Interconnect

The (slightly lilac-shaded) grey cable is the correct one for this application.

Chord offer a cable using the same conductors, but with a slightly different (or an additional?) outer sheath.

cheers, Martin

quote:

---

Originally posted by MartinC:

Please guys, if we are going to get technical let's get the information correct.

---

OK, then:-

quote:

---

Date: 26-Jun-99 11:15
Author: julian vereker
Subject: No
Snaics carry power as well as signal, this is not required for the connection between the CDSll and the pre-amp.

There is something of a compromise involved here, the Snaic needs to have low DCR because it is carrying the current to power the device, this implies a large conductor, and at the same time low capacitance is desirable for the sound quality, and this (sort of) implies small conductors.

There is no such compromise in a signal only interconnect, hence we use a different cable to make the audio CDSll to pre-amp interconnect.

julian

---

cheers, Martin

I can't resist taking this opportunity to suggest that the right cable should be used for the right job. If JV is lamenting the SNAIC for being non-optimum because it must serve the dual purpose of carrying power as well as signals (unlike the CDP to pre cable) then why the blazes didn't he use separate cables and connectors for the two? And I still haven't heard any objective reasons why the signals have to go via the psu in the first place. Why not a nice thick, low DCR cable from pre to psu and a nice thin, low C cable from pre to amp?

Sorry to go off-topic a little.
BAM

Firstly, the power and signal are run through the same interconnect to maintain a single common ground for both 24V DC and signal. The benefits of common grounding supercede the advantage of smaller conductors for the signal, in the SNAIC's case.

Why is this common ground thing such a big deal? Because the ground (0V) level is a system reference. If the reference varies - even by a little bit - by careless grounding practice, a certain amount of signal information becomes obscured i.e. permanently lost.

The power amp input cable does not need to be screened as it is handling higher level signals than those from source components. Therefore it is not necessary to use the screened cable (where the small signal conductor size is advantageous for low capacitance) which is used for the CD player to pre connection for example.

Chris

[This message was edited by Chris West NANA on FRIDAY 22 February 2002 at 10:10.]

"The benefits of common grounding supercede the advantage of smaller conductors for the signal"

Why?
Aren't shared grounds usually a bad thing? Electric current flows in a loop. The reason star grounding is advocated by Naim and others is that it avoids the problem of two such loops sharing a common conductor which, if the conductor has any resistance, will cause the two currents to create an additive voltage drop along the conductor. Mixing the two signals is usually very undesirable. With the Snaic the psu current loop and the audio signal current loop both share the screen. How can this be beneficial over the two loops being separated?

"If the reference varies - even by a little bit - by careless grounding practice, a certain amount of signal information becomes obscured i.e. permanently lost."

Why?
This isn't a mechanical system. The music signal is a RELATIVE voltage BETWEEN two points. Calling some arbitrary point in the system a "reference" is irrelevant because electrons only know about relative volatages.

The feedback system in the pre works to make the voltage between the pre-amp out and the star ground in the pre-amp as accurate as possible. That's all it knows about. The purest, least adulterated signal is between these two points inside the pre-amp and this is regardless of anything happeing in the external psu. So why isn't the most accurate way to get the music signal from pre to amp to connect the amp signal/gnd directly to the pre signal/gnd points and avoid the shared ground line in the Snaic and all that extra cable that isn't even the right sort of cable according to JV?

"The power amp input cable does not need to be screened as it is handling higher level signals than those from source components. Therefore it is not necessary to use the screened cable (where the small signal conductor size is advantageous for low capacitance) which is used for the CD player to pre connection for example."

