Wiring Woes and Auto Cable - Caravan and Motorhome Books

In electric speke, anything below 110 volts direct current is known as extra-low voltage. Unlike 240 volts (known formally as 'low voltage), no licence is required to work on extra-low voltage installations. Many people tackling this work do it very successfully, but one area that catches people out is cable sizes.

All electrical cable resists the flow of current, but the larger the diameter of the copper within the cable, and the shorter its length, the lower that resistance. We need to minimise resistance because it reduces the voltage that refrigerators, lights etc need to work properly.

Because all cables resist current, all cables will incur voltage loss. The highest acceptable voltage loss in 12/24-volt campervan or motorhome wiring is 3% of the battery voltage. That is: 0.36-volts, and 0.72-volts, at 12 and 24-volts respectively. Professional installers specify cable for 5% loss because it cuts cost – at the expense of people whose electric fridges then never work as well as their makers' claimed. And, for years wonder why.

Tables of cable sizes for various current flows (at 3% drop) are readily available. It would seem easy then to select the right cable. But there's a HUGE trap.

Automobile cable utilises a rating system that appears to correspond (but doesn't) with appliance manufacturers recommendations and ratings. It does not correspond with other cable size standards either, nor is it possible to make meaningful comparisons. Which is a pity because otherwise it's good cable.

Sold by auto parts stores, auto-electricians, etc, auto-cable's rating is of its outside diameter, not the conductor size. That is it includes the thickness of the insulation – and this varies. All this absurd rating thus tells you is what size hole the cable can pass through.

Appliance manufacturers and non-auto electrical trades mostly use one or other of two systems (ISO and AWG). The former rates cable directly in terms of the cross-sectional area (in sq mm) of its copper conductor: thus ISO 2.5 cable is a conductor area of 2.5 sq mm. The main alternative, AWG, uses a numerical system (of which AWG 0-18 is likely to be used in caravans and motorhomes). Comparison tables, that relate AWG gauges to cross-sectional area are available.

Compounding potential confusion, both ISO and auto-cable ratings use similar numbers. For example, a typically recommended cable size is 4.00 mm.sq. This may be spelled out in full, but may often just be noted as 4 mm cable.

An unwary buyer asking for 4 mm cable at an auto-parts store will almost certainly be sold cable marked 4 mm, but with a conductor area of (typically) 1.80 sq.mm.

If this is used instead of 4.00 sq.mm cable, the voltage drop is more than doubled. This is also the reason why some solar installations disappoint, and why batteries are often grossly undercharged. Countless vehicles are wired this way!

As if the above were not bad enough, auto-cable is usually rated also in amps. But that rating does not indicate what the cable should carry. Instead, it warns of the highest current that cable can carry before its insulation melts (this rating is often known as 'ampacity'). And it does not take voltage drop into account. If it did, that drop may exceed 50%.

As noted previously, appliance makers usually specify the minimum-sized cable that may be used. Some supply the actual cable. Heavier cable must be used if the cable has to be longer than that specified. But this is of no help if wiring a complete vehicle. There, one must assess the maximum current that will be carried by each cable and specify the size accordingly.

Tables of cable sizes for various lengths of cable and current carrying capacity are obtainable from cable manufacturers. They are published in books such as my own 'Motorhome Electrics'. There is a simplified version in this feature.

There is 'rounding up' discrepancy in most published AWG/ISO conversion tables (not in mine). Use one AWG gauge size larger (ie. numerically lower) than the conversion shows. For example, ISO 2.5 is AWG 13 (not 14 as usually shown) etc. Using ISO gauges is easy. The actual gauge number (eg. 1.5, 2.5, 4.0, 6.0, 10 etc) is the actual cross sectional area of the conductor, in sq.mm.

Another very easy way to work out the voltage drop for any given length of cable, current flow, and cable size is the simple formula shown below. I always use this in conjunction with my published tables. This formula is simply:

Voltage drop equals (cable length (in metres) X current (in amps) X 0.017) divided by cable cross-section in mm.sq.

For example: 10 metres X 5 amps X 0.017 = 0.85. Divided by (say) 2.5, the voltage drop is 0.34 volt. This is just acceptable.

The above, and wire-tables indicate voltage drop across a single conductor. For chassis (earth) returns, the resistance of the return path can be ignored. Where there's another conductor for the return path, the total cable length must be taken into account. In other words, if there's a separate conductor for the earth return that's ten metres of cable, so you do the sum as if it were ten.

Single-core ISO-rated cable in sizes from 1.5sq.mm. – 16 sq.mm is obtainable from electrical wholesalers, and some marine electrical suppliers.

