Schematic for naim amp

ALW

Member
2001-06-12 10:15 am
UK
True enough

As topologies go, they don't get any more generic.

That's true, strange though how almost every other amplifier I've heard just doesn't get the musical juices flowing in the same way these do.

So many manufacturers, so many designers, so much more advanced circuitry....so little music ;)

Ever heard a Krell - makes me want to go to sleep!

Andy (Naim user).
 
Re: True enough

ALW said:
That's true, strange though how almost every other amplifier I've heard just doesn't get the musical juices flowing in the same way these do.

That they do. I'd be interested in a schema too. One of my NAP160s has a bad ocsillation that we are having trouble sorting out. I must say thou that the SE EL84 amp i popped in its place does some things musically the NAIM doesn't.

dave
 
...musical juices flowing...

I can second that.
Although I heard many amps in the last twenty-something years, and some of them did certain things much better, no others amps did deliver music in general with the same ease and sheer power than the NAIMS.
Sweet ? Sometimes.
Spacy and etherical ? No.
Musical pace and drive ? Yessir.
Boring (like so many others)? Not at all.


But DIYers or cloners beware: There are more details to consider than just the schematic. There are several threads here at the forum btw.

Klaus (NAIM user - originals and DIYversions :) )
 
Fussy Amplifiers

Hi Dave, my experience of Naim amps is that they require highly inductive speaker cable, and I have read that this is intentional.
The Naim cables that I have seen look like 300 ohm tv ribbon cable only way oversized - standard transmitter cable maybe.
Each leg consisted of not very many 0.8 mm or so strands insulated and seperated by stiff black plastic dielectric.
I think it was a Nap-160 that did not like 10-pair telephone cable (reasonably capacitive and virtually zero loop inductance) and oscillated and sounded loaded down a bit, got a bit hot but did not break.
This is the only amp to misbehave with this cable IME.
These amps are very load critical/dependant IME, and when sitting right they are quite nice and when loaded wrong they go wrong.
Short cables when bi/tri amping might be inappropriate.
If it is board level ask me more.

Eric.
 

ALW

Member
2001-06-12 10:15 am
UK
Fussy amps

Naim choose not to fit the standard inductive Zobel at the o/p, using their own, relatively inexpensive, 'speaker cable to fulfil the same role.

Since speaker cable is a necessary evil, it's a good solution, providing you are happy to hear the amps as the manufacurer intends, and not tune by using cables.

The cable needs to be low capacitance (it is this that causes stability problems). A minimum length of 3.5m / channel is required to guarantee stability.

The spec of NAC A5, their current speaker cable is: -

Capacitance - 16pF per metre
Resistance - 9 milli ohms per metre.

Any cable similar to this will be fine.

A.
 
Re: Fussy Amplifiers

mrfeedback said:
my experience of Naim amps is that they require highly inductive speaker cable, and I have read that this is intentional.

If it is board level ask me more.

I worked at a dealer when the amps i have were sold new (and grabbed them when they came up used)

It is at board level -- had 3 of these amps (1 went off to Oz and now i have 2). The amp was happy for years (use 12 guage ZIP cable) and then one day my wife came home and one channel of one amp was dumping its entire rail voltage at hifrequency into my acoustats -- she had to take the phone outside to find out how to stop the thing. Tramatized the cat that was in the house. When i got home i pulled it out, went and got the backup from downstairs, plugged it in and everything was fine again (if they had been dynamic speakers i'd be getting new drivers i'm sure).

I took it to a real tech (i get lost with SS as of yet), where when he 1st hooked it up, it worked fine for a bit before doing its thing. Replaced ALL the caps as per NAIMs suggestion but it still isn't fixed. Any help would be most appreciated.

dave
 
Naim Amp mods

Hi,

yes the standard Naim configuration is sensitive to cables and oscillation, and has a bad treble.

All can be cured, if the hi-frequency behavior is analyzed, and the compensation changed to state of art.

I did it. It cured the dry treble and made the power amp stable with hi-capacitance cables.

The mods are 2:

first remove the two anti-oscillation styrophene caps from VAS output to rails and reroute them to ground, where they belong to (they are going from VAS input to ground afterwards, to make it clear). After this, the power amp are no longer that sensitive to power rail distortions.

second change the Miller compensation to feed forward compensation, like used in the Leach amp: reroute the Miller cap (22pF if I recall correctly) from the base of the VAS stage to feedback network and voila ... sunshine.
(the cap then goes from VAS output to -input of differential input stage then).

Last but not least, you could apply some very small bias to the power transistors, like 10mAmps. This will smooth the treble region, too.

When I posted that mods to the Naim forum, that post had been cancelled ;-(

regards,
Hartmut from Munich, long time Naim owner, transistor and tube DIY fanatic, tonearm collector and so on.
 
?

Hartmut,

thank you for your suggestions.

I tried to figure out what you mean but I do not understand what you are talking about:

1st tip: What two anti-oscillation caps ? to rails ? There aren't any.

