100 Volt Audio, Band I or Band II?

Hm, this is a slight red herring question Dutch, but I see where you are coming from. The line voltage is 100V RMS at full audio output, but this situation is anything but continuous, about the only thing which might get close is continuous loud organ music. Many systems never reach this voltage, particularly where the content is the odd speech message. The systems usually have transformer coupling to the amplifier and actual loudspeakers, so are more or less completely isolated, a sort of class 2 I suppose. Normally the wiring is done with sheathed cable, largely for tidyness, and I see no particular reason why this should not run on tray along with sheathed power cables. I would not like it in conduit or trunking with power cables, but not particularly because of insulation or safety, but because the PA wiring will not follow the same route very often. Whether it is band 1 or 2 is very debateable because it does not carry continuous power, consider it like an ethernet or TV aerial cable, and in reality it is very safe. Mike says very small cables work, but remember that large systems (such as I used to use at the Longleat rally many years ago) may well have pretty big amplifiers (1 kW or more) and the current will then be 10A on audio peaks along with commiserate voltage drop, and will cause harmonic distortion and signal compression. My favourite was my homebrew amp with 2 off 4CX250s, and very effective it was too but very heavy along with a similarly heavy 2 kV power supply. Transistors have greatly eased by back for a long time now, particularly when the PSU is switched mode too! The provisions of BS7671 for circuit protection are clearly unusable, and the idea of an AFDD or even RCD is certainly peculiar and well outside their specification! I suggest a sheathed cable, round is probably neatest, and tied to whatever support you have or the smallest size basket if necessary. There is no point in distributing a CPC.

David Stone said:

Whether it is band 1 or 2 is very debateable because it does not carry continuous power

I wholly disagree with that statement, and offers an unacceptable level of risk particularly in cases where SELV/PELV voltages are reduced for safety, for example in special locations.

Let me provide an example that more clearly illustrates the read-across to "mains". Consider a variac driving a 1:1 isolating transformer. Just because I can turn the voltage down (and keep it down most of the time), doesn't mean the output of the isolating transformer can be considered safe to run with SELV/PELV, Ethernet or coax ... because I can turn it up.

David Stone said:

consider it like an ethernet or TV aerial cable

No, no, no - these have separate installation requirements and are addressed by other standards entirely. They are definitely in a different category to 100 V line.

• Ethernet indoors is definitely Band I, effectively designed to meet SELV/PELV (although the product standard for some equipment is subject to mis-alignment of IEC 62368 series with IEC 61140 meaning potentially not meeting SELV/PELV requirements in all cases). Ethernet cables should be installed according to BS 6701 and BS EN 50174 series.
• TV aerial system installation should be carried out in according to BS EN 60728-11. Although probably Band I, the major difference here is earthing and, dependent on building in which the system is installed, lightning and surge protection (see BS EN 60728-11 and BS EN 62305 series). Because of this, routing in separate containment to Ethernet is often considered, especially where the infrastructure connected to the Ethernet is expensive (reduce risk of disturbances or worse from lightning and surges).

Hmm not really, the impedance and the duration are both  limited and that  really does matter for safety - the variac analogy is poor, because variacs do not change fast enough. If you must have an analogy, consider another, where there is the world of difference, say an electric fence, delivering a few half cycle of 10kV pk at perhaps 100kHz, and much safer than the same voltage derived from a mains transformer.

The peak energy of an electric fence, a few joules delivered in 10s of microseconds, is perhaps ten or tens, maybe even a hundred  kW, but it is not dangerous, as it is not capable of being dissipated in  the victim for  long enough to do lasting harm.

(similar arguments underpin 30mA RCDs and disconnection times, but those speeds are pedestrian compared to what we are considering here.)

Nor would I worry about 10A transient on a cable of 1 strand of 0.2mm2 or so, coming back to the jumper cable  if again it is only there for a small fraction of the time, the impedances are high - as I said above 1k ohm per 10watts. I think we can lose a few tens of ohms in the wiring, this is not power distribution where losses matter to the percent, it is signal distribution, where a loss of a dB or 2 will only just be noticed by the listener- the problem is not loss of signal but getting it hot and dripping the pvc off the wire if you overdo it. In some situations getting very close to that limit is acceptable, and jumper cable is well cooled dangling in free air. (If electric fences are tame, consider things like the shrike  if you like, they get wired up in jumper and phone cable, and  read down that datasheet link to the end for the voltages and energies, yes it hurts if you touch one, a lot actually, but the thing that makes it lethal is the final application, not the peak voltages or currents. )

As related example we have no twitch about multi kA transients on the wires to surge arrestors, again because they are not there long enough to do damage.

Duration and frequency matter .  And BS7671 is not written with that in mind.

Mike.

Good quality audio cable is recommended. Perhaps even L.S.F.

https://www.canford.co.uk/News/625_100V-vs-low-impedance-loudspeaker-systems-what-you-need-to-know

Z.

Thanks Mike, that is the reason I wrote the above. I have never heard of anyone coming to any harm from 100V speaker systems in 40 years of such things, and that part is very important. This is the problem with feature creep in BS7671 and the other standards, there is not some blanket applicability because it is electricity, the details matter completely. Take telephones, for example, the 80V RMS 16Hz ringing signal is undoubtedly more dangerous, but the source impedance is fairly high and no one cares if the phone has accessible connections. If BS7671 were applied to the existing telephone system, and the LVD and the rest, any maintenance would be impossible because isolation is also impossible. Much the greatest risk from the telephone system is falling from a height, not even working in manholes full of water!

There is also a great deal of "folklore" about audio cables, and various people seem to think that are in some way "magic". I worked for a company (SSL) that had a great deal of interest in such things, particularly whether cables had an audible "footprint". We spent a great deal of effort in measuring and double-blind testing a lot of cables and found no evidence there is at all, at any signal level or current. Some of the HiFi magazines at the time loved solid silver cable, oxygen-free cable, and other strange things but we found no evidence or difference. This did not stop us from using proprietary linear-crystal oxygen-free cables, but it certainly proved that certain articles at the time of cables incorporating "diodes" and other non-linear things were simply nonsense. The insulation is irrelevant to the linearity.

I have to disagree.

The issue isn't really about dry condition, but special locations (and reduced voltages etc. of SELV/PELV).

The point being, that you simply can't carte-blanche run 100 V line cables with all SELV/PELV and assume it will be OK.

David Stone said:

This is the problem with feature creep in BS7671 and the other standards

Works both ways. Been involved in looking at power over Ethernet in recent years ... and the issue there is the standards don't consider special locations, yet the use-cases people want for the technology definitely do feature special locations.

Hmmmmm ....

gkenyon said:

you simply can't carte-blanche run 100 V line cables with all SELV/PELV and assume it will be OK.

That I agree with.

But the guidance in BS7671 is not really appropriate - it is written by power people who think milliseconds are fast and microseconds are non-existent. high speed data, audio and pulsed power are very different disciplines, yes there is some overlap, but not a blanket cover-all.

Would you put an HRC fuse or an RCD on a 100V audio line - no, would you treat a 100V power cable to bare terminals and stripping and installing while alive, I hope not (some Americans might).

Is 100um of insulation enough to hold off occasional flights to 100V when there is no sustained arc hazard, yes,... Safe on a mains cable, not really.

Mike.

If you think about my original reply, I did say electrical separation and Class II are most appropriate.

If you think about my original reply, I did say electrical separation and Class II are most appropriate.