I've got a gas fire with a suspected faulty thermocouple. Symptoms are I can light the pilot but it goes out as soon as I release the knob. I've tried holding the knob down for about 30 seconds with no luck; previously it only used to take 3 seconds or so to catch.
I've taken the fire surround off and had a look at the thermocouple lead and gas valve, and it looks like an easy job. As far as I can see, I don't need to touch any gas piping, so there's no danger of a gas leak. And my understanding is thermocouples are fail-safe devices, so if I don't install it quite right It'll just cut the gas supply off.
But are there any gotchas that I should be aware of? I've heard it's not allowed to DIY gas, but I'm assuming that's only if I'm disturbing a gas pipe or something that will actually be carrying gas.
My plan of action is:
Turn off gas Remove old thermocouple Try to source a replacement Test old and new thermocouples with a blowtorch and multimeter Fit new thermocouple Turn gas back on Er, that's it I hope.
Any obvious problems with my cunning plan, and is it actually "illegal" as some people are telling me?
Don't use a blowtorch you could wreck it. They only sit in a small flame. Thermocouples are "fairly" standard. I am sure someone on here will help you to identify a universal one that will fit and answer your other questions. Realistically, you don't even need to turn off the gas - no thermocouple = no gas flow.
A thermocouple the most likely suspect for these symptoms and I don't think this holds (m)any significant hazards: You're right, if there's anything wrong, it simply won't work, so not much of a risk. The most difficult aspect is probably sourcing a suitable thermocouple. If you're lucky, a 'universal' one should fit nd you can get them in a place like Wickes for a few quid. If you need a specific one, you may need to see a heating supplies place. I probably wouldn't bother with the test step, but just slap it in and see if everything works but, if you do, don't use the blowtorch on a roaring blue flame as it might be too hot. The only way I think it could be construed as 'illegal' (and even then that's not really the right term) would be if you bodged it to work outside the manufacturer's specs, ie to bypass the thermocouple. Then Darwin can come into play before the law....
Before committing to changing, take a look at how the pilot flame plays over the thermocouple. It's common for crud to alter the pilot gas flow and move it away from the thermocouple. Check the pilot, clean up the pilot shroud and if all looks normal then your replacement schedule sounds fine.
The only illegal bit would be if the OP was paid to work on gas, or if he was not competent to do the work
Changing a thermocouple is such a trivial, fail safe operation, it would be truly difficult to show that any serious outcome was a consequence of the OP's actions.
I would be surprised if a universal thermocouple didn't fit.
It's illegal if you are not competent to do it, or if you are doing it for gain (e.g. business), or if the property is rented.
The other thing to check first is that the pilot flame is correctly playing on the thermocouple tip. There's normally a picture in the servicing instructions showing the pilot flame size and position. If dirt has got into the pilot jet or air hole(s), the flame may be the wrong size and/or pointing in the wrong direction, and missing the thermocouple tip.
Note that even if you are competent to change the thermocouple, if you are not generally familiar with gas appliances, you may well completely overlook some other serious fault which would be obvious to someone who is more familiar with them, such as poor combustion leading to sooting up and carbon monoxide production. Such sooting up is often the cause of dirt getting into a pilot light, as described above. So even if you change the thermocouple yourself, that's in no way a substitute for having the fire properly inspected and serviced. Do you have a carbon monoxide detector in the room? If the fire is room-sealed, then it's likely you already broke the room seal in getting to the thermocouple, and that needs properly sealing again.
There could be several reasons why your pilot light doesn't stay on - and a failed thermostat is only one of them. In my experience, a more likely cause is that the pilot jet is partially blocked and/or a lump of crud is diverting the flame away from the thermocouple, and not heating it enough to hold the pilot valve open.
It's not illegal to sort that out yourself without being qualified as long as you are "competent" and as long as you're not doing it for someone else for payment.
I've often fixed similar symptoms on my gas fire by dismantling the pilot jet assembly and cleaning it up, and blowing through the jet itself with compressed air. If you don't feel confident about doing that, it's better to "get a man in".
I'm not sure that you would be able to test a thermocouple with a multimeter. It only produces a few millivolts - so you'd need a fairly sensitive meter. It drives a small current into a coil of hundreds of turns of fine wire in the gas valve in order to provide enough magnetic force to hold the valve open. That coil is another potential source of failure, and is something else which you may need to test. I imagine that it would have quite a high resistance, so your meter would need to be able to distinguish between a high resistance and an open circuit.
Incidentally, I wouldn't use a blow-torch on the thermocouple - a cigarette lighter would be more appropriate.
The solenoid will have a resistance matched to the thermocouple to maximise the power transfer. It will not be of a high resistance consistent with hundreds of turns of fine wire.
