This is one area that many builders seem to dread and sometimes even choose to contract out. Personally, I have absolutely no concerns about doing this because of my background with electronics. Like many of the assembly steps, this is an area where it is important to use the right materials and tools and then take the time to do it carefully. If done properly, the results will be trouble free and last dozens of years; done improperly it will plague the craft for the rest of its life. These are just some of my thoughts to get someone started in this area. They're definitely not a complete guide and if someone is not familiar with doing wiring then I'd suggest they get some good reference material such as the AeroElectric Connection. Since AC43.13 is the bible that an inspector should hold your work to, I'd strongly suggest going through this material at least once and consulting it if there are any doubts. Another excellent document is NASA Technical Standard 8739-4.
The other alternative to wiring the craft is to contract someone else to do it. While this is plausible, it will probably require more time (and obviously expense) to do properly than the builder first anticipates. If one is contracting this out, then they should be sure that they're using someone who is experienced with aviation wiring ... while an AME (A&P) might possibly do a good job, typically an avionics person will make a much neater and reliable installation. While an ordinary electrician could possibly do the job, there are a lot of differences in the various practices, tools and requirements.
The third alternative is to do some of the basic power wiring oneself and then contract an avionics person to make the harnesses for things like radios, transponders and GPS. This can be done either directly on the physical panel or by making an accurate diagram of the various equipment locations and then having a custom harness built remotely ... basically all the connectors would be pre-wired and probably the only connections left would be for power, ground and possibly switches.
The CHR Construction Manual claims that they wired a Commuter IIB in twelve hours. While that may have been possible for a very basic installation, I would challenge whether it meets the standards and needs of current craft. I am also aware of helicopter kits that have been wired by licensed technicians and took several hundred hours to do the job properly. Doing it properly including a radio(s), intercom, transponder, antennas, etc. will take a lot of time.
FUSES and CIRCUIT BREAKERS
GROUNDING & NOISE
WIRE ROUTING and SUPPORTS
Like many jobs, the proper tools can make this a relatively easy job with reliable results; the wrong tools will introduce flaws that will compromise the results and eventually lead to problems. Compared to the tools used in the overall project, electrical tools are not a very big investment and they should last a lifetime of light duty use. These are the basic ones that I consider to be essential although there are lots of others, such as cable cutters, that are nice to have available.
I think everyone who has done any amount of wiring has used a wide variety of inappropriate tools to cut wire, whether it be dull pliers, X-Acto knives or even an axe. One of the essential tools is to have a good pair of wire cutters; the ideal ones are flush side cutters that have the very sharp edges and are only used for this purpose. Alternatively, a pair of good nail or toe clippers can be used for the smaller gauges of wire ... perhaps 16 gauge and smaller.
While flush cutters are great for the smaller gauges of wire, they're not nearly as effective with the larger gauges especially anything larger than about 18 gauge. Regular side cutters will work fine for 16-10 gauge but they will "squish" the wire somewhat. For the larger sizes one should ideally use cable cutters with the curved cutting surfaces. Other tools could be used but they tend to mash the ends of the wire. I actually tried these heavy duty cutters on some #22 wire and they cut it equally as well as the flush cutters above, if not better. This tool also worked fine on the heavy #2 battery / starter wire.
Although the insulation on aircraft wire such as ETFE (tefzel) is quite thin, it is also very tough. The only way I know to consistently, reliably and quickly strip the insulation without "nicking" the wire is to use a proper wire stripper. The better tools have one set of jaws that grip the wire plus a set of blades with split circular holes for each wire size that are designed to nick (or grip) the insulation and then pull it off as the handle is squeezed. Alternatively, the insulation can be removed by first nicking (not cutting through) it with a sharp X-Acto blade, but I don't generally recommend this technique unless one has experience doing it. If one or more strands of the wire is cut and falls out, I'd recommend re-doing that end as it is indicative that the cut was made too deep and other strands are probably also weakened. Technically, even if a strand is only nicked then one should consider re-doing it. There are some interesting pictures and discussion about wire strippers here.
Unfortunately, there are a couple of places where the builder will probably have to resort to the sharp blade removal technique: namely multi-conductor shielded wire, heavy cable and coax cable. There are specialized strippers available but they're expensive and seldom used. My advice would be to go slowly and carefully while remembering that in order to remove most insulation it only needs to be scored rather than cut. Flexing the wire on the score line generally causes the insulation to cleanly sever itself. If removing long pieces of insulation (i.e. 2" on the end of mulit-conductor wire), one can score the insulation circularly then lengthwise and only just cut through the last 1/16" or so at the free end. The cut end allows a starting point to grip the insulation and then peel it back along the score line.
There are a variety of these for various different connectors and some of them are quite expensive. The one required crimper is for PIDG terminals and I'd highly recommend that one gets a ratcheting handle double-crimp tool. These are extremely easy to use and will repeatedly produce a reliable crimp. Although there are some very expensive ones that are excellent and designed for production environments, there are also good quality ones for about $40 that are definitely more than adequate for our needs. While I don't recommend them, the cheap automotive style single crimpers can be used but they require careful positioning for each of the two crimps and it is difficult to consistently obtain the required crimp pressure. One can look here for some interesting pictures of what happens inside a crimped joint.
Anomoly: I've recently done some serious (i.e. hard) pull testing and found an interesting anomoly. My tests with #20 wire and red terminals all exhibited a very strong pull resistance. However, #16 wire in blue terminals only had about 1/3 to 1/2 of the pull resistance. I've tried a couple of different terminal types and have even tried using automotive crimers to try over-crimp the joints in case my ratcheting crimers weren't accurate ... they were all weaker than the #20/red combination. The only test with #16 wire that produce an equally strong result was when I put it in a red terminal. Hmmmm ...
Tip: I've found that some brands of the PIDG terminals have a bump on the wire side that cause them to try shift in the crimpers as the crimp is being started. I've found that I can normally apply one or two "clicks" of the crimper with just the terminal inserted but without the wire in the terminal. This allows my free hand to carefully hold the terminal in position while the crimpers start to grab terminal and just begin to form the crimp without inhibiting the later insertion of the wire.
Some connectors have what looks like two "U" shaped openings on them for the wire and the insulation. The generic term I've heard for these is a "B" crimp since when they're properly formed that is what the result looks like. I ended up using these on the connectors for my ICOM radio, new extension leads for the Westach gauges and on the Whelen strobe wires. Note that there are different sizes crimp jaws for different terminals and wire sizes. These kinds of crimps and tools definitely take some practice in order to get really good results ... the biggest problem I've found is that the terminal may want to rotate when the crimp is first started and I don't have a third hand for the wire, crimper and terminal. I normally insert the terminal without the wire and just start the crimp using my free hand to hold the terminal and prevent it's rotation. I then insert the wire and continue the crimp while using my free hand to hold the wire in position and also trying to prevent any rotation. If one starts slowly and feels rotation, the pressure can be released and the terminal repositioned. The second crimp is for strain relief on the wire's insulation ... with soft PVC insulation this is also a B crimp but on Tefzel wire this is usually just a pair of circular hoops around the wire.
If one is using D-Sub connectors then there are three options for the pins & sockets: solder cup, B-crimp and mil-spec crimp. While the Daniels crimpers (AF8 or AFM8) do a beautiful job on the mil-spec connectors, the tools and positioners are very expensive for just the occasional use. I used these relatively cheap crimpers and although the results weren't nearly as perfect, they were totally adequate in electrical and pull tests. I'm not sure if it's just my crimpers, but I found that they created more of a diamond shaped crimp which can make extraction much more difficult.
There are a FEW places that require require the soldering of wires. The ideal tool is a bench-top soldering station that has either a temperature adjustment or interchangeable temperature controlled tips. These are probably too expensive ($100+) to just be bought for this project, but if one has access to these they do work great and I generally use 700░F 1/32" tips. The other option is to use a good quality small iron of about 25 watts and these can be obtained for perhaps about $15. The use of a soldering GUN is definitely *NOT* recommended in most cases as they produce way too much heat over a broad area unless they are very carefully teased.
If one is using heatshrink tubing then they'll need some kind of a heat source. I've even been known to use the side of a soldering iron which does work but I don't recommend it. The ideal tool is a heat gun with a curved end piece that directs the flow around the tubing to be shrunk. A regular heat gun can be used, but one has to be careful of what other areas are being heated by the excess hot air.
These can be handy if one is using bolts with nuts to mount various things. Unless one is planning on doing a lot of this kind of work, it's probably best to just find a very cheap set of them.
I find that I use hemostats a lot more than things like needle-nose pliers. The one potential problem is that some of them have very sharp serations on the jaws. A piece of silicone tubing or even tape will quickly solve the problem. Other hemostats have smooth sides to the jaws. Another very handy use of hemostats is for installing adel clamps, especially when two or three are clamped together. The adel clamps can be positioned and closed by hand then a hemostat used to hold it in that position while the other clamps are postioned and likewise held closed. It is then a simple matter to install a bolt through all the clamps, start the nut and then remove the hemostats ... works great for me.
I assume every builder has a lot of these and the only reason I mention these is if one is doing avionics racks they may require a very long philips driver. I also found it handy when working on the avionics racks to have a 90░ driver.
I'd hate to count the miles of wiring I've done in the past,
but until I started doing this project I'd seldom used "aircraft"
wire. Now that I've started using it, I must say "I
LIKE IT" and will continue to use it in other
endeavours. Wherever possible, I'm only using Mil 22759/16
unshielded and Mil 27500/18 shielded wire; this specification
implies extruded ETFE (Tefzel) medium weight insulation, 150 Deg.
C, 600 Volt, tin coated copper. This is the wire that is sold by
the better aviation suppliers and while it's a lot more expensive
than what's available in the local auto/hardware store, its more
than worth it to me as it is MUCH better for our purpose.
Needless to say, I'm converted.
N.B. "Aircraft" wire is a very broad statement and that's why I qualified it with the Mil numbers, but there are many other numbers that are equally acceptable. Places like WireMasters have some online reference documents for those who want to see the myriad list of options.
