Thermocouple / millivolt gas valves - how do they work?

[DTJ]

Just out of interest I was reading about thermocouple or millivolt gas valves as used in gas hot water systems and domestic gas space heaters.

Details on how they work is a little elusive and gets buried in all the domestic air conditioning FAQ pages. I looked at a gas heater some time back and the over-temperature sensor was wired to the under temperature (pilot) sensor. I guess they wire them in opposition for that to work.

It seems the thermocouple is more likely a thermopile and the current it generates operates a  solenoid coil.

Quoted voltages are anything from 20mV to 800mV.


Any ideas of the current level involved or other details.

[mzzj]

There seem to be few voltage levels, this pdf gives some resistance numbers and from there you can assume some power and current numbers available

http://library.coburns.com/specs/CATALOG_Honeywell_Q340A1074-U.pdf

It lists thermopiles and thermocouples separately
Thermopile with 750mV output  and ~3ohm resistance would make 250mA short-circuit current or 125mA @ 375mV @3ohm load so you have about 45mW max available for the valve. Suprisingly little for electromechanical valve!

[DTJ]

Thanks for the link. They are interesting little devices. I'm impressed the valves can operate on so little power.

I'm surprised how little info there is around on them considering how many are used about the place.

[ArthurDent]

A lot of gas stoves use 2 heat generating thermo sources. The thermocouple is in the pilot flame which must be lit for gas to enter and if the flame goes out, this low(20 millivolt) signal goes to zero and the main solenoid valve closes cutting the gas supply off, which is a safety measure. The link below is a video of how it works.

A thermopile is a bunch of series connected thermocouples that generates maybe 800 millivolts and this is also in the pilot flame. The thermopile controls another valve that is in series with a switch type thermostat to turn the main burner on and off to control the temperature output of the device.

[Renate]

I too have been amazed at how little electrical power actuates these switches.
I don't know exactly what pressure your valve is using, but my propane uses 11" of water, which is hardly any pressure.
The funny thing is, the electric gas valves in an RV (for fridge, heat, hot water) take about 3W (12V, 48ohm).
On the other hand, my oven has the standard thermopile valve in it.

[soldar]

I too have been amazed at how little electrical power actuates these switches.

Not really. You open the valve by pushing with your finger. Then the solenoid only has to hold the valve open. 

I have replaced quite a few of these. My water heater went without one for over a year until I found a replacement. In the meanwhile I just put a wall wart with a resistor. 

[Renate]

You open the valve by pushing with your finger. Then the solenoid only has to hold the valve open.

That is true for the pilot.
But there are furnaces that uses 750 mV for the room thermostat.
That does have to switch on and off.

[soldar]

That is true for the pilot.
But there are furnaces that uses 750 mV for the room thermostat.
That does have to switch on and off.

I have never seen a thermocouple singlehandedly control a thermostat. Thermostats either were bimetallic or have electronics.

[Renate]

I have never seen a thermocouple singlehandedly control a thermostat.

Huh? I don't know what you're saying.

I'm saying:
On the wall a bimetallic thermostat with a knob for the occupants to control the temperature.
The open circuit voltage is 750 mV across this thermostat.
It is being fed by a thermopile on the pilot of a gas furnace.
An electric gas valve is in the series circuit.
https://www.honeywellstore.com/store/products/750-millivolt-heat-only-thermostat-ct33a1009.htm

[soldar]

I have never seen a thermocouple singlehandedly control a thermostat.

Huh? I don't know what you're saying.

I'm saying:
On the wall a bimetallic thermostat with a knob for the occupants to control the temperature.
The open circuit voltage is 750 mV across this thermostat.
It is being fed by a thermopile on the pilot of a gas furnace.
An electric gas valve is in the series circuit.
https://www.honeywellstore.com/store/products/750-millivolt-heat-only-thermostat-ct33a1009.htm

OK, let me see if I can explain better.  A thermocouple does not have the power to directly open a normal gas or water valve.  AFAIK power always has to be supplied externally, either by electrical power or by the pressure of the fluid itself.  If there is no electrical power being supplied then it is probably a diaphragm valve with a tiny pilot valve where the pressure of the fluid itself is what actuates the valve. If there is no fluid pressure the valve will not open. The thermocouple only supplies a tiny amount of power, in the mW range.

The video linked to by ArthurDent explains it pretty well.

This other video explains one type of setup where the batteries in the thermostat supply power to the main gas valve which is still a pilot valve. So for the main gas to flow both the safety and the main pilot have to be activated.

https://youtu.be/riaN_j5tmX4?t=0

[jmoschetti45]

I have never seen a thermocouple singlehandedly control a thermostat.

Huh? I don't know what you're saying.

