Jouni Valkonen <[email protected]> wrote: Mats calibrated the thermometer and indeed he saw the 1 degree offset in > calibration. Temperature of icy water was measured 1.0°C. Therefore this > error is known. >
I discussed that with him. I do not think that was a 1.0°C error. He was seeing actual 1.0°C temperature variations. This was not an artifact. The instrument was working fine; he was using ice water with ice cubes in it, which has a lot of thermal gradients and streams of cold water. I have done that sort of thing often. You have to use finely crushed ice slurry or ice water that is poured off, with no ice, warmed up to maybe 5°C. Plus you have to vigorously stir the water. Otherwise you will get actual 1 or 2°C variations. An instrument of this type, of this quality will NEVER have an error as large as 1.0°C unless there is something seriously wrong with it. When there is something seriously wrong, it will probably display randomly -53.4, +10, +3 . . . or more likely it will display ERROR. Or nothing; blank. These things do not break a little bit. They are itty-bitty computers, and like all computers they tend toward all-or-nothing performance. Also they have internal boot up testing during turn on, and they tell you when the thermocouple is busted. As I mentioned, the Omega HH12B can adjust the OFFSET by +/- 4°C but at normal temperatures up to the temperature of a roasting turkey I have never seen the two thermouples separate by anything like that. I assume that correction is needed at much higher temperatures when you change the resolution from 0.1 to 1. You never need it at these temperatures. The Delta T ability depends on precision, not absolute accuracy. In other words, the TCs might think the temperature is 22.3°C when it is actually 20.4°C. That's inaccurate. But they will both agree later on that it has risen by 4.8°C, and they will be right. Or if you put the second one in a colder sample, the Delta T difference they show will be right. If they don't agree, you can tweak the OFFSET to bring the 2 together, or press REL (relative) to force the 2 thermocouples to agree and then start your T1-T2 measurement. Here is the manual: http://www.omega.com/Manuals/manualpdf/M4613.pdf - Jed