Hm, interesting. Looks like the coil resistance is only about 0.05 Ohms. The OP will still have difficulty trying to test the thermocouple and coil - but not quite for the reasons I quoted.
Thanks for that information. I'd assumed that the gas valve would have a fairly high impedance, but it looks like that's not the case. So I guess I'll be limited to measuring voltage with my multimeter.
But if I can find a replacement for ten quid or so, then I'll probably just try swapping it.
For info, a working spare thermocouple here will generate 20mV when placed in the blue flame of a gas ring. I tried it on current too and got 200uA with the multimeter connected across the thermocouple terminals on the lowest current range of my meter (4mA fsd).
The voltage/current rose rapidly on placing the thermocouple in the flame and reached its max after about 30s. Expect a bit less if you are just using the yellow flame of a lighter.
You would at least be able to check if the gas valve coil was open circuit by multimetering it, zero ohms would imply that at least the coils were intact.
The reading was probably limited by the impedance of the meter. The Wiki article which someone cited seemed to suggest that it should be able to drive 200mA (milli-amps as opposed to micro-amps) into a matching impedance.
I suppose that zero ohms could also indicate a short circuit - but that seems less likely than a break. So, if it reads zero, it's probably ok.
No problem, if you are competant and not making a charge for your services. Besides, you are not touching the gas side.
It could the thermocouple, or it could be the fine wire of the solenoid coil which has gone open circuit. I have replaced both in the past, but the coil is inside the valve which would mean disconnecting gas lines.
He can check the thermocouple and the solenoid for being open circuit with a meter and that would be the limit of any electrical testing he could perform.
I suppose he could use an AAA or AA battery to try to hold the solenoid and mimic the thermocouple?
I still feel its best just to change the thermocouple and fingers crossed. Thermocouples are generally the weakest link of a pilot light!
Thanks for everyone's help. I've got the themocouple out now, and on inspection it seems that the tip is burned through. Resistance reads more or less open circuit (several megohms, but not constant).
I'm pretty sure the problem is a dead thermocouple, so I'll be off to see if I can source a replacement.
Each thermocouple junction will only produce a few millivolts (15 to
75 millivolts afaicr from reading some of the literature) but can produce amps of current into a low resistance load.
Assuming a thermocouple voltage of 50mV and a ten turn winding of circa 18 SWG enamelled wire making up the gas valve solenoid with a resistance of circa 0.05 ohms, this would get us an ampere turns value of 10 which should provide ample holding in force when the brass shim used to prevent residual magnetism from making the solenoid stick is only a mere 5 thou' thick as opposed to the more typical 20 thou' brass stud thickness on a typical telephone relay.
Afaicr, a typical telephone relay needed a few hundred ampere turns worth of magnetic force to operate but would hold in to a few dozen ampere turns worth of magnetic force even with the larger non magnetic gap.
In this case, the solenoid is designed for hold up, not operation from the fully released condition since the 'operating force' is provided by manaul operation, not magnetic.
The big problem with trying to test such extremely low voltage, low resistance highish current circuits (and amp or so at anywhere from 10 to 50 millivolts) with a typical multimeter (whether analogue or digital) is that when it comes to measuring current, the current shunt in the meter is designed to drop the 200mV required by the ADC to show full scale. You might do a lot better to choose the 10 or 20 Amp scale (usually a different positive lead socket) and hope to see a 0.5A reading when shorting out the thermocouple with the meter (20 milliohm shunt for a 10A scale with a 200mV ADC meter). Trying more 'senstive' milliampere settings with such meters will just display lower milliampere readings due to the lack of voltage from the thermocouple.
A more sensible way to use the meter is to put it on the volts range and measure the millivolt output and use the meter's low ohms scale to check for the very low thermocouple and coil winding resistances which, for all intents and purposes with an ordinary meter will look indistinguishable from a short circuit. However, when testing the thermocouple, repeat the test with the meter connections reversed to be certain that any small thermoelectric potentials aren't causing a false over-read of what should look like a short circuit (the reversed connection, in the presence of some thermoelectric effect will likely give "negative resistance readings on both types of meter alike).
Well, with such extremely low resistances, it will make some difference so it's best to avoid having to coil up a yard or two of excess co-axial pyro cable on a 40cm run if you can possibly avoid it.
As regards gas safety, since the connection to the gas safety valve is purely electrical in nature, provided you don't have to disturb any of the gas plumbing, replacing a thermocouple is entirely free of the risk of creating a gas leak hazard, assuming the gas pipework and connections are still in good condition to begin with.
Obviously, if inspection of the job suggests that there might be a problem in the gas pipework itself, you'll choose to call on the services of a gas safe fitter to service (or condemn) the fire rather than take any chances. More often than not, the gas pipework and joints are typically in excellent, if dusty, condition. Just be prepared to "Play Safe" if you have any doubts.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.