TWO unit conversion questions

[Chris]

PRIMARY QUESTION:
I have had no problem with unit conversions in the past, but have
suddenly had difficulty when I tried to convert temperatures. I tried
a brainless 32 deg F to deg C and got an "Inconsistent Units" error.
I get the same error no matter what temp units I move from or into.

The only inconsistency I can imagine is that the Nspire may be
thinking angles when it sees the degree character in the temperature
unit, but that is the syntax provided by the Catalog. Every other
unit out there converts easily within each unit family. Hmmmm....

Any ideas out there?


SECONDARY QUESTION:
I thought I remembered the Nspire collapsing complex units into
cleaner forms (like 1 kg*m/sec^2 = 1 N), but now I can't make either
my Nspire or my old 89 do this. Was I just dreaming that this was
once possible? TI folks, can such a feature happen in the future?

[Corey Andreasen]

The reason for this is that the F and C scales are not directly proportional. That pesky y-intercept caused by the fact that 32 degrees F is the temperature at which water freezes. The problem for automatically converting is that sometimes you want to consider a change in temperature, and in that instance you DO want a direct proportion.

For example, 50 F = 10 C. But an increase of 50 degrees F corresponds to an increase of about 27.8 degrees C.

Corey

[Nelson Sousa]

to convert temperature values you must use the tmpcnv; to convert
differences in temperature (temperature increases for example), use
deltatmpcnv.

Both functions are available through the catalog.

As for converting complex units, TI-Nspire does it fine (so does
TI-89), as long as they're consistent. Check that you entered the
units properly. Remember that grams is gm, not g and that seconds is
s, not sec.

If you can't find the error post the exact unit conversion you're
trying, maybe we can help.

Cheers,
Nelson

[Chris]

Perfect ... Thanks.
I was using _sec, not _s, so the Newtons now appear as predicted.

I was also able to use tmpcnv to get my desired results, but I'm still
a bit annoyed. Corey is correct in that temperature conversions are
not directly proportional, but they are all still linear conversions.
It is _FAR_ more intuitive to think that all units would convert the
same way (esp. since any two members of a "unit family" are linearly
related). So why treat temperature differently? The tmpcnv command
does the job, but why should a user need to think of that????

Thanks for the help and listening to my $0.02.

[Nelson Sousa]

because 0 degrees Celsius are not 0 degrees fahrenheit. that's a
somewhat complicated question to solve using only the TI-Nspire ;)

But seriously, it's because temperature conversion is a totally
different question. The user should actively think about what he's
doing, avoiding the mistake of converting a temperature value of 10
degrees when he actually meant an increase in 10 degrees.

You can multiply units by other units, but that doesn't make any sense
with temperature units unless you're using Kelvin (or Rankine). 1
kg*ºC doesn't have any meaning, whereas 1 kg*K does.


Nelson

[Corey Andreasen]

Because it matters whether you are measuring the temperature on the scale or computing a difference in temperatures. It would be problematic if you had some setting where you were looking at a difference in temperatures and wanted to convert to the other scale, but the nSpire did so as if the differences were measures of temperature. I gave a numerical example of a reading of 50F converted to C as opposed to a difference of 50F converted to C, but I didn't give a setting.

Think about a simple example like a cooling time lab. You could measure the actual temperatures in either scale. But if you wanted to subtract the room temperature (to shift the scale in order to fit an exponential decay curve), now you are working with differences in temperature and the conversion is different. And when you subtract measurements you get the same units. There's no way for the nSpire to automatically know which conversion you need.

It's nice that nSpire has a command to do the conversion, though. I didn't know that. But then, we could fill a large book with what I don't know about the nSpire!

Corey


----- Original Message ----
From: Chris <chr...@westminster.net>
To: tinspire <tins...@googlegroups.com>

[Chris]

Ok, that's the reasonable explanation I was looking for (differences
in temperature vs an actual temperature measurement having the same
units, but different conversion meanings). Thanks, much, Corey!