Thanks for such a detailed response, and yes, I do get the general idea (even if not all the technical details). I had indeed noticed your “mid-air” cabling in previous posts.
1 Like
The only flaw with this idea is that this from a signal perspective would only apply if the return from the loudspeaker was to Earth potential as opposed back to the amplifier. Otherwise the stray capacitance to ground would cancel itself out on the return. Indeed I have undertaken measurements elsewhere for other use cases using parallel conductors for the out and return… the signal integrity only breaks down noticeably when you vary the interaction of one of the cable directions compared to the other with respect to the impedance as seen by the amplifier of the load.
So this idea of stray capacitance to earth affecting the speaker cable impedance and therefore signal is very much loudspeaker cable geometry dependent. A parallel cable like NACA5 any stray reactance would cancel, some sort of coaxial speaker cable could be vulnerable.
I suspect another possibility is the loudspeaker cable may capacitively load the amplifier outputs to ground, the coupling is out of band of the signal, but this may affect the amplifier performance some way… but I feel the impact of this would very much be amplifier design dependent however.. and good designs should be immune to this in my opinion.
Hi @Innocent_Bystander ,
Very many thanks for your thoughtful assessment of the DIY Copper Pipe loudspeaker connections. I had tried to think about most aspects that you have raised - with the exception of insulation. I decided to make the lack of insulation a USP (‘unique selling point’). Of course not in any way practical or safe!
But then there are the aftermarket upselling possibilities to think about. I can envisage the adverts now…“Any RAL colour (of insulation) you desire!”
Plus, just think of all the endless discussion on HiFi forums in general about which insulation colour ‘sounds the best’.
On more serious matters and your other technical observations:- I did try to form a helical spiral for the copper pipe pair, it is just not quite so obvious from the photo. I got at least 180 degree rotation in the forming - not as tightly spiralled as Chord Sarum T - but nonetheless a worthwhile twist.
The Golumb vs Fibonacci choice came from my desire to differentiate my DIY Copper Pipe from other offerings in the marketplace. For customers ordering the longer lengths I would probably do the spacing starting from each end and work towards the middle, so it would be two ‘Golomb’ sequences meeting at the point of maximum spacing in the middle of each conductor length. This ‘dual’ spacing sequence might be marketed as an upgrade variant (from the entry level)!
Thanks for your thoughtful assessment.
However - now I have posted the impedance data - do you have any additional thoughts regarding the DIY CuP connections in comparison to standard Naim NAC A5 loudspeaker cable?
ATB
E of E
I see your chosen spacing has caused the CuP to exactly match NACA5’s capacitance, but is interesting to see NACA5 has greater variability in inductance with frequency (assuming all measurements were repeatable) but the reason for that is not immediately obvious. However, and prefacing this by saying I have no experience of comparisons of speaker cables with such similar characteristics, I doubt the magnitude of inductance variability are of sufficient magnitude to cause any significant audible difference between these two cables.
1 Like
Hi I_B
Thanks for your observations and comments. This project is a series of discoveries and learning opportunities for me. I became aware of the likely similarities of the impedance measurements of a Copper Pipe pair as a loudspeaker cable to NAIM NAC A5 when testing the LCR values (posted on my ‘Audio Grade Solder’ thread approximately at post 100).
You have been very helpful in all of those discussions - which I greatly appreciate.
Link to the conclusion of those discussions is available here: Audio Grade Solder.
One of the reasons for the variability in Inductance (L) may be that the consistency of where the conducting groups of copper strands sit within the insulation (of NAC A5) is quite poor. For the unterminated (white colour insulation) NAC A5 cables that my HiFi retailer provided to me - in some places the copper wires were practically breaking through the insulation!
I am not sure the above would fully account for the degree of variability vs frequency in the L measurements? Perhaps it is also likely to be differences in the consistency of the very short loopback connection I used for the CuP and the NAC A5 tests?
I think I should ‘own up’ and say that last year I thought I could get ALL the data I needed using one instrument - the LCR meter. I was mistaken. To correctly measure the Shunt Conductance (G) of a cable I need a different type of meter. But I feel the missing information of Shunt Conductance (G) probably may show up the reasons why some commercial cables seem exorbitantly more expensive than others - for providing the same nominal characteristic impedances…
Whether a higher value of Shunt Conductance (G) would make any difference to how a loudspeaker cable might sound in a given system (if all other the comparable loudspeaker cable data were sufficiently identical) - who knows?
