its the question that drives me crazy, how and why it works the way it works.
As IanFinch said, it will be always about that "narrow" operating point, and in order to "headshot" it we basically must understand everything.....
I need more resources to make it happen. Lots of modules must be build. Cables, grounding, everything on the same plane, coupling and so on .....
Some think that we are totally tuning in to "AC impedance" 230V/120V if we get it to that "narrow" operating point ---- that would be worst of all things if thats true
EDIT: Watching movies just with 1 speaker and 1 amplifier unit is incredible, even simplest "gun-shot" gives u that in-ur face feeling lol. Mind-blowing.
As IanFinch said, it will be always about that "narrow" operating point, and in order to "headshot" it we basically must understand everything.....
I need more resources to make it happen. Lots of modules must be build. Cables, grounding, everything on the same plane, coupling and so on .....
Some think that we are totally tuning in to "AC impedance" 230V/120V if we get it to that "narrow" operating point ---- that would be worst of all things if thats true
EDIT: Watching movies just with 1 speaker and 1 amplifier unit is incredible, even simplest "gun-shot" gives u that in-ur face feeling lol. Mind-blowing.
Last edited:
That does seem incredible, rensli. It is hard to get exciting dynamics from mono amplification and typical music program but movies generally do have enhanced dynamic range for their dramatic sound effects that are intended to make you recoil, squirm or jump out of your seat 
while the movie plot goes from one shock to another. We probably feel invigorated after these dynamic blasts and shock effects - like we're getting more for our money too.
while the movie plot goes from one shock to another. We probably feel invigorated after these dynamic blasts and shock effects - like we're getting more for our money too.
The cascoded VAS should reduce the Early effect, and the 2-BJT current sources on it and the LTP couldn't help but improve linearity.
I do wonder what the current mirror in the LTP will do to the sound though. While it will certainly provide better gain, I suspect the old, unbalanced version had a lot to do with the NAP250 sound....
(And if you're going to use a current mirror, why not at least a Wilson?)
Its gain and phase margins are woeful without the Naim preamp's filtering. After that, I suspect Vereker's bizarre "phase compensation" networks on the pre-drivers are tuned by ear for Naim's preamps.
Here is quotation from JV in his last interview, June 2012:
It probably served Naim's purpose to have some mysterious circuitry in the top models, other than just regulated supplies anyway. Otherwise, they had not much more going for them than a bigger power supply and box. Any technician might have exposed the sham that the rest of the design was essentially the same as any other NAP model and that would not be good for business or consumer audio generally.
The resulting component details remained almost unchanged over many years and through many different models, so there is reason to believe the scheme was a clever notion with little practical substance.
It probably served Naim's purpose to have some mysterious circuitry in the top models, other than just regulated supplies anyway. Otherwise, they had not much more going for them than a bigger power supply and box. Any technician might have exposed the sham that the rest of the design was essentially the same as any other NAP model and that would not be good for business or consumer audio generally.
You can try, but i think it would be for the worst....
Set up an experiment.
Take 2x10mF capacitors, make PSU out of them, take a naim NAP 1 unit, and make an amplifier out of it.
1. Try it with 680R/560R and 680R/390 for negative.
whether u like it or not, just let it be for an hour and listen to it with various programs.
2. Try 33/27R for positive and 560R/270R for negative and give it a time to adjust itself and then listen
3. 560R/560R for positive and 1K/1K for negative.
You can repeat it with 2x4700uF caps now and u will get some "idea" what are they doing to acoustics.
unfortunately, the resistor type plays here very important role too...
Simplest answer would be:
Positive shoulder values are like "valve", the higher the values, less dynamic output. (MORE dynamic PSU = higher resistor values) and vise versa.
Negative shoulder is "heavily based" on NPN VAS transistor type and HFE, basically on a "whole" negative network. Its like, you are adjusting the Q factor, less values, the sharper it gets(depends on PSU and everything else, wiring, CAPS, everything
go figure right......
Last edited:
I think you are more likely to perceive real differences with switched resistors such that there is no delay and less opportunity for psychological effects to corrupt the experiment, albeit a sighted comparison by probably only one listener.
It's easy to disappear down a rabbit hole of imaginings in what I call workshop experiments. We need others to corroborate our experiences such that they will prove true for most listeners, most of the time too.
It's easy to disappear down a rabbit hole of imaginings in what I call workshop experiments. We need others to corroborate our experiences such that they will prove true for most listeners, most of the time too.
Naim have always designed the Naits to a budget and to sound worse than their flagships. I can see 6 inferior circuit decisions in this schematic.