Did you mean this? The voltage signal out of the CDP is some 2V peak (I don't recall exactly). The amp has a gain of some 27x. For normal listening levels the voltage signal from pre to amp will be considerably smaller than 2V peak.

quote:

---

Mixing the two signals is usually very undesirable

---

Concerning the Snaic, we're talking about DC voltage - a steady and constant voltage - which does not interfere with the AC music signals in the same cable. Using the same ground conductor in this application is beneficial. It is not unusual to combine DC power and signal in the same lead, e.g. "phantom" powered mics use a DC 48V power supply sent from the mic preamp via the signal lead....

quote:

---

The music signal is a RELATIVE voltage BETWEEN two points

---

Yes, and having a consistent OV ground as one "point" means you can have a music signal voltage relative to that reference, that isn't compromised as it passes through the amplification chain by non-signal related variances in the ground potential (a.k.a. noise).

quote:

---

The feedback system in the pre works to make the voltage between the pre-amp out and the star ground in the pre-amp as accurate as possible

---

A practical hi-Fi system does not revolve around one feedback controlled gain block. The signal has to get through the entire system without getting screwed up. With the system architecture we've had to date, when the preamp supply is separated from the preamp, it is best placed between the pre and power amps such that both amplifiers refer to the same system ground. The bottom line is that a Naim system sounds best in this configuration (if you want it to sound like a Naim system).

quote:

---

The voltage signal out of the CDP is some 2V peak (I don't recall exactly). The amp has a gain of some 27x. For normal listening levels the voltage signal from pre to amp will be considerably smaller than 2V peak.

---

Peak output for a Naim preamp is about 7.5V but typically far less.
CD players have a peak of 2V (higher than most source components) but again typically far less. The argument for screening a CD interconnect is perhaps less strong but probably still wise in practice. The reasons why Naim use an unscreened input lead for 250's and above are not solely related to signal level - and I'm not opening that can - but as I said, it is not necessary to have screened cable for that interface.

Regards

Chris

quote:

---

Originally posted by Chris West NANA:
With the system architecture we've had to date, when the preamp supply is separated from the preamp, it is best placed between the pre and power amps such that both amplifiers refer to the same system ground.

---

Chris,

hints of changes to this with the new generation preamp and split-rail power supplies?

cheers, Martin

No hints intended....I'm just keeping an open mind regarding the future.

Regards

Chris

quote:

---

hints of changes to this with the new generation preamp and split-rail power supplies?

---

IIRC the 552 was wired to the 500 via the power supply.

Paul

Later

Chris

having you around to answer all those questions re: cabling/earthing etc.. is a real breath of fresh air in this forum.

many thanks -- and i find your answers very useful.

enjoy

ken

quote:

---

having you around to answer all those questions re: cabling/earthing etc.. is a real breath of fresh air in this forum.

many thanks -- and i find your answers very useful.

---

I have to applaud you for taking a stab Chris, but I don't believe you've answered any of the questions in the manner that is needed - not criticism, just observation.

To digress slightly, the whole idea of using unscreened cables anywhere in the low level signal paths of a system is a total anathema to precision measuring equipment (which is what a HiFi is).

Is it any wonder that Naim systems can suffer RF?

I should also add that the cable screens should be just that - electrostatic screens and should never, in any properly designed amplification system, be allowed to form part of the signal path. This is basic good engineering practice that you will find in any and every reference text book you care to mention. Unless Naim have some new take on the laws of physics there are better ways to do the system screening, with, I suspect, consequently higher RF immunity and ultimately better performance. Unfortunately it's not something that can be practically done with the systems in their current form - a ground up re-configuration is needed, there's too much to change and too many legacies to support.

The Prefix is a classic example - we have a low noise amplifier with lots and lots of gain, bandwidth well outside of the audio band and unscreened connections at its input!.

If you wanted to design a radio receiver front end it's a pretty good place to start - the antenna is built in

Now there are probably good reasons for doing this, of which cable capacitance, and the practicalities of finding suitably flexible cable for these connections are just two, but with the huge number of people who find themselves unable to use this product because of RF issues I'm suprised a revisit hasn't been performed. With a limited market to start with it makes little sense to shrink it further, and the knock-on effect of customers finding out after they've purchased it that it doesn't work makes the customer service work harder!

Martin,

quote:

---

he mentions using smallish conductors as being better for signal transfer, then goes on to make the connection leads to Naim power amps, which only carry the signal, from mains cable!

---

Without a screen the mains cable will have low capacitance.

quote:

---

Concerning the Snaic, we're talking about DC voltage - a steady and constant voltage - which does not interfere with the AC music signals in the same cable. Using the same ground conductor in this application is beneficial. It is not unusual to combine DC power and signal in the same lead, e.g. "phantom" powered mics use a DC 48V power supply sent from the mic preamp via the signal lead....

