Testing Thermocouple Accuracy

[Randy W]

A while back someone told of a test for checking the accuracy of a thermocouple. If I remember right a cone was placed in the kiln and fired. When the cone bent, the pyrometer was adjusted to a certain setting.

I searched in the archives and didn't have any luck. Does anyone remember that test?

I've had a number of pieces break on me lately, usually the next day after firing.

http://f1.pg.photos.yahoo.com/ph/74dodg ... pg&.view=t

Thanks, Randy

[Tony Smith]

Cones are not the best indicator of temperature since they respond to heat work (temperature and time combined). You're better off going through Bullseye's "knowing your kiln" process and using the temperatures determined by following that process for annealing... but before you do that, remove your thermocouple and inspect it. Make sure there is no obvious damage to the tip of the thermocouple or the wires. Make sure the wires are clean and making good contact with your controller. Check that the screws are tight. If in doubt, replace the thermocouple.

Tony

[DonMcClennen]

Hi Randy
I feel your pain and frustration.
I have a pot melt about 12" dia that I fused on a double rim plus a clear base bringing it up to 20"dia and about 3 layers thick. A week later it broke in half exactely like yours. I know it was not the schedule-- I suspected a small piece of non compatible glass..but a glass blower I talked with suggested the problem might be the inclusion of opal yellow in the melt. He claimed opal yellow is what they call a "warm" glass, meaning it melts at a hotter temp. than other glass.At this point I'm not sure I want to spend any more time or moneyon this frustrating adventure.

[Bert Weiss]

Cones are not the best indicator of temperature since they respond to heat work (temperature and time combined). You're better off going through Bullseye's "knowing your kiln" process and using the temperatures determined by following that process for annealing... but before you do that, remove your thermocouple and inspect it. Make sure there is no obvious damage to the tip of the thermocouple or the wires. Make sure the wires are clean and making good contact with your controller. Check that the screws are tight. If in doubt, replace the thermocouple.

Tony

I'm with Tony on this 100%. If you are using Bullseye clear do a slump test using BE tested clear. Directions are in the archives or Fusing Book 1 by Lundstrom.

[Tony Smith]

Hi Randy
I feel your pain and frustration.
I have a pot melt about 12" dia that I fused on a double rim plus a clear base bringing it up to 20"dia and about 3 layers thick. A week later it broke in half exactely like yours. I know it was not the schedule-- I suspected a small piece of non compatible glass..but a glass blower I talked with suggested the problem might be the inclusion of opal yellow in the melt. He claimed opal yellow is what they call a "warm" glass, meaning it melts at a hotter temp. than other glass.At this point I'm not sure I want to spend any more time or moneyon this frustrating adventure.

How do you know it's not your schedule? just curious...

Tony

[BobB]

Like Tony said Excellent tech notes here.

http://www.bullseyeconnection.com/pdfs/ ... tes_01.pdf


http://www.bullseyeconnection.com/getsmart.html

[Stuart Clayman]

Paragon once told me to put the thermocouple in a glass of ice water and see what it read. Then put it into an 250 degree oven and see what it read at that temp. I ran those tests and also then ran the Kiln thru a cycle with a analogue proymater to see what the temp was doing.

[Barbara Muth]

I suspected I was not annealing correctly (though my schedule for annealing was dead on) so I ran the "know your kiln" and then ran a cycle with a pyrometer in addition to the thermocouple and graphed the results (comparing the two temps every 20 minutes or so). I was delighted to see that I was heating evenly and that my kiln was annealing at the right temp. But it meant that my breaking problem was due to firing too quickly on subsequent firings.

For those who have wondered, I have now sucessfullt re-fired pieces. But I am now ramping up at 150 dph for 10 inch square two layer pieces. This is really slowing down my production time. Guess I need a bigger kiln!

Your break doesn't have the classic s annealing shape. Hopefully others will chime in on what they think is the cause. In the meantime, posting your firing schedule and the glass you used could help. It's a shame, the pattern and color on that piece is quite attractive.

Barbara

[Tony Smith]

Paragon once told me to put the thermocouple in a glass of ice water and see what it read. Then put it into an 250 degree oven and see what it read at that temp. I ran those tests and also then ran the Kiln thru a cycle with a analogue proymater to see what the temp was doing.

