Cabling and Grounding for Audio, Video, Speakers and Power. Part 1
My point of view based on 40 years of professional experience in both the Audio and Video industries.
Before I launch into what must be one of the most hotly debated topics in audio (and video) I would like to put the following ‘on the table’.
I have worked in the professional recording, broadcast and systems integration industries most of my life and spent almost a decade as a professor in one of the most respected Electrical and Electronic Engineering Universities in Europe, teaching electronics, audio and recording technology to both engineering and music students.
During this period of some forty plus years I have researched and tested the performance of cables, demonstrated their effects to students and been responsible for the design and building of some of Europe’s and America’s most prestigious and largest recording studios and TV, and radio broadcast facilities. All the way from the early days of analog SD TV (NTSC & PAL) to some of the most advanced 4K all IP based facilities of today.
NEVER have I seen any special or significant attention paid to the use of esoteric audio, video or power cords. Nor have I EVER heard a client (or their CTO), who make their living by providing the best and most cutting edge technical facilities, or my CTO, ever say, ‘oh yes! and by the way, please use vendor XX or YY cables as they sound so much better’!!!
Good engineering practice, is just that, and good engineering practice when executed correctly in the building of a facility, negates any needs for such extreme measures (and ridiculous costs) in the deployment of esoteric cabling. My own company uses literally hundreds of thousands of standard 18, 16 & 14 SWG IEC cables for equipment power connections, being appropriately sized to the current draw of the connected device. As for exotic cabling, Klotz or Mogami is as far as we go. Belden providing most of all our other cabling needs for A/V, radio and TV broadcast and all production facilities.
(This ignores the entire move of the industry to an all digital infrastructure using all optical and CAT6 infrastructures for device interconnects and facility distribution and monitoring. Baseband analog I/O is on its way out, big time. I can’t wait to see what these Esoteric cable manufactures come up with to make these digital connections sound or look better!)
A Quick Discussion on the Topic
With that out of the way, onto the topic in hand that often creates a line in the sand between the ‘enthusiasts’, esoteric cable vendors and academically trained professional engineers like myself.
I’m NOT going to provide mathematical evidence, pages of physics or measurements for my following comments (maybe a few well deserved links). It is not necessary. Why? Because I have been there and done it, both as a professional engineer and an A/V enthusiast. I’ve heard, and demonstrated, cable differences and could always associate them with a technical issue (or trick), so there has never been any black magic science required.
I’m NOT saying that cables do not influence the performance of their attached pieces of hardware. THEY DO, and can be clearly demonstrated that they do (my students were often amazed at my demonstrations). However, their effects are only observable/audible when the surrounding technical infrastructures and design is subpar or if the STANDARD cabling used is just not capable or suitable of meeting the demands of the connection. e.g. bell flex for speaker connections.
All of the audio and video integration and system design companies I have worked for over the years use ‘standard’, well designed, professional cables from companies like Belden, Klotz, Mogami etc. for a facilities entire build out. Utilizing those power cords that are sent with the vendors equipment. This cabling is all at the ‘pointy end‘ of the signal chain, so if there is anywhere that you need the very BEST connectivity it is during the production process. So why doesn’t it exist there? It’s not required because the engineering design of these facilities negates its requirement.
What do I mean by engineering design? A great deal of thought and effort (and cost) goes into creating the electrical infrastructure and grounding systems for professional technical facilities. This process includes:
- Oversized (or at least correctly sized) AC power distribution systems.
- Heavily filtered and stabilized power supplies from central UPS systems, some of which I have installed providing multiple 500+KVA supplies.
- Carefully designed power distribution systems providing correctly sized feeders to all distribution boards and multiple 20 and 30 amp feeds to each console and rack.
- Separate technical facility, office and HVAC power feeds.
- Separate safety and technical grounds to ensure that electrical interference and noise are ‘sunk’ to ground rather than being injected into a devices signal chain.
- Balanced connectivity to reject interfering signals. i.e. analog audio and networks. (The use of balanced distribution for digital audio are few and far between, read on)
- Correct use of gauges of cables for speaker interconnects. Many speakers these days being active, require just balanced line level connections. Except of course 70 volt distribution systems.
- Carefully designed very low impedance technical grounds. Installed to prevent any chance of hum loops.
- ONLY one bonding point between the technical and safety grounds, keeping all lightning grounds (a whole separate topic) well away from such points.
- Some facilities even have equi-potential ground matts inserted into their concrete foundations so that if a lightning strike enters the building, it produces the minimal possible potential difference across a room, helping to minimize possible damage to electronics.
The list goes on, and YES you can do most of this at home to support your own home theater. It’s NOT difficult if you are handy and abide by a few electrical codes. I carried out most of the above when I built my own piece of audio heaven, see here.
