular Naim buyers, don't understand much electronics. The system approach guarantees that someone wouldn't connect his beloved preamp, producing 10V output, to the 0.4V Naim power amp input, for example. In addition, they expect from you to buy one more of their products, can't blame them for that ... :)

Glad to hear that your clone sounds right. I like mine very much, but I can't stop myself to experimenting. I tried different driver transistors lately. For sure 44D11/45D11 created oscillation in my design, compared to MJE243/253 used before. I'll try very soon ZTX753/653 but I'm not sure how they will behave at higher power demand (they are only 1W transistors in TO-92). I'm curious, though. Finally, I can always insert the TIP41/42 as supplied in the 'original' kit.

Regards
 
Well everyone, I have received my clone kit (the one from gigawork on ebay). I decided to use 2sc2240 for the input transistors instead of the supplied BC550 transistors. However after purchasing these and looking at the data sheet, I noticed the 2sc2240 has a different pin-out than the bc550 (2sc2240 pinout is bce, while bc550 pinout is cbe).
So I have bent the pins of the 2sc2240 transistors so they will be compatible with the board.
I have also insured that the first transistor has a higher hFE than the second to make setting the DC offset easier. Please note the first transsitor is actually the inner transistor on the board not the outer one.
The hFE values for the transistors are about 500 for the first transistor and about 400 on the second.
I will keep you guys updated as I make futher progress.
I am also 80% through completing my first JLH amplifier (40 Watts per channel) so I will be definitely be comparing these 2 amps sound wise.
 
Hi again,

The following information is the final addition to my previous post #241. I think it will give you the last drops of performance that you can extract from this circuit.

I tried almost everything possible – I've played with driver transistors types, ground configurations, changes in protection circuit, changes of the input and output components, different types of feedback capacitors and so on, but these last modifications gave me the most significant improvement. Most of them are as a result of reading the Douglas Self book about audio amplifiers, which is, IMO, about 90% relevant to this circuit. He helped me better understand how this amp actually works.

Right now I basically have the original kit schematic with coil/resistor at the output, increased values of protection resistors R32 and R33, ZTX 653/753 for Q4 and Q6, MJE 243/253 as drivers and non-polar capacitors in the feedback. Everything else is as supplied originally with the kit or according to my first post (#241)

What's new? :

1. I changed the Baxandall diode type from 1N4148 to 1N4005, because I found the 1N4005's Vfrd curve is closer to the Vbe curve of the output transistors. You can check it in the diode spec sheets by comparing with your output transistors spec sheet.
To prove my point, I also found this link with distortion measurements of some amp with 2SC5200 output transistors, and as you'll see the Baxenadall diode type 1N400x add the least distortion and the 1N4148 the most: http://www.wimdehaan.nl/hybrid/mugen/measure/

2. If you don't use exactly 40V power supplies you may want to compensate the current through the diodes D6 and D7 by changing the value of resistor R20 accordingly. This will provide the intended correct voltage drop provided by the diodes and therefore the correct current through Q1 and Q2. As a result Q1 and Q2 will have better current balance, hence less distortion.
You can additionally change the balance between Q1/Q2 by changing slightly resistor value R21. I personally have a multiturns trimmer in parallel with 680 ohms resistor there, but I'll tell you that if you use the right current through the diodes D6,D7 resistor R21 will stay slightly below but around 620 ohms. With right Q1/Q2 balance you'll notice also very low output DC component - 3mV or less, in my case for both channels.

3. This upgrade is thanks to Mr. Douglas Self insight into feedbacks' paths. He says: “ Class-B output stages are a maelstrom of high-amplitude halfwave- rectified currents, and if the feedback takeoff point is in slightly the wrong
place, these currents contaminate the feedback signal, making it an inaccurate representation of the output voltage, and hence introducing distortion.”
Well, according to his explanation this circuit board has the feedback taken from the wrong point! The right way to do is to take the feedback from the output coil and to send it directly to the common point of C12 and R9. Originally, on the kit's PCB the line between C12/R9 and the output coil is “contaminated” with the protection circuit components and also the path is taken from a place where the “maelstrom” happens. The connection should be physically moved on the coil and also the effect of the additional components must be minimized.
The easiest way to do it correctly is to remove the common ends of C12 and R9 and to connect them on the component side of the board. From there take a good quality wire, about 1 in. long and solder them to the output coil wire (again on the component side). I heard improved sound when I did it, and it also has electrical benefits – for unknown to me reasons the balance between Q1 and Q2 has improved too.

That is for now. There is not much else to do in this amp.
Enjoy your music collection!
 
