Slightly OT - Furnace problem

I would replace the thermopile, it is possibly getting weak and unable to produce enough power to open the gas valve with the slight resistance in the t'stat. It is possible the gas valve is faulty too, but the thermopile is cheap enough to try. Millivolt systems can be a bit touchy considering the minute voltage produce by the thermopile. Greg

"Greg O" wrote: I would replace the thermopile, it is possibly getting weak and unable to produce enough power to open the gas valve with the slight resistance in the t'stat. (clip) ^^^^^^^^^^^^^^^^^ Shorting the contacts where the thermosat plutgs in turns the system on, so I don't think it is in the thermopile or valve. I would clean the contacts in the thermostat socket. You may have a little oxidation there.

With a new t'stat? I am still betting on the thermopile! I work on this stuff almost every day and have seen this problem often. They get weak after a few years and act flaky. Greg

Just because it's a new stat doesn't meen it's a good one. If shorting the wire turns it on, my bet is on the 'stat.

"Greg O" wrote: With a new t'stat?

^^^^^^^^^^^^^^^^^^ A wire connection across the thermostat socket turns the furnace on, so the thermopile seems to be working. It's possible that the new thermostat is bad, but all I am saying is dirty contacts between the thermostat and its receptacle would cause this. It's so easy to check, and costs nothing--then if that doesn't fix it, check and compare resistance across the old thermostat and the new thermostat.

What is the millivolt reading, at the gas valve, with the stat wires jumpered together, and then with the t'stat connected? My bet is the millivolt reading will be low either way, indicating a bad thermopile. I have seen it many times! Greg

nothing--then

There is no thermostat socket. The wires from the furnace are connected directly to screw terminals on the thermostat.

Jerry

OK, this comes back to the original question which, perhaps, I didn't express as clearly as I might have. What should the open circuit voltage from the thermopile be? What should the current be with the thermostat switch closed? You suggest that the voltage across the gas valve when it is energised is low. What should it be?

Jerry

Open circuit on a thermopile should be well over 200 millivolts, which you are getting.. Closed circuit, with the furnace running should be "close" to

200 millivolts, maybe 175 minimum. 80-90 is way to low, 200 is the number I look for. That is why I say thermopile. I have seen open circuit voltages over 200, but with a load applied the thermopile can not produce the power needed so the voltage drops way off. It is still possible you have a bad gas valve that is overloading the circuit. Again, thermopiles are cheap, toss one on then see what your results are. I don't normally recommend throwing parts at something, but when the parts cost less than the diagnostic time I will do it. Greg

What voltage does it say on the valve?

80 or 90 millivolts is the drop I see across the thermostat.

One of my problems is that the furnace is located in the crawl space under the house and it is extremely difficult for me to get to it. I had a friend here yesterday who is much more agile than I, but we decided the thermostat was bad & replaced it. Everything seemed to work. Then, after he left, I tried it again and it didn't work...

The other problem is that, while I've had considerable experience working with control systems in general, I'm not familiar with a thermopile powered millivolt system and am trying to learn what to expect of the various parts of the system.

I measured the contact resistance of the new thermostat and found it to be

1.5 ohm, which I suspect is excessive. My friend took the old thermostat with him for his kid to play with, so I don't have it to measure.

Jerry

A millivolt system is pretty much like any other. The thermopile and pilot flame produces the electric current to power the system. The thermostat is simply a switch to turn the system on and off, and the gas valve is the load. Now they confuse things slightly by connecting both leads of the thermopile to the gas valve, but the one lead is just parked there on a dead terminal. Power comes out of the one lead of the thermopile, through the t'stat, then the gas valve and then back into the thermopile through the other wire. Greg

1.5 ohm is excessive. See recommendations for wire runs to stat in a millivolt system: 60 feet max (round trip) of #18, which is about 0.45 ohm.

The thermostat has an anticipator heater in series with the Heat contacts, and the thermopile has to run that, too. It has to be set right, if you leave it in the position for a 24V system the resistance will be too high.

You can totally bypass the anticipator (set the resistance to minimum) to get the heat back on, but the furnace is going to overshoot the temperature setting on every heat cycle.

If you manually short across the Heat and Common terminals to get the furnace to kick on, you are bypassing the anticipator heater in the thermostat and masking the bad thermopile in the pilot light. The gas valve is more sensitive than it needs to be to accommodate the higher resistance in the anticipator.

The thermopile is just a bunch of thermocouple junctions in series to get 750mv of power, and they do go bad over time just from the erosive effects of the constant pilot flame - only takes one bad TC in a series circuit to knock the entire system out of whack. And they're cheap.

-->--

This is way too high. If you only have 200 mV to start with, then an 80 mV drop JUST at the thermostat is way too much. This may be due to the resistance of the bimetal spring, depending on how the thermostat is constructed. Or, the high current has already fried the contacts. My guess is it really is NOT designed for millivolt systems, and was mislabeled by cheap overseas manufacturers who could care less if their stuff works. You may have to buy a name brand thermostat at a heating/ cooling supply.

You have to be very careful making milli-Ohm measurements with a standard DVM. Just short the probes and what reading do you get? If the thermostat really had 1.5 Ohms resistance, it could NEVER work. Too bad you didn't measure it before the first use, to have a reading to compare to.

Jon

(measurements

I re-visited the "anticipator" setting on the thermostat and decided that the labeling was a bit ambiguous. So, I set it per the "other" possible interpretation of the instructions and measured the resistance. It was over

1Kohm!

So, with the ohmmeter connected, I tweaked the anticipator until I got minimum resistance (about 0.2 ohms), then connected it back up. It seems to work, now.

I realize that I may just be masking a bad thermopile. But, for the time being, at least, I have heat... And on a Sunday morning, I'll settle for that.

Thanks to all who replied for your suggestions and for filling in a few gaps in my knowledge.

Jerry

I don't think a stat designed for use with a millivolt system has an anticipator. Those are in 24-volt stats.

If it's a 24 volt system and a milivolt stat was installed it would likely work - ONCE.

This was the only simple, manual thermostat the local hardware sold. The packaging specifically stated it was for both millivolt and 24 volt systems and included instructions for setting the anticipator for both. When I set it per the instructions, the resistance was fairly high. So, I put an ohmmeter on it and set it for minimum resistance. That seems to be working.

Jerry