Chimney flue temperature

I am interested in measuring the flue temperature at the top of our house chimney. I propose to record temperatures on a micro computer (BBC MicroBit) analog port.
If I use a thermistor, on a long twisted pair copper wire, would the results be affected by the temp of the copper wire changing, or would I need a resistance bridge or something. The MicroBit ports run at 0 - 3 volts (Max 5v).
What would be the best sort of circuit / components for this application?
BobH

Look up Wheatstone Bridge......

And also look up thermocouple!

Mark

Interesting one. I don't think running a long line to a thermistor is a good idea because it will pick up mains hum.

I'd suggest putting the Microbit in a box on the chimney and use the long line to power it, not to measure small voltages. I've not used one but the Microbit is fitted with Bluetooth. As Bluetooth is a serial radio link it should be possible to transmit the thermistor measurements to a computer in the house. I've done remote signalling with an Arduino and a Bluetooth module.

The thermistor needs a stabilised power source. The microbit has a 3V output that could be used via an external resistor to power the thermistor. The value of the resistor would depend on the thermistor used. I couldn't find a specification of the Microbit's ADC: measuring small voltages with it might be tricky, and you may need an op amp to beef things up.

Dave

0.25mm diameter (i.e. thin) copper wire is (according to the IEC) about 400 ohms per metre so even if you need a 50 metre there and back run the total resistance will only be 40 ohms i.e. less than 1% compared to your thermistor assuming it will be at least five thousand ohms. It's drift with temperature will be even tinier so don't expect it to significantly affect the accuracy compared to other factors (such as how accurately you place the thermistor in the plume)

Beware of using a low-resistance thermistors, these have a significant self-heating effect even when run off only a few volts, higher resistance ones ( a couple of tens of Ks) are significantly less prone to such errors. Something like the microbit will be quite happy taking readings from a high impedance source as, presumably, you are talking seconds between readings, not milliseconds.

Neil

<edit> seeing the comments above about hum etc. the simplest answer is  a twisted pair should stop any mains hum issues and you can also add a capacitor to smooth the signal, but use a low-leakage type. You can also filter the signal digitally by taking the average of any arbitary number of readings (make it a power of 2, for example add 256 readings then lose the lowest byte) - it may seem wasteful, but the 'cost' of multiple readings when you only need a results at a large time interval is effectively nil.

Edited By Neil Wyatt on 19/01/2017 20:27:10

Frances
I run a log stove, and I am curious as to how much heat goes up the flue or to be more precise, what is the exit flue temperature compared with the bottom of the flue.
Dave
The display is part of the MicroBit, so I didn't want to climb the chimney to see the readings.
The following link may show the ACD properties;-
**LINK**

The appropriate technology for this is 1-wire. There is probably some support software already written for it on the microbit.

Get a laser thermometer. Cost from $10 - $30. I bought one which measures temperature up to 550 deg C. Just aim at the target and read the temperature. No wires required.

Paul.

Edited By Paul Lousick on 19/01/2017 21:06:07

The display. Understood. What I'm suggesting is that you ignore that and program the microbit to broadcast the thermistor readings over a Bluetooth radio link. If you have a Bluetooth capable laptop or tablet you can "pair" with the Microbit wirelessly. A bluetooth link appears on Windows as a com: port to which you can read with a Terminal emulator like putty or hyperterminal. In effect you transfer the display to another computer.

ADC Link. Yes it's there! The microbit ADC is in the range 0 to 1023, I guess across 3V. That is it reads from 0 to 3V in steps of about 3mV. That should be OK for a thermistor & resistor (I'm theorising!), for example this one from Maplins has a resistance of 7.73k at 100C, 150k at 20C, and 534.9k at 0C. The max temp is 125C

I've no idea how hot flue gas from a domestic chimney gets. For what it's worth, I read somewhere that smoke from the funnel of a Dreadnought Battleship emerged at about 400C. It could make working in the spotting tops very uncomfortable!

Dave

Before choosing what type of sensor you really need to know what the maximum temperature might be at the point you are measuring.

If its less than 200C the most accurate would be a PRT (Platinum Resistance Thermometer). A 3 or 4 wire PRT100 can be wired with microscopically thin wire and as long as all wires are the same length remains accurate.

Over 200C a thermocouple is better and more likely to be robust.