Besides all the above bullets points, most equipment (home or professional), other than power amplifiers or active speakers, draw an almost constant current as their signal I/O is line level voltage driven, requiring but milliamps or microamps of current. The AC power draw to such a device does not therefore significantly vary based on the signal level. So worries about the impedance of an AC power connection in some way impacting a varying AC current draw is not relevant. (Varying AC currents in power feeders can induce currents into the grounding cables contained in the same cable. This induced ‘noise’ can impose itself on the system’s signals if the grounding system and its impedance are poorly designed.)
It should be pointed out here that there is a considerable move by manufactures to move away from analog power supplies to switch mode power supplies (SMPS) in order to save costs and reduce heat. Another argument for this move is that the transformers associated with analog power supplies radiate high magnetic fields which can induce unwanted noise and hum into the surrounding electronics. While this is true, it can easily be mitigated with the use of mu-metal screened toroids, more costly, but it totally fixes the problem, and I have used this technique in EVERY power supply I have ever built over the past 35+ years.
These switch mode power supplies can introduce all sorts of out of phase current demands on the high voltage connections and in large facilities can cause significant neutral currents which have to be taken into account with oversized neutral conductors and K rated transformers. Switch mode power supplies are notorious for creating high frequency switching pulses that can get back into the main supply and, of course, into the equipment supply and grounding. Therefore a great deal of attention needs to be applied to your grounding schemes if your equipment is loaded with these supplies. Modern switch mode supplies are very heavily filtered on both their inputs and outputs to reduce these problems, but they are not 100% effective, and the supply can still radiate electrical RF interference and should be placed inside metal screening boxes.
Effects of power supply noise on analog and digital signal processing is another topic, but with modern day switch mode and analog designs, together with screened power supplies and a correctly grounded design, should not create a visible or audible problem, nor should it be measurable on any audio or video signal path, but I digress.
Look at any high-end audio equipment (excluding active speakers), you will rarely see any switch mode supplies in them, most being analog based. (There are always exceptions to the rule, for example, Emotiva, who use high efficiency switch mode power supplies in their power amplifiers like the XPA-7 Gen 3.) Some playback devices E.G Oppo 205, even use a separate analog power supply just to feed the analog audio sections, acknowledging the above issues, but still use a screened switch mode supply for the video and audio digital data signal paths, which are FAR MORE tolerant to noise. There is NO comparison between your eyes ability to resolve video information and your ears ability to resolve audio information on the best playback systems. Your ears are several magnitudes better at detecting audio anomalies than your eyes at detecting video anomalies, hence all of this rhetoric.
It has often been said that engineers have yet to produce measuring hardware that can measure sound parameters to the same level that your ear can analyze it. A young really well educated pair of ears can detect changes in perceived audio that instruments have difficulty resolving. It is well known that changes as small as 0.5dB or less can be detected by listeners, depending on its frequency, and that distortions or frequency components below the noise floor can be resolved by experienced listeners. Oh yes, and while we are talking about the ear, it’s not phase sensitive to steady state phase changes, only phase that changes with time; another great student demo showing the non-audible effects of varying the phase of the harmonics of a square wave. If you ears were sensitive to fixed phase shifts life would be quite miserable! However, phase/timing errors are audible on transients, like a kick drum for example, reducing its impact and punch.
Power and Grounding
Power Cables
- Low power equipment
This covers all pre-amps and players of all types. These devices draw very low currents typically being 50 watts or 0.5 amps or less, even my Denon AVP A1HDCI only draws a maximum of 200 watts or 2 amps. So even if your power cable was so poor that it was one ohm DC resistance the maximum voltage drop that you could generate would be 2 volts. Typically, 14SWG power cords barely register 0.002 ohms per foot.
All pretty irrelevant as I have seen my house AC supply vary by +-3 volts on a regular basis. So any good quality multi-stranded, plain copper, 12SWG power cords will provide you with power connectivity that will serve you just fine. Unless the cords are ridiculously long. In my case none or more than three feet. (Note: My UPS’s are single conversion sine wave ‘green units’ that only have there filtering active until the power is lost, at which point they kick in to protect the equipment and projector bulb.)
However, if as mentioned earlier, the equipment contains a switch mode supply, you have a different problem. High frequency noise from poorly designed supplies may find that the impedance of thin power cables to be high and prevents them being absorbed within your houses main supply and effectively reflects them back to the device. Or, your grounding is so poor, that the electrical noise becomes superimposed on the signal. This problem will NOT affect digitally connected devices over HDMI or ANY other digital links, but potentially may effect other associated analog connections as any noise can be coupled to other devices via any common ground connection such as a signal shield. This would potentially then inject noise into a receivers/pre-amplifiers DAC system and impact the final sound. Noise interfering with digital signals would have to be so high as too impact at least the first bit level or impact the signal digital clock, increasing jitter. Firstly, if the noise is that high you have a host of other problems that you need to address and secondly, many digital audio and network connections are either fully balanced to reject air born interfering signals or optical and immune to air born interference and grounding loops.