Having a thermocouple read correctly (or incorrectly) over a 218° range isn't a good indicator of what its accuracy might be over a 1450° range. I ran three side by side thermocouples last night in the same kiln. All three were within 5° at room temperature and while two of them were close start to finish, one (the one connected to the controller) was off by 40° at 500°. By the time the temperature got to 1000°, the three were only off by 10°. At 1450°, the three were crossing over each other staying within a 5° range as the kiln cycled on and off.

If you doubt your kiln temperature, get a good thermocouple from Omega and a handheld thermocouple meter and compare the controller to the meter.

Tony

[Bert Weiss]

Paragon once told me to put the thermocouple in a glass of ice water and see what it read. Then put it into an 250 degree oven and see what it read at that temp. I ran those tests and also then ran the Kiln thru a cycle with a analogue proymater to see what the temp was doing.

Having a thermocouple read correctly (or incorrectly) over a 218° range isn't a good indicator of what its accuracy might be over a 1450° range. I ran three side by side thermocouples last night in the same kiln. All three were within 5° at room temperature and while two of them were close start to finish, one (the one connected to the controller) was off by 40° at 500°. By the time the temperature got to 1000°, the three were only off by 10°. At 1450°, the three were crossing over each other staying within a 5° range as the kiln cycled on and off.

If you doubt your kiln temperature, get a good thermocouple from Omega and a handheld thermocouple meter and compare the controller to the meter.

Tony

Tony makes a good point here. The temperature range that counts for us is 700 - 1000ºF. The rest of the temperatures are not relevant except that you know how the glass reacts at a given reading in your kiln.

With a slump test, exact temperatures are not necessary, as long as you anneal soak around 100º below slump temp and go slow enough for the next 300º.

There is a big difference between cheap and good thermocouples. A good thermocuople is encased in an inconel sheath so that the business end of the unit is not exposed to atmosphere. It takes a few seconds longer to react, but will last a lot longer in good working order. The inconel (a high temp stainless steel) reacts much faster than a ceramic shield.

[KellyG]

Had the exact same break on a pot melt last month. Used BE opal teal, lt. green and YELLOW. Fired AFAP to 1700, held apprx. 45 minutes. Annealed at 980/45 min. then 940/45 min. Slow to cool. Did some cold work a couple of days later, then put aside for a week or so. This piece sat in the shop for about two weeks, then one day I saw that it cracked just as Randy's (straight down the middle).

I cut it up on tile saw and will use the pieces at some point. I'd like a answer to the cause before I attempt a 2nd melt.

Puzzledgirl

[Tony Smith]

You've got a problem with your annealing schedule. You should be doing a controlled cooling to at least 750° (no more than 120°/hr). I would be conservative and do 45 minute soak at 960° (or 980° doesn't really make a difference since all you are doing is bringing the piece to an equilibrium temp) followed by 75°/hr to 750° and 120°/hr to 450° depending on the geometry of the finished part

Tony

[Bert Weiss]

Had the exact same break on a pot melt last month. Used BE opal teal, lt. green and YELLOW. Fired AFAP to 1700, held apprx. 45 minutes. Annealed at 980/45 min. then 940/45 min. Slow to cool. Did some cold work a couple of days later, then put aside for a week or so. This piece sat in the shop for about two weeks, then one day I saw that it cracked just as Randy's (straight down the middle).

I cut it up on tile saw and will use the pieces at some point. I'd like a answer to the cause before I attempt a 2nd melt.

Puzzledgirl

I think 920 is my Bullseye soak, Graham starts at 950. 980 is too hot. Do a slump test. Soak 100ºF below slump temp.

[Hohenshelt]

Testing the accuracy of a thermocouple is not an easy process. A thermocouple works by having two different metal wires coming together in the firing chamber. They produce a known milivolt signal ( thousandths of a volt) at a given temperature. There is an international standard for thermocouples. Most common type in glass kilns is K-type. It is most accurate above 600F and below 2000F. Most are designed with an inconel sheath to protect the wire inside and increase the life of the metal wire inside. Checking the accuracy or calibration of the thermocouple is difficult without a certified test meter. The meter interprets the milivolt signal into a temperature. For instance 25.1 milivolts may be 1080 F. That is all digital controllers do is take the milivolt output from the thermcouple and look at a set table for that thermocouple and displays the temperature associated with that milivolt output. Over time the milivolt signal from a K-type may drift enough to produce up to a 30 F difference in temperature. Checking the accuracy with an analog pyrometer is only good for gross errors as its accuracy is poor. Thermcouples will drift to a point then fail. Gross errors or erratic temperature readings are usually the result of loose connections in the thermcouple connection wiring.