In theory, the less powerful the better the sound can be. The former CEO of Roxanne told me back in the day that Naim's first prototype Nait sounded too good. So they had to degrade the design for product/price positioning reasons!
2 cents from me.
PRaT is not a "sound effect" per se. It's what live music has. IMO a useful question is not why Naim adds PRaT but rather why so many other designs are missing it.
The base networks were not chosen by ear. That's really far-fetched. JV's approach was to measure everything.
PRaT is not a "sound effect" per se. It's what live music has. IMO a useful question is not why Naim adds PRaT but rather why so many other designs are missing it.
The base networks were not chosen by ear. That's really far-fetched. JV's approach was to measure everything.
Interesting.....wouldn't we all like to see just what that first Nait prototype design was. It's also been interesting over the years to see the sustained popularity and preference for Nait2, 3 though. These may have been dumbed-down NAP designs but strangely, over many years, I've only seen one rough sketch of the Nait2 amp and preamp here at DIYaudio.
Here is a more recent LT Spice draft of Nait2 by "Nobeone" (website hi-fi AF) who documents his refurbs and upgrades in grand style: https://www.hifiaf.com/wp-content/uploads/2020/06/49952781032_f57ea6336e_o.png
Here is a more recent LT Spice draft of Nait2 by "Nobeone" (website hi-fi AF) who documents his refurbs and upgrades in grand style: https://www.hifiaf.com/wp-content/uploads/2020/06/49952781032_f57ea6336e_o.png
Member
Joined 2009
Paid Member
Brian,
Would you be able to hazard a guess at which design elements that are rarely seen outside of this amp contribute the most to the unique sonic signature ?
For example, would the choice of the high Cob Zetex transistor at the Vas be a particularly key element, which could be introduced into a ‘blameless’ amplifier to give a recognizable hint of the NAP 140 sound ?
Would you be able to hazard a guess at which design elements that are rarely seen outside of this amp contribute the most to the unique sonic signature ?
For example, would the choice of the high Cob Zetex transistor at the Vas be a particularly key element, which could be introduced into a ‘blameless’ amplifier to give a recognizable hint of the NAP 140 sound ?
Grasshopper,
It's not like every transistor is independent of the other parts it is connected to.
If you took part of the suspension out of a Ferrari and found a way to install it in your Ford F-150, would it make the F-150 perform more like a Ferrari?
However, you can study the Ferrari to understand why that suspension was chosen. You can study the many characteristics of a ZTX 753, one of which is Cbc nominal value, and consider how they affect the system.
Blind monk.
(The short answer: no I don't think the ZTX 753 contains a secret sauce that you can transfer to another design)
It's not like every transistor is independent of the other parts it is connected to.
If you took part of the suspension out of a Ferrari and found a way to install it in your Ford F-150, would it make the F-150 perform more like a Ferrari?
However, you can study the Ferrari to understand why that suspension was chosen. You can study the many characteristics of a ZTX 753, one of which is Cbc nominal value, and consider how they affect the system.
Blind monk.
(The short answer: no I don't think the ZTX 753 contains a secret sauce that you can transfer to another design)
Last edited:
High capacitance transistors for a linear AC voltage gain would seem wrong but they can produce some interesting, if not useful sound effects. So the ZTX653/73 complements as VAS transistor and current source or sink should appear to be a bit of a problem here but instead they, and their SMD cousins or equivalents have become essential to mainstream Naim amplifier sound.
A thought experiment: Let's pull the VAS transistors and replace with low capacitance types - maybe the latest Toshiba TTC003/4B pair or video drivers like C3503/A1381. Add a little fixed capacitance to stabilise if you need to and see how harmonic distortion and the sound effects fade away to leave you with bog-standard, near textbook audio.
In the real world, you the customer and the original designer actually want less than ideal linear amplification since the significant dynamic capacitance of the transimpedance amplifier is clearly where much of the PRAT and other controllable sound effects can arise. It's not silly to suggest this when those transistor types are still in use >40 years later, albeit as SMDs. Does no one ask why this is so, when every other transistor type in there can be changed on a whim or price shift?
When I was still doing a little repair work and coaching kids in kit assembly, I noticed that the usual VAS substitute transistors (e.g. 2N5551/5401 pair) were just not doing what the Ferranti/Zetex/Diodes pair did. The sound was flat but according to my 'scope's FFT at least, quite clean - cleaner than the few Naim products that I had checked around that time. So I couldn't help but think as Bigun suggests, that there was some secret sauce here. I was sure, even long ago, that there had to be a cute trick or two inside Naim's little boxes too.
