Advice on SNAICS please

I think the currents involved are too small to be significant. You could test this by looking for an induced voltage in the 'fifth' wire of a 4 way SNAIC.

Of more interest is the voltage dropped in the signal earth between preamp and PSU as a consequence of it also being the return for the DC.

Paul

Interesting point Paul. I would need to be convinced that the induced currents on the DC power lead were insignificant.I am learning that supposedly small effects can in fact be crucial... I hadn`t thought of what you mentioned, i.e the earth. I guess you have a good point. In addition, the two Din plugs must introduce some (small) impedance into the earth connection....possibilities for RF pickup here?

laurie S

Laurie,

If I can borrow from an earlier reply to this question;

"The DC power supplies passing through the SNAIC interconnect contain virtually no AC components. Since only AC can be coupled to adjacent conductors in the cable the power supplies cannot interfere with the signal. The power supplies also have minimal source impedence (less than 0.1 ohms) and this prevents supply modulation by the very small signal currents."

With regard to DC modulation by the AC signal current, the audio currents passing along the snaic are typically a few microamps. This coupled with the fact that the power supply has a very low impedance, which reduces crosstalk from audio to power conductors in a snaic, means that the effect is negligible.

The currents in the snaic earth do cause a very small dc level shift of 2 to 3 mV between the preamp and power supply. This will have no consequence on signal purity. On the issue of contacts in the earth return, this gives us no problem. We use din connectors because they have very low contact resistance and are self cleaning when inserted and so provide the very best results.

Richard

Jekyll,

I suspect it is having a single earth cable for power, DC return and both signals. It would prevent earth loops, drifts etc.

It has been a long time since my electronics classes so I could be completely wrong.

regards
Gus

Richard...thanks for the reply. You at least confirm that my reasoning is correct...though you are stating that the SIZE of the effect is negligibly small. On this basis I can understand the logic of having signal and power in the interconnect...the convenience gains outweigh the (allegedly negligible) disadvantages. However, within the PSU, preamp (and CD players) surely there are no "inconvenience" penalties for running DC power leads away from signal leads...the boxes have plenty of spare room in them! There may indeed (I speculate)be some worthwhile sonic advantage in doing this. Have Naim checked this one out?

regards

Laurie S

Laurie,

I ran your comments past Roy George and he feels that your assumption that there may be some advantage in separating the signal from the power, for instance in a power supply, is not correct. The signal, power and related earth should run together so that external disturbances, such as transformer fields, affect all equally. This means that the signal, power and earth show no relative differences even in the presence of strong external interference.

Richard

Richard

That’s Rubbish is that.

Its just like saying a cricketer would benefit from being hit simultaneously by a ball in the nose, mouth, throat and bollocks as opposed to just the bollocks.

PS Note my frequent use of the word BOLLOCKS.

Fatcat

Fatcat,

You're wrong!

Whilst Naim systems do not benefit from balanced or differential inputs, and hence there will be unequal pickup in the signal / 0V cables primarily due to differing impedances, the increase in inductance, by increasing loop area worsens interference pickup significantly.

It is standard good design practice to keep loop areas to a minimum - that's why Naim's internal wiring looks like it does, and why every preamp / PSU / amp wiring harness has the same wires in the same positions. It's not arbitrary.

Amplifiers measure an input relative to some point. If both these points have the same interfering signal you have a common-mode signal that results in no change in output at the amplifier, assuming good common-mode rejection (not really the case here, but the star-earthing gives some common-mode immunity).

If just the 0V reference or the signal varies you get an output i.e. an error.

Electronics really doesn't correlate with cricket.

The bit I dispute is that the 2-3mV offset present in the SNAIC 0V really is DC only, and does not affect signal purity.

Laurie,

The other thing you will note is the AC parts of the PSU's are usually well-seperated from the DC / signal parts, or the wiring runs at right angles reducing coupling due to different field orientations.

Andy.

My head says, 'ooh that sounds like the difference between an ANOVA and an ANCOVA', in that it sounds like a way of ensuring any variance within these signal path (if that is the correct phrase) due to error is controlled by ensuring that the same error variance occurs to the reference voltages. Is that the gist of it? Do these sources of variance really effect the signal, earth and power in the same way?

Apologies for any glaring errors, I had alot of other things on my mind when I should have been paying attention at 6th form .

The other bit has to do with the design of the cable itself. I recall reading that the optimum cable would be different for the signal and the power so putting both in the same cable requires a compromise.

I'm raising the issue just so someone more versed than I can state it properly. It may well be that the cable compromise is a better design solution to solutions that separate the two signals.

- GregB

Insert Witty Signature Line Here

Andrew

What you are saying is in the most part true. But your third paragraph fails to take account of the original question being asked. Splitting the scaic into two cables and subjecting one cable to interference will not effect the other. I think what you are trying to say, which I agree with, is if the cable carrying the signal from the pre amp to the power amp is split up and one part is subjected to interference this will have an adverse effect.

Fatcat

quote:

---

The other bit has to do with the design of the cable itself. I recall reading that the optimum cable would be different for the signal and the power so putting both in the same cable requires a compromise.

---

You're right Greg, for power supplies one wants low impedance, i.e. large chunky wires to minimise PSU degradation.

For signals one generally wants low capcitance - capacitance is related to distance between conductors / screens and generally uses thinner conductors (for practicality).

As it's not usually practical to make multi-cored cables with both styles inside, you have to compromise.

I found this with the cable I use to power the CD5 - I made a cable from 15A 3-core mains cable (as there's no signal requirement) and it sounds much better than a black SNAIC, because it has much lower impedance - in effect the CD5 sees a much better PSU as it's not degraded by the cable impedance as much.

Andy.