As Simon say you need little inductive load towards the high frequrncy. That could be build in the amp, old series does not have it, and thus use more inductive cables. However it also depends on your speakers, more precicely your tweeter network. If you measure your speakers impedance along the frequency responce and you find out they are slightly inductive you do not need additional inductance from the speaker cable and can go for zero inductance cables or short cables. I use dyton audio impedance measuring device, comes pretty usefull for other things too, like measuring your cables, drive units, etc.
My system at present 552 dr 300 dr 333 streamer xps dr BW 804d4 was thinking about chord signature xl or Sarum also will probably go for 350’s
[quote=“tgw56, post:9, topic:34111”]
hi TGW56, looking at your speakers impedance , more specifically the phase shift in high frequency range in your case is bellow zero and that indicates you have capacitive load and with NAP300 you need inductive cables, how long I cannot say without measurements of cables and speakers together, but would say don’t go below 5 meters. these speakers are quite difficult to drive at 70 Hz, the phase is -70degrees. Disclosure: I am not an audio engineer (i have MSC Nuclear Power Engineering) and audio is a hobby for me, so would be good if someone else could confirm that!!! The graph is taken from Sterephile, using same Dayton Audio’s DATS v2 that I use home.
Contrast that with Naim’s own NBL, yet they still sound better with 5m of A5 than 3.5m.
3 Likes
I loud speaker system is typically a complex load in terms of impedance so will contain capacitive and inductive elements at different points in the frequency band pass.
My observation of that chart, is that looks quite a challenging loudspeaker to drive for an amplifier. But regular largish gauge speaker cable like NACA5 would be fine, though I would avoid very long lengths, or extremely short lengths for most amps. For older Naim amp designs, 4 to 5 m is often a sweet spot.
Do remember with loudspeakers if a critically measuring the impedance, you need to it with a signal generator when it’s generating audio in the room, as the room coupling will affect the reactive elements of the impedance. Looking at a speaker impedance graph passively decoupled from the room it is in could be a little inaccurate. Not sure Stereophile magazine did this ! Possibly accounts for it sounds better than it measures statements you see from time to time, as well as some speakers needing a bit of loudness to ‘wake’ them up.
What a superb speaker the NBL was! Those bounds of being well within ±22.5 degrees is remarkably narrow - in fact I don’t think I’ve ever seen it before. (I can do 37.5 deg all day with a closed box but this is significantly narrower and shows they didn’t skimp on box size.) It’s a shame so few people “got” them.
The cable won’t really affect what’s going on down at 70Hz and certainly not in terms of inductance. I’d be a little more concerned at the top end where it’s about to go through 0 degrees just above 20kHz. It might be nice to have this already inductive but so much of this is a matter of chance, not least because you can get ringing on the output that has nothing to do with the amplifier becoming unstable (and this is something you can unpick and show that the frequency of ringing is exactly what you’d expect from the values of the inductance and capacitance). I’d like to see a more comprehensive set of measurements from JA (I assume this is JA) before thinking to act on them, perhaps including square wave performance. I think I’d just trust Naim when they say their amplifiers are stable. We haven’t seen a Naim amp blow up of its own accord for at least 4 decades - certainly I haven’t - and back then people could have been using almost anything to connect the speakers. And NACA5 is such an extraordinarily good cable that I would want to use anyway. (I have been told on good authority that Chord can do a tiny bit better but with quite a serious price hike.) I don’t think it’s a huge leap of faith to just go with the Naim recommendations and, when it comes to choosing length, I would bear in mind the lengths of cable they themselves would be using when listening/auditioning their own products. I’m going to guess that starts at 5m, but it might be 8m in a large listening room. I think I’m right in saying that the general feeling was no one could tell the difference between lengths above 5m, though I’ve never owned NACA5 longer than 5m so can’t confirm this personally. But it makes sense as the cable has a very low resistance of about 0.02 ohms round-trip, IIRC, so you’re just adding a fraction of that as you go up in length.
1 Like
“The cable won’t really affect what’s going on down at 70Hz and certainly not in terms of inductance.” Indeed, but what makes you think I have said that? Can you quote me?
Please excuse my lack of knowledge of these curves. Could someone please tell me which curve is which. I assume one is for impedance and another for phase.
Warm regards,
Mitch in Oz.
Solid line is impedance, dotted line is phase.