I believe that the very early products used different VAS transistors again, probably with similar properties to the ZTX types so guys, prove me wrong about this by thought or deed. I'd be only too happy to hear if there really was much else apart from classy workmanship, that was significantly special about the original, as in 20th century Naim products. Some nominate the PCB designs, wiring, transformer, capacitors, resistors, cables, even the connectors to make sense of the sound but I would say, that's interpreting an image - a remote visual impression with our imagination working overtime, like when we have no clue
A thought experiment: Let's pull the VAS transistors and replace with low capacitance types - maybe the latest Toshiba TTC003/4B pair or video drivers like C3503/A1381. Add a little fixed capacitance to stabilise if you need to and see how harmonic distortion and the sound effects fade away to leave you with bog-standard, near textbook audio.
In the real world, you the customer and the original designer actually want less than ideal linear amplification since the significant dynamic capacitance of the transimpedance amplifier is clearly where much of the PRAT and other controllable sound effects can arise. It's not silly to suggest this when those transistor types are still in use >40 years later, albeit as SMDs. Does no one ask why this is so, when every other transistor type in there can be changed on a whim or price shift?
When I was still doing a little repair work and coaching kids in kit assembly, I noticed that the usual VAS substitute transistors (e.g. 2N5551/5401 pair) were just not doing what the Ferranti/Zetex/Diodes pair did. The sound was flat but according to my 'scope's FFT at least, quite clean - cleaner than the few Naim products that I had checked around that time. So I couldn't help but think as Bigun suggests, that there was some secret sauce here. I was sure, even long ago, that there had to be a cute trick or two inside Naim's little boxes too.
I believe that the very early products used different VAS transistors again, probably with similar properties to the ZTX types so guys, prove me wrong about this by thought or deed. I'd be only too happy to hear if there really was much else apart from classy workmanship, that was significantly special about the original, as in 20th century Naim products. Some nominate the PCB designs, wiring, transformer, capacitors, resistors, cables, even the connectors to make sense of the sound but I would say, that's interpreting an image - a remote visual impression with our imagination working overtime, like when we have no clue
Ian,
I played alot with these pairs, basically i tried any given transistor on the market.
They all can be configured, just "swapping" does not work. Changing a transistor = changing half of the circuit.
In other words, "BLANK CLEAN PAGE" - as clean sound as possible, with opened up dynamics.
If you get to that point, naim sound is possible.
We need precise layout example for a given transistor models, resistors and capacitors.
I played alot with these pairs, basically i tried any given transistor on the market.
They all can be configured, just "swapping" does not work. Changing a transistor = changing half of the circuit.
In other words, "BLANK CLEAN PAGE" - as clean sound as possible, with opened up dynamics.
If you get to that point, naim sound is possible.
We need precise layout example for a given transistor models, resistors and capacitors.
Hi rensli, I only suggested an experiment to identify the source of the sound effect(s). As Brian reminded us, it's not like any transistor is independent of other parts it's connected to and we need to look at how modifications affect other behaviours like overall linearity, stability and noise. In other words, I think we have to keep within the bounds of the original design and just make and test incremental changes to learn anything useful about it.
Basically, the model amplifiers being compared just need to differ in the Cob or Cbc of their "transimpedance" a.k.a. voltage amplifier. You can't just trim a fixed capacitor and hope that will do anything as this is a dynamic capacitance that tracks the constantly changing signal voltage, as you find in transistor junctions, varactors, tuning diodes and so on. It means a major sound effect is primarily a fast and complex tremolo that more or less tracks the signal voltage. There could be more effects there too.
This would horrify a hifi enthusiast but I can only say that few listeners seem to have any idea how bad their amazing quality hifi system really performs by the time the audio reaches their brain and they perceive music. Don't be surprised by ugly measures that turn into pleasant sound and vice-versa when you play with amplifier voicing and sound effects. I know of only a handful of people who are really successful at it.
Basically, the model amplifiers being compared just need to differ in the Cob or Cbc of their "transimpedance" a.k.a. voltage amplifier. You can't just trim a fixed capacitor and hope that will do anything as this is a dynamic capacitance that tracks the constantly changing signal voltage, as you find in transistor junctions, varactors, tuning diodes and so on. It means a major sound effect is primarily a fast and complex tremolo that more or less tracks the signal voltage. There could be more effects there too.
This would horrify a hifi enthusiast but I can only say that few listeners seem to have any idea how bad their amazing quality hifi system really performs by the time the audio reaches their brain and they perceive music. Don't be surprised by ugly measures that turn into pleasant sound and vice-versa when you play with amplifier voicing and sound effects. I know of only a handful of people who are really successful at it.