Welcome! Here are the website rules, as well as some tips for using this forum.
Need to contact us? Visit https://heatinghelp.com/contact-us/.
Click here to Find a Contractor in your area.


Richie Member Posts: 1
It seems easy enough to replace a thermocouple. As of late, I've been going back on calls to replace lots of tcouples that are 1 to 2 years old. In some cases, a couple of months.
I have been in heating for 15 years and this passed year I feel as though I know nothing..
Does brand matter? Please don't respond with "hand tight 1/4 turn". Give me that at least..What could I be doing wrong? They seem to burn out from the bottom and melt. I check to make sure the pilot tube is not leaking under the pilot head.


  • lchmb
    lchmb Member Posts: 2,997
    I use

    I use the husky brand and don't seem to have to many go bad. I won't mention the hand tight as you asked but how does the pilot look with the unit running? Is the flame too sharp? Is a burner tube firing at the pilot with a sharp flame? Are your gas pressure's good? Is the pilot assembly in a whole being damaged? Are you seeing this on all brands of unit's or just one or two in particular?
  • David_5
    David_5 Member Posts: 250
    Here's what I check

    A new thermocouple should produce 25-30MV in an open circuit( connected only to a meter). A thermocouple should produce about half the open circuit reading in a closed test( connected to gas valve, holding pilot coil). The gas valve should drop out about 10MV less than the closed circuit reading after the pilot flame goes out. This works out 25-30MV open circuit, 12-15MV closed circuit, 2-5MV reading when the valve drops out. The thermocouple should glow dull red when in the flame properly. Bright red will reduce the lifespan. I have also noticed on systems with an ECO the spade connectors get corroded and cause a voltage drop. Replacing spade connectors helps to bring the voltage drop to a minimum. I Have Timmie to thank for all of this info, I have experienced fewer thermocouple problems since being more thorough.

  • MIKE6
    MIKE6 Member Posts: 102
    Best t lead

    The best T lead I've found is the Baso K-16 Husky through Johnson Controls.It has more meat to it and will get you out of the yearly change out.BTW some manufacturers recomend yearly replacement on the package.IMO Honeywell is the worst.
  • J.C.A._3
    J.C.A._3 Member Posts: 2,981
    Sad to say....

    But I agree with Mike. The Baso/Husky thermocouples seem to last way longer than most. Be warned, they don't fit in all appliances! If you've got a "push through" fit like all of the TeledyneLaars I've seen lately, they won't go into them fully. They have a larger base than the Honeywell universal Q series. Chris
  • Here is a good step by step

    procedure. For more get in touch with me at [email protected].

    I am often asked about troubleshooting a thermocouple on gas systems. This will be a permanent reference that will give a step-by-step procedure.

    A thermocouple is a device used to satisfy pilot safety on many 24 volt gas systems. The thermocouple is a device made up of two dissimilar metals. They are joined together at the tip (Hot Junction). When heat is applied to that hot junction a small millivoltage is created. This develops because of temperature difference between the hot junction and what is called the cold junction. The flame has to envelop the upper 1/2" to 3/8" of the thermocouple and the tip should glow a "dull red". If the flame is adjusted to a sharp flame it will glow "cherry red" this will cause the tip to be welded and eventually the thermocouple will fail. The flame should be adjusted to a soft blue flame, not roaring or lifting. The normal millivolt output is 25 to 35 millivolts, on some you may even get up to 35.

    The other part of this safety pilot system is the electromagnet (power unit). It is if you will the LOAD and we can say the thermocouple is the SOURCE. The electromagnet is made up of a coil of wire and "U" shaped iron core.When the thermocouple is heated and the millivolts generated the coil will be energized and create a magnetic field. The magnetic field will cause the "U" shaped iron core to be magnetized, it in turn will hold open a seat allowing gas to pass through.

    When this system malfunctions it typically causes the pilot to go out and the gas will not flow. The first thing that should be done when arriving at a pilot outage situation is to do some visual checks.


    Once those things are addressed it is a good idea to take some millivolt readings. It should also be mentioned that many times it is the policy of some to replace the thermocouple on a call and clean the pilot. It is not a bad thing to do, however it is statistically about 85% of the time it is the thermocouple giving the problem. It is the other 15% of the time that taking readings can solve other problems.

    You need a multimeter with a DC volt scale as the millivolts generated are DC volts. There are four readings we are going to take they are

    OPEN CIRCUIT - this is taken with the thermocouple disconnected and the meter leads attached to the outside of the thermocouple and the other meter lead attached to the tip of thermocouple. The pilot-on-off knob will have to be held manually to take this reading. This measures the output of T'couple the readings must be above 17 to 18 millivolts.

    * CLOSED CIRCUIT - This measures the millivolts used by the coil in the electromagnet . A rule-of-thumb is this reading should be roughly half of the open circuit. It is taken using an adapter screwed into the magnet and the thermocouple screwed into the adapter.

    CLOSED CIRCUIT LOAD - This reading is taken the same as the previous reading except the burner is now on. With a proper flame this reading should be about the same as the previous reading. With a lifting main burner flame or excessive drafts or chimney pull, this reading may reduce from previous reading (flame being pulled away from the thermocouple). With the cold junction being heated this reading may increase. If the "cold junction" is heated excessively it will break down.

    DROP OUT - This is the final reading. It requires the pilot to be blown out. It measures the ability of the magnet to hold under reduced MV input. A good unit should drop out below 6 MV's - normal is 1 to 2 MV's. The allowable "drop out" time is 180 seconds yes three minutes. It is more likely to be a minute and half to two minutes. There will be an audible "click" when the magnet shuts down.