I have completed one channel of my nap 140.
I hooked it up to bench power supply and measured dc offset at various rail voltages.
The lab power supply I have only goes to 30V so I have used a curve fit to predict dc offset at 40V (ie dc offset of approximately -35mV).
While this is a satisfactory result, I am thinking of choosing another transistor (same model number) with lower Hfe for the second LTP transistor to reduce dc offset further.
I am currently using Hfe of 500 and 400. I am thinking of using 500 and 300.
Anyway I thought this might help people with dc offset problems.
Have a look at the excel attachment for the results.
 

Attachments

  • dc offset measurements.zip
    2.6 KB · Views: 428
Tube Shunter nap140 boards

hello all

I have just received some of these boards and apart from a few missing/wrong components they seem pretty good.

Anyway i have built them according to the pcb/cct diag and they nearly work, at least there was no smoke :D


i read somewhere in this thread that there were some reversed pcb legends but mine <seem> ok, does anyone know if they have been modified ?

The strange thing is they produce music with no problem but I am having a bit of difficulty in setting the bias across R29, I get 0mv regardless of tr1 setting ? any ideas.
I have used a BC546 at Q5 - not got round to using the supplied TIP41 - could this be the problem ?

Thanks in advance for any suggestions.

ta

Graham
 
Hi Graham,

Stupid question but ... did you check whether the BC546 is correctly oriented on the board? ... TIP41 has different pin out and you should have two of the pins of BC546 crossed.
My first guess would be that the problem is in Q5. My second, that Q11, Q12 are not connected correctly and you're hearing only sound from your driver transistors Q9 and Q10.
Do you have power from your amp, or just low level output? Do you have voltage across R30 or it's 0mV too?
The circuit should work fine as is. The wrong pin configuration of some of the transistors are probably corrected. It doesn't mean there are not other errors. Why don't you check where all path are going?! Are you sure in all resistor values?

If you find where the problem is, don't forget to turn back CCW the TR1 before turning on the power supply, or you may blow the output transistors at start-up.

Regards
 
Hello chaps - thanks for the replies

I have attached a picture of both boards just after they were assembled, R31 & one C9 missing.

Q1/2/5 are all BC546 and they are all oriented according to the board legend.
The sound was reasonably loud (in mono), I was using an ipod as the source which has an adequate o/p level, if not as high as a pre-amp - so I think it is more than just low level - I will check again though (when I can - gotta go to work !). It's too bad that work is allowed to get in the way of the real business of life.

I didn't check the voltage across R30.
When I built it I was extreme in caution when checking all resistor values, I checked each one with a dvm - if any are wrong i shall cry !
I have also been through the orientation of all trans (compared to datasheets) and I <thought> they were correct - still happy to be proved wrong though !

The o/p trans are MJL3281 and are (l-r) BCE which seems to match the board tracks ok.

Another useful piece of info is that the dc offset at the speakers was very low - compared to some reported in other posts, I think it was around 5-6mv.

When I get back to my bench I will start from scratch with checking and applying any suggestions made here.

there is a pic here : (and others if you go up the directory list)

http://people.exeter.ac.uk/gar202/p...kBoxAmp/Phase_2_Naim140/TS_NaimNAP140_008.jpg

thanks again

Graham
 
Graham,

I think you have a working amp if your DC offset is so small and you have output power.
Now, are you sure you really see 0mV across R29? Sometimes strange things happens with the DMM probes.
In this situation one reason to have sound but no DC voltage (mV range) across R29 and R30 can be that the output transistors are deeply in class B (in class C to be more precise) and you can't move them towards conduction with the available trimmer range. This is not very likely, unless the trimmer is defective. So, check that your trimmer is OK. Turn it with power off and make sure that you see change in resistance between the mid pin and some of the other two. If it's OK, check the resistor values R15,16,17 to be on the safe side. And definitely check the paths on board.
All this provided that you have real output power.
If somewhat there is not watts of output power or you hear too much distortions, your problems are in the area of the output transistors.
Troubleshooting this way is more of a guesstimate, though. Would be more helpful if you post the circuit with measured voltages in each circuit node.

Regards
 
Hi

I am not going to be able to do any more on it until the weekend - but I have had a bit of a go today.

Neither amp responds to changes in TR1, although I can see changing resistance across R17 as I adjust TR1. I checked resistor values R15,16,17, I also checked the tracks to Q5 and they are all correct, but it does look like the problem is in the Q5 area.

Fortunately the current for both amps on both rails seems to naturally settle around 20mA with no signal which would explain the operation of the amp and no smoke !

They do sound pretty good though, quite clear with lots of bass, I am using slightly rubbish speakers though.


Graham
 
Hi

I agree - I will definitely be setting it up to run at about 35 mA as soon as I find out why it doesn't adjust !

I can't leave it as it is - you never know how stable it will be, and having sacrificed one (cheapo) bass driver recently..... (not on this amp I might add)

Tube Shunter did supply a tip41 for use Q5 instead of the BC546, perhaps he knows something.....

Graham