So what do I like about this wire ... everything. It took me a while to get used to working with it as it has a relatively thin but extremely tough insulation which is very temperature resistant compared with most other wire that I have worked with. Try comparison soldering like-sized PVC and tefzel wire and the results are very obvious; the PVC insulation melts, gives off gas and withdraws while there is essentially no change to the tefzel. Although the wire itself is copper, the strands have a tin coating that resists oxidation and makes it much easier to form reliable crimp connections or to solder onto. At first it felt like a negative when I realized how stiff this wire was, but I now appreciate this feature. This allows one to create wire bundles that are quite rigid and will stay in place, yet they're made up of stranded wire to resist breakage if and when they flex.
The insulation on this wire is normally white but is also available in various solid colours plus in white or colours with various colours of stripe(s). I had thought about using colours as an aid to visually identify various things, but changed my mind on the rainbow approach and initially settled on mostly white with identifying labels on many of the ends. The biggest problem with colour coding is that there always seems to be an exception that defeats the whole original purpose. There are some defacto standards and I've chosen to use these in some places both because of pre-wired components and also because they eliminate some of the labeling requirements. In fact, I'm actually using a lot more coloured wire than I'd originally planned, but just because a wire is white does not imply that it's not one of the standards i.e. I used white not red wire for the landing light power lead due to availability. Coloured wire is very useful on multi-pin connectors, especially higher density ones such as avionics, where even a single coloured wire can usually be used to quickly identify the orientation. Another potential problem with using coloured wire is that unless one has a convenient source for it, it seems like I'm always running out of some size/colour combination and having to order more of it. The standards I've chosen to use in many places are:
Note: The Davtron clock uses an orange power wire which is normally connected to the "always hot" bus (instrument bus in my implementation due to the battery backup version) and my Over Voltage Protection (OVP) modules use an orange wire for their input which they can also short to ground in order to force the breaker to trip. I've used orange instead of red wire for the "always on" ELT remote switch and the Dynon Keep-Alive circuits as well as in some of my electrical tach (with backup) circuitry. Thus orange is essentially used for special power inputs, especially those that can be "hot" with the master switch in the OFF position.
The other potential exception was in the CHR-supplied horn module which has a two conductor shielded power lead that uses white for ground and white/blue for power. Since I've re-wired it's plug to avoid a bunch of splices, I've also changed it to follow my wiring conventions on colours and shield grounds.
Tip: As a quick and dirty approach to colour coding wires when only white wire is available, I've sometimes used 1/8" or 1/4" automotive-style striping tape to wrap a colour band on the wire near the end(s). This is sort of a variation on the use of the type of pre-printed marking labels that just wrap around the wire. Note that on small wire sizes these may require a bit of clear heatshrink to hold the bands in place permanently.
At the risk of offending those with any amount of electrical experience due to over simplification, it should be noted that physical wire size is the opposite of the numerical value i.e. 20 gauge is much smaller than 12 gauge. Once one gets larger than 1 gauge, the wire size goes to 0 and then 00, 000 and 0000 which is sometimes written as */0. Although the odd-numbered sizes are defined, in practice they're not very common. I tend to use the terms "gauge" and "AWG" (American Wire Gauge) interchangeably.
There are two approaches to selecting wire size (i.e. AWG). First there is the brute force way which is to always use a size that is much bigger than required. While this works fine electrically, it just adds unnecessary weight, increases costs and makes the various wire bundles larger and more difficult to deal with. The second and more proper way is to size each individual wire to its requirements. By using the charts in AC43.13, one can identify the required size based on factors such as load, length, continuous vs. intermittent use, bundled vs. unbundled, allowable voltage drop, etc. This sounds like an onerous task, but one quickly creates some "rules of thumb" that meet all the requirements and allow for quick selection using a limited number of wire sizes in order to reduce the number of sizes stocked. In my own case, I'm GENERALLY using:
The four smaller sizes are quite conservative(by about one wire size), whereas 8 AWG is specified as about 45 amps continuous in a bundle or about 70 amps continuous in free air. Since 25 amps is pushing the limits for 12 AWG in a bundle due to temperature rise, I chose to go with 10 AWG for the landing light circuit. Of course there's always a few exceptions such as on the circuits where I've used fusible links; here the fusible link wire size determines the size of the rest of the wire.
Avionics (i.e. radio interconnects, encoder, etc.) wiring GENERALLY uses 22 AWG (or even 24 AWG) wiring as it is carrying low-level signals over short distances. Note that 22 AWG wire should actually be able to carry up to about 5 amps in a bundle.
Anecdote: I was looking at a large breaker panel that had been wired into an experimental aircraft and noted that all of the wiring was 14 AWG even though there was only one 20 amp breaker, a couple of 10 amp breakers, several five amp breakers and mostly 1 and 2 amp breakers. The resulting wire bundles were over 1-1/2" round and very stiff even though the panel was hinged for access and there was limited space for routing. Using properly sized wire would have reduced the bundle size to less than half it's size. After using expensive breakers and more expensive larger wire, the builder then chose to save a few pennies by using cheap vinyl single crimp terminals.
There is also the choice of shielded versus unshielded wire. Generally, anything to do with the radio microphone and headphone circuits, including an intercom and jacks, uses shielded wiring. I'm somewhat paranoid about EMI/RFI so I also tend to use shielded wire for all low level signals (especially if they have long runs) and things like the strobe and the alternator output which can have significant ripple and noise before the battery acts as a big capacitor/filter. Also, governors may be susceptible to noise so I used all shielded wire there as cheap insurance. I have seen wiring where the shield is used both as a power ground and also shielding, but I do not like this technique as I don't believe it is nearly as effective as a shield. I'm only using the shielding for EMI/RFI suppresion and use single-ended grounding, preferably at the load end. The exception to this (I know, another one) is for my ELT remote switch where I've pre-wired it for use with either the current 121.5/243 MHz unit that I have or for later installation of a newer 406MHz unit ... since I didn't have the appropriate five conductor shielded wire, I followed the manufacturer's diagram for the use of four conductor shielded wire.
A slightly different use of shielded wire is on my "always on" power feed for the ELT remote indicator and Dynon "keep alive power". Here I'm using a piece of shielded wire on the short run from the battery input on the solenoid board to breakers mounted on the console switch panel; this is being done primarily for abrasion resistance on the unfused wire and the shield isn't even connected to anything.
As if there wasn't enough different types of wires and choices, I've used one other type. On the frame ground that runs from the frame to the engine, I'm using a heavy flat braided cable. This is commonly available in automotive stores and is designed to allow a lot of flexing without breaking the cable. Since this carries the entire current for the starter and from the alternator, it's important to make sure that the bolted ends make both a good and a reliable connection.
I will also be using thermocouple extension wire to make extension cables for things like CHT and EGT probes which go to a powered gauge. A thermocouple operates by measuring the very small voltage difference developed between it's junction (i.e. the probe) and it's reference junction which is where it transitions to a different type of wire such as standard copper wire. By using proper thermocouple wire to make the extension cables, the reference junction is moved to a known place, such as the instrument case, where the electronics can compensate for ambient temperature fluctuations. The use of this wire is dependent upon the actual instrument being used and the installer should consult the instrument manufacturer if it's not clearly detailed in the installation manual. While shielded thermocouple wire is available, as a minimum one should use twisted pair wiring for extension cables due to the small voltages and the potential for picking up noise. Also note that any connections between probes and thermocouple extension wires can create a dis-similar metal junction which can inadvertantly create a thermal reference joint. Although special connectors are available, I understand that standard connectors can be used IF they are small and both are at the same temperature gradient (i.e. close together).
While technically it's not wire, another wiring alternative is
the use of bus bars which are generally either coated copper or
brass. The actual size can range from small relatively flexible
pieces (i.e. .032" x 1/4" x L) to much heavier bars (i.e.
1/8" x 1" x L) depending on the current carrying
requirements. These bus bars can be used in several different
ways to solve potentially difficult wiring situations and I ended
up using them for the following:
- On terminal strips to form a power (or ground) distribution bus
- For power distribution to multiple breakers or fuses that are in close proximity
- For short, heavy and/or difficult runs between items such as contactors
Note that these bus bars are *not* flexible and should only be used between componenets that are solidly mounted. In the case of battery terminals, bar stock could cause a failure around the mounting posts. Bars from lighter stock can have a bend put in them to absorb any vibration between components (basically an upside down U with legs).
Observation: For those that move their battery to the extreme front of the frame, this requires a fairly long run of very heavy cable. While Tefzel wire can be obtained in these sizes, it is extremely stiff and hard to "bend" around corners. I have seen experimental aircraft use heavy welding cable which is cheaper, quite flexible and much easier to route but I haven't checked into the type of insulation that is used and whether it emits toxic fumes if it is exposed to a flame. Another alternative for those that are anal about weight could be Copper Coated Aluminum wire ... about 3/4 the weight and a lot more expensive.
If one is installing radio(s), a transponder, an ELT or a GPS then they will require some kind of coax cabling and the proper mating connectors. RG-58* was the standard for many years and will work well. However, I'm using the newer (and more expensive) standard RG-400 which has less power loss and is available from most of the aviation suppliers that stock wiring supplies. RG-142 is basically the same as RG-400 but has a silver covered copper covered steel center conductor (i.e. solid) versus stranded silver on copper. There is also an RG+142™ for very high performance applications.
Tip: If one is choosing to use a variety of sizes and colours of wire, it quickly becomes a mess to keep track and store each of them. In the smaller amounts that an amateur builder usually requires, these are normally shipped as loose "coils" rather than on spools. The first problem is keeping track of the actual wire size ... I find it easiest to first double check the markings when the wire is received and then attach a tag to the coil (small piece of poster board on a twist tie). Afterwards it's much easier to just glance at the tag rather than try to read the MS markings. The second problem is keeping all the coils in a convenient place, both for storage and useage. I went to the local pizzeria and got a bunch of new (i.e. clean) boxes which make perfect storage boxes for the various wire sizes.
