The Rate of Change Thermometer would show the temperature by sliding a small bar up and down a vertical scale. The slope of the bar would indicate the rate of change of temperature over the period of five or ten minutes. If the bar slopes up then it's getting warmer, down and it's getting colder.

How it could work (1): The expensive (and pretty) analogue way. Two bimetallic strips indicate the temperature. One is bigger and thicker than the other and is therefore slower to react to temperature changes. The two strips are geared up so that the lightweight pointer is moved by a large amount for every small change in the strips. The Thermometer is factory callibrated by the position the strips are attached to the pointer.

How it could work (2): The cheap digital solution. Put a digital thermometer, a small signal processor with a clock, a stepper motor and a few cogs, gears and assorted bits of metal in a box. Assemble as seems fit.
-- st3f, Dec 09 2002

And a resistor. Them being futile, and all.

The analogue method is really clever. I like it. Is it your own?
-- egbert, Dec 09 2002


Does it move or run away by itself? If it doesn't it should.
-- TBK, Dec 09 2002


Rotating-disc thermographs have been baked for a while in industrial plats, but this sounds nice. A scrolling false-colour display would be the thing.
-- 8th of 7, Dec 09 2002


If the rate of temperature change over the last five or ten minutes is enough to show up on a bar graph, I would suggest closing the door. With central heating/cooling, your house should be comfortable all day, and this device would only serve to settle kids' squabbles.

I would go for a charting thermometer, though, especially for my greenhouse.
-- dalek, Dec 09 2002


The analog version is quite nice.

Does this really indicate rate of change or is it really the direction, trend, of change?

Does the necessity for such an instrument really matter, [dalek]?  The novelty of the idea is enough for me.
-- bristolz, Dec 09 2002


//bris: Does this really indicate rate of change...//
ish.

//dalek: ...last five or ten minutes..//
You may have a point there. In which case, I'd make the big bar bigger and go for half an hour instead.

//dalek: I would go for a charting thermometer...//
No problem. It's all a matter of taste anyway.

egbert: yes.
-- st3f, Dec 09 2002


"ish?" You no like my comment?
-- bristolz, Dec 09 2002


bristolz: as described, the slope would indeed indicate the rate of change (as well as the direction, of course).
-- dalek, Dec 09 2002


Bris: I could have said, "The slope of the pointer approximates rate of change at a point in time between the mean time taken by each bar to react on a scale defined by the mechanism of the thermometer." I thought, "ish." said all that more succinctly. It wasn't a critisism. Meh.
-- st3f, Dec 09 2002


Yeah, I get it now.
-- bristolz, Dec 09 2002


A few more ideas:
1. Use one bi-mettalic strip, but make the pointer very bendy and put it in clear gel. The end of the pointer will take longer to move, and so it will show a curve.
2. Use two mercury thermometers, with different time constants (i.e. one big, one small). Very similar to the proposed idea.
-- Ling, Apr 08 2004



random, halfbakery