Photo of the oil pressure switch N.O. contact inverting relay box, which will allow delivery of +12VDC to the EI SC-5 clock's engine monitor circuit, which requires logical positive voltage when the oil pressure's N.O. contact closes to ground. Interestingly, in the relay's N.O. position, the other pole of the relay delivers +12VDC output in the N.O. position too, which is inverse from the O.P. warn circuit. In other words, when oil pressure develops, this relay can both turn +12VDC circuits ON via its N.O contact (like the EI SC-5 engine monitor line) as well as turn +12VDC OFF via its N.C. contact (ON with no oil pressure), if needed. Importantly, although the relay is triggered by the B&C pressure switch, it does the latter independently of the critical primary oil pressure warning circuit, which is the only thing on the B&C pressure switches N.C. circuit. The relay is a HASCO SSD-103PU-12 SPDT 3A 12VDC with a nominal 320 ohm coil, that draws a measured 39ma current.

Throughout the wiring of the airplane, I've used several D-sub connectors for connections to avionics, to a custom chassis into which I've put relays and monitors, as well as for inline connections of bundles of wires. I've put together a collection of random notes about D-sub connections that may be useful to someone who's doing some of this work for the first (or second or third) time, which can be found in the Library, Project Documents, Design, Electrical section. Feel free to add comments with other suggestions, tips or tricks you may know about.

Following are some random notes on use of D-sub connectors in the panel wiring:

There are pros and cons to using D-sub connections, however, in my opinion, in most cases, most of the considerations are positive. When connecting to an avionics chassis or custom relay box with multiple wires, there's really no other choice. There may be some controversy when it comes to inline connections. On the down side, any added connection introduces another potential point of failure. However, on the upside, D-sub connections with high quality gold-plated pins are highly unlikely to corrode, and they provide a high denisity connection that makes disconnection, and therefore service & maintenance, much more efficient. Standard D-sub pins will carry 5A with ease, which covers the vast majority of circuits to and from the airplane panel. Although the old-school option is a hard wired or solder-spliced connection, a modern panel may have several complex avionics instruments with hundreds of connections. When (not if) it becomes necessary to remove the panel for any reason, the person who has hard wired all those connections, or even several of them, won't do it again.

Excellent high quality pins, plugs, receptacles and tools are available from B&C Specialties. At Steinair, they have all that and hoods (backshells) too. Mouser Electronics, a major online distributor, has all the above and specialty items, such as weathertight connectors, some with relatively shallow depth hoods (backshells.)  Bob Nuckolls' iconic Aeroelectric Connection web site has short articles about how to crimp D-sub pins and how to extract a D-sub pin from a connector.

When using D-sub connectors for inline connections, such as panel interconnects or disconnects, don't be stingy with the wire. Leave enough extra wire length on both ends that you'll have a chance of not only reaching it from under the panel, but also getting at the retaining screws, in case you need to open the connection for troubleshooting -- without having to remove your whole panel just to get it apart -- and back together.

When testing and troubleshooting D-sub's make a couple of jumper wires about 12" long, with a male pin on one end and a female pin on the other end. Then you can stick the appropriate mating end into the plug you're testing, and clamp the other end of the test jumper to your voltmeter. If you don't use real pins to check things, you'll have a heck of a time getting most VOM's probes to reach the pin inside the connector, especially without doing some damage to the connector and/or pins inside it.

Inline connections require female chassis nut hardware on the female side of the connection. It's highly recommended to also use removeable (blue) Loctite on the threads of the mounting stud of the chassis hardware (not on the actual pin-to-pin or retaining screw connections.). When joining the two sides of an inline connection, screw the retaining screws into the chassis nuts until they are snug, but not too hard. If you tighten too much, you may break the chassis nut's seat against its D-sub plate, causing the retaining screw and female chassis nut to just rotate. You'll have to take it all apart, and retighten the chassis nut side. One source for chassis nut parts is Altex Computers & Electronics.  Be sure to get lots. The parts are tiny, and you're bound to lose some.

The D-sub hoods, or backshells, come in various metal or plastic forms. I prefer plastic. Although the metal ones are undoubtedly more robust, they are very heavy. If the harness is properly secured, and handled carefully during maintenance, there should not be a lot of strain on the connector, and plastic should suffice.

When crimping pins onto small wires, such as 24AWG, be careful not to push the wire into the pin too far. It is possible to get the insulation inside the pin, in the crimp area, and you'll have a weak connection, if any.

When using the small red & white D-sub tool, the red end is used for insertion, and the white end is used for extraction. When inserting wires that are 22AWG and larger, you usually don't need an insertion tool. Be gentle pushing the pin in and it will usually "snap" into place without the insertion tool. With 24AWG wires always use the insertion tool, so you don't apply too much force and break the wire at the crimp connection between wire and pin.