There are a wide variety of connectors that may be used (or required) in aviation wiring. There is a right way to select and use them reliably or there is a wrong way which is cheap and will most likely cause problems at some later time. To my way of thinking, it is much easier to pay a little bit more up front and do it right once and only once rather than risk having to repeatedly troubleshoot the wiring or have the wiring snagged during inspection causing it to be redone. While most of the connectors described here are required to hook up items like breakers and avionics, I have also seen builders who choose to put removeable connectors at a lot of other places such as wherever two wires must be joined. My personal choice is to minimize the number of removeable connectors since every one of these is an opportunity for corrosion and/or mechanical failure. I'd rather install a butt splice or do a proper solder joint. If I ever determine in the future that I'd prefer a removeable connector then there should be just enough slack in the harness to allow for this.
SOME of the connectors used are:
Ring Terminals - crimp on, sometimes referred to by AMP's name of PIDG (Pre-Insulated Diamond Grip)
I'm only using MS-25036-xxx type terminals which are double crimp types with nylon insulation. Unlike the cheap automotive terminals which have vinyl insulation and a single possibly non-ridged crimp, these are the much higher quality mil-spec terminals. There are actually two metal bands inside them; the inner one is similar to all terminals but has ridges which tend to grip and hold the wire as it is being crimped while the outer metal band crimps onto the insulation to provide strain relief (i.e. extra pull-out protection) and also to prevent hinging (leading to breakage) where the insulation terminates. There are some good illustrations and text located here. Like the various other terminals, the insulation colour denotes the wire size it is intended for and there are a wide variety of hole sizes that should be matched to the task. RED = 22/20/18, Blue = 16/14, Yellow = 12/10. Although these can be crimped with cheap single crimpers if one is very careful and performs the two crimps individually, I'm only using a ratchet-type double crimp tool in order to get consistency and reliability.
Although it is normal practice to only insert one wire into each terminal, it is possible and acceptable to insert multiple wires into a quality crimp connector to create either a daisy chain or for use on busy studs such as grounds. However, it does take more care to properly match the terminal size and being cautious that each wire is fully inserted before crimping. The resulting connection should then be carefully inspected and pull tested. While it might seem logical to untwist the individual strands from multiple wires and reform them as one bigger twisted wire, it is my understanding that its more preferable to just leave each wire as-is. I've done this in a few places and the connections appear to be reliable and strong. I've also chosen to limit this technique to a maximum of two wires per terminal.
Tip: I've used a bit of 24 AWG wire for fusible links and technically this wire is too small for use in the standard red terminals. By stripping the wire about twice as long as normal, the stripped end can be folded in half back onto itself and thus doubling the wire size. I've only done this a few times but all of these connections have withstood a serious pull test.
Butt Splice - crimp on
These are used to join two pieces of wire end to end using a crimp-style connector. While there are a variety of different types, I'll only use the MS-7928/5-x type. These are of the double crimp type with translucent nylon insulation which allows one to see into the window and verify that each wire is properly seated against the stop. Note that I received one order of these from a well known large aviation supplier which were actually of the cheap single crimp type with solid vinyl insulation ... they were given away to someone who likes to use this style for non-aviation purposes. As an alternative to these, one could use solder seal splices, but these are more expensive and require a very careful application of the correct amount of heat to achieve the proper connection. One of the advantages of the solder seals is that they're slightly shorter and contain a heat-sensitive glue that makes the resulting splice weatherproof.
Here we go with another exception. I obtained some butt splices that have the windowed inner metal sleeve but not the outer metal tube for the double crimp. Instead of the normal nylon insulation, they have a heavy double-wall translucent heat shrink tube which extends well past each end. The idea here is that one does a single crimp (preferably using ratcheting crimpers) on the metal sleeve and then shrinks them in place with a heat gun. The double wall heat shrink has an inner "glue" layer that should then make the joint weatherproof. Because the heat shrink is relatively thick, it provides the wire support that normally comes from the secondary metal tube that crimps around the wire's insulation. I'll add some feedback once I've properly tested one of them. I was thinking that these kinds of splices may work well on the required exterior splices, such as the fuel level senders, and would eliminate the need to add additional weatherproofing to standard butt splices. It's my understanding that these were originally developed for the marine industry as weatherproof (i.e. waterproof) connectors.
Update: Although the double-wall heatshrink on these connectors works very well, I found that I had to be very careful with the crimps. I tried my normal jaws (i.e. PIDG) in my ratcheting crimper, non-insulated terminal jaws and also an "automotive" crimper. The non-insulated terminal jaws seemed to work the best but I still found that I had to carefully inspect and pull test each crimp. In my opinion, these splices are much more finnicky than the regular PIDG splices.
These are the familiar blade style connectors and are also available in a mil-spec double crimp type. When selecting these, one needs to select the male or female side in addition to the blade size and wire size. There's also a style where the full connector is covered by the nylon covering and another style where the connecting area is uncovered which is useful in some applications, such as relays, with tight entries. There has been some negative words written by others about these terminals in the past, but when installed properly they can make a very reliable and durable connection. For those that denounce these terminals in aircraft, perhaps they should look at the connections on a Cessna split-type master switch or on a Hobbs meter ... enough said. If one really wants another opinion, they can read this article. One real nice feature of these connectors is that they're relatively easy to install in tight spaces ... certainly a lot easier than trying to get a screwdriver in there and making sure the joint is secure.
These terminals were new to me and are a crimp-on connector designed to allow a connection between two wires to be opened up without tools for maintenance or installation/removal. Both sides of the connection use the same knife terminal and like other terminals, they're generally available in red and blue types for different wire sizes. When new, they seem to make a very solid connection but I'm not sure how tight the connection is if they're opened/closed repeatedly. It's important when making the connection to verify that they're fully seated and since the blades are not insulated they may require heatshrink or some other form of insulation over the formed connection. My dual-point CHT probe for the LASAR ignition system came with this type of connectors and I'm seriously considering changing the Westach CHT probe connectors to them.
Cap - crimp on
Sometimes one finds that they have either an unused wire on a pre-wired device or that they want to pre-wire a device for future expansion. The unused end of this wire should be protected from electrical contact and there are a lot of ways of doing this, such as with heatshrink tubing. There is also a proper crimp-on end cap connector that is eventually just cut off.
Solder Shields & Splices
When working with shielded wire, typically one end of the braided shield is connected to ground to form a drain. On smaller gauges, if a couple of inches of the outer insulation is removed at the end then its possible to bend the wire over where the outer insulation ends and extract the center conductor through the side of the braid thus creating two wires; the center conductor and an unshielded braided wire. On heavier gauges of shielded wire or multi-conductor shielded cables, this is much harder to do and quite often the braid is unravelled and then twisted to form a wire. Solder shields are a simple and neat, but expensive, way to solve this time-consuming process. They consist of a short length of transparent heatshrink tubing with a band of pre-fluxed solder in the center and a band of hot melt "glue" at either end to form a sealed connection. If a drain wire is pre-installed then they're usually called "solder shields"; without the drain wire they're usually called "solder splices". A solder splice can be used to make a solder shield by adding a pre-tinned drain wire, but this is a bit more difficult as there is no pre-formed pocket for the drain wire in the shield and it must be carefully positioned. In addition to butt-splicing two wires, solder splices can also be used to make a "Y" splice where two wires are joined to a single (typically larger) wire. Then again, a solder shield could also be used to "tap" a wire ... there are a lot of possibilities.
To install solder shields, a band of about 1/4"+ of the outer insulation is first removed at about 1"-3" away from the end of the cable and the solder shield is then slid over the the cable such that the solder band is centered over the uninsulated band of the wire's shielding. The solder shield is then heated to the point where the solder melts to form the joint and the shield has shrunk with the "glue" oozing out the ends. The connection to the drain wire is then complete. Simple after you've done it a couple of times, but you do have to be careful that the joint has become hot enough to actually melt the solder and "glue" plus fully shrink the tube. If you're adding your own drain wire rather than a pre-installed one, the end of the wire should be prepared and pre-tinned before inserting it. The better quality solder shields contain a "thermochromatic temperature indicator" which is a fancy way of saying that the solder band changes colour when heated to the proper temperature.
Solder shields and splices are available in several sizes (i.e. diameters) that must be matched to the task. Typically the "-2" size is appropriate for smaller unshielded wires such as 20 or 22 AWG and the "-3" size is suited for slightly heavier wire or shielded 2 conductor 20-22 AWG wires that may be found in the radio / intercom / headset areas. Since each manufacturer has their own reference numbers, this should be double checked.
These are the connectors that are primarily used in avionics but are now finding expanded use in various other aircraft electrical items such as lamp dimmers. They come in a variety of sizes from 9 to 50+ pins and configurations (standard 2 row, High Density 3 rows, gold, tin, etc.) that must be matched for their intended purpose. Since the mating connector is usually supplied with avionics, the biggest difference for the installer is the type of pins or sockets that are used on the individual connector. Originally these only came with a solder cup type tail where each individual wire is soldered into it's respective position. For aircraft use, these solder joints should each have a relatively long piece of heatshrink over them to act as a strain relief. There is also a kind of shell that uses stamped pins and sockets and is typically used in computer applications. The actual connector shell is supplied without any pins or sockets installed and these are first crimped onto the wire and then individually inserted into the shell. If a position is not required for a connection then it is typically left open. These stamped pins should be crimped onto the wire with a properly sized B-crimp open barrel tool.
For avionics use, the more commonly used D-Sub connector is of a type that uses machined mil-spec pins or sockets and this is the type that I would recommend. These are very high quality but present a potential problem as they require a specialized tool to form the proper 4 (or 8) indent crimp. The industry standard is the Daniels AF8 or AFM8 with the appropriate turret or positioner, but these are quite expensive and different settings or positioners are required for each type of pin. There is also a cheaper tool available for just the standard D-Sub pins and sockets that doesn't have all of the features of the fancier tools ... so far I've had good success using one of these but the crimps aren't nearly as "perfect" as the more expensive tools. For those that are uncomfortable working with these pins, most avionics shops will fabricate a harness, or as an alternative one might be able to find an avionics shop or technician that will just crimp these pins if one was to supply them the pre-cut wire and pins. I got a local avionics shop to crimp the non-standard pins/sockets onto pre-cut wire and the 8 indent crimped results were much better than my cheaper tool ... the price was certainly right and in hindsight it would have been just as wise to have them do all the crimps (hint: like many, they have a Friday get together that requires refreshments).