2 Likes
The DCR of 5m of this cable is about 0R02. The voice coil will change resistance, probably within a cycle, by that amount at some not unreasonable listening level. Besides, the variance of the voice coils will probably be 10x that amount.
Dear @Christian_Thomas , I feel you may not have read the full post from start to end. Power amps need little inductive load the high frequency to avoid oscillations. Of course drive units have enough and if you run active you are OK, but if you have passive XO there are usually capacitors in the network and the overall high frequency load may become capacitive instead of inductive, like in the particular case we are talking in this post. To avoid having discussions like this most audio system manufactures put little coil at the output stage. However the inductive element deteriorates the overall sound quality and on top of that most speaker cables have inductance anyway, so thus Mr. Julian Vereker has decided not to use inductive load at the output stage, but use the inductance of the cable (part of PRaT philosophy) and this Naim speaker cables have one of the biggest inductance per meter if you make the effort to compare some.
1 Like
In front of “these” there is a full stop. It separates two sentences.
A capital letter would help if you consider it that important. I’m not sure whether you wish to lecture me on my grammar, amplifier design or what Julian Vereker did (or in this part of the amplifier, the 0R22, what John Linsley Hood did first).
I don’t know what your 2nd sentence is trying to say but the story is this:- The capacitance of the dominant pole shows up as an inductive output impedance and, as such, is at risk of resonating with capacitances showing up in the load. The risk is highest in the stopband region where the capacitances of the devices dominate the transfer function and especially where they can change value substantially depending on conditions. The usual inductor is there to keep the load looking inductive and therefore circumvent this issue. One of the alternatives (and they are not always required) is to introduce a zero in the form of a series resistor to extend the “normal” behaviour of the amplifier to a higher frequency and have the available loop gain fall below unity before the problem conditions take hold and to meet stability criteria. Either will work, though I’m dying to hear why “the inductive element deteriorates the overall sound quality” if it doesn’t kick in until perhaps 70 or 100kHz.
Yes, I know what the inductance of Naim cable is.
But the absence of a Zobel network hasn’t been the case since the chrome bumper days. You can use any cable with DR amps and New Classic and nearly any with Olive and pre DR Classic.
Don’t take my word for it. Take the words of Steve Sells at Naim. There’s more from Naim on this topic in that thread but at 6500 posts long, this it what I could find easily.
@Richard.Dane I was sure there was mention by you of reviewing the FAQ to make the topic on cables less absolutist and more in line with current information from Naim.
1 Like
FZ, it’s my understanding that there are Zobel networks but it’s the added inductance that is left out with Naim power amps.
As for the latest amps being able to use any speaker cable, while it may be true that they are even more robust these days and won’t blow up with any speaker cable, it does not mean that they will perform optimally and sound at their best with any cable; so Naim’s recommendations still stand.
Yes, updating the FAQ is on the to do list, but I wanted some more feedback from Naim first.
2 Likes
When naim has open up on cables from the NC, and with full balanced to 350s, I have use All nordost cables from power cords all the way to speaker cables , it’s just sound so much openness , depth , huge soundstage, clarity , micro details, punch , synergy , decays are beautiful , precise placing of all instruments, rich music output ,vocal are very very bodied and layered , and many more added values that makes the whole presentation of music to another level that naim has wanted the NC to be .
3 Likes
This is what I did with the cable works to my 332 and 350s and rest of my components . Fully utilised all balanced connections that naim has brought in to the NC
2 Likes
Haha! You may be using the term Zobel Network correctly. Too many people refer to the Boucherot cell as the “zobel”, which has made the description very imprecise.
Yes, I think the whole question is completely outdated and actually, even pre 1980, I doubt Naim fared any worse than any other amp manufacturer. Certainly I’ve seen more blown up Musical Fidelity amps than I have Naims; by some margin, too.
A while back I did some modelling of the core NAP250 circuit (ie. without the response shaping filters around it) and was surprised by quite how nicely behaved it was in that rolloff at the top end. There are a LOT of limitations to this sort of modelling, and it’s much more sensitive to the choice of components IRL than it is on the screen, (and I don’t have a model for their output transistors) but the first impression was quite how textbook it looked. Nice smooth rolloff with no discontinuities all the way down. This region can be something of a mountain range in some amps and you generally don’t get a nice looking result at the first try! 
That’s with the circuit that does the rounds on the internet, which is as old as the hills. I don’t believe that much has changed since but the point is that it never was a dog.