ATA 26-18, Duct Overheat

767-200 with a right wing duct overheat system that would not test.







The duct overheat systems (Left, Right, Body) use detector loops and/or overheat switches that provide a ground signal in an overheat condition.



During a system test, integrity is verified by providing a ground signal that must pass through the entire length of loops and wires.



Signal path during test.




This aircraft had a broken wire on a terminal at one of the overheat switches in the pylon. The switch(s) only use two connections. A single wire hooked to ground and a double wire connection used in the loop path. If the single (ground) wire was broke, the system would test normally, but that switch would have no ground path for warning if it happened to overheat. The broken wire was on the double wire connection. With the signal path open, a proper test could not be confirmed.





The PRSOV (Pressure Regulating and Shut-off Valve) is located in this area. It is VERY difficult to remove this valve and it is almost certain that the terminal in question was broken during a valve change.

A note about the functionality of these systems: Even with a broken loop path wire, in an overheat condition, a warning will still be shown. Each end of the loops are connected together at the test switch. The card only needs to see a ground. That ground can come from either side of a broken loop system. This same theory is used on engine fire systems.




767-200 WDM 26-18-21

How did we approach this problem???

Pulled the plug behind the test switch panel. Grounded the wire going to card. Received both the overhead annunciator and EICAS message. (Card working properly)

Shot switch on the panel itself. (Switch working properly)

Shot each end of loop from panel plug. (15 Meg Ohms - Open loop)

Instead of dropping the underside leading edge panels to expose the loop connections themselves, the switch area was looked at first. One quick visual of the terminal in question revealed what the problem was. It was verified by leaving both terminals off the switch. A jumper to ground on the good terminal wire gave the warning. No warning was seen on the bad terminal wire. Some insulation was scraped through and the jumper attached..... received the warning.

New terminal installed............. Problem solved.

Excellent work Mark!

Same concept used for "Wheel Well Fire Loops"
Only one aircraft checked, but I had about 330 ohms across pins 15 and 16 at fire test panel.
This value is "way" out of tolerance.
Total loop resistance should be less than 20 ohms.