Sidenote: These mil-spec crimp pins/sockets can make an effective one-pin connector without the use of a shell if one uses the appropriate heatshrink over them.
The D-Sub crimp connectors have a special insertion / removal tool for the pins and sockets. I've found that when using #20 or #22 wire that the wire is stiff enough to allow insertion without the tool and one can feel the "snap" as the pin or socket locks into place. Smaller wire may require the use of the red end of the tool to insert the wired contact. The white end of the tool is used to remove a pin or socket and is inserted from the wire end until it pushes back the tabs holding the pin or socket in position. This takes a little practice and its much more preferable to just try and make sure that the orientation and wire lengths are correct before inserting them in the first place. There's some documentation and better illustrations of the tool here.
D-Sub connectors should be used with a backshell which will provide additional support for the wires. These are available in a variety of materials and configurations; straight exit, 45░ and 90░. The metal shells are heavy and I believe just the metalized plastic ones are acceptable for our needs. I also like to use a piece of heatshrink or vinyl tubing around the wire bundle where it exits the backshell to pad the bundle to the hole size and also to avoid any rubbing action directly on the wires. While there are proper rubber bushings available, there are a wide variety of inner and outer diameters ... it seems I never have the right combination for a one-off wire and connector combination.
Another thing to carefully watch out for is the pictorial pin-out wiring diagrams. I've seen these where sometimes they are showing the view from the pin (or socket) side and other times from the wire side of the connector. I've even seen these where they're showing the mating connector rather than the connector being wired. If there is any doubt, the connector shells usually have the actual pin numbers stamped on them beside the pin/socket locations, but these are VERY small and often difficult to see without a magnifying glass. As an example of the this problem; the Garmin diagrams that I have show the connectors from the perspective of how you'd be looking at it as you are inserting wires (i.e. the rear) whereas the ICOM diagram that I have shows the connector from the front side!!!
Tip: When wiring a D-Sub connector, I first verify the orientation and location of the number one pin. I then use a felt pen to put a mark on the wiring side of the connector to indicate this and save any guesswork.
There are several different shapes and sizes of connectors that use the so called B-crimp contacts. These contacts are typically of a stamped material and before assembly they have a couple of tangs or wings that form a "U" shape to receive both the stripped end of the wire and typically another "U" to grip the insulation for a strain relief. The contacts are first individually crimped onto the wire using the appropriate opening in a B-crimp tool and then the contact/wire assembly is inserted into the housing. I have seen various butchered installations of these contacts using just needle-nose pliers and I highly doubt that the result is very reliable in the long term. Even the cheapest B-crimp tool can be used to make effective connections if one goes slowly and makes sure the crimp is being properly formed. When using PVC insulated wire, the strain relief crimp is normally also done as a B-crimp. I have heard that its best when using tefzel wire to just make a circular crimp but this may require the tang(s) to be slightly shortened. In some connectors there is also room for a bit of heatshrink tubing over the strain relief crimp if one is having doubts. For those that are paranoid, I've also seen/used a TINY amount of solder on the crimp to try ensure reliability. Only a very small amount of solder is required and the joint should have extra flux added, otherwise it is possible to have a poor joint or to destroy the contact. The wire should also be pre-tinned. *IF* I choose to add solder, I use a pair of hemostats just outside the crimp on the contact side to act as a heat sink when soldering, especially on the Molex type contacts that have a spring-like action.
The most common use of B-crimps that a builder may encounter is if they use one of the Amphenol Mate-n-Lock connectors or a Molex connector. My ICOM IC-A200 radio uses a Molex connector and I believe this is exactly the same connector that is used on Bendix-King radios. N.B. The ICOM wiring diagram shows the connector from the FRONT side and not the side that you'd be inserting wires from. The sockets used by Westberg are also of the B-crimp variety, as are the ones used by Whelen on their strobe systems.
There are two common types of the male 50 ohm RG58 / RG400 BNC coax connectors that can be used by a builder; the crimp-on style and the UG88U solder/clamp style; the different styles are illustrated here. The crimp-on style should only be used with a proper ratcheting crimper. Perhaps I'm a bit from the old school as I prefer the UG88U style where the tip pin is soldered in place and then the body is screwed together and clamps the braid. I've used these for a lot of years and have never had any problem with them except to remember the assembly order. Jim Weir did an excellent article in Kitplanes magazine about how to properly install this type of connector. I've found that some of the aviation suppliers don't fully clarify which style of connector is in their catalog and even those that do so will sometimes ship the wrong type. After receiving an order of the crimp-ons when I ordered solder/clamp-ons, I finally just ordered a bunch of Amphenol 31-2 connectors from Digi-Key. Because of availability, I also ended up using one crimp-on connector ... Amphenol part # 112514.
Although there are 90 deg. adapters with a male connector on one end and a female connector on the other end, I think that some of these may be of dubious quality for avionics use. There is a 90 deg. connector that is designed to be used at the back of avionics trays that is very low profile and where the coax enters one end and the other end is a male BNC connector. It would appear that Gulf-Coast-Avionics and Pacific-Coast-Avionics stock these as part # 9-30-10 for $20 each at the time of this writing. I'm sure they may be available elsewhere and possibly for a cheaper price, but I haven't searched for them.
Heavy Cable Lugs (i.e. cable of 8 AWG or larger)
Most lugs (terminals) on the end of heavy cable (i.e. 2 AWG) are of the open-ended hoop type that are then compressed against the cable. I've seen a lot of different quality-related things in these lugs; ridged vs. non-ridged, open slot vs formed and brazed, etc. Of even more dubious quality is some of the installation methods I've seen, which appear to be anything from hitting them directly with a hammer to way over-compressing them with what may have been a screwdriver blade or a blunt chisel. Then there are the multiple crimps, cross-wise vs. length-wise crimps, etc. There are a couple of different relatively cheap jigs that should be used with these and contain either a built in stop or a gauge to determine when they've been crimped enough.
On the #8 and #2 cable that I used, I chose to go with closed-end solder-on lugs with dual-wall heatshrink over the connection which makes for an environmentally sealed cable end. Both the cable and these connectors are a bit of a pain to install, but I'm happy with the results. In order to properly heat the #2 cable and connector, I use a small butane torch as it takes a lot of heat to bring them up to soldering temperature, whereas the #8 cable ends were applied using a heavy soldering gun. And yes there is definitely some wicking action, but these cables are already heavy enough that they just about feel like solid wire. Because they're already so stiff, I feel that proper supports on the cable should prevent any hinging action.
Note that its important to start with lugs that have the proper hole size for the cable that is being used. If the cable hole is slightly too large, the cable can be padded out with short "spikes" of copper wire rather than just using more solder.
Also note that the one thing I don't like about dual-wall heatshrink is that the glue acts as a lubricant and it continues to shrink and slide after the heat source has been removed. This is especially true on heavy lugs and cable that retain the heat. The solution I use is to just apply a bit of heat to tubing and watch the results ... if there's not enough shrink then I can add a bit more heat.
Circular or "Cannon"
Although this style of connector is commonly called a Cannon connector, technically they're circular connectors built to certain military specifications such as MIL-C-26482 since they're now manufactured by several other vendors such as Amphenol. Although they're somewhat bulky, these are a top-of-the-line connector that are reliable, weatherproof and come in an extremely large variety of sizes and orientations plus pin configurations and shapes. If you're buying one of these to mate with an existing connector then its important to carefully select the correct mating connector. If buying a new socket/plug assembly, then its often easier to talk to someone who has experience with these connectors or spend a bit of time carefully reading the manufacturer's literature.
None of these are required for a stock Safari, but the one place a builder will require one is if they are installing an R22 governor. My cargo hook also has one of these since it has a connector on the outside of the craft for the electrical release and the hook is meant to be easily removeable. I have also seen some fixed-wing craft use these on their instrument panels so that just one or two connectors can be opened up and the whole panel can be removed. For the Safari, I think this just adds a couple of points for potential failure and I'm choosing not to go this route.
If one is installing one of these connectors for the first time, then I'd recommend to go slowly and fully understand the assembly sequence of the connector. I've worked with some of the solder pin types and they work extremely well, but its important to to use all the parts in order to get a proper weatherproof connector. Sometimes the pin numbering/lettering can be a little confusing, but again if one takes the time to try understand this it then becomes straight forward.
Although I'm not using them, one alternative to expensive circular connectors might be audio XLR connectors. These are commonly available at various audio and electronics outlets and have a locking tab. They're available as both inline and bulkhead mounts, but I haven't checked to see if they're available in weatherproof versions.
The Westberg instruments supplied in the Safari kit have a pin which receives a push-on socket that is crimped to the end of the wire. The sockets appear to just be tin plated and have a piece of loose-fitting vinyl tubing that slides over them after installation. It seems that this tubing needs a drop of RTV, lacing cord or some other secondary retention mechanism to hold it in place. The supplied wiring, at least on the one in front of me as I write this, is E118871 CM/CL3 20AWG. I did a quick search and couldn't find the exact specifications on the wire, but it would appear to be PVC jacketed and I seem to remember that it has teflon insulation on the wires. I'll be making my extension cables from "aircraft" wire and since most of the signals being dealt with are low level, I'll probably use Mil 27500/18 shielded wire.
Wicks part # 183 is a male/female connector to make extension cables for Westach instruments, but at $0.50 for each set this can add up really quick and I assume they're still the plated variety. I'm going to investigate whether I can get quality sockets from one of the electronics suppliers where they can usually be ordered in larger quantities. Perhaps the Molex .093" sockets or AMP Commercial Mate-n-Lok sockets will work and they are also available with gold plating for better reliability. Instead of the Westach supplied vinyl tubing, I'll be using heatshrink to act as insulation, extra clamping force and a strain relief. For the heatshrink on the connections at the instruments, I know how to use a soldering iron barrel rather than hot air to shrink the tubing with minimal heat transfer.
