Maybe a bit OT, but very strange.
On triggering, the PIR security lamp lit up and immediately went out. I assumed its R7 linear halogen bulb (230v 120W) had blown. I opened the lamp to replace the bulb, but it looked ok. As far as I could tell, the filament was exactly the same as the replacement bulb I was just about to put in. I replaced it anyway as it was 3 years old. The new bulb came on normally when I restored the power.
On examination the old bulb looked fine, but I thought I'd check it anyway with an old Avo Multiminor I had handy. Instead of being open circuit, it had a resistance of around 300K. I checked another new spare bulb, and that was around 50 ohms.
Not trusting the battery in the Avo, I checked the resistance with an electronic multimeter (I confirmed that was reading ok first with a 560k resistor). The first reading on the bulb was around 9M, but it wasn't steady; it increased every time the meter read. Over about 30 seconds it increased by 100k or so every reading. I removed the leads and reapplied them - not realising I had reversed them. I could not believe my eyes - the meter read -400k!!! Yes, "negative" 400k. And, like the first reading, it was not steady, this time decreasing about 10k every reading. When it got to zero, it slowly continued but this time showed "positive" ohms. Note these readings were very variable - perhaps depending on how hard I pressed the probe points on the bulb connectors. One way they could start between 6 and 10M, the other way between -400 and -180k, but they were consistent in increasing or decreasing over time. Repeating everything with the Avo just showed an open circuit either way round. For info, the multimeter applies 0.3v to measure resistance.
Any idea what is going on? Is it possible that some odd sort of thermocouple was created between the tungsten and filament support when the bulb blew?