---

That's not quite correct Chris - the SNAIC +24V rails and 0V conductor (shield) carry an AC signal since the preamps do not have constant current drain. The +24V rail modulation is coupled by cable capacitance and is probably inconsequential - the current in the shield is modulated by the pre-amp current drain which varies with the signal passing through the preamp.

Combining such currents with the signal return from the preamp is a big no-no ni any precision measuring system.

Naim seem to get away with it as the DCR of the screen is low and the current modulations are fairly small.

quote:

---

Yes, and having a consistent OV ground as one "point" means you can have a music signal voltage relative to that reference, that isn't compromised as it passes through the amplification chain by non-signal related variances in the ground potential (a.k.a. noise).

---

Chris, here's your starter for one: -

Where is the single 0V connection or reference in an active system - which PSU do you fancy?

The preamp is connected to it's PSU, the crossover to another (joined with a SNAIC). The amps are connected from here. Where is the common 0V reference point that they share?

The 0V reference in a PSU is 'generated' by the regulators, but only in as much as the regulator maintains its o/p relative to its ground point. This ground point is floating, being connected to the reservoir capacitor -ve's and the transformer centre tap.

At some distant point (along a SNAIC, and an interconnect with relatively high DCR and variable impedance at RF) it is tied to absolute (mains) earth via the CD player or turntable.

There seems to be benefits to doing this, but I suggest they come primarily from improvements in RF screening rather than the fact the system now 'knows' it's connected to earth.

Andy.

P.S. I bet that the 'Burndy' interconnect from NAC552 to its PS has more than one 0V return in an attempt to address exactly these shared current path issues.

[This message was edited by Andrew L. Weekes on SUNDAY 24 February 2002 at 09:48.]

[This message was edited by Andrew L. Weekes on SUNDAY 24 February 2002 at 09:50.]

[This message was edited by Andrew L. Weekes on SUNDAY 24 February 2002 at 09:51.]

[This message was edited by Andrew L. Weekes on SUNDAY 24 February 2002 at 09:52.]

[This message was edited by Andrew L. Weekes on SUNDAY 24 February 2002 at 09:55.]

[This message was edited by Andrew L. Weekes on SUNDAY 24 February 2002 at 09:57.]

hmmm... dont understand.

i am finding the information that Chris is providing very useful as he is providing FACTS -- i.e. how cabling is arranged in a typical naim system. i hope he will continue (i.e. providing factual info on what IS)

you and others are challenging the validity of naim's designs in this regard. also very valid and interesting.

i would find it hard to believe that naim havent thought of all the issues you raise-- could just boild down to a differences of approach. what do other hifi manf do (e.eg linn??) and how do their equipment sound?? fwiw, my naim system sounds great

This is basic good engineering practice that you will find in any and every reference text book you care to mention

not a criticism, but i well remember julian making the point that he found very little useful for good hifi design in text books that he had read. now, i don't have any hifi design experience to challenge that. but i respect his viewpoint.

all good stuff this...

enjoy

ken

In a passive setup the preamp and power amp are referenced to the PSU.

But strangely in an active system, supposedly a more revealing and critical setup, the power amps are referenced to the active crossover (a pre-amp, in effect with different filtering!).

Make sense of that if you can.

Andy.

quote:

---

i am finding the information that Chris is providing very useful as he is providing FACTS -- i.e. how cabling is arranged in a typical naim system. i hope he will continue (i.e. providing factual info on what IS)

---

It's the why is we have a problem with

quote:

---

not a criticism, but i well remember julian making the point that he found very little useful for good hifi design in text books that he had read. now, i don't have any hifi design experience to challenge that. but i respect his viewpoint.

---

I'm not going to argue in detail, since he cannot defend himself, and I hold Julian's contributions to this forum in highest esteem, even if only for the fact that he did take the time and effort to contribute technically.

I think there's plenty of good starting points in text books, but little detail or correlation between theory and application.