2nd tip: base to FB network or to - of diff amp ?

3rd tip: bias ? they are already biased at about 10ma.

Can you be more specific and post a detailed description or schematic please?

Thanks,
Klaus
 
Naim mods explained

Locating the mentioned caps in the Neil McBride schematic:

the Miller cap C6 is that around Q5, it is 47pF there. It is 22pF in Naim original schematics.

I rerouted that cap, which going from collector of Q5 to base of Q5 originally, to going from collector of Q5 to -input of differential input stage.

two anti-oscillation styrophene caps from VAS output to rails are missing in that schematic; if they were in the Neil McBride schematic, they should be from collectors of Q5,6 to each power rail. The values in the original circuit are 470pF or 560pF. Neil McBride left them out, wise guy ;-)

You can either leave them out, too, or reroute them going from collectors of Q5,6 to ground.

Bias: all Naim power amps, that I have measured so far, with generic Naim power transistors have zero idle current, only those with that Sanken outputs (early chrome bumper case NAP140 have these) have about 10mAmps.

Warning: check with oscilloscope for oscillations! I am not responsible for malfunctions, fried tweeters, blown output transistors, and so on.

Do modifications only, if you really know, what you are doing ;-)

regards,
Hartmut from Munich
 
Elbow Grease

Hi Hartmut, these Naim amplifiers seem like a suspect way to spec an amplifier for the market place given the variance in choice of loudspeaker loads and the cable connection IMO/IME.
I understand that your modifications help to improve stability, and what is your sonics critique of the result that you now have.
I recognise that as original these are 'fast' sounding amplifiers but I have never been in love with them.
Amplifier strong load dependence smacks of idiosyncratic resultant sound IME, and whilst such an amplifier/load can be tuned to give a very satisfactory portrayal, a clean and non load dependant amp stage can be much superior, and much less tweaking hassles.

Dave, it sounds like you have a temperature dependent problem - dry solder joints and marginal electrolytics can be the problem.
Another cause of oscillation problems that I have found in the past are things like small value ceramics going intermittently open circuit/or leaky, also front end small signal transistors going bunky (leaky/noisey/open circuit) and this can be time, temperature or mechanical shock dependent.
Hair dryer and freeze spray happily/sadly may or may not reveal the culprit.
Swapping semis or ceramics from channel to channel may reveal the suspect component but may temporarily cure the duff component also.
Blanket changes of electrolytics (I understand that you have done this) and a very critical solder joint inspection and as required resoldering can be effective.
On older (and a lot of newish) gear, for QA, no workshop warranty returns/blowups, and long term customer sonic and reliability satisfaction I usually just blindly (a bit of thought and diligence are involved) blanket resolder both channels and psus, all connectors and controls, solvent (isopropyl alcohol) wash the boards, contact treat all connectors and controls, and then run it up on a variac and hey presto a 'brand new' amplifier - the result can be even better than factory with modern low esr electros IME.
This process takes some time, solder, solvent and efforts but is the only sure fire method to return an amp to channel matched operation and ultimate reliability - the result is more than the sum of the individual efforts.
By the time you have swapped a number of components from channel to channel you have rendered the item sonically non factory anyway.
One note here - renovating all the solder joints that count with new solder will change the sonics somewhat - if you are keen you can do a solder type A/B sonics comparison - seriously.
There are further sonics effects and run up procedures- email me if you want more info.

Works for me,
Eric.

Given the stability/load requirements of these amps maybe the zip cord that you are using is sailing too close to the wind and this needs to be checked also.
 

ALW

Member
2001-06-12 10:15 am
UK
Bias settings

Bias: all Naim power amps, that I have measured so far, with generic Naim power transistors have zero idle current, only those with that Sanken outputs (early chrome bumper case NAP140 have these) have about 10mAmps.

I respectfully suggest that either the amps you were testing were faulty / maladjusted, or you were measuring incorrectly, since all Naim amps, irrespective of o/p device, have some bias current. Why would Naim fit a bias pot otherwise?

There's a resistor in series with the o/p that can be mistaken for the emitter resistors ;)

I always set mine for min THD.

A.
 

ALW

Member
2001-06-12 10:15 am
UK
Eric

The speaker cable isn't a variable as it's part of the amplifier design. You try other cables at your own risk and the manufacturer warns against voiding of warranty.

Once you accept the above, the amplifiers are not, in any way, load dependant, any more than any other amplifier.

They are unconditionally stable, providing a min. of 3.5m of Naim cable is used.

The only limits are the normal ones of power etc, thereafter.

A.
 
Re: Eric

ALW said:
Once you accept the above, the amplifiers are not, in any way, load dependant, any more than any other amplifier.

They are unconditionally stable, providing a min. of 3.5m of Naim cable is used.

I know that back in the days of working in the hifi store, the NAIM NAP250 was the only amp we could find that did a decent job of driving the wicked ugly load (especially a stacked pair) of Dayton Wright XG-8s ESLs.

dave