The ends of the wires on my CHT probes have the Westach pin push-on connectors whereas these probes normally have either a ring or knife terminal. I think the ring terminals are really awkward unless one is using a terminal strip which would be buried either way at the back of the interior or on the exterior where it would be exposed to the elements. The other more common installation method for probes with ring terminals is to also put a ring connector on the extension cable and then use a bolt/nut combination to make the connection (note that the stainless star washer goes between the connectors to ensure good electrical contact). I'm thinking that I'll seriously consider changing the Westach connectors to knife connectors. Note that the wires on thermocouples such as the supplied CHT probes should *NOT* be shortened as their length contributes to the calibration. The teflon wire on the other Westach thermistor-based temperature senders can be shortened without affecting their calibration.
STUDS and/or BOLTS
There are several places where one may require a bolted connection to a terminal with a hole in it; battery cables, starter, alternator, CHT probes, terminal strips, etc. It should be noted that AC43.13 (Section 15, 11-185+ in version -1B) has some very good guidelines and tables that deal with bonding and the various materials that should be used for bolts, nuts and washers. When using steel bolts, one trick to obtaining a good bond is to use external tooth lock washers which tend to dig through any surface contamination. The MS27212-* terminal strips use stainless steel washers and nuts to avoid metal incompatibilities.
These can be used for power distribution, tie-in points, test points etc. and are available in a wide varitey of configurations and sizes. Although they are installed on the CHR-supplied solenoid board, personally, I will not use the European style which have a screw exerting a clamping action on a bare wire. I've seen too many problems with their installation such as wire strands sticking out of their entrance and loose connections. I much prefer the terminal strips that use crimp-on ring connectors, such as the common barrier strips or MS27212-* terminal strips. Note that if the bus bar is removed from all or part of an MS27212-* terminal strip then it effectively makes a tie-in or test point.
I debated whether to add extra test points and/or connectors in my various wiring. Eventually I came to the conclusion that the fewer the number of connections that I made then the wiring would be more reliable with less potential points of failure. If one looks carefully at various connectors and thinks about what it would take to install a disconnect at a later time, it is apparent that a small piece of wire might actually have to be removed in order to install the two sides of a connection. In the only places where I envision future changes, such as the ELT, I've left a bit of extra wire in a service loop to make the upgrade task easier. Otherwise, I've generally routed the wire as appropriate with the fewest number of connections possible.
General Use Removeable Connectors
I have seen where some builders install removeable connectors at nearly every place where two wires connect (i.e. instrument sensors to extension cables). While this is very convenient from a maintenance standpoint, it also creates extra potential points of mechanical and electrical failures. I'm choosing to minimize the extra use of removeable connectors and using permanent connections instead. If I later decide to change them, there should be just enough slack in the harness to install a removeable connector (note that in most removeable connectors the physical wire ends are separated by perhaps 1/4" - 1/2"). In the one or two areas where I suspect that I might want to do this, I'm purposely leaving a little extra slack (i.e. a service loop) to make this task much easier. For things like temperature senders, I'm shortening the wires a bit and then crimping or soldering them to extension cables. If the sensor ever has to be replaced, the new sensor will have more than enough length to account for the old cut off splice.
There are a lot of good quality removeable connectors available and this is not a place to save a few pennies. Companies like Molex make a wide variety of reliable connectors that are available in locking varieties and a large assortment of contact sizes and configurations. Gold plated contacts cost a few cents more than tin but are definitely more corrosion resistant. For connections that are outside the cabin area, I would recommend using weatherproof connectors that are available from automotive suppliers or the larger electronics distributors. While one may not plan to fly in the rain, it's hard to predict if or when the craft may be exposed to the elements.
There is an ongoing debate about the use of soldering in aviation wiring. In general, soldering of wires should be avoided in aircraft wiring with the reason being that some solder may wick up the wire under the insulation which then effectively changes stranded wire into solid wire. This creates a hinge point where any wire flexing will be concentrated and eventually lead to breakage. In some cases (such as indicator lamps) soldering is required and these joint should be covered with heatshrink tubing which is kept longer than in electronics wiring; perhaps 1" over the wire's insulation. In addition to supplying electrical insulation, this adds extra support to the wire to try prevent breakage. "Cannon" style circular connectors are not a concern since they have a built-in wire support that prevents wire movement for the last inch or so. All that being said, myself and avionics technicians that I know often use soldered joints to join wires instead of crimp splices. When properly done they provide a very solid joint that is less bulky than a crimped connection.
Good soldering techniques are also important. For those that aren't familiar with this, its easy to practice on some scrap wire and check the results; the finished joints should be shiny without an excess of solder. It's important to start with clean surfaces to be joined. I tend to strip the wire just before soldering and then avoid touching the bare wire with my bare hands if possible. The other mating surface should be cleaned if it shows oxidation and may require a BIT of extra flux if its questionable; I've found the "pens" of flux to be extremely easy and convenient to use. A good resin-core solder (60/40 or 63/37) that's not too thick is essential; I prefer 22 gauge which is about .030" (or 0.75mm) for most jobs and 1/16" solder for heavier applications. When using fine solder, it's easy to feed more of it into the joint, whereas with thicker solder its somewhat harder to control the amount of solder that may melt and puddle into a joint. I prefer to always "pre-tin" both surfaces before joining them by applying a small amount of solder to each of them separately; this also allows one to see how clean the surfaces are and how well the solder adheres. Final soldering is then a simple matter of heating both surfaces up to temperature and possibly adding a little extra solder to fill any voids. If everything is set up properly, this should only take 3-5 seconds and will not overheat the surrounding areas, cause the joint to crystalize or cause excessive wicking.
A couple of other tips for those who haven't done a lot of soldering:
There was a recent discussion on a forum about whether crimp terminals should have a bit of solder applied to them after crimping. After observing a certified avionics technician's work, a lot of reading and various pull tests, I'm in agreement with the rest of the aviation community that the extra bit of solder is not required if one has used proper Mil-Spec terminals and they have been installed with the proper tools. Also, this crimp / solder technique is not described in AC43.13. My one possible exception is for B-crimps where I don't have a variety of special (i.e. expensive) crimpers. This is evaluated on a case-by-case basis and is performed with special cautions (minimal solder & heat sinking) if I choose to go that route.
Soldering is a relatively easy process to learn but it does take some practice and attention to proper techniques. If one can't tell the difference between "shiny" and "dull" or is just plain ham-fisted then it's probably best to get a truly competent friend or just hire someone else who does know the difference to do any soldering tasks. I was recently involved with trying to track down some electrical snags (stuck mic, an open circuit, etc.) on a Safari that had twenty-one hours on the hobbs. I was truly shocked by what we found ... even more so when I found out the supposed qualifications of one person involved with it's build. This work would have failed even the most basic junior high school shop course let alone to be installed in an aircraft by someone who was supposed to be trained and competent. As to whether it was fully tested before being delivered ... I highly doubt it. Some of the things we found with just one switch and circuit:
FUSES and CIRCUIT BREAKERS
One of the common misconceptions about aviation fuses and breakers is that they're there to protect the connected devices. WRONG ... they're there to protect the wire! Whether it's due to chafing/shorting or an improperly sized breaker, when a wire carries more than it's rated current it will start to heat up to the point where it can start a fire. Whether this is the spectacular arcing that some of us have had the [mis]fortune to witness or just a slow heating like a toaster element is the result of several different factors. Regardless, it is something to be carefully avoided and AC 43.13.1B dedicates a lot of space to discussing the proper procedues to select the appropriate wire size for each run.
In general, the aviation recommendations that I've seen state that a breaker should only be reset once while in flight. If it trips, there is a reason which needs to be thoroughly investigated which can best be done on the ground where there is no safety of flight issue. My feeling is that unless a breaker is critical, it's probably better on a helicopter to just land and check it out rather than resetting the breaker ... the only real critical item might be an electric tach or a radio / transponder while in controlled airspace. Because of this, I'm starting to come around to the philosophy that fuses(instead of breakers) really aren't that bad and in a properly wired craft shouldn't present a problem ... I can't remember the last time I've changed a fuse on an auto. Another issue with this philosophy change is that non-critical fuses really don't need to be easily accessible in flight and it should be okay to mount them in places like the solenoid board under the seat.
The principle of melting a piece of wire due to over current can actually be used to build a fusible link which is essentially a piece of lighter gauge sacrificial wire encased in a silicone covered fiberglass sleeve. These fusible links can be used on wires that don't normally have a fuse but should be protected: i.e. the starter switch, amp meter etc. A pictorial and discussion on these devices can be found here. If one is making up a lot of these, they might want to try out a piece of 7453K86 tubing from McMaster-Carr.
The Safari Instrument Kit comes with P&B W31 series combination breaker / toggle switches. These are a great space and weight saving device but I've now heard of a couple of issues with them regarding reliability, both mechanically and thermally. I chose to replace the critical ones (master, alternator and LASAR«) with separate Klixon 7277-2-x breakers and individual toggle switches (AN3021-x / MS35058 series). These circuits are electrically identical but I felt that the change would be more reliable.
I received a couple of electrical items that either contained or recommended using inline fuses. While these work well and I understand the recommendation, I chose not to use them. The problem that I perceived is that these fuses end up being located behind panels and in order to check, replace or disable them, it is necessary to physically remove panels. While these were not flight critical components, I still view this as an inconvenience. I chose to either add these items to existing breakers or to install a new breaker for them. In hindsight would I make these changes again? Probably not but I would try to group all fuses in one general area or fuse block so one doesn't have to hunt all over the place for them. Remember that the builder might not be the person flying or maintaining the craft.
The power feed to the breakers can be supplied by an individual wire but normally a row of these breakers is fed by a bus bar (typically brass). One interesting trick I've seen for this is a U-shaped (actually sort of a flat bottomed V) in which two rows of breakers were fed from a single brass strip. This certainly cuts down on the number of power feed wires required but one does need to be careful about sizing the feed wire for the sum of all the breakers. I chose to group my power feeds in an effort to reduce noise but in hindsight I really don't think this is necessary and it adds extra wiring over just using bus bars.
