Re: TI-92 bug


[Prev][Next][Index][Thread]

Re: TI-92 bug



I did not get your results. I got  2*e^35i  ( for exact and auto modes )
and   -1.80738-.85637i ( in approximate mode )
My TI-85 gave -1.80738-.85637i also.
By the way the first TI-92 I had gave some big errors doing problems
with exponetials. TI replaced my calculator with a new one. The original
one I bought was one of the eariest versions. Yours might be an earlier version
also.

Good Lock

Gary Wardall

>Hi,
>
>I think I may have found a bug in the TI-92 when calculating with
>complex numbers in polar format, in degree mode. (No error occurs in
>radian mode).
>It seems to me that the error occurs because of a bug in the equation
>the TI-92 uses when calculating complex numbers in degree mode. However
>the equation used by the calc in radian mode is allright.
>
>The following example shows that a wrong result is calculated:
>First, in MODE, set the calc to: Angle:DEGREES , Complex format:POLAR ,
>Exact/Approx:APPROXIMATE.
>Enter the equation: 2*e^35i      ( i is the complex constant sqrt(-1) )
>Now you get: 2*e^35i = 2*e^-154.65i   , which is wrong.
>The result should obviously have been the same as entered 2*e^35i (the
>calc should not modify it).
>
>Heres a more precise description of why the error occurs:
>
>Set the calc to: Angle:RADIAN , Complex format:RECTANGULAR ,
>Exact/Approx:AUTO.
>
>Typing  a*e^bi  you get  a*cos(b)+a*sin(b)i  which is right.
>(This is right because the calc is in radian mode)
>
>Set the calc to DEGREES:
>
>Typing  a*e^bi  you get  a*cos((180/pi)*b)+a*sin((180/pi)*b)i  which is
>wrong.
>This should have been:   a*cos((pi/180)*b)+a*sin((pi/180)*b)i
>
>The error occurs because the equation has 180/pi instead of pi/180 which
>it should be.
>
>It would sure be nice to have this little, but serious, bug corrected in
>future EPROM upgrades, since I use complex calculating in degree mode a
>lot as an electrical engineer.
>
>Some people say that it isn't right to calculate with complex numbers in
>degrees and that it should be done in radians. However it works fine on
>my HP48 (and all other calcs I have tried).
>
>I would very much like to hear your response on this. Thanks.
>
>
>Sincerely,
>
>Peter Maigaard
>Århus Technical College, Denmark
>sm@ciconia.dk