Auto-cable is fine but only if the conductor area is also provided (for some brands it is). Check that out for yourself, as the seller is not likely to understand your query (let alone the vital need to know).

The very best cable for caravan and motorhome wiring is so-called 'tinned' copper. It's not actually tinned, but rather copper electroplated with a nickel alloy. It is obtainable from boating electrical equipment suppliers, but is hard to locate otherwise. It's worth tracking down.

Finally I might mention that I recently heard from a member who had just finished wiring a complete motorhome, unfortunately using mostly 2.50 mm and 4.00 mm automobile rated cable. The good news was that he had planned to use a 12 volt system but had not yet bought the bits and pieces. The simple solution was to switch to 24-volts. The voltage drop was thus halved – still a bit high for comfort, but the only alternative to totally rewiring.

There is increasing evidence that some RVs may have part or all 12/24 volt wiring that is substantially lighter than that intended by appliance makers. This will introduce a voltage drop up to 55% greater than intended - but in some circumstances it may be up to 70%. Where the original cable was over-specified this may not matter, but will cause a considerable loss of performance if that cable was originally specified to run close to its intended load.

The problem is caused by confusion between two cable rating systems that, in the smaller sizes, have apparently identical 'numbers', but are in fact specified in different ways. I earlier believed this to be a mainly (and recent) Australian phenomenon, but it increasingly appears to be a far wider problem than I suspected - and one that may have existed for many years. It does not appear to affect USA-built RVs as that country uses a different cable rating system (AWG).

Manufacturers of 12/24 volt appliances specify the cable that must be used to connect an appliance, or supply that cable (where the cable is supplied and used there is no problem: unless it is lengthened).

This cable is almost always specified using the internationally recognised ISO (International Standards Organisation) rating system. The ISO system rates cables in terms of the cross-sectional area of the copper conductor - using the following increments. 1.0, 1.5, 2.5, 4.0, 6.0, 10, 16, 25, 35, 50, 70, 100 sq mm etc. Thus an ISO 4.0 sq mm cable is a cable that has a cross-sectional area of 4.00 sq mm. This is the cable used for mains wiring in Australia and is also fine for 12/24 volt RV use.

The auto-electrical industry however uses what is known as auto cable, and this is the cable that you will be sold if you buy it from almost anywhere except a general electric wholesaler. It is that sold by auto parts suppliers.

Unlike ISO-rated cable, auto cable is rated in terms of its overall diameter (ie. insulation and all). This insulation is typically 0.5-1.0 mm thick (but varies not only from maker to maker, but varies from one type of auto cable to another from the same maker). Thus the nominal auto cable size gives little indication of the amount of copper it contains (and hence current carrying capacity). The most commonly used auto cable sizes (in RVs) are 2-3 mm, 4 mm, and 6 mm.

The following Table shows the actual sizes (in the all-important square mm) of '4 mm auto cable' sold by various makers in March 2003.

Specified correctly, auto cable is a perfectly good product. The problem, from an RV point of view, is that auto cable, sold in some numerically identical sizes as ISO-rated cable, is being widely mistaken for the ISO rated cable usually intended.

In many instances auto cable's size in square mm is also marked on the drum (but in small print). But as very few people are aware of the situation, this marking generally has little relevance to them - and may only be noticed if one buys the whole drum.

There is also confusion about the current ratings usually attributed to most makes of auto cable.

These 'current ratings' are fire safety ratings. They relate only to acceptable temperature rise: what the cable can carry before the insulation becomes unsafe. Further, these 'ratings' often assume that current flow is not continuous.

One (imported) 4 mm auto cable of 1.85 sq mm is rated by its maker at an extraordinary 60 amps. Another 4 mm, 1.85 sq mm (Australian-made) auto cable is rated at a far more realistic 10 amps. But in no case are these ratings directly related to acceptable voltage drop.

This matter came to light during 2002. During that year I became increasingly aware that complaints of (mainly) gas/electric fridges working fine on 240 volts, but not on 12 volts, needed closer investigation.

There were simply too many to be random faults: there had to be some common cause.

I looked into about 50 instances of this (often via email/telephone) and found that that all but two were due to 4 mm auto cable having been used in place of that specified. This typically introduced about 0.8 volt drop between battery and fridge (the odd two were caused by corroded connectors) and were fixed completely by installing adequate cable (6 mm auto cable is usually fine - it is typically 4.9 sq mm).

Early such 12-volt fridges typically draw 12.5 amps (current models draw about 15 amps). Fridge maker usually specify a maximum voltage drop of 3% (0.36 volts). For this example assume a fridge that is three metres cable run from the battery (i.e. six metres of conductor).