Tip: If using multiple Klixon 7277-2-x breakers mounted in a vertical plane (i.e. one above the other), you should leave more than the minimum clearance to allow for easier access to the terminals and to avoid any possibility of shorting. Although they can be mounted as close as 1" on vertical center, I would recommend 1-1/8" as the minimum and preferably even a bit more.
Tip 2: I didn't like the idea of two separate switches controlling the landing light circuit (i.e. ARM and ON) and I decided to keep just the ON switch on the collective or cyclic and to use a regular breaker instead of the ARM breaker/switch. Since the largest Klixon 7277-2-x breaker is 20 amps and my switch panel had already been built, I chose to go with a P&B W23-X1A1G-25 breaker which is basically the same physical size as the W31 series breaker / toggle switch that I was replacing. It should be noted that the W23 series breakers use a 3/8" mounting bushing versus the 15/32" bushing on the W31 series so I had to fabricate a spacer. Also, the line and load terminals are reversed on these two series and just to further increase the mounting discrepencies, the W23 series breakers don't have anti-rotation flats or a tanged washer slot.
GROUNDING & NOISE
One of the wiring issues to be very conscious of is ground loops. Although we tend to think of all metal attached to the frame as being a common ground, there are slight variances in it's potential that can induce radio noise and other gremlins. In general, all electrical devices should ideally terminate at a single common ground point. Of course, there are exceptions such as the starter and alternator that internally ground to the frame and it should be okay to ground some purely resistive or on/off items such as the tail lamp and chip detectors to the frame, but all electronics, instruments and other electrically noisy devices such as the strobe lamp should terminate at a single common point (i.e. ground bus). I've added a single ground bus in close proximity to the power bus that will in turn be connected by a single heavy wire (#8) to the frame where the negative battery cable attaches.
Although I'm a firm believer in the single ground bus concept, there are effective ways to distribute secondary ground busses. One of the places where a builder might consider this is for an avionics ground bus which is only used for the radios and intercom. There are multiple grounding requirements for this equipment plus the shields and a ground bus in close proximity to the devices will allow for convenience and short wiring runs with a single heavy wire running back to the main ground bus.
A not so obvious and subtle source of ground loops is from electronic "boxes". Both the Robinson governor controller and the LASAR« controller have their metal boxes internally connected to their negative power inputs. If these boxes are mounted on a piece of metal connected to the frame then they'll create a ground loop. Since I plan to mount these boxes on a metal plate, I'm actually using insulating shoulder washers for their attachment bolts. I still need to carefully check my avionics boxes and decide how to handle their grounding requirements.
Another area to pay attention to is the bonding of various frame elements that are bolted together. After paying careful attention to their wiring, builders often forget that bolted pieces are relying on the electrical connection of two painted/coated pieces joined by a smooth bolt. Needless to say, this can create a less than perfect electrical joint. If one looks at various production craft, they will see short wire or braided jumpers spanning bolted joints and going to their own bonding bolt. This is done to ensure the electrical reliability of the joint and to try reduce the elecrical potential difference between the various components.
One possible solution to the painted surface issue is to use a stainless external tooth washer (MS35335-xx) between the bolt/nut and the painted surface. The teeth of the washer should cut through the paint and provide multiple points of electrical continuity. In areas such as the main ground lugs and the external power plug which require a high current load through a bolt on a painted surface, I've chosen to remove a bit of the paint under the washer and use DC-4 to seal the area once the bolts have been permanently installed.
The problems with using the frame as a ground reference or the use of multiple ground busses is manifested when using sensitive sensors since steel and wire have finite and measureable different resistances which can throw off the accuracy of the sensor/gauge. For the Westberg thermistor sensors, I will be grounding the one lead of the sensors right at the gauge ground in order to minimize any differences in ground potential. I will also be building my extension cables from shielded cable to minimize any induced noise. One unique ground wire quirk that can show up is if one uses an instrument with thermocouples (i.e. CHT and EGT) which are available in "grounded" and "non-grounded" types. In order to prevent / minimize ground loops and to try negate the effect of steel vs. wire resistance, these instruments are normally grounded right at the engine case which allows for consistent and accurate readings regardless of thermocouple type.
Shielded wire is normally used to minimize the effects of EMI (Electro-Magnetic Interference) and RFI (Radio Frequency Interference), both on the lines than can be generating the noise and more importantly, on lines for devices that are sensitive to interference. Lines that generate noise are typically those that have a rapidly changing signal, such as antennas and strobes, plus any lines that supply devices with motors. The headphone and microphone circuits are sensitive to noise pickup as are most devices that use very low level signals, such as some of the instruments. Note that twisted pair wiring for instrument sensors can also help to reduce the effects of RFI and many of the supplied sensors come with twisted pair leads. If one is building their own extension cables then as a minimum they should be similarly twisted. Multi-conductor Mil 27500/18 wire is already twisted and thus has both kinds of protection. I've recently read several articles on the over-use and abuse of shielded wire that's just adding weight and not doing anything useful. While I agree with parts of the articles, I also still tend to use shielded wire in many places since there are other issues such as proximity to noisy wires and devices that aren't considered. A few ounces of extra weight can save a lot of time and effort tracking down an interference problem and then correcting it.
Another consideration for noise reduction is the bus layout. Typically a bus such as the main power bus has a feed line at one end with various wires for devices (i.e. breakers) connected to the bus. When connecting the device wires, one can either just connect them randomly or consider the order of the wires relative to the location of the feed line. Although it may seem counter-intuitive, by placing the wires for the noisy devices closest to the feed end of the bus then the other devices further from the feed will not "see" nearly as much of the noise. The same methodology can be applied to the ground bus. Alternatively, the feed could be placed near the center of the bus and device wire placement could be used to create a "noisy" side and a "quiet" side on the bus.
The use of small capacitors can be used to filter out some noise both on the feed to sensitive devices and when used near a noisy device to prevent it's transmission. For example, I've placed a 0.1uF capacitor on my instrument cooling fan right where the wires enter the fan. One can also place a similar capacitor on the feed line to something like a radio in order to filter high frequency noise and also something like a 47uF or 100uF tantalum capacitor to filter low frequencies and voltage fluctuations. In extreme cases, one may have to resort to adding a small inductor and a capacitor to create an L-C filter.
There are a wide variety of indicator lamps that can be used for signalling various conditions. My kit contained three MS25041-2 indicators which have both a built-in press-to-test feature and a dimming function. For the 12 volt Safari they use a #330 bulb and I chose to add extra matching indicators. Other types of indicators with just a bulb inside a lens assembly or various LEDs (Light Emitting Diodes) can also be used but the dimmer must be based on an external circuit.
One of the issues that comes up when using other types of indicators is how to implement a test circuit to verify that the bulb is working ... there are a couple of different approaches to this. Many implementations just supply a single pushbutton to test all the lamps and verify that they light up. A slightly different arrangement is used on Robinson helicopters where there are pushbuttons on some of the circuits closer to where the sensor is located which also test more of the wiring. Alternatively, some kinds of indicators such as the alternator, governor and low oil pressure really don't need a test circuit since by default they're illuminated when the master switch is first turned on.
When trying to implement a lamp test circuit, a lot depends on how the individual indicators are wired to their sensors. Most of the aviation circuits I've seen are based on one side of the indicator being wired directly to a source of positive voltage (+5V, +12V or +24V) and the sensor then completes the circuit by grounding the other side of the indicator when it is to be activated. Its relatively easy to implement a circuit for this scenario but it does require the use of diodes to allow a single test switch to illuminate all the indicators while each sensor only illuminates one indicator.
A similar technique can be used if all the indicators are wired such that their sensors provide a source of power to the indicator and the other side of the indicators are all grounded.
The difficulty comes in when a panel has a mixture of ground completing sensors and power completing sensors. In this case, one can use a single switch that has a double set of contacts, active electrical components or two separate test switches. Also note that the above diagrams are for basic switch-type sensors and that electronic sensors may require extra isolating diodes. As a result of the diode voltage drop, this technique does cause the indicators to be slightly dimmer when the test switch is used, but it really shouldn't be noticeable.
WIRE ROUTING and SUPPORTS
At first glance it appears easy to just string the various wires between their respective attach points. However, this step should be carefully considered and various paths for wire bundles planned. Sometimes it is better to route a wire over a slightly longer path so that it can get the extra support from a bundle and this is one area where neatness does count since it will make the result much more reliable. When multiple small wires are properly laced together, they make an extremely rigid bundle and each wire (and connector) gets most of it's support from the whole bundle rather than just a single connector and tie points. Although the planning takes extra time it is a very wise investment.
Interesting sidenote: I'm aware of one aircraft inspector who reportedly looks at the wiring as one of the first items. His philosophy is that if the wiring is neat, orderly and properly done by the builder then it's a very good bet that the rest of the craft is also well constructed.
One should try to separate antenna coax cables from general wiring. One should also try to isolate high noise wires (i.e. strobes, P-leads and alternators) from other general wiring, especially avionics circuits and other low level signals. If it is necessary for high noise wires to cross low-level signal wires, then ideally they should cross at 90║ which helps to prevent cross-talk.
Due to pre-wiring connectors and general layout issues, when actually doing the wiring it is quite normal to terminate a wire at one end and then leave the other end to be terminated later. I'm using labels on the ends of my wires but these are time-consuming and a waste for the unterminated end which will probably be trimmed. The solution is to use a small piece of masking tape on the unterminated end and clearly label it as to the wire's purpose. When the wire is finally terminated it can be trimmed and a proper label installed. Trust me, if this isn't done then there will come a time when one starts to scratch their head and try figure out the purpose of an unlabelled wire ... been there, done that and never again. It's amazing how often tomorrow's task gets deferred and it can become weeks before one returns to an area.
Tip - Aviation wire can be expensive and it's amazing how much is eventually used. When making harnesses and initial wiring runs one can guesstimate the required lengths plus a bit of extra while still conserving wire. However, inevitably some wire is too short and must be spliced (which reduces reliability) or replaced which can take a long time once it has been bundled or inserted into a special connector. I prefer to guesstimate the length plus a bit of extra and then still add another foot or so depending on the length of the run. While this certainly wastes some wire, I feel that the absence of splices and not having to re-run wires is well worth the extra expense.
