Re: Thermometer - Today's Math Challenge

Please show calculation method. It's a spiral. My take is the spiral should be 30.7% less LENGTH, not turns.

Anyone remember how to calculate spiral LENGTH ?:-)

...Jim Thompson

Working ...

reduce sensitivity from 180 degrees per 90 degrees F to 180 degrees per 130 degrees F

ie multiply sensitivity by 9/13 therefore remove (1 -9/13) * 9.25 turns ie remove (4/13) * (37/4) turns ie remove 37/4 turns = 2.846 turns

If the scale is linear (equal angular change per degree temperature change) then this would work.

Analogy might be a spiral tube with compressed air following Hooks law.

Not convinced that you are right. I would say that there is a greater angular deviation for the inner turns.

then this would work.

deviation for the inner turns.

Which is why simple math calculations will not work. Nipping it back until the desired result is found is the best way to get there, AND have it be calibrated for the range ends when you are done.

then this would work.

deviation for the inner turns.

Practically, one makes a calculation, then makes a mark on the bimetalic strip, then calibrates - and if it is ok snips ONCE.

I think that your aesthetic sense is challenged.

feel)

degrees temperature swing ?)

The ONLY way to make the thermometer reliably and reasonably 'correct' is full swing calibration cycles. Sorry. Only in mass production can one zero in on a reasonably close nip point to make a repeatable mass producible product that does not require a full swing calibration. They still get tested and certified functional in one though.

Based on questionable field measurements? I don't think so.

snips ONCE.

This prior calibration step method interests me. I know of a couple of ways, but all would be subsequently tested AGAIN after any cut is made, so you are not saving anything. The time the extra cut cycle or two takes is unimportant when accuracy is the goal. The true calibration is the placement of the indicator.

I think your common sense doesn't exist.

I worked in thermometry... you?

s ok snips ONCE.

e of

rectal no doubt ?

snips ONCE.

No. If one wants a fully corrected, calibrated-across-its-range scale, it would be the last job on the list to be done. The simple beginning and end point method I gave by nipping back incrementally will get you there quicker.

The beginning of range and end of range has already been specified at

snips ONCE.

Not quite. Jim needs to modify the degrees rotation per degrees temperature multiplier adjusted first. After that, setting the indicator to the correct position for a single temp is trivially easy.