Horace Heffner <[email protected]> wrote:
> Putting a metal thermocouple up against a metal surface sounds like a > prescription for variable but systematic error, depending on vibrations, > touching the wire, humidity, etc. The steel nut can short out at least some > some of the potential. This means requiring a high bias. However, if the > short is removed or reduced, then the bias is too high. When playing with > the bias in my spreadsheet I settled on 0.8°C. However, it looked as if only > one bias was not sufficient to fit the numbers. > Let me again suggest that you borrow or purchase a thermocouple and test it yourself. I expect you will find this is not a problem. I have put my Omega HH12B thermocouples against pipes, dry surfaces, wet surfaces, under water and in various other places. They seem to work fine. I myself am well grounded, and I have used the thermocouples as a fever thermometer under my tongue. It shows the same temperature as the fever thermometer. It works a lot faster. Omega and the others who sell these things realize that the instruments will be used in a wide variety of environments for many purposes. I expect they realize that if the standard thermocouple probe they ship with the instrument did not work on pipes, that would be a problem for many customers, since this is a recommended configuration, widely used. So they would provide another kind of probe. I got an extra sheathed probe which I use to check the temperature of roasting turkeys in the oven. Probably the off-the-shelf probe that came with it would not be good for that. Mallove and I used to dump the probes directly into a chemical cells which is probably not a good idea. There are many problems with Rossi's test. You need not invent any more problems such as the notion that thermocouple probes do not work on pipes or that thermocouples routinely have errors of 1 or 2°C in the range of 0 to 100°C. If you will try using a laboratory grade meter you will find they do not have the problems you imagine they might. - Jed
