kamidon@zip.eecs.umich.edu (Keith Amidon) (12/17/90)
I don't know if anynoe has run up against this yet, but there seems to be
a significant (at least for me) problem with the units functions on the
HP48SX. I have rev D ROMs, and what I get is that:
1_C^.5/F^.5 != 1_V^.5
^ ^ ^
| +- Farads +- Volts
+- Coulombs
Now, unless I am off my rocker, this is wrong. The really frustrating thing
is that it appears to work properly when using the NMOS Transistor equations
in the Equation Library, but if I reconstruct the exact same equation with
the exact same units outside of the library, it doesn't work. Does anyone
know what is going on here? Any help would be GREATLY appreciated....
By the way: What I am working on is a more detailed version of the eqs
in the equation library for semiconductor devices.
Anyway, thanks in advance....
Keith
kamidon@zip.eecs.umich.edu
-- kamidon@zip.eecs.umich.edu (Keith Amidon) (12/17/90)
In regard to my previous post on unit conversion problems on the hp48sx:
I investigated this a little further, and what is happening is that:
UBASE: 1_V = 1_kg*m^2/(A*s^3)
1_C = 1_A*s
1_F = 1_A^2*s^4/(kg*m^2)
1_C/F = 1_kg*m^2/(A*s^3)
The above are all correct. However, here is where things start getting
weird:
UBASE: 1_C^.5 = 1_A*s <-- 1_C^.5 == 1_C ?!?!?!
1_F^.5 = 1_A*s^2/(kg*m) <-- why no SQROOT() for kg?
1_V^.5 = 1_kg*m/(A*s^2) <-- why no SQROOT() for kg?
why no SQROOT() for A?
why (s^3)^.5 == s^2 ?
It is possible to have an SI base to the.5 power as verified by:
SQROOT(1_kg*m^2/(A*s^3)) == 1_kg^.5*m/(A^.5*s^1.5)
This takes the SQROOT of UBASE(1_V) and gets the right answer. I am totally
clueless as to why the calculator would behave in the way it is. Does
anyone else know?
Again, thanks for the help......
Keith
kamido@zip.eecs.umich.edu
-- akcs.dnickel@hpcvbbs.UUCP (Derek Scott Nickel) (12/18/90)
The HP 48SX, by design, only stores unit exponents as integers.
Derek S. NickelTDSTRONG%MTUS5.BITNET@CUNYVM.CUNY.EDU (Tim Strong) (12/19/90)
On Tue, 18 Dec 1990 02:45 CST >...The above are all correct. However, here is where things start getting >weird: > > > UBASE: 1_C^.5 = 1_A*s <-- 1_C^.5 == 1_C ?!?!?! > 1_F^.5 = 1_A*s^2/(kg*m) <-- why no SQROOT() for kg? > 1_V^.5 = 1_kg*m/(A*s^2) <-- why no SQROOT() for kg? > why no SQROOT() for A? > why (s^3)^.5 == s^2 ? >... > >kamido@zip.eecs.umich.edu >-- I believe quite som time ago it was mentioned that the HP48SX cannot handle fractional exponents in the units correctly. It just something that wasn't built in to the calculator. At least I believe thats how the discussion went. At the time I believe someone posted a work around technique. Unfortunately, I don't have a copy perhaps someone else does and will post it. Any takers? ====================================================================== ___ :__) _ _: _ _ TIM STRONG <TDSTRONG@MTUS5.BITNET> : \ (_: (_: (_: : Michigan Technological University ======================================================================
Ian.Goldberg@f646.n250.z1.fidonet.org (Ian Goldberg) (12/20/90)
The unit 'ga' has the same problem. UBASE(1_ga) is 9.80665_m/s^2, but
UBASE(1_ga^.5) is 3.13155712067_m/s. I got around this problem by always
UBASEing before using ga. Now I use CONST(g) on the Library Card instead.
BTW, how do you know what revision ROM you have?
Also, how do you write programs in ML?
--
EGSGate Fidonet Gateway, Toronto (egsgate.fidonet.org)
...!{uunet, moore, lsuc}!eastern!egsgateakcs.ciarn@hpcvbbs.UUCP (Dan Ciarniello) (01/13/91)
Keith,
I don't know why UBASE gives the results that you state in response #1,
but your original problem is not a problem.
1_C^.5/F^.5 is indeed 1_V^.5
A dimensional analysis of Q=CV (Q = charge in coulombs, C = capacitance
in Farads, V = voltage in volts) bears this out.
[C] = [F][V]
or [C]/[F] = [V]
[C^.5]/[F^.5] = [V^.5]
(I guess that you must be off your rocker :-))
Hope this helps,
Dan