Andy.

definitely agree...

enjoy

ken

quote:

---

you and others are challenging the validity of naim's designs in this regard. also very valid and interesting

---

AFAIK most Naim pre/psu/power assemblies can be connected two ways, either via the psu or from the 4 pin output on the pre direct to the power.

And AFAIK most people who've tried this prefer the Naim way, so the validity of this form of connection isn't being challenged.

The question is 'why' does it work?

One can see the historical logic of it, the SNAPS has two independent power supplies, the signal comes out of the pre into the SNAPS out of the SNAPS and into the NAXO, the SNAPS providing power to both. (I've assumed some internal bridging wiring in the SNAPS, which I think was recommended by NAIM, otherwise you run a 4 pin-4 pin from SNAPS out to SNAPS in...)

Paul

The question is 'why' does it work?

you've hit the nail on the head paul. i guess i have been misreading those posts -- it sounded to me like the validity of the connection strategy is being challenged. absolutely nothing wrong with that by the way.

as i said earlier, at my low level of understanding, the most useful output from these posts for me is the description of WHAT IS, rather than WHY IS (i'm saving that for later...). there is a lot coming out that i simply did not know...

hence my applauding chris west's posts...

many thanks paul...

enjoy

ken

Onwards! Mr Tibbs, whilst they do have some things in common, I would say analog preamps are very much a different species from CD players. Goose and Gander they are not. For one thing, in a CD player you have to make digital and analog circuits work together in harmony. In Naim preamps the digital bits (control circuits etc) can be effectively isolated by putting circuits on standby whilst the music is playing. In a CD player you have to make everthing work concurrently if you want music to come out. AFAIK this entails a different approach to the grounding and power supply arrangements.

Andrew, I wouldn't draw too many comparisons with precision measuring equipment. Designers of measurement amps have been known to stumble badly when applying the same principles of design to Hi-FI systems. Where Hi-Fi systems are concerned, experiment and trial and error (The "Black Art" factor if you like) along with experience and the exact sciences all come into play. Look at cable directionality, we could attempt to explain that one away 'til the cows come home.

OK, active systems by necessity have two psu's linked together by one Snaic for signal and ground. You can still consider them (in combination) as central to the front and back ends of the system. The psu for the SNAXO is not hanging off to one side.

quote:

---

But strangely in an active system, supposedly a more revealing and critical setup, the power amps are referenced to the active crossover.

---

I don't think it would be a wise move to run six amp signals back through the SNAXO psu. In any of these systems its a question of whats best for the performance of the system as a whole, all things considered.

Regards to all

Chris

quote:

---

IMO they don't actually work in harmony and can be considered as completely separate processes.

---

I hear you Mr. Tibbs, but what I mean by "in harmony" is that within the CD player, both analog and digital circuitry have been designed to work optimally whilst inhabiting one box, and for the most part one circuit board. I agree that the digital and analog processes are separate and only meet electrically at the DAC - at which point both
"sides" of the DAC are sharing the same ground.

IMO, I suspect that if the analog output signals were run through the Burndy to the XPS, for the sake of using its analog psu ground reference, that there would be more to lose overall by mingling audio signals with digital supply lines in the Burndy (through crosstalk). I know for a fact that the exact ways in which wires are routed inside the CDS2 are critical to keeping noise and jitter to a minimum, and performance to the max.

regards,

Chris

[This message was edited by Chris West NANA on MONDAY 25 February 2002 at 03:19.]

Perhaps another case for simpler solutions.

Dave Dever

quote:

---

...or you'd need to use two separate leads, one for analog signals, one for digital signals--and the analog signal would have to pass through the XPS near to the digital regs.

---

There has to be another reason though - the CD3.5 / CD5 also have the PSU 'hanging' on them, analogue o/p being taken from the CD5, yet it would be so easy, in this case, to take the signal via the PSU.

I think it has a lot more to do with connection simplicity, the Prefix o/p passes throught the PSU, for example, because it's a logical, neat connection scheme (that may sound better as a result?).

The same interconnect could have been used with a CD5, maybe cost is the factor - a new interconnect would have to be supplied.

I just cannot see the logic for different connection schemes for different sources otherwise.

Andy.