Fire/smoke alarms - advice please

Seems like you need a smoke sensor/fire alarm which operates only when you aren't in there. I could easily program the smokes on my burglar alarm to work that way, but I don't know if there's a simpler solution. Might also be nice to have them automatically cut the power (except lighting). Such systems are used in computer rooms and doubtless other industrial premises. The power cutting can have a delay so that someone can acknowledge they're acting on the alarm and hold-off the automatic cut, if cutting the power might have significant impact in the case of a false alarm.

I would really expect a kiln to have a thermal fuse.

- according to the manual

the only input to the controller is the thermocouple. No thermal fuse (it's a bit different to a kettle or a washing machine

The connections at the top of the diagram (as in 's', 'op' etc), are connections into the microcontroller board. There's an output to the safety relay from the controller board, but it's not clear what the logic is that drives that output.

I'm probably expecting too much - but when I worked in process control there was an emaphasis on fail-safe, and checking for 'potential fault' conditions (like 'we've had the mains on this element for the last hour and the temp's only gone up by 20c')....

Ionisation and optical sensors are probably of little use as I suspect the kiln can give off sufficient fumes (not a lot are needed) to trigger them both during overnight firing.

The usual solution where hot equipment is involved (big commercial deep fat friers for example) would be to string a wire along the shed roof with fusible links in the wire

above likely problem areas such as the kiln. The wire is spring tensioned with a standing load of 10 - 40lbs so when the link breaks they open a water valve to an overhead sprinkler system or similar. Their great advantage is they don't depend on electricity so if the kiln malfunction trips the workshop power the fire protection still works.

I suppose that's a solution - certainly fits the 'fail-safe' criterion...

Not a lot of fumes from this process - but point taken about the sensitivity of conventional sensors...

Starting to wonder if an alternative solution would be another layer of electronics control - simulating a thermal fuse and killing (& latching off) the power to the kiln in the event of over-temperature...

'Twould mean drilling an extra hole in the kiln for the additional thermocouple - and a box with transformer / contactor / some electronics installed between the kiln and the wall-socket.

The controller checks for thermocouple o/c, and reversed, but not (apparently) for 'illogical' conditions....

Again, probably overkill, but the process control gear I worked on had electromechanical 'watchdog' circuitry in case the uP went awol....

...I'm sorely tempted to reverse-engineer the controller & do it right!

Adrian

I suspect you could do it very simply by nothing more than hanging up a =A33 battery ionisation alarm. I'd expect you're only gonig to get bother from it if you have quite a lot of combustible junk get into the kiln, which shoudlnt ever happen. I'm no expert on kilns though.

NT

We had a dry powder or aff ( memory fade) extinguisher in the forwarder with a plastic hose running around the engine compartment. The idea was that the hose ruptured at the hottest point and the chemical emptied through the hole.

AJH

Arduino, and learn to program it

(Or wait for me to finish mine)

I've got a Pic-Axe development board somewhere about the place - seems there's a stand-alone chip (or maybe two) that translates from thermocouple-to-analog.....

Having got that far - there's a great temptation to turn it all into a controller

Adrian

I _think_ I'd want it to link into the house alarms..... (but I can see ructions if 'something in the kiln' sets the house noisemakers off at 2am!)

Generally, there's not a lot in the way of fumes - I tend to fire on some stuff called Kiln Paper, which produces a slight smell as it fires

- but I guess the answer is to try it & see....

Thanks Adrian

- according to the manual

the only input to the controller is the thermocouple. No thermal fuse (it's a bit different to a kettle or a washing machine - the 'switch' temperature would need to be somewhere around 900c....)

The connections at the top of the diagram (as in 's', 'op' etc), are connections into the microcontroller board. There's an output to the safety relay from the controller board, but it's not clear what the logic is that drives that output.

I'm probably expecting too much - but when I worked in process control there was an emaphasis on fail-safe, and checking for 'potential fault' conditions (like 'we've had the mains on this element for the last hour and the temp's only gone up by 20c')....

