Practical Wiring - Professional Motorsport: Concentric Twist Layer Design
Watch This Course
$299 USD
-OR-
Or
8 easy payments of only $37.38
Instant access. Easy checkout. No fees.
Learn more
Concentric Twist Layer Design
17.02
| 00:00 | - Concentric twisting is the name given to the bundling technique we use to gather our wires together as they run throughout our motorsport wiring harness. |
| 00:08 | It is one of the most obvious differences between a motorsport wiring harness and a club day or modified street car wiring harness. |
| 00:15 | The technique entails that we organise and run our wires in distinct concentrically wound layers between the harness branch points and connector terminations. |
| 00:24 | Each layer is twisted around the one before it with the direction of the twist alternating as we build up the layers. |
| 00:31 | This is a complicated and time consuming process but it gives the finished harness some very desirable properties. |
| 00:38 | The finished harness will be very flexible, as when it needs to bend to run around a corner, the layering of the wires inside ensures that they all stay in place and do not bunch up. |
| 00:49 | The twist given to the wires ensures that they are all evenly stressed as the harness is bent. |
| 00:54 | If the wires were run in a straight lay fashion, and the finished harness was bent around a corner, the wires on the outside of the bend would be more stressed than those on the inside. |
| 01:04 | Concentric twisting results in our wiring harness having a tightly packed circular cross section. |
| 01:10 | Which means that this cross section is as small as possible for the number of wires making it easier to install in the vehicle. |
| 01:17 | The circular cross section also makes the job of sheathing the harness much easier as our DR25 shrink tubing will be much easier to slide into place over the uniform harness cross section. |
| 01:28 | This section of the course will outline the guidelines that I use to plan the layers in each section of a harness. |
| 01:35 | This includes the direction of twist and the individual wires in each layer. |
| 01:39 | We will not talk about the physical concentric twisting process now, but this is covered in detail later in the practical skills section of the course. |
| 01:47 | The simplest specification detailing the construction of a concentrically twisted wiring harness calls for the core of the harness to be a single wire, with six wires then twisted around this as the first layer. |
| 01:59 | Each additional layer is then made up of the number of wires in the previous layer plus an additional six. |
| 02:05 | Giving 12 in the second layer and 18 in the third and so on. |
| 02:09 | This specification works well if you're constructing a wiring harness out of only a single size of wire. |
| 02:16 | But in an automotive application this is almost never the case. |
| 02:20 | You will have different sizes of wire to handle different current levels, some wires will be shielded and some will need to be in individually twisted pairs. |
| 02:28 | Tyco Electronics have a publically available document that outlines their codes of practice for hand laying a concentrically twisted cable. |
| 02:36 | And this is a much better resource for our application as it allows for the construction of cables with varying sizes of wires which are called a hybrid cable. |
| 02:45 | We've provided a link to this document below the course module as it is a great resource but in my experience we do still need to vary from it to get the best result possible. |
| 02:55 | With the power supply, grounding sensors, actuators and network communication sections of our harness designed, along with our routing plan, we now have all of the information required to know the exact wires that will be passing through any part of our harness. |
| 03:10 | This is the information we need to determine how those wires will be layered in the concentric twist. |
| 03:16 | I prepare a document that lists all of the harness sections, this will be every section that ends in a connector but also the sections that span between branch points. |
| 03:26 | I then list under each harness section the details of the wires that will pass through that section. |
| 03:33 | We can then go through each harness section at a time and organise the wires in that section into layers. |
| 03:39 | So when we are later constructing that harness, there's not guess work required. |
| 03:43 | This raises the question of what rules do we apply to organise these wires into their layer groups? There's an element of experience needed for this task to know which layering will give the best outcome but I'll outline now how I approach the task and the rules that I've developed over the process of constructing many of these harnesses. |
| 04:02 | The first group we create for each harness section is called the core. |
| 04:05 | This is the centre of the harness section around which all the other layers will be wrapped. |
| 04:11 | The core should contain the largest and least flexible cables in the harness section. |
| 04:16 | But typically no more than five cables total. |
| 04:20 | This is the largest number of cables I find feasibly possible to tidily twist together to form a core. |
| 04:26 | I've used the term cables here instead of wires as the harness section may contain cables that have multiple wires within which we look at as a single element for our layer design. |
| 04:38 | A common example to demonstrate what I mean would be a core containing three twisted shielded pair cables and two unshielded twisted pairs being engine speed, engine phase, and knock signal, and two CAN bus trunks. |
| 04:51 | This is 10 wires in total, not including the shields, but only five cables. |
| 04:57 | The desing process for our core differs from the subsequent layer design in a couple of important aspects. |
| 05:04 | In the core it is acceptable to have cables of different diameters twisted together. |
| 05:09 | As long as they twist tidily and still have a roughly circular cross section. |