Tip 2 - When laying out a wiring bundle, I find it easiest to first work with any significantly heavier wires (i.e. a #12 and a bunch of #18-22). These are much stiffer, harder to route and will provide support as additional wires are added to the bundle.
On some devices, such as the altitude encoder, one has the choice of using either RS-232 or parallel connections. Most certainly the single (or dual) wire RS-232 connection(s) is/are much easier to install and route than the ten (or eleven) wires for the parallel connection. However, one needs to make sure their various components are compatible both physically and logically before going this route.
When reading various articles, one may come across the term
"service loop". This is usually used in reference to
instrument panels where enough wire is left so that the panel may
be [re]moved after installation to gain access to the back side
of the instruments and doesn't necessarily refer to a true "loop"
of wire but rather just sufficient excess length. The potential
requirement for excess length occurs at all connectors so that
the connection can be opened up and the component removed without
having to cut wires. This also occurs in subtle places like ring
terminals where too tight of a wiring run can make it impossible
to remove the terminal from the stud.
More: I read an interesting discussion about wiring where one of the posters discussed leaving enough wire on each panel connection to allow for three cuts and new terminals to be attached. This certainly would make it easier to change various instrument types or avionics at a later date. I guess there's always a tradeoff between leaving enough free wire just in case one eventually changes a device versus the larger bulk of the excess wire ... it seems that if I leave just a bit of excess and later make a change, I'm usually still an inch short and have to do the splice anyway. Most of my instrument panel connections have enough excess (an inch or two) so as not to create strain but if I changed to a deeper instrument (for example), I'd need to splice the wires. My radio and intercom wires don't have a lot of excess and I can see how that may cause problems if I change these items as much of their wiring is shielded which is much harder to splice properly.
AC43.13-1B details many proper ways of holding wire bundles in place. A simple and safe approach is to always use adel clamps around the wires/bundles and to put enough supports in place that the wires can't touch anything else. The idea being that the wires are only supported by the rubber inserts in the clamps and don't have an opportunity to chafe against any structure. Note that there are two distinctive styles of adel clamps; standard and wedge. The wedge style is the preferred one for wiring as they prevent the pinching of wires in a bundle ... unfortunately my kit from CHR came with only the standard style of clamps.
There are places where its sometimes hard to install adel clamps. If these are identified before painting then its sometimes possible to weld a tab in place to receive the bolt for a clamp. Where these locations are near a piece of aluminum sheet, its sometimes possible to use various combinations of angle stock to create a mounting base for a clamp in the correct location. There are also various shapes of nylon standoffs that can be used to attach a wire bundle to, but again one must be careful about their shape and whether the contact area creates a chafing point that must be protected. I never use the standoffs with the double sided foam tape, unless the tape is removed and an alternate attach method is used, since all the foams I've seen will harden with age and release their bond. If I use nylon standoffs, I prefer the style that has a hole for a bolt or screw so that it will be permanently attached and I also check each of them for any molding flash that can create a sharp edge. I believe some of the Hysol products could be used to "glue" them in place, but I haven't seen the long-term effects of this.
I've seen "nylon" tie wraps used for a lot of things around wiring and some of them raise serious doubts for me. While they are extremely handy to use, they can also cause problems if improperly selected or applied. Most of the cheaper ones are made of a nylon-like material that has no UV resistency and should not be used on the exterior where they will eventually harden and crack. In general, I do not believe in using these to clamp a wire to a frame member. The problem is that the rest of the wire can chafe against the frame member unless it is covered with spiral wrap, mesh or some other form of abrasion-resistent covering. There is a technique where the cable tie is first installed very loosly and a second tie is installed around it and between the frame member and the wire bundle effectively forming a figure 8; this does work, but I prefer not to use this technique. Another alternative is "Figure-Eight" ties (McMaster part # 6705K13) ... not the most common item but they would be much easier to use than the two tie method. If one still chooses to use tie wraps, it is important not to over tighten them as they can crush insulation and also the braided shielding in shielded wire and coax. Crushed coax braid will definitely result in a degradation of antenna performance.
Instead of tie wraps, I've chosen to use lacing cord. This cord does take a few minutes of practice to get used to, but it is a well established aviation practice. Before applying the lacing cord it is necessary to keep the various wires orderly as they are placed in position and connected to their respective ends. One trick I've found is to use the coated wire tie wraps that sometimes come with garbage bags or gardening supplies. These wraps are gently applied to hold the wire(s) in place until the final lacing and the wraps can easily be opened and re-applied while adding the various wires in the bundle. Alternatively, temporary tie wraps can be carefully used and then removed during the lacing process ... I'll admit to using hundreds of these temporary ties. Another trick is to use a cut rubber band wrapped several times around the bundle and then held with a single overhand knot. In addition to the AC43.13 techniques, there's an article here that describes some lacing techniques.
Tip - Lacing cord comes in large spools (mine had 500 yards on it) and it is highly doubtful that one will use the whole spool. While each wrap/knot may only need an inch or two, it's much easier to cut an 10"-12" piece to allow for easy positioning and tying of the knot. I've found that the use of hemostats and/or tweezers makes this easier in restricted areas.
Tip 2 - After tightening the knot in lacing cord there are tag ends that need to be cut off. Instead of using cutters, you can use a soldering iron to "melt" the cord and sever it. This has the added advantage of melting the ends of the strands together and minimizes fraying. I'd suggest using an old tip on the soldering iron since it does leave a residue. I've used this technique a bit but have gone back to using a small sharp pair of scissors and leaving a tag end of about 1/4". The melted ends leave a hard bump and it can sometimes be hard to get the soldering iron into the correct position.
Another bundling technique is the use of expandable sleeving and there is a wide variety of materials, sizes and colours available. Here is an example and some types of this material are available from McMaster-Carr as well as the larger electronics suppliers, Terminal Town, etc. In general, I prefer to not use this material in areas like the instrument panel unless there is a concern about possible chafing. However, it is very effective to use it in places like the wires to the fuel tank probes and tail lamp / chip detector where there is a fairly long run with minimal supports and lots of opportunity for chafing. A small peiece of heatshrink tubing over the ends will lock the sleeving in place.
I've also seen builders use spiral wrap around bundles of wire. The big advantage of this material is that it can be installed after all the ends of a bundle have been connected. Even though my Safari Kit contained a couple of different sizes of this material, I prefer not to use this material except in areas where I know a bundle must be supported against an area of high vibration and then only for very short pieces. Before using this material, one should check whether the individual product is UV and ozone resistant ... much of the cheaper material will quickly deteriorate. The other thing to be very careful with is to ensure that when the material is cut that it doesn't produce a sharp edge that can nick the underlying wires.
As an alternative to spiral wrap, I've also used Tygon tubing. I had one area where there could possibly be a bit of contact between a wire bundle and the frame. I used a piece of 1/4" ID Tygon tubing about 1-1/2" long that I first split lengthwise to allow it to be slipped over the bundle. It was then held in place with normal lacing cord wraps at either end. It seems to work quite well.
There is a kind of tape that is often referred to as "self-fusing" and is typically silicone based. This tape is extremely handy both as insulation and to bulk up wire bundles to prevent chafing, such as the exit of D-Sub shells. However, I now have some serious concerns about it's long-term viability. I recently watched an avionics tech remove a D-Sub shell from a connector that had been in use in a panel for only a year or so and had used this tape to build up and protect the wire bundle. I was amazed to see that a lot of the tape looked to be "squished out" for lack of a better description and that wire insulation was now touching the shell in multiple places. Although I have several different types of this tape, I'm choosing to not use it.
Note that one type of tape that is not in my wiring supplies is standard vinyl electrical tape. In my opinion, this tape has absolutely no use in aviation wiring. The problem is that over time, and especially when exposed to various fluids, the adhesive on this tape starts letting go. One then gets a flopping tag end and possibly unprotected wires. Properly sized and typed heatshrink should be used instead.
Example of rat's nest wiring on a fixed wing's panel ... certainly not the way I'd recommend
- When creating wire bundles, it is much easier if these can first be created outside of the craft, laced together and then installed. Unfortunately, this is usually only practical for avionics interconnects.
- When creating a wire bundle in place, it is best for neatness and minimum size that all the wires run parallel without twisting around each other. This is easier said than done. I've found that if one uses a cable tie (about one click shy of tight) near the common end of the bundle, it can be slid along the bundle and will force it to align. If it is slid about 2 or 3 inches at a time, one can apply lacing cord or permanent ties just behind it. Note that the free ends will do a fair amount of twisting, but this can easily be handled if the free ends aren't terminated.
- Adel clamps work very well to hold wire bundles in place but sometimes the attachment point is not at the correct angle or offset. The solution is often to use a bit of angle stock to make a 90░ bracket. The first time I needed one of these, I fabricated one ... the second time I got smart and made a few extras. I find that it really slows me down when I've figured out how I want to make an electrical run and then have to stop and cut/drill metal.
- At the bottom of the console, I have a fiberglass conduit to route the wires from the side of the radio stack to the hole that is approximately in the center and just under the stack. Each of the bundles was first wrapped in kynar heatshrink and then threaded through the conduit. Once the panel was in place, I still had a few wires that needed to be threaded through the conduit, but it was getting pretty tight. I first threaded a scrap piece of #20 wire through it but knew that tape or heatshrink might not hold any wires attached to it. By stripping about 6" of insulation and soldering on the wires to be pulled about every 1/2" or so, it made a very strong connection. The new wires / bundle were then pulled through the conduit. Since it was getting quite tight and awkward, I also added a new "pull" wire which was left in place to assist in any future wiring requirements.
- I'm using firewall pass throughs that have multiple holes in a soft rubber-like center material. On one side these holes are quite visible and easy to access but on the other side they're recessed in a tube about 3/4"+ and can be difficult to see and find when threading a wire throught them. By first inserting a short length of brass tubing (about 3/32" dia. x 4" long) into the correct hole, the wire can easily be inserted into the tube and then threaded through the fitting. Once a couple of inches of wire have been fed through the tube, the tube is removed.
