Bulent Aliev wrote:

I was just thinking: a sensor from a house smoke alarm positioned at the cooling air exit point, can give a first alert signal? Any smoke has to come out there.
Buly

They're smoke detectors, and the smoke has to reach a certain threshold before the alarm trips.  There won't necessarily be a lot of smoke in a gas fire.  Any smoke that is made will be heavily diluted by the fresh air used for cooling.

There was a recent thread on the Aerolectric list that discussed detection techniques. Infra-red detectors were recommended, and they're supposed to be fairly cheap and will feed into one of those miniature monitors so that you can see what is happening.  In my opinion, it's more weight and complexity than it's worth.
I didn't like the first idea that was offered; run a twisted pair of thin wire all around the engine compartment.  The theory is that a fire will burn through the insulation, shorting the wires and tripping an alarm.  My theory is that the wires would get oxidized and covered in melted plastic before they could short, preventing a good contact and defeating the alarm.  I would modify the design to be a single wire of the thinnest possible gauge, made from that cheap twisted aluminum stuff instead of real copper.  It would always be powered when the engine was running, and if anything cut the wire for any reason an alarm would go off.  I'd have no idea what cut the wire, but I'd know something was amiss, be it a fire or parts leaving the engine.  Either way the alarm would mean, "Get on the ground. NOW!!"  I'm going on the assumption that the death rate on precautionary landings is very low.
--         ,|"|"|,                                    |
----===<{{(oQo)}}>===----        Dyke Delta         |
       o|  d  |o          www.ernest.isa-geek.org  |

--
Homepage:  http://www.flyrotary.com/
Archive and UnSub:   http://mail.lancaironline.net/lists/flyrotary/