I'm saying:
On the wall a bimetallic thermostat with a knob for the occupants to control the temperature.
The open circuit voltage is 750 mV across this thermostat.
It is being fed by a thermopile on the pilot of a gas furnace.
An electric gas valve is in the series circuit.
https://www.honeywellstore.com/store/products/750-millivolt-heat-only-thermostat-ct33a1009.htm

OK, let me see if I can explain better.  A thermocouple does not have the power to directly open a normal gas or water valve.  AFAIK power always has to be supplied externally, either by electrical power or by the pressure of the fluid itself.  If there is no electrical power being supplied then it is probably a diaphragm valve with a tiny pilot valve where the pressure of the fluid itself is what actuates the valve. If there is no fluid pressure the valve will not open. The thermocouple only supplies a tiny amount of power, in the mW range.

The video linked to by ArthurDent explains it pretty well.

This other video explains one type of setup where the batteries in the thermostat supply power to the main gas valve which is still a pilot valve. So for the main gas to flow both the safety and the main pilot have to be activated.

https://youtu.be/riaN_j5tmX4?t=0

My day job, when I'm not doing EE related things....

In a thermoCOUPLE setup, you're talking a few mW at best (barely enough to even measure). This will require some other sort of power (typically 24VAC) to open the main coil in the valve for the burner. The few mW are only used to hold the pilot valve open. After you press the button down and light it, a very small electromagnet will be enough to keep it held open.

In a thermoPILE setup, the above is the same except the 24VAC for the main coil. This will be anywhere from 500-800mV from the thermopile. Run it to a thermostat, etc to control the burner. The main valve (at least as I was taught back in the day) is called a "pick valve". A tiny valve opens, and the pressure from that assists in a larger valve opening.

Next time I have a broken one lying around, I'll try and remember to gut it, I've never actually looked at the innards of one. Got a pile of 24V ones behind the garage...

[soldar]

I have replaced a few thermocouples, valves, etc. Easy to buy cheaply on eBay. They are not complicated at all. The Wikipedia article explains it quite well. The main valve is actuated by the pressure of the fluid itself and the electric signal controls a very small pilot valve. Water valves for washers work on the same principle.

Thermocouples might give some voltage in open circuit but that falls when in-circuit. I see them more as current sources than voltage sources.

[German_EE]

"I too have been amazed at how little electrical power actuates these switches.
Not really. You open the valve by pushing with your finger. Then the solenoid only has to hold the valve open."

This has been one of those 'Ah!' moments, I've wondered for years how those things operated and now I know.

Now a few words of caution.

A few decades ago I was living in a shared house with some students, the idea was that we purchased a wreck of a house then fixed it up enough to sell it at a profit when we graduated, using the profits to pay off our loans. We did all of the repairs except for the gas pipes because in the interests of safety we needed a professional here. So, the guy comes in and fits new copper pipes for the gas throughout the building, he then goes down into the cellar and pressure tests the installation, the conversation went something like this:

German_EE    How dangerous is this stuff? I've soldered copper pipe for the water and it was easy, gas should be the same, right? 

Gasman         Well, suppose there was a leak right here and the gas kept coming out until there was a layer 30cm (12 inches) deep throughout the cellar at floor level, then someone turned on the light and there was a spark. You'd take out your house, two houses to the left, two houses to the right and the house across the street. Sign here please.

I signed, we paid the bill, everyone was happy.

[soldar]

Well, suppose there was a leak right here and the gas kept coming out until there was a layer 30cm (12 inches) deep throughout the cellar at floor level, then someone turned on the light and there was a spark. You'd take out your house, two houses to the left, two houses to the right and the house across the street.

Anyone who can reliably sweat or thread connect pipes for water can do it for gas. It is pretty much the same. And if you pressure-test the system you would discover leaks anyway.

I have done gas pipe work, minor repairs, and never even pressure tested because it was too much hassle for such a small thing. I would use the soapy water test to check for leaks.

Gas works at much lower pressure than water too. You could check the gas piping installation by connecting it to the water system and checking for leaks.

Like everything, like electricity, gas is not dangerous per se, it is dangerous when you do not know what you are doing.

If it was a natural gas installation the guy was funning you. While propane and butane will sink and present a serious danger in boats, natural gas will rise and not accumulate in the bilges.

[johnkenyon]

Like everything, like electricity, gas is not dangerous per se, it is dangerous when you do not know what you are doing.

The problem with gas is that it's the ultimate Dunning-Kruger job.

For most DIY people, the gap between self-perceived competence with gas and reality is mahoosive.

This is magnified (in the UK any way) by a dearth of reference material detailing "how to do gas".
The information flow is restricted to prevent the unqualified from having a dabble, and people then think that a lack of documentation means that "it must be obvious", and they plough on regardless.

[mikerj]

For most DIY people, the gap between self-perceived competence with gas and reality is mahoosive.

I'm not sure that it's any worse than people's self-perceived competence with electricity.  It's the consequences of their meddling that are potentially much worse.