The high temperature capability of a thermistor will depend on how its packaged so its range is similar to a PRT, but its a much lower cost. Thermistors are very non-linear compared to PRTs and thermocouples and you might need to use some maths or a look up table to display accurate values.

Solid state sensor will be encapsulated in plastic so that limits the upper temperature but if you use an intelligent sensor it will give a ready calibrated digital output without any interfacing or analogue circuitry.

PRT and thermocouple signals need analogue circuitry but in general are the most accurate, however they will end up more expensive than other sensors.

I no longer have a wood burning stove but I have a feeling that the flue exit temperature will be in hundreds of degrees and a thermocouple may be the best choice.

Ian P

I agree with Paul above.

**LINK**

£14.99 from Maplin.

Hi Bob,

I am interested why you would want to measure this temperature and what would be the relevance?

I live in an old stone cottage. The large stone fireplace has a multi fuel stove fitted for the past 12 years with a 5 or 6" dia stainless liner fitted above the stove. I usually burn coal in the stove.

I would expect on mine the exiting (from chimney top) gases of combustion to be quite low in temperature unlike a miniature steam loco.

On the few occasions I start off with a pure wood fire there is often an inability for the liner to draw leading to smoke in the front room, and frantic opening of windows downstairs. I am sure this is due to the lower temperature of wood burning, and inability for the smoke to rise due to temperature drop as it goes up the liner.

Once the stove and liner warms up, the stove performs very well.

I dont think I get any down draughts were I live.

Cheers,

Julian

Amount of heat? You will also need to know the flow rate - ie the amount of gas - and salso some heat loss going out of the chimney will relate to the latent heat of water. Wet logs need a lot of drying and all that heat of converting water into steam adds up.

It's about 400 ohms per 1000 metres isn't it Neil?

John

Why am I curious about exit temperature, because I AM !! I draw off heat from the lower end of the stainless flue pipe (150mm dia) and circulate that around the house. This is done by having another stainless tube 210mm dia around the lower end of the flue pipe. Air is drawn between the two pipes by a fan which sucks, then distributes via insulated ducts around the house. I test for CO at each outlet. But I would like to know if this has much affect on flue gas temperature and could I lengthen the "air heater"
BobH

Sounds interesting and though not model engineering it is 'making stuff'. I have often wondered about this as a way of triggering a re-fuel alert. As you suggested at the beginning a bridge type arrangement would balance the copper loss and allow calibration before installation using the maplin thermometer.

Julian, your problem IS a downdraught caused by a cold night chilling off the chimney while the house is still warm. Burn dry paper rings for a few minutes with the door closed and bottom air only. An engineering solution might be an adaptor for a fan (hairdryer) to push air in the bottom vent with other vents closed.

If I was starting from scratch and building the whole lot myself I would use a 4 wire Platinum resistance thermometer with a local interface chip to either Bluetooth or wired serial comms with the wired connection out of preference because that way I would not have to change batteries.

I'm staring at a project I have on the bench at the moment which uses just such a set up. RTD to MAXIM INTEGRATED PRODUCTS MAX31865ATP RTD interface chip with serial output to PIC18F45J10 driving a DDM4 4 digit display.

If you were doing it at home without all of the kit I have available at work you should look around for stuff that already does what you need functionally and hack it about.

Our lab building does cleaver things with inlet and outlet air (the building is essentially sealed) We have 4 towers each containing 3 pairs of ducts carrying inlet air and exhaust. There is a chamber containing a very large thick disc constructed like a honeycomb. Think of fitting a large number of drinking straws into a section of pipe. The disk is arranged to have part of it's flat surface in the air path of the inlet air and another part in the exhaust. A motor driven spindle causes the disc to rotate slowly. So hot exhaust air heats part of the disc and is cooled before exiting the building and the now hot part transfers the heat to in inlet air as it rotate into that air stream. We get around 8 to 10 degrees temperature change in the inlet air on a good day which amounts to a big energy save.

regards Martin

Oh and it works 'tother way about in the summer.

We also have 4 deep well heat pumps. Good innit.

Martin

Most folks don't realise that the funnels were generally double skinned and vented so that air goes in at the bottom and out the top to cool the outer layer.

Even with this it could be unbearable to work near the funnel on some ships when they were driving the boilers hard.

Neil