| 05:13 | In our example above, the shielded twisted pair cables will have a larger diameter and be less flexible than the unshielded twisted pairs. |
| 05:22 | But a suitable core can still be formed from the five cables. |
| 05:26 | The other aspect is the amount of twist angle imparted to the cables. |
| 05:30 | More commonly called the lay length. |
| 05:32 | This is the distance a cable travels down the harness section as it makes one complete revolution within the core. |
| 05:39 | For the subsequent layers of the harness, this lay length is partially dependent on the diameter of the previous layer around which we're working. |
| 05:47 | But for our core there is no previous layer. |
| 05:50 | You'll find when it comes to build time, that the cables in your core will form a natural comfortable lay length. |
| 05:56 | And this is likely to end up being longer or less of a twist angle than the subsequent layers. |
| 06:02 | This is absolutely fine and the important part to focus on is that the cables in the core form a tidy bundle. |
| 06:09 | We will discuss this further and demonstrate the process in the construction skills section of the course. |
| 06:14 | I have mentioned that the core should not be built out of more than five cables, and while this is a rule, I stick to reasonably firmly there's one common scenario where it is broken. |
| 06:25 | If our core is built out of large diameter cables there'll be open spaces between those cables that is still within the overall circular cross section of the core. |
| 06:36 | These spaces can be an excellent place to run small 24 gauge wires. |
| 06:42 | If fitting them into a layer later would be difficult or require many filler wires to be added. |
| 06:47 | The core of our harness sections will change along the length of the complete harness as it's possible the cables in the core will need to branch off to their destinations, meaning we will need to reorganise the core for the remaining harness sections. |
| 07:01 | A common example of this happening is with the shielded twisted pair cables we use for engine speed and position sensor signals. |
| 07:08 | As they're a relatively large cable they often form the core of our harness for many harness sections from the ECU towards the engine. |
| 07:16 | However as we get further out towards the ends of our harness, these shielded twisted pair cables will branch off, meaning we now need to choose another core. |
| 07:25 | If available we would then choose the next largest cable or cables, often being power supply wires to high current devices like radiator fans, a water pump, or a fuel pump for example. |
| 07:38 | Once you have determined the core of each harness section, you can begin designing the layers that will wrap around the core. |
| 07:44 | For this step we will choose the next largest conductor size and work outwards towards the smaller conductor sizes with each layer containing only a single conductor size. |
| 07:56 | As an example, it's common to have 20 gauge conductors as your next largest conductor size after your core cable. |
| 08:02 | This is due to the pins size limitations of the motorsport connectors through which the harness may need to pass. |
| 08:08 | This actually helps us in our layer design as instead of needing to deal with a lone 18 gauge power wire for a high current device, we instead have two 20 gauge wires of which there are most likely others helping us to fill out a layer. |
| 08:22 | When talking about the core of the harness, I mentioned the term lay length, and how for the core it is typically quite long. |
| 08:30 | For every layer of our harness, except for the core, we're aiming for a lay length of between eight and 12 times the finished diameter of the layer that we're currently building. |
| 08:41 | This lay length has been determined to ensure the finished harness section will have the mechanical properties that we're looking for. |
| 08:47 | Primarily flexibility. |
| 08:49 | To attain this lay length we undertake a simple calculation and use a lookup table. |
| 08:53 | We take the finished diameter of the previous layer and divide it by the diameter of the individual wires in the current layer that we're working on. |
| 09:03 | This result is then crossed over to a look up table giving the number of wires of that diameter which will twist around the previous layer to give the correct lay length. |
| 09:14 | As the desired lay length is a range, varying from this number slightly will usually still result in an acceptable harness section. |
| 09:21 | If possible I like to lean towards the high end of the lay length scale, as I find it still results in a nicely flexible harness but is often easier to construct. |
| 09:31 | This method of determining the layer structure is outlined in the moulded cable assembly section of MILSTD 339 which is well worth a read if you intend on making many concentrically twisted harnesses. |
| 09:45 | The lookup table which we have included below is sourced from this standard. |
| 09:49 | Obviously to use this method, we need to know the diameter of the previous layer and the diameter of the cables in the layer that we're currently designing. |
| 09:58 | This is usually not a problem as we can calculate the finished diameter of the layer we're designing by adding twice the diameter of the wires in this layer to the diameter of the layer underneath. |
| 10:10 | This means that as long as we know the diameter of the core, we can work outwards designing the subsequent layers without needing to physically construct the previous layers. |
| 10:19 | While there are methods outlined in MILSTD339 for calculating the diameter of different core constructions, I find it easiest to just build a small section of the core out of offcuts of the same cable to be used and measure it. |
| 10:35 | When doing this, measure the diameter of the core in multiple locations and average your result. |
| 10:41 | Often you will find yourself in the situation where you do not have enough wires available in a harness section to complete a layer. |