- High power equipment
This covers all power amplifiers and active speakers. Most audiophile power amplifiers use analog power supplies with variations on Class A or Class AB power amplifiers. Whereas most active speakers use Class D power amplifiers with switching mode power supplies, just like my Genelecs. Why? Because designing high power speakers with conventional built in amplifiers and analog supplies would produce very large amounts of heat which would be difficult to dissipate. For both devices, power cables and feeders must be as low an impedance/dc resistance as possible, as the instantaneous load continually changes with the music level. With switch mode supplies the resulting high frequency noise must be shunted to ground at the earliest opportunity and all power cabling feeding these devices must have a low impedance/dc resistance. In my case each speaker gets its own 20 amp supply over 12SWG copper cable from the rooms dedicated 100amp distribution board. The board being fed by a multi stranded 6SWG plain copper feeder directly from the houses main electrical connection.
- Power Phases
All American homes are supplied with 240 volts which is centered tapped to give the house two 120volt supplies being 180 degrees out of phase. Be advised that using both phases in an AV setup to distribute the load is NOT recommended. It increases the possibility of yet more current loops due to earth leakage beside being potentially more dangerous. In my case I use a single phase to provide a 100 amp dedicated supply, that is assuming that your house connection can support this. My own home having a 200amp service.
Grounding, Technical Grounding and Bonding
There is only one way to correctly ground technical equipment to obtain the optimal reduction and prevention of noise and hum; a technical grounding system. Which, with a little effort on behalf of the reader, can be created by any competent enthusiast.
So what is a technical ground? It is a separate ground system that is used in parallel with the safety ground and has a very low impedance, being bonded to the safety ground ONLY at the point of entry to the building. The electrical distribution system that is part of a technically grounded system having all connections between the neutral and ground, as found inside distribution boards, removed. This ensures no ground loops within the power system and ensures that the technical ground provides the shortest and lowest impedance path for the absorption of all electrical noise.
A technical ground may be created in numerous ways depending upon the ground conditions and how low you want its impedance (resistance) to be. The simplest method is to drive several 6 foot ground rods into an area that you know remains damp, bonding these rods together using 6SWG cable and then taking a 6SWG cable from the bonding point back to your systems main grounding point, often found on the rear of the receiver or a copper buss bar in the rear of the rack, to which the chassis of each item of equipment should be connected. Do not rely on the rack bonding to the equipment using the rack screws. Paint will often inhibit this connection and rack screws use isolation washers to prevent facial damage to the equipment ears. In my case I have a copper buss bar bonded to the rear of my rack to which all grounds are attached. It is also usually common, in most technical facilities, to isolate the rack from a rasied floor – not normally an issue for residential users. Raised floors sit on metal supports and code requires that these have to be bonded to the facilities safety ground. Any inadvertent electrical connection from the rack to the floor would produce a ground loop.
Typically, in professional facilities, rows of racks are all bonded to one central isolated copper plate using 6SWG copper wire, all these plates are then bonded using 0 Gauge back to the central facility bonding point. In some facilities, based upon the clients’ requests and technical requirements, these bonding cables maybe stranded copper, flat ribbon and in some extremes braided flat ribbon. The latter two having very high surface areas to minimize skin effect.
What is skin effect I hear you say? It is the migration of electrons to the outer layers of a conductor, effectively reducing its cross-sectional area and increasing its impedance. The effect is generally only apparent at frequencies in excess of 100KHz, well above any signal that would pass down an audio cable. However, it does apply to electrical noise in ground systems in both analog and digital video systems and especially in data centers. This is one of the arguments for the use of cables that are composed of many smaller strands, as this creates a much higher available surface area and therefore lower impedance. Another great demo to students, using bell flex or Litz wire driving a standard electrical bell.
You will have noted by now that the term IMPEDANCE has cropped up and replaced resistance. The resistance of a conductor is its ohms law value that opposes the flow of a direct current (DC). That is a current that always flows in one direction. Impedance is the opposition to alternating current (AC) flow (audio, video, data and noise), and primarily depends upon the conductor’s capacitance and inductance per unit foot, but is still measured in ohms. Capacitive impedance goes down, while inductive impedance goes up with rising frequency. The combination of the impedance and resistance of a conductor has a significant effect on the passage of a signal as the cable starts to behave as a filter. Such that what goes in doesn’t come out! At audio frequencies and with large gauge and/or multi stranded plain copper cables exhibiting low capacitive coupling and inductance per unit foot, together with low source impedances, these issues should be in-audible.
We now have an idea of how to correctly ground equipment and supply it with power. In the next post in this series I shall discuss the various types of cabling, connectors and materials issues.
I take no responsibility for those of you who mame or kill themselves or break electrical codes if you take on a technical ground or high voltage upgrade! If in any doubt get a licensed and knowledgable electrician.