Take care to record and annotate everything on both sides of the connection, including ultimate destination of wires and the wire colors. When using jumpers across D-sub pins, whether inside a custom chassis or on either side on an inline connection, be sure to record which pins are jumpered, and which ends of a jumper has the actual circuit wire. Once the D-sub hoods (or backshells) are closed up, it's hard to tell. If you don't have this data, you may not be able to tell whether you have a jumper problem, or an actual circuit problem in a wire out beyond the D-sub.

A view of the LED annunciator construction. The four on the left are finished, covered with multiple layers of heat shrink to connect the Fresnel lens barrel with the wiring. The two on the right have just the colored Fresnel lens barrels and white spacer rings installed.

This picture shows, clockwise from top, 1) a "hobbyist" LED/resistor/wires assembly from SuperBrightLEDs.com, 2) another one where the resistor and red wire have snapped off the LED bulb, 3) red Fresnel mounting lens, 4) green Fresnel mounting lens, 5) black retaining rings, 6) white spacer rings, 7) one of the fully encased LED assemblies from Radio Shack, which does mount relatively flush, and 8) one of the fully encased LED assemblies from SteinAir. The latter two alternatives are mechanically robust, but suffer from a relatively narrow viewing angle, and the SteinAir model is not flush.

 [...a work in progress, as I plan to add more photos, etc.]

Friend and fellow Rebel builder Terry Dazey was the latest of several people to inquire about details of the LED indicators. My response below covers most of the details of the why, what, and how. Terry asked:

Maybe I missed a link in your web Rebel photo description, but I am curious what is the part number and where did you purchased the small, colored LED annunciator lights as shown on your left wing panel? I like 'em.

Well, that's a story. After way too much cogitation and research, and after trying several things, I ended up with the design and parts I did because I had the following requirements:

A) I wanted a semi-flush appearance, B) I wanted wide angle visibility, C) I wanted a couple of the LED's to be self-blinking, and D) I wanted them to be replaceable, because even LED's do sometimes burn out.

Although these goals were met, the tradeoff, at least as I did it, was to give up getting out-of-the-box-robust mechanical strength, and I had to beef up the parts I has chosen with lots of heat shrink. In detail....

First, the basics: in a nominal +12 VDC circuit (more likely north of 13.8 VDC with an alternator running) an LED must have a resistor ahead of the LED, in series with it, to act as a current limiter. Without the limiting resistor, the internal resistance of an LED is too small to self-regulate at that voltage, and excess current will promptly burn out the LED. LED's come in different colors, which are derived through use of different elements (metals and minerals) in the LED. The current handling (internal resistance) characteristics of the different materials also vary, which means different colors of LED's will often require different values of external resistors. Common values range from 470 ohms to 1K ohms or more.

A bare LED has small stiff wires coming out of a plastic bulb. The wires are fragile, and can be easily broken off where they enter the hard bulb. It's not easy to solder resistors on one of these wires without breaking it, or overheating the LED, much less solder a longer wire onto the resistor. Consequently, manufacturers have come up with various standard, and not so standard ways to mount the fragile LED wire leads, including use of external ceramic "sockets" into which the bare LED wires are plugged, and encasing the whole thing is a larger volume of plastic, etc. In addition, there are different ways the physical mounting or encasing can be built to include the necessary (at 13.8 VDC) resistor. Most amateur builders will not buy bare LED's, but rather one of these kinds of preformed, pre-wired LED assemblies that include at least a resistor and extended wires. [See photo below for views of the parts I used, as well as a couple alternatives.]

I used the preformed "hobbyist" 5MM LED assemblies available in various colors from SuperBrightLEDs. The trouble with those is mechanical -- the resistor is soldered right onto one of the stiff and fragile LED leads, then attached to longer flexible stranded hookup wire. The resistor and hookup wires continue to hang onto the very fragile junction between themselves and the glass "bulb" of the LED -- there is no external casing to cover and connect both the bare lead, and the resistor and stranded hookup wire extensions. Until the LED assemblies I chose are beefed up with multiple layers of heat shrink, such that I could finally enclosed both the mount (see below) and the heat shrink laden wires together as one unit -- the fragility remains. SuperBrightLED's pre-wired LED assemblies include blinking RED ones. (http://allelectonics.com also has blinking ones in other colors.)

Now for the assembly. Looking at the edge of the panel, with the finish side to the left, and proceeding left to right, first there is the plastic, cylindrical (Fresnel) lens, cylindrically shaped with one end open. The lens "cylinder" goes through the panel opening, and snaps in place. On the back side of the panel, a white plastic ring "spacer" is slipped over the lens barrel. It's sole purpose is to take up space. (See photo below) Then the LED assembly (LED, resistor & extension wires) is gently pressed into the back of the lens barrel. Fnally, the black plastic ring retainer is press fit over the wires and the last part of the lens barrel, and clamps the lens around the LED body, locking it in place so it won't slide out of the lens barrel. Additional layers of heat shrink are then added carefully, until the fragile leads are protected all the way from LED body to the extended hookup wires. It's good to order several extra LED's of each color... some may break in this process, and they're not terribly expensive.