Nevertheless, I now believe such differences could be reliably simulated and visualised (using transient analysis) of circuit simulations of combined amplifier-cable-loudspeaker models in LTSpice.
ATB
E of E
3 Likes
Fascinating stuff, as always EofE! This and the audio grade solder threads hopefully will provide a useful resource in the future (though maybe there will be sufficient for a scientific paper that could be published for wider reading and reference in due course…).
BTW, I don’t need thanking every time for my input - I am very happy to contribute what little I can. You are the one deserving of thanks!
1 Like
The idea of a matched impedance loudspeaker cable - which I think you first suggested on my ‘Audio Grade Solder’ thread here - was and is (to my thinking) very sound.
I recall that there have been some discussions/disagreements on the forum in the past between technically competent persons regarding such suggestions. IIRC mainly due to the fact that the transmission line length of typical loudspeaker or other audio cables is ‘insignificantly’ short relative to the likely wavelengths of the electrical signal contained within the music waveform.
IMO, transmission line theory in the case of loudspeaker cables cannot always be ignored as the signal reflections (at the mismatched impedance interfaces) still occur. It is whether those reflections are significant in some way to the reproduction of sound from the loudspeaker and subsequently experienced at the listening position. It is my contention that they (the energy reflections at such mismatched interfaces) are significant and need to be thought about.
I plan to provide more information and opinions on this topic over the coming weeks.
In the meantime, here is a bit more detail on the components I used for DIY CuP loudspeaker connections (should one wish to construct one’s own):-
Here are the clamps that I used: Parker ROPD108X (double-tube hydraulic pipe clamp).
Here are the other bits, the spade terminals the PEEK M6 bolts and nylon nuts.
Here is the finished article.
PS: The next iteration of the Copper Pipe loudspeaker cable would have an air gap slot in it (along its length) to mitigate residual skin effect losses.
Those spades look familiar!
I await your muses on cable reflections with interest… And wish you luck/fun cutting a longitudinal slot in the tubes!
Meanwhile an idle thought: from the pic it looks as if the red terminal of your speaker is on the right with respect to the photo, and that of speaker to the left. If the spacing of the two sets of terminals the same or very similar, have you considered turning the amp around, raising it up, and connecting direct to the speaker with rods or tubes that fit the internal banana diameter, to all intents and purposes resulting in no cable (other than the connectors themselves and internal wiring)?
Hi Simon,
I have taken a bit more time to provide a reply to your technically astute and considered input. There are a few places on the forum where your have provided insight on these topics in the past. One particularly insightful example (IMO) is from a 2021 thread discussion on NACA5.
(Specific link to the full text here)
I agree with your observations in your post above that some amplifier designs (those including some global feedback from output) and certain amplifier output circuit architectures may be more susceptible to excessive load capacitance between the ‘out and back’ conductors.
Included in your response, the additional work to which you refer…
I assume is in your ‘radio ham’ hobby area and hence on RF carrier (plus signal modulation) to antenna feed?
I personally do think that the issue of capacitance to ground is relevant for audio even if the capacitance to ground can be ‘balanced’ to be identical for out and return paths - I bet it is not.
Furthermore, in the case of NAIM NAC A5 having inconsistent conductor centring within the insulation, I would almost guarantee that the capacitance to ground experienced by each conducting element is not (balanced). Whether such variations or inconsistencies introduce sufficiently audible distortions at the listening position - only the listeners ‘ears’ and ‘audio perception’ can be the judge.
FYI I am in the process of constructing LTSpice models to represent the amplifier-cable-loudspeaker subsystem of my HiFi to sufficient accuracy to model effects such as the various cable characteristics (including different capacitance to ground for each conductor), variations (of in loss) in dielectric and conductors and the observed currents and voltages for transient response at both the amplifier and loudspeaker terminals in addition to the usual AC analysis.
I would appreciate further comment from yourself once I post on these things if you have the time and inclination.
Thanks in advance,
E of E
I was have the inclination to engage with yourself. 
Yes the measurements were in the area of RF engineering which can sometimes be a more exaggerated version of audio engineering.
Regarding inconsistencies in the NACA5 construction affecting audibility in terms of unbalanced leakage capacitance to ground… I guess it’s theoretically feasible, but I would have thought in practice such aspects would be very much masked out by more significant perturbations such as humidity, temperature, air pressure, mains voltage and even background RF from denergising ionospheric layers.. in other words background noise.