[lyner]

I know this isn't a do it yerself repair site but there are some brains out there and someone may have a suggestion.

I have a very basic convector heater (with one of those zig zag elements which you see in all the modern types). On the mica (type) board which supports the element (2kW btw) there is a bimetal thermal safety switch. This has started misbehaving and cutting out a very few seconds after the heater. is switched on. Its cutout temperature has clearly changed for some reason. There is the same free flow of air around it as always - no blockage.

I either have to sling it out or replace the cutout but it will have to be a fuse rather than a bimetal, re-setting device; no problem.
For ~60p, I can get a thermal fuse from Maplin but I can't decide what value of temperature it should blow at. There is a huge range of types available - from about 40C to 200C.
I can't think which one to go for. 40C is clearly too low and 200C could be too high (as it's the maximum available).
Any ideas? Or has anyone had and solved a similar prob?
Educated guesses?
"Measure the temperature" I hear you cry. - I may have to, I suppose, but that has heath and safety implications.

[LeeE]

I think you have to be very careful here because it may not be the thermal safety switch that is faulty; it could be that the heater is actually over-heating.

If you really want to have a go yourself you should remove the switch and test it outside the heater.  With it disconnected from any electricity supply, you could try dipping it in warmed/hot/boiling water to see if it is actuating; if actuates in merely warm or (hand)hot water it could indicate that it's faulty.  If it doesn't actuate in boiling water then it could be ok and the fault may lie elsewhere.  It may also be worth googling the part number of the switch to see if you can find it's specifications.

I think it could be very dangerous to replace it with an entirely different type of cut-out.

[Bored chemist]

"I either have to sling it out or replace the cutout but it will have to be a fuse rather than a bimetal, re-setting device"
Why?

Does the old one have any markings on it that might identify the temperature it's meant to work at?
Does the insulation on the wires to it have anything written on it that might be a temperature rating?

[lyner]

The system is very simple and I think there can't be anything wrong with the rest of the heater. The case is not distorted. The holes (top and bottom) are unobstructed. There are two elements in parallel and they both appear identical - even down to the slightly glowing tips of the sawtooth-shaped sections.  You can see when the contacts open, very soon after switch on. The present cut out is a shiny metal frame which is riveted to the element 'wafer' and there is no number or markings on it. Bringing a hot soldering iron near to the unit makes the contacts spring apart and they go back together with a click when it has cooled down. There is no sign of distortion in the unit, either. 
I, initially, thought it must be the thermostat or timer at fault but eliminated them as suspects by intelligent use of the meter - also the elements are the same resistance.

I think I might well just buy two fuses from Maplin - one medium and one high temp. If the medium temp one doesn't blow under normal conditions, I'll leave it in . If it does blow, I'll put  the hotter one in. I suppose I could test the system with a fire blanket covering the top and using a range of fuses. But that sounds too much like a big job.
that's why I asked you gents for your experienced views.
I can't imagine that the air around the sensor would get much above 100C because it is underneath the element and constantly rising upwards, under normal operation. The element is large and covers a big internal volume so it doesn't appear to radiate much. Once the top is covered, the body would soon fill with very hot air and get to the cutout.

BC "Why?" Because I can't see anything other than one-shot fuses available as spares.

[RD]

Someone may have turned the thermostat down to a low temperature so the heater reaches this temperature very quickly,
 i.e. thermal cut out is ok, but thermostat is set too low.

If you want to measure high temperatures Maplin sell thermocouple sensors which attach to a digital multimeter,
 these can measure temperatures of up to 800 ^o C.  (£3-49, multimeter not included [] )

http://en.wikipedia.org/wiki/Thermocouple#K

BTW I just happen to know that the temperature 1cm above the barrel of a 25 Watt soldering iron (held horizontally) is ~100 ^o C

[lyner]

Someone may have turned the thermostat down to a low temperature so the heater reaches this temperature very quickly,
 i.e. thermal cut out is ok, but thermostat is set too low.

As I wrote, above, I have eliminated the thermostat, timer and elements using a meter. The thermal cut out goes open circuit - and you can see it operating!

I could try the thermometer idea. though, thanks. When I can get down to dear old Maplin. . . .

[RD]

BTW I just happen to know that the temperature 1cm above the barrel of a 25 Watt soldering iron (held horizontally) is ~100 ^o C

I know this because heatshrink sleeving contracts at about 90-120^oC , and  I have used the heat from said iron to perform this task.
 (i.e. a bit of sleeving could be used as a crude temperature indicator: if it has shrunk then it has been exposed to temperatures above 90^oC ). 

[BenV]

I could try the thermometer idea. though, thanks. When I can get down to dear old Maplin. . . .

Strange how, because of the context, I read that as Thermo-Meter, rather than the more usual ther-mometer.

[lyner]

Well ther e'safunnything!

[techmind]

...bimetal thermal safety switch. This has started misbehaving and cutting out a very few seconds after the heater. is switched on. Its cutout temperature has clearly changed for some reason. There is the same free flow of air around it as always - no blockage.

You wouldn't really expect such a basic switch to dramatically change its cutoff point - unless perhaps it got mechanically bent or taken way over-temperature. The other possibility is that the contacts have got damaged/corroded so that the dirty contact of the thermal switch itself is generating heat. You could try some switch-cleaner or very fine emery-paper...?

I can't quite visualise the construction of your heater, but fan heaters and convector-heaters can gradually accumulate dust on the element which although not that visible does restrict the airflow and lead to over-temperature. I suggest blowing the dust out, or very carefully vacuuming the innards - when it's cold, switched off, and unplugged of course.

[lyner]

It's a really bog standard convector. About 2000mm by 1500mm by 150mm deep with the elements stretched horizontally near the bottom. It ia totally clean. The bimetal separates very soon after switch on. It is closed when it starts.
I have actually replaced it with a one shot fuse which is supposed to blow at 150deg C.

I was as surprised as you, techmind; it is too small to be bashed without being bent off the board and God knows what G force would have been required to bend the strip. But I saw the malfunction operate several times. The 'snap - back' action of the bimetal worked consistently but at 'luke warm temperature.
Sorted now! It has been removed and the rivet holes now hold the thermal fuse with screws. If I weren't so mean, I would test it with a fire blanket draped over and then buy another.