Every working day starts with us sitting in a briefing room where the shiftleader tells us what the previous shift has been up to and what work needs to be done and who will tackle it.
Today, the aircraft that I was assigned to had come in with a defect. The problem was (boeing 737 pg) that when the flight crew selected the flaps to up, the green full extend light on the overhead panel (LE devices position indicator) and the amber LE transit light on the P2 main instrument panel remained on.

The green light on the overhead panel should have gone off because these lights indicate a other then normal situation (dark cockpit philosophy) and the amber light on the main instrument panel is a warning to the crew that leading edge devices are moving (or not stowed).

The fault had been reproduced on the ground and at that time it was noted that the LE devices (Leading Edge devices = krueger flaps and slats = lift creating devices on the forward side of the wing) were actually working correct and that we were facing with a sensor problem rather than a mechanical problem.

Some colleagues of ours on the line had quickly done some troubleshooting and ordered a rod end switch for us, this switch is in the actuator that operates the leading edge slat (position #3 was faulty, this is a slat and not a krueger flap, see the flight controls section if u want to know more about krueger flaps and slats).
The switch that they had ordered sends a signal to the FSEU (flap/slat electronics unit) to indicate that the LE devices (Leading Edge devices = krueger flaps and slats). There seemed to be some discussion as to how the system actually works.

Some engineers say that if the FSEU detects a fault or one sensor different then the others for a duration of time then it will give a indication for the most extreme situation to be on the safe side.
Other engineers say that we should change the easiest thing first and if that fails, we work towards the more difficult.
Personally I find it best to find out as best u can what had actually caused the problem and try to positively identify that part as a faulty unit.

I went to the books and found this schematic:

Now this schematic may need some explaining.
These are the three sensors that are on the slat:

And the complaint was that these lights remained on while they shouldn't have:

If u know your logic schematic's this tells u that the failed sensor is not the lock sensor (keep in mind that these schematics are simplified and as a result are often incorrect, note that the signal from the full extend switch that goes through the inverter is never high as long as the mid extend sensor is NOT in mid extend, therefore making it impossible to get a transit light on whenever the slats are not in mid extend, this is offcourse impossible and we know what they meant).
By that logic if the failed sensor would be the lock sensor, the amber LE transfer light on the main instrument panel and the amber transit light on the overhead panel would have come on, this is however not the case. The the amber LE transfer light on the main instrument panel did come on but combined with the green full extend light on the overhead panel wich rules out the lock switch and points the finger at the full extend switch.

I have replaced this switch a couple of times before and I can tell u that this switch is an absolute nightmare.
there is no access at all, I ordered the full extend switch from the stores, grabbed my camera and got at it.

This is the panel that I need to take off to get access to the sensors.

When I get the panel off I have a look at the switch but I can see nothing unusual (sometimes obvious damage can be the cause like bent brackets and such).

But I see nothing out of the ordinary.

This picture doesn't quite capture the lack of access but I can assure u that I wil have to work with extention tools, I won't be able to get my hands in there for sure.

At this time the new switch arrives, this is what it looks like outside of the box.

I use my extension tools to remove the bolts that hold the sensor in place.

When the bolt is undone I can remove it with some pliers.

Right, I'm off to a good start, the last time I had done this job this particular bolt had cost me near 10 minutes to remove. This time it cost me close to 3 minutes.

I need to keep changing the tool that I use to get to different bolts that hold the switch.

Right, that's both the bolts out that hold the switch.

I pull the shims out from underneath the switch. If I keep these carefully together I may not need to adjust the distance between the new switch and the target when the job is done and the gap is measured.

After that I remove 2 P-clips that I have reasonable access to.

Now it get's hard, the next P-clip is all the way at the topside with absolutely no access.

The only way to get that one off is to remove all the hydraulic piping and even then it's maybe.
I leave it for now and start to disassemble the other side of the wire loom.

I've cut away all the tie laces that hold the loom together.

This is what it looks like from the bottom.

I decide that I'm not going to remove the top three P-clips that are impossible to reach, I will attempt to temporarily splice the new switch onto the old wiring and pull them all through through the P-clips.
I cut the old switch off (staggered to keep the splices from going through the clips at the same time).

There, the splices are made and I am ready to pull them through.

But unfortunately, it doesn't work. The first P-clip is too tight, I cannot pull the wiring through. If I pull too hard I damge the other wiring, then we are up the proverbial shit creek without a paddle, the other option is that I snap the wire at the splice, in this case we have entered the proverbial world of trouble.
It seems that I am now between a rock and a hard place.

I decide its time for plan B, I need to splice the wires in with red splices, these are smaller splices and are actually thinner then the wire itself. If the splice is thinner then the actual wire, it would slide easier through the clip.
I strip the wire core down to a smaller size and crimp the red splice in.

This is a good shot, the splice that I have in my hand now is a size 26-22 gage splice. The other splices are 20-16 gage splices. U can see that the red one (26-22 gage) is thinner and is actually thinner then the wire.

Then I put some lubricant onto the wire to make it easier for feeding it through the p-clip.
I must have had some good Karma today because the wires fed through with ease now.

At this time my shiftleader comes to ask if it is possible to get the aircraft out of the hanger to get the new A-check aircraft inside so that my colleagues can get started on it.
Having pulled the wires through the P-clip I feel confident enough to agree and send the aircraft outside in one hour and 45 minutes. This means that the time is now a large factor.

These are the old splices, I will cut these ones out because we are only allowed to have three splices maximum in this particular wire and offcourse it's best to replace the old splices alltogether.

I cut the old splices out.

I splice the wires. Once again, allways put the heatshrinks over the wire BEFORE crimping the splice in place.

Using a heat gun is offcourse better then using a cigarette lighter to crimp the splices ;-)

There, the splices are made, now I need to reconnect the P-clips and the new switch.

As usual, I discover that it is easier to remove things than it is to install them again.

After some strugling I manage to do up the P-clips and the sensor.

After that I get to tie lacing the loom up at the splice field.

Time is now up and the towtruck comes to tow the aircraft outside.

I inspect the area once more. The aircraft is towed outside, I close up the panels that I took off and I check if the system works properly again.
The system passed the tests, the defect is fixed and hopefully this sensor will work for a long time to come because replacing this sensor is not my favorite pass-time.
As for now. This job is finished and the airplane is ready for it's flight in the morning.

Back to:

What do engineers actually do?