| 10:48 | When this occurs, we add ballast or filler wires of the same diameter to bring the number of wires into the acceptable range. |
| 10:57 | These ballast wires serve no electrical purpose and will be terminated within the rigid moulded boot at either end of the harness section. |
| 11:05 | If you strike a situation where you have two layers next to one another, made up of the same size conductors put any required filler wires into the outer layer as in the event of a crash or a misshandling of the harness, these filler wires may protect the more important conductors beneath. |
| 11:21 | As an example, we'll look at the main harness section heading away from the ECU for a six cylinder engine. |
| 11:28 | Included below is a list of all the wires in this harness section. |
| 11:32 | We'll go through them now and sort them into their layers. |
| 11:36 | The largest cables in this harness section are the 22 gauge twisted shielded pairs, used for the engine speed and position sensors along with the two knock sensors. |
| 11:45 | These four cables will be twisted together to form the core of our harness. |
| 11:50 | A quick test build with offcuts of this cable gives a finished core diameter averaging 6.5 millimetres. |
| 11:58 | Our next largest cables are the 20 gauge power and ground wires for the internally ignited ignition coils. |
| 12:05 | A measurement of the diameter of these cables gives 1.27 millimetres. |
| 12:10 | If we divide our core diameter of 6.5 millimetres by 1.27 millimetres we get a result of 5.12 Crossing this over on our lookup table, we arrive at a result of 18. |
| 12:25 | This means we need 18 20 gauge wires in this layer to result in a full layer with the correct lay length. |
| 12:33 | We only have 12 20 gauge wires in our design however so we'll need to add in six ballast wires to complete the layer. |
| 12:41 | We can now add twice the diameter of our 20 gauge wires to our core diameter to get the finished diameter of this layer. |
| 12:49 | Two multiplied by 1.27 millimetres plus 6.5 millimetres gives us a finished layer diameter of 9.04 millimetres. |
| 12:59 | All of the remaining wires in the harness section are 22 gauge. |
| 13:03 | Ideally we will be able to run them all in the next layer needing very few if any filler wires to complete the layer. |
| 13:10 | In practice the design does not often work out this way, and we will often need to adjust our design to suit the situation. |
| 13:17 | The process begins the same as for the previous layer. |
| 13:21 | We take our 22 gauge wire diameter of 1.12 millimetres and divide our previous finished layer diameter by that. |
| 13:30 | 9.04 millimetres divided by 1.12 millimetres give a result of 8.07 Crossing this over to our lookup table we arrive at a result of 27. |
| 13:42 | This means that 27 22 gauge wires will form this layer. |
| 13:47 | Unfortunately looking at our harness section, we have 34 22 gauge wires to contend with. |
| 13:54 | This would mean another layer after this one made up primarily of ballast wires. |
| 13:59 | Sometimes this is the only option but usually there is another way to approach the problem. |
| 14:05 | In this instance I would choose to move six of the 22 gauge wires, most likely the coils signal wires, to the previous layer. |
| 14:12 | Also increasing their wire size to 20 gauge to match that layer. |
| 14:17 | This increase in wire size will have no ill effect on the signal and makes better use of the previous layer as well as these six wires can replace the ballast wires. |
| 14:29 | This leaves us with 28 22 gauge wires for our current layer with an ideal number of 27. |
| 14:35 | Deviating from our lookup table result by one conductor is not going to cause any issues and still result in a layer with an acceptable lay length. |
| 14:44 | To summarise the design of this harness section layer, our core will contain four twisted shielded pair cables. |
| 14:52 | Our first layer having 18 20 gauge wires and our final layer having the remaining 28 22 gauge wires. |
| 15:00 | Calculating our final cable diameter gives a result of 11.28 millimetres. |
| 15:05 | An added benefit of knowing this cable diameter in advance is that we can now specify that this harness section will be sheathed in 3/4 inch Raychem DR25, as this is the largest size that will shrink to fit this section. |
| 15:20 | As you can see, the concentric twisting technique results in a very small harness section diameter and this can be a huge help when needing to fit the harness into tight spaces. |
| 15:31 | It should be reasonably clear from the discussion in this section that building a concentrically twisted wiring harness is a complicated task and will require practice and experience to become proficient in. |
| 15:42 | Presented here are the reference guidelines I use when constructing a harness but there are no completely set in stone rules. |
| 15:49 | Often you will find yourself in situations where you will have multiple possible layering schemes. |
| 15:55 | By undertaking detailed planning on the harness construction from beginning to end before you lay your first wires, you will be able to evaluate how these options will affect the rest of the construction down the line and choose the best one. |
| 16:09 | My experience building harnesses like however has shown me a relatively common pattern to the layering scheme. |
| 16:15 | It should not be taken as wrote but more of a guide to get your started. |
| 16:19 | The core is usually built out of the shielded cables being crank and cam triggers and knock sensor wiring. |
| 16:26 | The first layer is then the power supply wires to the high current devices and twisted communication bus wires, typically CAN and ethernet. |
| 16:35 | At this stage the remaining wires are usually all 22 gauge and are organised into layers according to where they branch off the main harness path. |
| 16:45 | Harness sections which branch off to sensors and actuators most often have four or less wires, and these are simply twisted together as if they were forming a core and then sheathed when the harness construction is finalised. |