    A normal set of readings

    OC- 30 millivolts
    CC- 15 millivolts
    CC(load) -15 millivolts
    DO- 1 millivolt

    The best way to be able to diagnose these readings is to use MILLIVOLT CHARTS these can not be displayed here but I can provide them if you e-mail me.

    I will add a little more to this later but for now this will be a good start
  • J.C.A._3
    J.C.A._3 Member Posts: 2,981
    Thanks Timmie

    Yet another refresher course! Chris
  • gasfolk
    gasfolk Member Posts: 392
    Technical at White-Rodgers claims

    that their thermocouples should last 8 to 10 years, 5 at the inside. And if they are burning out after 1 to 2 years? Then, they say, "you are toasting it."

    Besides a too hot pilot flame, excessive heating of the cold junction, and mechanical damage, does anyone know another specific hazard for thermocouples?

  • Thermocouples

    from different manufacturers vary as to their dependability. I find White-Rodgers to be on the low end of dependability.

    The only thermocouples I recommend are made by Johnson Controls. The K15 and K16 series are the best. If you are having durability problems then use the K16RA which is a nickel plated high ambient or corrosive environment thermocouple. The Husky will fit most applications and for those that it does not the Slim Jim will fit.

    To answer gasfolks question. The adjustment of the pilot flame to envelope the upper 1/2 to 3/8 of the thermocouple is important, the flame should be a soft blue flame not roaring which will cause the tip to glow a "dull red" versus "cherry red".

    The combustion condition (excessive temperatures) in the chamber is also an issue and this will require a combustion test and draft measurement to insure that excessive temperatures are not being applied to the pilot. In some cases on water heaters it may be necessary to alter the pilot adaptation to get better quality performance. This however should not be done unless you have had proper training.

    The possiblity of the equipment operating in a depressurization environment will certainly lead to thermocouple failure. In addition if the equipment is flued together with a "fan assisted" furnace or boiler this can lead to problems. There are solutions to this also but training is required.

    The thing that I find is often a problem is the environment in which the equipment is operating. Many times corrosive chemicals and airborne contaminants are being drawn into the air gas mix and a checmical reaction takes place. This again requires attendance at a training session by a professional combustion person to help you to see the various affects this will have.

    Last of all the failure to put all the doors and covers back in place on equipment. The failure to do this will cause an alteration in combustion air and the flame stability is affected.

    The design of some equipment is also a problem. When there is high deamnd for heat (very cold weather) the temperatures that are created in the chamber have an adverse affect on the pilot and thermocouple system. The addition of the K16RA thermocouple can offer some assitance toward extending the life of the thermocople in this situation.

    Insufficient air for combustion and dryers operating in close proximity to equipment also lead to problems.

    Last of all and this is not directed at any one in particular but just plain lack of service personnel and installers knowing what they are doing.

    My book "Circuitry and Troubleshooting" addresses many of the things in question here.

  • gasfolk
    gasfolk Member Posts: 392

    Any estimate of the average expected life of a good quality thermocouple?

    Also, where can one get your book(s) and the upcoming schedule for your seminars?


  • Under normal conditions

    a thermocouple can last 8 to 10 years. I have in fact seen them last 15 years.

    Some of the cheap models do not last through one heating season.

    I will e-mail you inforamtion on manuals and schedules.
  • Dan_16
    Dan_16 Member Posts: 11

    Those eco switches can also throw you off as there tempermental too overtightening and dirty contact.
  • Mitch_4
    Mitch_4 Member Posts: 955
    additional test

    There is also the "Pull in test" after doing the drop out take the erading and multiply by 3. so if your drop out was 2 mv, pull in should be 6. Light the pilot, and when you read 6 mv closed cct test, release the pilot stat over ride button. The electro magnet in the pilot stat should hold it open. If not the pilot stat may be going on you.

  • Dale
    Dale Member Posts: 1,317
    One more thing

    Besides what Timmie and the other guys have said, I'll add this, we also use the huskys for hot pilots like conversion burners, we use HW for the softer smaller pilots like water heaters, especially the target type burners, where you are lucky to get 12mv with the tear drop pilot. Also, if you service fireplaces some (with an additional powerpile for the main valve) use a quick dropout thermocouple, that has a reduced size tip, do not replace it with a standard as the drop out time is too long. Some unvented space heaters use this special tcouple. Timmie covers this in his excellent millivolt systems course.
  • Arthur
    Arthur Member Posts: 216

    I had a case of a pilot going out after the cylinder got hot or sometimes the next day, The magnet held in quite quick within 20 sec indicating a good t/c, Volt reading were all well within the guide lines mentioned. finally in desparation I renewed the magnet assemby and have had no more problems since, I had alsways thought that if the magnet either worked or they didn't and if it held in to start with it was ok,
    I sometimes test the pilot while the burner is out by connecting a short pipe to the pilot connection (an old pilot pipe and nut) then slip a hose on to the pipe and to the pilot pipe (make sure it doesn't leak) then jam the valve knob down with a short screw driver and then observe how the pilot burns on to the t/c. and take reading if needed.
    Solved many a problem this way saves time in taking the whole burner assemby in and out several times.
    An old eco butterfly wire is a handy way to connect to test Volt readings test.
  • jrc2905
    jrc2905 Member Posts: 98

    Could you send me some informaion on your book?
  • To the

This discussion has been closed.