Using Cortex to provide external temperature to boiler logic

[stachoo]

Hi everybody
I would like to use use temperature information from my Idratek external temperature sensor (LTS) to as an external temperature input to my boiler (ACV Prestige 50).

Normally my boiler should get external temperature reading from a resistive NTC temperature sensor (12KOhm at 20 degrees) and use it to adapt the circulation temperature of the radiators circuit to the external conditions.

Does anybody know a way to provide such output from Idratek or should I simply connect a regular sensor directly to the boiler?

Maybe an Idratek light dimmer (SLD/QLD) could be used?

Thanks for Your input

[Karam]

If your boiler system is expecting a connection to a NTC thermistor of a particular type then it will internally have some way of linearising the characteristic of that thermistor. Different thermistors have different non linear characteristics. So one theoretical problem will be how to create the appropriate non linear signal for the boiler input. Secondly if the boiler is expecting a variable resistance then you may have to mimick a controllable variable resistance (perhaps via an analogue optoisolator). Connecting an analogue voltage output from an IDRATEK QAO directly to the boiler may be feasible if the boiler is capable of accepting a voltage input but you'd also have to be careful about issues with the 0V reference between boiler and the IDRATEK system. You'd have to ask the manufacturers about this method. Assuming it were possible to connect a QAO output to the boiler either via an isolator or directly then you'd have to create a mapping table to convert different temperatures into appropriate resistance/signal values (assuming you knew the expected characteristic). This perhaps could be done by using multiple temperature threshold triggers.

It may be that just very crude knowledge of external temperature would be sufficient to at least improve matters between winter and summer. In which case you might just use a couple of relay outputs connected to a suitable optoisolator together with suitable drive current resistors in such a way as to switch the output between 4 resistance values known to represent 4 temperature points. Maybe even 2 would be sufficient.

So in theory it may be feasible but in practice it would probably be much easier to just fit a NTC thermistor.

[Andrew Millne]

So in theory it may be feasible but in practice it would probably be much easier to just fit a NTC thermistor.

This sums up why I love Karam's posts, he always gives you the full (very full) picture.

[Karam]

.. A bad habit which I'm evidently failing miserably to curb