- Before lacing a bundle or installing the final cable ties, it's best to make sure that all the required wires are in the bundle. It only takes a few extra minutes to add a wire to bundle before it is laced but it takes a lot longer to remove the lacing, add a wire and then re-lace it. This is a lesson I learned the hard way on one of the longest runs that I have ... I overlooked two wires and it took a long time to add them neatly afterwards.
AC43.13-1B (Section 16) calls for the marking of all wires at both ends and possibly other points. If one ever has to try working on an unmarked wiring bundle, they'll quickly realize the value of this extra fabrication step. The method I've selected is to use clear heat shrink tubing that then bonds a paper ID label to the wire. One of the tubing materials I use is clear 3M FP-301 (Flexible Polyolefin, Self-extinguishing, Flame Retardent) which is available in bulk and various ID sizes from Digi-Key and probably many other electronics suppliers. The paper ID labels are printed on my computer in a small bold font and then cut into a strip that is inserted totally within the tubing. The wire is then inserted into the tubing which is shrunk into place. I certainly can't claim credit for this technique and I've seen it often enough that I don't know for sure who originated the idea. One problem with this is that I may be at the hangar and performing wiring when I don't have access to my computer. Although I try to anticipate the requirements, if I don't have the proper labels with me I then use a portable Brother P-Touch« unit to create the required label.
It's interesting to note that I've been using several different brands of clear polyolefin heat shrink (including 3M, Alpha, NTE and Kynar) and some of them are somewhat clearer than others. So far, the closest to true clear that I've used is the NTE brand. Although most of these materials will shrink at least 2:1, it should be noted that they increase in thickness as they shrink and it's best to use the smallest size tubing that works with the wire size being used. There might be a few places where this extra thickness can be used to enhance abrasion resistance. I've found that I'm using mostly 1/16" with some 3/32" and other larger sizes.
For those that want a really professional looking
wiring harness, it would appear that the K-Sun
LABELShop« 2001XLST has the ability to create printed labels
on special heatshrink tubing. I haven't physically seen the unit
or results, but I have seen pictures of harnesses that used this
device and they are impressive. It would appear that the labels
and tubes are similar to the tapes produced on the Brother
P-touch« series which I have used and been impressed with.
Obviously it is not a cheap solution as both the unit and the
tubes are expensive, but I have seen sale pricing from various
distributors on this unit.
Update: It looks like K-Sun has brought out a new lower cost printer (BEE3) that will print on their heatshrink tubing. It can also print on standard tapes that can be used for various placards. Given it's low cost ($79 list at this time) there is no longer any reason why a builder shouldn't get one. This would be a lot faster and more convenient than the paper label inside clear heatshrink method. Heck, for that price CHR should make a deal with K-Sun and include one with each Safari Instrument Kit. Like ink-jet printers, these printers are cheap and the manufacturer makes their real profits on the supplies.
The ultimate solution to labelling wires is to use something like a hot stamp machine which is specifically designed for this purpose and puts a marking directly on the insulation every couple of inches. Unfortunately, these are priced well out of the homebuilder's league and they can cause problems with thin insulation on smaller gauges of wire. However, if one has a very friendly avionics shop then it is possible that they have one of these machines and might consider marking various pieces of wire before installation. Note that only high quality stamping machines should be used since there have been cases where the stamping has weakened the insulation and caused electrical failures.
Another alternative is to use snap-on markers. These work well for adding labels after terminals have been installed on wires since they don't require any kind of sliding over the end of the wire. Personally I don't use these for several reasons: they typically are used in numeric form which requires an accurate diagram to cross-reference the function and I always seem to run out of some number / letter at the most inconvenient time and have to purchase more.
One of the problems with labelling, other than using direct transfer, is that it creates a significant "bump" in the wire, especially on smaller wire gauges. For a single wire this isn't an issue but when one has a large bundle with a lot of terminations near each other, this can create a noticeable increase in bundle size. I've slightly modified my strategy and on some ends where the source/termination is obvious, I've chosen not to label them. For example, the use of standardized colours for avionics wiring means that the function is clearly identified by colour at the connector end. Also, around my breakers all the wires are laced together and a terminal can only go to one place and the breaker itself is labelled on the panel side. The only time this becomes an issue is if the bundle were to be cut apart and a lot of termination screws removed; in that case each of the wires would need to be labelled even with something as basic as masking tape and writing.
The other thing I've done is to strictly follow the convention that the lower side of all Klixon breakers and P&B W31 switch/breakers go to the source bus and the upper side goes to the load. I've also followed this convention with plain switches located on the switch panel where the lower side is the unswitched power and the upper side goes to the load; in the case of the master switch, the lower side is the unswitched ground input. Just to complicate things and make an exception, the single P&B W23 breaker that I've used is marked in the reverse of this so it's lower side is the load and the upper side goes to the positive bus.
This always seems to be the last item that people address and the first item that they forget to update. Obviously every builder gets to decide on the amount and detail of their documentation, but I believe this is very important when dealing with wiring. Quite often I see people deferring it to the end and it becomes another round-tuit that never gets accomplished. While the craft may be perfectly wired for the time being, I find that I can't remember the exact details at a later time when I'm either trouble shooting or wanting to make a change. More importantly, I try to make my documentation such that another person can figure out exactly how I've done it ... what happens if the ship is later sold? The installation instructions usually show a wiring diagram, complete with optional circuits, but I want my diagrams to reflect the actual implementation in this craft.
In the case of wiring, I strongly believe the documentation should be done before the task is started which then creates a simple road-map to follow during the actual process. By making extra copies, each wire can be marked off as its installed. It is then a simple matter of looking at a sheet of paper to see if everything has been done, but this doesn't relieve the need to do a visual check that all the required connections have been made. These documents should also be double checked during the wiring process and notes made of any changes. I'll admit that I failed to do this on an indicator lamp and used my normal #22 wire for this function ... after it was partially laced into the bundle I realized that this should have been a #20 wire since it is fed by a #24 fusible link. Trust me, it takes a lot longer to change a wire than it does to just double check a diagram. The final step of the process is to go back and update the master documents with any subtle changes that were made during the actual wiring ... I don't know about others, but I try to double check the basic logic during wiring and find that I often make subtle changes.
The other thing that preparing this documentation makes me do is to think about the various wiring alternatives and try to pre-plan for future upgrades. For example, I may eventually install a Dynon D10A EFIS system and I've chosen to pre-wire a connector plug for it and install the required fuses and switches. This also made me note that it can take an RS-232 input from the GPS unit which must be switched with the Dynon-PC connection. Also the Dynon unit has it's own altitude encoder and I looked at the various permutations of encoder-transponder-GPS-Dynon to see if it was plausible and appropriate to wire both encoders into the system such that if one of the encoders were to fail then I could make a simple configuration menu change to continue obtaining valid Mode C information while the other unit was removed (possibly at a later date) for repair.
While trouble shooting the electrics in another Safari, another reason for accurate documentation and labelling became obvious. Rather than run the ground wire for the passenger side radio push-to-talk switch back to the common ground bus, the person doing the wiring chose to run it to the landing light switch which also used ground to turn it on. Unfortunately they chose the wrong side of the landing light switch which meant that the PTT switch would only work when the landing light was ON !!! Obviously it really hadn't been tested.
Anecdote: I was looking at a partially-built complex experimental aircraft that had been acquired by a new owner. There were hundreds of wires that had already been routed and installed but the labelling consisted simply of a number on each end with absolutely no further documentation. The new owner is now debating whether to rip out hundreds of hours of work (plus hundreds of dollars of equipment) in order to start all over again or to go through the laborious task of creating a brand new wiring diagram and then trying to identify whether suitable wires exist. In either case, the new owner is facing a huge time penalty simply because the previous owner didn't take the time to prepare and forward suitable documentation.
After going through the wiring process, it's certainly tempting to just turn everything on and do a "smoke test". While this certainly works, it can also be very expensive. My preference is to do a much more controlled and systematic approach to this testing.
Before even considering adding power, the first thing to check is that all screw-type terminals have been properly tightened and that all connectors are firmly seated / secured. It's quite easy to just "temporarily" put a connector in place for routing and then get interrupted and never get around to final tightening of a screw or nut. I'll admit that I found one of these on the double check and I know a couple of other builders who have had very expensive experiences due to this same problem. Since I try to pre-wire as much as possible on the bench, this double check of all screw terminals and connectors is the last thing I do before physically installing a device. If a screw terminal is to receive a second ring terminal or if the connection generally requires future attention, I add a piece of twist tie material or an unclipped tie wrap so that it is obvious at a later time. One can also add a dot of torque seal to studs or screws to indicate that they have received final tightening.
If there is any doubt about special circuits, I tend to test them individually as they're developed and wired. This is especially true for "black boxes" and other custom digital circuits. Even relay-type circuits can get complex and it's often easier to test them on the bench rather than on the craft.
Once everything is ready to test, all switches are turned OFF and all breakers are pulled so as to allow the incremental testing of one circuit at a time. First up is the MASTER circuit and any GROUND POWER circuits. Lights are easy circuits to test and one can slowly progress to the more complicated (and expensive) circuits. Instruments and avionics are among the last to be tested. At each step, only one circuit at a time should be fully tested and if there is any kind of an anomaly then it should be investigated and fixed immediately. Unfortunately, the alternator and charging circuit can only be tested once the engine is started and in this case I'd recommend a thorough test with only the essential circuits enabled ... a bad regulator can instantly destroy sensitive electronics and the charging voltage should be checked before enabling other circuits.
If problems are discovered, then it's usually a matter of tracing the circuit physically and electrically (with a volt/ohmeter) to try find the source of the problem. Each case is somewhat unique and it's hard to give general guidelines except to first follow the power circuits and then the signal circuits. This is one area that a builder may want to seek help if they're not comfortable working with electrical circuits.
<NEXT> Engine and Related Equipment
<BACK> Avionics, Instruments and Electrical
Last updated: September 22, 2011