The LED mounting pieces came from Mouser Electronics. Representative part numbers are:

Red lens     593-3000R (other colors have a different last letter)
White spacer ring     593-SPC125
Black plastic retainer ring     593-RNG268

All these individual parts are very cheap. Get extras. The biggest expense is shipping. I should also note that it's not advisable to use a plain white LED behind these colored lenses. A plain white LED will tend to wash out the color of the lens a great deal. Better to match the LED color with the same lens color. All that having been said, as it turned out, the only place I needed blinkers was the oil pressure and stuck starter LED's, both RED.

There are several other, more out-of-the-box-robust assemblies available, including ones from SteinAir, RadioShack, etc. Unfortunately, the LED's with hard plastic cases from SteinAir sit quite high off the panel, definitely not flush, and have a limited viewing angle. Radio Shack has some nice plastic-encased LED's that are nearly flush, but they also have a narrow viewing angle. I wanted the co-pilot to have almost as good a view of these lights, especially the annunciator LED's above the left EFIS, as the pilot. That requirement led to the Fresnel lenses I used -- which also happened to lie very flush. The photo below shows a collection of the parts I used, and the RadioShack and SteinAir alternative styles. Of course, there are others too.

This picture shows, clockwise from top, 1) a "hobbyist" LED/resistor/wires assembly from SuperBrightLEDs.com, 2) another one where the resistor and red wire have snapped off the LED bulb, 3) red Fresnel mounting lens, 4) green Fresnel mounting lens, 5) black retaining rings, 6) white spacer rings, 7) one of the fully encased LED assemblies from Radio Shack, which does mount relatively flush, and 8) one of the fully encased LED assemblies from SteinAir. The latter two alternatives are mechanically robust, but suffer from a relatively narrow viewing angle, and the SteinAir model is not flush.

For more info, search on "LED" on the web site. Most of the relevant info will be on the first page of search results.

The left side panel went in today. Very snug with all that wire in back, but it's in. It turned out there wasn't an actual wiring fault type of short in the A/P wiring after all, just two loose wires out at one of the servo cable ends that were touching. They had been "de-pinned" fom their D-sub connector for purposes of stringing wire through wire grommets. The next task is to go through the testing checklist again, making sure all the circuits still work and haven't been dislodged or shorted somehow during the press-in reshaping of the wire bundles. The ELT remote needs to be installed, and a few more mounting screws added, but other than that, it's mounted.

I've decided to install and leave fuses in the circuits that have both power switches and wiring that is completed all the way to the device -- no bare wire ends -- such as alternator switching, E-BUS ALT FEED, EFIS 2 ALT FEED, etc. Those rest of the circuits, those having either unterminated wiring or no switch will have to wait for their respective gear or other loads, for obvious reasons. Getting the fuses in to stay, where I can, will reduce some of the thinking/remembering that will be required to have power where it's needed down the road, and not have it where it may still be unsafe. It will also cut down on wear and tear of the fuse sockets. When I get the last of the testing done on this panel, I'll lace up more bundles behind that I've left open for possible changes, and get this thing mounted where it's supposed to be -- clock and all. Yea!

Moments ago, I finished connecting up the left side panel. Whew! Including both sides of the panel disconnects, there are 126 point -to-point connections. The only unconnected quick tab there in the lower left is the low voltage test port. The three white wires on the right are indicator feeds to the LED annunciators on the center panel. Fortunately, there is very little actual wiring needed to connect the center and right panels, as most connections are via harnesses and plugs that are already done.

All that wire makes moving the panel quite stiff. After bundling the wires a bit more after testing, it will be a chore to slowly get the panel into place, when that time comes. For now, it's time for comprehensive and careful testing, recording the results. Obviously, it's not yet possible to check functionality of all the outlying gear, but I can check virtually all power & switching circuits to their current logical end, along with all the relays and most panel-mounted indicator LEDs. I plan to disconnect the new battery from the battery bus, tape up some loose wire ends in the fuselage and wing struts that may be energized, then run tests using an outboard, variable voltage power supply, with a small 1A fuse placed in each circuit, one at a time. After nearly six months working on the panel wiring, a moment of truth has arrived.

A recent visitor wondered if the all the quick tabs are adequate connectors for this purpose. "Won't they slide off?" The answer is that they are used extensively in commercial, certificated aircraft, because it takes somethin on the order of 200 G's to dislodge them. In my opinion, they are more reliable than ring terminals on screws, which can and do loosen with vibration. Bob Nuckolls has written a short article about Faston Quick Tabs.