advertisement


Which brand of Capacitors do Naim use?

Pete MB&D said:
I believe JV made an error, but you can't just up grade and amp with out admitting to making a mistake and recalling all the 250's, the 160's also suffered from the same mistake have a Google and see.

The 135's could also do with the smoothers turning round so the terminals are next to the rectifiers.

Pete
Click to expand...

Oh I don't know. In the old days Naim were far more happy to "upgrade" things than they are now. They happily upgraded 42's and 32's to the .5 variant so why not allow slight upgrades to power amps.

High end audio was seen as a much more experimental thing in the 70's and 80's than it is now.

The big change in philosophy seemed to accompany the introduction of the olive range to my mind, and since then that same philosophy has become more extreme.
 
Great post Martin - esp the bit about winter boredom :)

IMO the wiring layout in the 250 is not the problem with it at all. The cables are all fairly heavy gauge and comparatively short, so there is going to be very little in the way of error voltages generated across them.

The real problem with the 250 lies in its regulator having a relatively poor HF response. You can fix this quite effectively and easily by adding some capacitance across the power terminals of the amp boards. See here for the detail -- http://www.pinkfishmedia.net/forum/showthread.php?t=164025&highlight=Naim+NAP250

Mr Tibbs

(only slightly bored ATM)
 
Mr Tibbs said:
IMO the wiring layout in the 250 is not the problem with it at all. The cables are all fairly heavy gauge and comparatively short, so there is going to be very little in the way of error voltages generated across them.
Click to expand...
Inductance dominates gauge. Doug Self and others showed how this sort of error caused high levels of hum and distortion a long time ago in Wireless World and other publications
 
davidsrsb said:
Inductance dominates gauge. Doug Self and others showed how this sort of error caused high levels of hum and distortion a long time ago in Wireless World and other publications
Click to expand...

There is no audible hum with the standard wiring and distortion levels in these amps has always been low.

Mr Tibbs
 
Mr Tibbs said:
Great post Martin - esp the bit about winter boredom :)

IMO the wiring layout in the 250 is not the problem with it at all. The cables are all fairly heavy gauge and comparatively short, so there is going to be very little in the way of error voltages generated across them.

The real problem with the 250 lies in its regulator having a relatively poor HF response. You can fix this quite effectively and easily by adding some capacitance across the power terminals of the amp boards. See here for the detail -- http://www.pinkfishmedia.net/forum/showthread.php?t=164025&highlight=Naim+NAP250

Mr Tibbs

(only slightly bored ATM)
Click to expand...

I assume this is the same for the 135 too, about the poor HF Response?
 
colasblue said:
Oh I don't know. In the old days Naim were far more happy to "upgrade" things than they are now. They happily upgraded 42's and 32's to the .5 variant so why not allow slight upgrades to power amps.

High end audio was seen as a much more experimental thing in the 70's and 80's than it is now.

The big change in philosophy seemed to accompany the introduction of the olive range to my mind, and since them that same philosophy has become more extreme.
Click to expand...
If there was an error, they would have been recalled, lot's of manufacturers have done this over the years, do you honestly think they would risk their company reputation for such a trivial thing to carry out.

This is the first time I have read anywhere about this "error"

If the error does indeed exist, what effect would this have on the sound of the amp.

Meant for PeteMB&D
 
It's not dangerous, its not terrible, it does work as it is, it just could have ben done better.

Certainly doesn't justify a product recall.

@ragaman

IS YOUR HEAD ACTUALLY SO FAR UP YOUR ARSE THE ONLY WORLD VIEW YOU GET IS THROUGH YOUR OWN NOSTRILS?
 
Ragaman said:
This is the first time I have read anywhere about this "error"

If the error does indeed exist, what effect would this have on the sound of the amp.
Click to expand...

Well, it does exist: finesse in layout is a more real issue in terms of getting 'ultimate perforemance' than actual design, beyond fairly simple designs. No its not recallable - see my previous post for a wider context from a disinterested outsider.

There are no shortage of bits of kit out there where the layout has been done straight off datasheet and performance actually realised is considerably less than is possible (if you want other examples that are well-known - the Rega Planet CD player and the Squeezebox have pretty simple-to-rectify issues long ago discussed here. Many many others out there... hey, I even found the Parrott MKi9200 in my car has a layout designed for the amp section by someone who obviousy does not understand analogue design sensitivities and cloned the IC datasheet, not very well. 10p's worth of added capacitor yields better performance, sufficiently improved you can notice it at low volume in a car on the move.. ie SNR 20dB or less...)

Specifically with the 250? more IMD than the design is capable of, lends a shoutiness at higher levels that is easily rectified. Most of those on the upgrade treadmill aspire to 135s instead though, which being monoblocs elide the issue. Its not a field for rational discussion at times.
 