Does seem a bit odd that something with a micro-controller doesn't have a watch dog.

I think a fixed temperature alarm, mounted where it isn't going to get a blast of hot air when you open one of the kilns... They trigger when the "ambient temperature" gets into the high 50's C. I don't know how long the temperature has to be above the trigger point before they trigger. This sort of alarm is the type that building regs now require in some kitchens. An oven is a low temperature kiln so one assumes that these fixed temp alarms can cope with the blast of hot air from an oven at 250C being opened. OK 800C is considerably higher but do you open them when that hot?

Well if you want the high tech solution get a computer and webcam, remove the webcam IR filter and run the free ISpy software which includes a flame sense motion detection option :-)

Dead easy for kilns, as you don't care about cold junction compensation.

House alarms are for other people, outside the house. For this, you probably want something independent, with a (smaller) sounder inside the house. Control panels are cheap, but then so are Arduineaux.

Doesn't work though if your ignition source is an IR laser. I'm trying to emulate the Violet Fire system from Concorde, and I'm looking in the UV.

There may be some kind of inbuilt safety-circuit - there's something on the controller board that lives-up the 'safety relay' - but I don't know what the logic is behind it.. Not a watchdog in the 'uP's out to lunch - better stop everything' sense...

Sounds like a plan. It gets hot in there in the summer - but 'hot' as in

I guess it depends on what I'm trying to achieve.... whether I want to know that everything's operating correctly, or that the shed is in danger of catching fire imminently.... I think I'd prefer the first option, with a built-in 'power-kill' relay, if things seemed to be going out of spec.

Having said that, I shall install a new controller board and it'll never fail again, ever ......

Adrian

May be a bit too 'hi tech' - more stuff to go wrong at the critical moment.....? Though I do have a spare webcam.....

Thinking about it - the most likely failure mode (if the kiln locked 'on') would be for the elements to burn out. I 'think' the fairly massive firebrick chamber would likely contain the heat, without getting hot enough to set fire to anything around it....

The Kiln Manual says

------------quote---- ERROR MESSAGES The following errors messages may appear in the display of your controller. Err 1 Error 1 occurs during a ?ring segment programmed to increase the temperature. When the kiln is maintaining the ? requires that the kiln maintain a minimum rise of 12°F (4.8°C) per hour. This usually indicates that the elements or the relay have failed. The kiln will shut off. Err F Error F occurs during a ?ring segment programmed to decrease the temperature. When the kiln is maintaining the ? requires that the kiln maintain a minimum fall of 12°F (4.8°C) per hour. This usually indicates that a relay has failed and needs to be replaced. The kiln will shut off. Err D Error D occurs during a ?ring segment programmed to increase the temperature. If the kiln temperature is 100°F (38°C) or greater above the set point for 15 seconds then the controller registers an error d situation. This usually indicates that a relay has failed and needs to be replaced. The kiln will shut off. FAIL This indicates that the thermocouple has failed and needs to be replaced. The kiln will shut off. tc-- This indicates that the thermocouple has been installed backwards. The thermocouple needs to be removed and reinstalled into the thermocouple block with the red wire inserted in the hole stamped ?-?.

-------unquote---

So I'd imagine that Err1 might/should have caught this particular fault. Something was badly wrong with the temperature sensing. The thermocouple is good because I've swapped that to the other kiln, and it's working fine. Wiring to the thermocouple checks out OK. At process temperature (c. 760c) the temperature display was indicating

The fault's now changed, and the controller is showing 'Fail' - indicating that it can't 'see' the thermocouple.

It would be useful to have the extra capacity of the new (faulty) kiln, so today I shall swap the controllers between the kilns and see if that makes the fault one work. If it does then I shall return the duff controller for replacement under guarantee, and ask the manufacturer a few pointed questions about kiln safety systems!

Adrian

I think the chips I was seeing on Google (Max6675) do all of that for you....?

Adrian