NCC200 LTP

[a.palfreyman]

There are two main changes from NCC200 to 220:
1) Quasi > fully complementary o/p stage
2) Change of R2 from 22K to 1K

The effect of #2 is not that subtle, it cleans up the sound and makes it sound less "dense" i.e. less of the Naim sound.

I would hazard a guess that this makes more difference to the sound than the changes to the o/p stages.

@Joppe - you could try linking/shorting R2 on your NCC200 boards and have a listen.

@S-Man - No, definitely not; I've long said look at the effect of the large R2, model it, and realise what it contributes to 'naimness' by imposing a definite, shall we say, harmonic profile on the output? If that's what you want, that's mostly how it is done! Move away from that and logically you are into finessing a classic 3-stage, LTP/VAS/OPS, Class aB amplifier much like any other.

(To me - like most things audio - beyond a certain competent minimum, it's mostly about personal aesthetic choice more than much else: and that is what DIY is for)


ETA: do I have to point-out that in such a DIY context, unfortunately the choice of what is 'competent' also tends to become somewhat subjective? That's likely a separate debate, entirely!

Totally agree Martin.

I have listened to the R2 effect, back and forth, quite a few times. It's an interesting subjective effect.
Setting R2 to to 22K (or 7.5K in a NAP200) gives a more dense sound which can be quite appealing. I have found I can listen to new music at modest levels all day (e.g. Radio Paradise) on these type of amps, due to some sort of tune-enhancement (or simplication?). On a more neutral amp there is more variation to the sound from one track to another, with some stuff sounding worse, but most sounding better. The really telling test is when you wind the wick up a bit and the denseness of the sound becomes tiresome.
Maybe this is why owned and sold NCC200s twice. I should have tried the R2 mod on them!

Having read these comments in the Qudos thread I decided to try the variation 2) and put 2k7 across the 22k resistor that feeds TR2 of the LTP of my NCC200s (giving as near as dammit 2k4). First hour sounded very bright (but have experienced this before so knew what to expect) and it settled down over the next few hours. There seems less "projection" in a 3D sense, but this is more natural and there was definitely a change in the tonal balance. Following day, played for several hours and found several benefits, pretty much as described above. I had been having problems with a lack of focus in the centre of the image (particularly with trumpet left and sax right which could tend to 'merge' in the middle of the image) and this was now much better. Instruments sound more natural and the soundstaging (particularly depth) seems more natural: piano notes on the right-hand could appear to 'jump-out' at you but not after this mod. I found though that some frequencies seem reduced in output so I can agree with the description of a "tuning" effect with the 22k in place. Will be giving this some time to see if my ears settle with this because to me the improvements far outweigh the negatives.
Made some rudimentary measurements:
Before mod: 1k (TR1 feed) has 0.54Vdc across it and 22k (TR2 feed) has 10.1Vdc across it indicating a standing current of 0.54mA and 0.46mA respectively.
After mod: 1k (TR1 feed) has 0.52Vdc across it and 2k4 (TR2 feed) has 1.15Vdc across it indicating a standing current of 0.52mA and 0.48mA respectively.
Could this "rebalancing" of the collector currents have an effect on the sound? I don't know enough about electronics to answer that. The only other thing I can think of is the ratio of 22k on TR2 and 27k feed-back resistor. These have an almost exact ration of 9:11 so could this be causing a tuning effect where notes of a similar ratio played simultaneously "affect" each other?
Best,
AP
Forgot to mention this had no vast effect on the bias current or the DC offset on the output

 

[S-Man]

Forgot to mention this had no vast effect on the bias current or the DC offset on the output

And it marginally improves the phase margin (if anyone was worried about stability).

 

[MJS]

I suspect that the fact that these have remained unbalanced over the years is part of the appeal of the sound of this amp circuit. If it ain't broke..etc.

I've never played with these resistors, or a current mirror here. It's on my (very long) todo list.

 

[martin clark]

+1 @MJS

By the time you start sticking a current mirror in there, I think you are into regular 'Lin 3-stage' amp (re)design...

 

[Arkless Electronics]

FWIW a current mirror increases OLG by 12dB and reduces THD from 0.06 to 0.02% @ 50W 1KHz according to a sim (already had the schematic in my simulator). Changing the LTP resistor from 22K to 1K appears to do sweet FA according to my simulator but I only tested at 1KHz into 8R and nothing else...

 

[a.palfreyman]

I suspect that the fact that these have remained unbalanced over the years is part of the appeal of the sound of this amp circuit. If it ain't broke..etc.

I've never played with these resistors, or a current mirror here. It's on my (very long) todo list.

You have to try it. Played for another 4 hours last night. The sound is opening out, more detail and some "confusion", especially in the centre now pretty much gone. I might have to play around slightly with my mid-tweet cross-over as some higher Freq sax notes can be a little forward, but have yet to try some rock (been listening to Blue Note jazz) as rock always seems a bit darker in terms of mix. There is some delightful cymbal work on The Sidewinder and Takin' Off that really shine through in a natural way without being obvious. I won't be reverting back to 22k.
I know little about electronics but been speaking to an Instrumentation Engineer at work. In the new year we may try and get some measurements done at 22k and 22k//2k7 (2k4) to see if there are any measurable differences. As he said it is easy to test with single tones but difficult to test with music which is much more complex and single tones may not show up any possible interactions between harmonics of lower Freq notes with higher Freq notes. In the meantime I might build a simple diff amp and connect across the l+r speaker terminals and scope the results. Only problem is it is sounding so good I don't want to mess too much with it as IME newly soldered components (even a single resistor) 'seem' to take at least 10hrs to settle in.
AP