Voltage drop = Length of conductor in metres x Current in amps x 0.017 divided by conductor cross section is sq. mm. For early gas fridges this is thus: 6 x 12.5 x 0.017 = 1.275. Here 4.0 sq mm cable results in a voltage drop of 0.318 volts. With 15 amps draw, the drop with 4.0 sq mm cable is 0.382 volts. This is a bit high, but marginally acceptable.

6 x 12.5 x 0.017 = 1.275 divided by 1.85. This is a voltage drop of 0.69 volts. Or, at 15 amps, of 0.83 volts.

Circuits running a few globes at 12/24 volts are often specified to run over 1.5 sq mm cable. Most auto cable of the closest apparent size (2 mm) has 1 sq mm of conductor, but some has as little as 0.5 sq mm. The latter will typically introduce a drop of a volt or so (or about 8%). Globe brightness of incandescent and halogen globes is proportional to the fourth power of the voltage so the effect of voltage drop is profound: e.g. 5% voltage drop causes 20% loss of brilliance. (For people who care about this stuff, the relationship is defined by the Stefan-Boltzman equation).

The larger 6 mm auto cable is less of a worry. This cable is typically 4.9 sq mm - and is a good substitute for 4 mm cable where 4.00 sq mm was intended. But it still pays to read the small print on the drum: plastic is much cheaper than copper.

In most instances the incorrect use of auto cable will degrade performance but is unlikely to present as a general safety risk (fridge cables may run hot - but not to the extent of being a fire risk. There may however be a specific safety risk.

There have been various magazine and website articles (and website correspondence) actually recommending 4 mm auto cable for electric brake circuits. Anyone following this advice is likely to have brake wiring with totally unacceptable voltage loss. Electric brakes are conservatively rated, but not to that extent.

This is a field that's a long way from rocket science and I seek no credit for apparently uncovering it. Any electrical engineer would have done the same (and probably much quicker because I'm a research engineer). But there appear to be hardly any electrical engineers working in the practical side of the RV industry. It might also have been spotted years ago if users' complaints re fridge performance had been taken more seriously. Seemingly they were not.

The extent of the problem is beginning to be realised following my original writing about it in the CMCA's 'The Wanderer' magazine Tech News column. This column has been picked up by global Internet search engines and recently became 'number one on the Google hit parade' (search for 'auto cable' + 'sizes'). It is too early to comment but the issue is definitely not confined to Australia.

I alerted the Australian auto electrical industry about this problem (via an article in Automotive Electrical & Air Conditioning News in January 2003). Feedback from that initial article confirms that few auto-electricians are aware of ISO cable ratings - and several rightly point out appliance makers often specify 4 sq mm cable simply as 4 mm cable (i.e. the square millimetre bit is implied!).

When faced with the 12-volt fridge problem etc, auto electricians quickly pick the cause and fit heavier cable (usually 6 mm auto cable) but dealing as they are with isolated cases, were not aware that the problem may be so common.

Electricians (ie. other than auto electricians) seem well aware that with appliance cabling, cross-sectional area is implied (ISO is the Australian mains-wiring standard). Several I've discussed this with are well aware that auto cable is quite different - but all assumed that this was common knowledge.

Since the publication of my original articles on this matter I have been commissioned to write a complete series on electrical installation in RVs. This will be published in Automotive Electrical & Air Conditioning News, and may eventually be expanded and published as a specialist RV book for the auto electrical industry late this year or early 2004.

The generality of this Update Note will be incorporated into future editions of Motorhome Electrics but it may be a year or two before the full story is known.

There no easy fix for the fridge cable. It simply has to be replaced. For lighting circuits that use incandescent globes, a relatively simple fix is to replace the fittings or globes with halogen equivalents of half the wattage. These give the same light for half the current and thus half the voltage drop.

If you have the problem and are already using halogens you can either replace the cable - or wait for the shortly forthcoming LED (Light Emitting Diode) cluster globes, a few of which are already on the market (but costly). These will draw so little current that the original voltage drop will be no problem.

I should stress that I am attributing no blame whatever to RV manufacturers. It is an unfortunate but understandable problem that remained unsuspected for years. Some people are unlikely to thank me for uncovering it - but there's no reason to perpetuate it.

If blame is to be attributed it should perhaps be directed at whoever thought up the auto cable method of rating, without considering its possible repercussions; and appliance makers who did not spell out exactly what they were specifying.

Do please note that the above column is copyright but may be reproduced subject to the following: