Thermal protection - High Tolerance 240°C

not clearly understanding what you really want -

Don't you think that a Honeywell"Dialapak" or Fanwel Electronic Indicating/Controlling Temperature Controller 0-300C , Pt 100 ohms + Pt 100 ohms thermocouple(T/C) should do the job.

They have +/- 0.01% accuracy.

In your case you need 2 sets. One to go on/off at 220C and the other as limite to cut off at 240C.

I do similar with;

a) Boiler Exhaust - 1 Set Temperature controller in Panel at 260C & T/C in exhaust to switch off gas burners for safety reasons.(I've 2 pressure controllers , one for on/off & other for high limit). This Exhaust safety is in event the exhaust rises due to tube scalling/failure.

b) Oven/Furnce - 2 sets. One to go on/off with desired temperature and another as a limit (boxed in panel/ office) to switch off the heating system (in case someone did hanky/panky with the on/off).

No, this isn't a controller. The system already has two of them! (main + backup). Thing is, there's radioactive stuff in the enclosure with the vapour, so the system's got to be super-safe. What's needed is a device to break the circuit (in the unlikely event of both controllers failing & applying full volts with no control).

I'm beginning to think the only way is to use a third controller (or maybe even two more) driving a Safety relay to break the primary circuit if it goes overtemperature.

No, this isn't a controller.

Exactly. JohnDG I'm an admirer of yours. I was surprised with your question and was thinking it cannot be JohnDG not to know such a thing.

What's needed is a device to break the circuit

Need to visuliaze to think more since "radioactive stuff".

I'm beginning to think the only way is to use a third controller (or maybe even two more) driving a Safety relay to break the primary circuit if it goes overtemperature.

Might work but unable to visuliaze your system.

Hope you can read this - dashed it out in AutoCAD. Can't show you real customer stuff, or say too much (confidentiallity). Hope this gives you a bit of a handle on the problem.

Cheers, John

I'm beginning to think the only way is to use a third controller (or maybe even two more) driving a Safety relay to break the primary circuit if it goes overtemperature.

For the time being I'll stick with this but send it out to Electrical Engineering, Communication & Electronics & Instrumentation and see if we get a different answer.

Good drawings. I suggest one modification--have the safety relay be double-pole so it breaks both sides of the circuit feeding the heating element (even if it is line-neutral).

Your comment about "radioactive materials" raises one other thing to watch: If the radiation is able to reach the temperature sensing elements (thermocouples, RTD's, or whatever), make sure that they are immune to damage from this radiation (or have some sort of method of verifying their reliability as time passes). I wouldn't want you to have an accident which could be avoided. Since there is a fairly good amount of accumulated industry experience with radiation effects on materials, you should be able to find what you need.

--John M.

Customer doesn't like my suggestion, as it's difficult to get the extra sensor wires out of the heater enclosure. He's going to have another think.

John M. - thanks for the point about relay contacts - I'd probably have got there when it came to the full design - this is just a conceptual sketch.

Your other point about radiation - it's very low level (too low to worry about materials degrading), but the radionuclide is in gas form, so they don't want a fire (or explosion!) which could rupture the containment.

Thanks all for your comments - all ideas gratefully received.

I will assume that your two original controllers are electronic and with a bit of luck, they are PICs or other programmable micro controllers.

You need to have the firmware changed so that each of them has a heartbeat signal (say once every second for example) that comes out on a signal wire and you need some simple electronics that watches both heartbeats separately and if either one fails (the sensing electronics could be something as simple as two 555 in a retriggerable format), the safety switch is pulled when one or the other times out.

The heartbeat can also be used in a further form, one controller must send a heartbeat pulse to the other controller, the other must return a heartbeat within a set period, if either miss, main switch pulled......allow say 3 beats to be missing from either first to allow for firmware interrupts etc to be handled properly....

The original controllers are broadly similar to those mentioned by Ducon in post #1.

It may be unlikely that there could be a firmware change made, but it would be a good idea to ask the supplier if a "heartbeat" already exists and is it accessible.

Some have it already and use a blinking LED for example to show correct working, that is the heartbeat.

Unfortunately, no amount of modification along these lines is going to make the system proof against the "idiot factor". The temperature set-point must be accessible to the operators, who could (for whatever reason) set a value too high for safe operation.

It is a requirement of the specification that an independent thermal cutout mechanism is supplied, which (as for any Safety system) cannot be overridden by an operator.

The customer is making life difficult unnecessarily, he doesn't want extra wire, but he does want safety....

Then the only method I feel is that assuming the sensors that are already installed are "Voltage" sensors, it cannot be too difficult to monitor the original sensor outputs, without loading them up, with a small PIC bases product that has the ability to shut down everything.

To monitor it, you just need a simple "Heartbeat" controller, which as I said before, could be something as simple as an 555 in Retriggerable format!!!

.....AND charge big bucks for it, I would!!!!

Yes, I'm coming to that kind of conclusion (i.e. monitoring the signal we already have).

Problem is, software solutions are (generally) a no-no in Safety applications. Even the couple of dozen bytes it would take in a PIC program would probably take these guys a few months to send out for independent verification - then they'd insist on at least triple redundancy, 'cause I can't see Microchip releasing PICs with any Safety approvals! As I said to the customer, I feel like I'm running through treacle on this project

Sorry, I cannot help further....

Ask Del to design something to do the job......

Thanks, anyway (to all ... )

Designing it is no problem - getting it accepted (and past all the Standards) is!

If I still remember this thread when the customer & I agree on a solution, I'll post an update.

Do that, I would be very interested.

It is a requirement of the specification that an independent thermal cutout mechanism is supplied, which (as for any Safety system) cannot be overridden by an operator.

You learn by experience - Back some 15 years I automated the burner system of an Aluminum Roll treatment Oven and the operators overrid the temperature setting. Next day it was "Molten Aluminium". So I had to put another Controller in the clients office enclosed in a glass box to safeguard.You may need to do the same.