Just pulled it down, it can get wedged on occasion due to my supersized brain

I only asked as I wondered if Naim may have engineered it this way, to sound a certain way to match the sound they produce as their products are not the most neutral around. Why I asked what effect it may have on the sound as I have read many times their products have an enhanced energy to the midrange especially.

Maybe you should squeeze your head out once in a while, no offence ;)
@colasblue
 
none taken - sorry if I caused any!

Just bear in mind that most of us in this room have ripped the guts out of their Naim kit and replaced them with what they consider to be better.
 
The regulated Naim amps (250 and 135) have a basic defect in the design of the regulator - it is "wrong way up". Any feedback based regulator has to have loop gain that falls as frequency rises, usually implemented by making the main gain stage into an integrator. The Naim mistake is to reference this to the input rail, rather than to earth. This means that at high frequencies, noise on the input is exactly copied to the regulator output. If it was earth referenced, it would be vastly attenuated.

Parts count would be unchanged; this is just a shaming mistake caused by lack of understanding.
 
Maybe this is the reason why when I choke regulated the front end of a NAP500, it improved to a very large degree-much more poise, control, clarity. The choke of course preferentially smooths out HF noise, while the smoothing caps and regulators smooth off LF noise.

PigletsDad said:
The regulated Naim amps (250 and 135) have a basic defect in the design of the regulator - it is "wrong way up". Any feedback based regulator has to have loop gain that falls as frequency rises, usually implemented by making the main gain stage into an integrator. The Naim mistake is to reference this to the input rail, rather than to earth. This means that at high frequencies, noise on the input is exactly copied to the regulator output. If it was earth referenced, it would be vastly attenuated.

Parts count would be unchanged; this is just a shaming mistake caused by lack of understanding.
Click to expand...
 
PigletsDad said:
The regulated Naim amps (250 and 135) have a basic defect in the design of the regulator - it is "wrong way up". Any feedback based regulator has to have loop gain that falls as frequency rises, usually implemented by making the main gain stage into an integrator. The Naim mistake is to reference this to the input rail, rather than to earth. This means that at high frequencies, noise on the input is exactly copied to the regulator output. If it was earth referenced, it would be vastly attenuated.

Parts count would be unchanged; this is just a shaming mistake caused by lack of understanding.
Click to expand...

That is astonishing. Any ideas why they did that. A low noise and interference free reference is fundamental to a series regulator.
 
john.luckins said:
That is astonishing. Any ideas why they did that. A low noise and interference free reference is fundamental to a series regulator.
Click to expand...

This isn't a problem with the reference, it is the Miller compensation in the gain stage.

Look at http://www.neilmcbride.co.uk/output-reg2.pdf

You can see that the dominant pole compensation is provided by C104 (C204 on negative rail). At high frequencies, the gain of Q105 falls, so it output tracks the raw supply at its emitter.

If instead they had swapped the polarity round, so that the Q105 gain stage was earth referenced, at high frequency it would just be steady. I expect you would get 20 or 30dB of extra rejection.

The reference supply is prefiltered by R101, R102 and C101. You could replace R102 with a two terminal constant current diode for a bit of extra rejection, but that wouldn't help HF performance.
 
Thanks PD, I'd never seen this regs circuit diagram before. Where would you put the comp cap in that circuit? From the base of Q5 to 0v presumably? With a series R again to set the 3db frequency?

John
 
john.luckins said:
Thanks PD, I'd never seen this regs circuit diagram before. Where would you put the comp cap in that circuit? From the base of Q5 to 0v presumably? With a series R again to set the 3db frequency?

John
Click to expand...

No it would still go collector to base, just that the polarity would now be reversed, and emitter connected to earth rail. Obviously LTP also needs to be reversed.

Series R is for pole-zero compensation - try to get a bit of extra bandwidth by splicing compensation rolloff onto natural openloop frequency response. Difficulty is that the pass transistor sees operating current vary from about 40mA to 10A, and Ft will vary by a factor of a few over this range, so it is not easy getting pole-zero to work cleanly in this type of circuit.
 
The Naim circuit does have poor HF rejection.
Designing a voltage regulator with good HF rejection AND stability into a wide range of loads is actually quite difficult. Too often a power supply is an afterthought, left to the trainee etc and cribbed from text books and application notes
 
davidsrsb said:
The Naim circuit does have poor HF rejection.
Designing a voltage regulator with good HF rejection AND stability into a wide range of loads is actually quite difficult. Too often a power supply is an afterthought, left to the trainee etc and cribbed from text books and application notes
Click to expand...

And high current / high power regulators are even harder!

At least in this application, a specific load is known, so the design only needs to be stable running a NAP circuit.
 
Would it be the front or back end of the nap amplifier circuit that is most affected by the regulator?
 
You must log in or register to reply here.


advertisement


Back
Top