The Museum of HP Calculators

HP Forum Archive 21

[ Return to Index | Top of Index ]

39GII bug in complex exponential
Message #1 Posted by Vincent Weber on 2 Sept 2012, 8:38 a.m.

Hi all,

I just got and my 39GII (firmware from May 9th).

I was puzzled to enter: e^i and to get: 1+0*i; actually I get the same answer with e^(x*i), x being any number.

I am ever more puzzled after I read a thread where Tim gave different (and correct) results.

Is there a new firmware that I'm not aware of ? Is it a bug ? or just my machine ??

Thanks and cheers,

Vincent

      
Re: 39GII bug in complex exponential
Message #2 Posted by Namir on 2 Sept 2012, 9:16 a.m.,
in response to message #1 by Vincent Weber

I have an Aug 23, 2012 firmware version. The accompanying emulator gives the correct answer to e^i. Ask Tim how to get the new firmware version.

Namir

      
Re: 39GII bug in complex exponential
Message #3 Posted by Charles C(UK) on 2 Sept 2012, 9:22 a.m.,
in response to message #1 by Vincent Weber

Mine has the same firmware and behaves in the same way.

My copy of the emulator, with firmware date string 2012/07/10, behaves much better and returns -1 - 1.26676153736E-12*i when computing e^(i*pi) - which is something like -1.

I would like that firmware installed on my hardware HP39GII but have no means to upload it.

            
Re: 39GII bug in complex exponential
Message #4 Posted by Tim Wessman on 2 Sept 2012, 9:33 a.m.,
in response to message #3 by Charles C(UK)

A little bird told me to wait just a little bit more...

TW

      
Re: 39GII bug in complex exponential
Message #5 Posted by Manolo Sobrino on 3 Sept 2012, 12:24 a.m.,
in response to message #1 by Vincent Weber

I'm starting to consider that these two widely known papers should be widely read...

* T. E. Hull, Thomas F. Fairgrieve, Ping Tak Peter Tang, “Implementing Complex Elementary Functions Using Exception Handling”, ACM Transactions on Mathematical Software, Volume 20 (1994), pp 215–244, Corrigenda, p553

Check p. 231 for CEXP algorithm.

(for free ftp://ftp.cs.utoronto.ca/na/reports/cmplx-elem-fcns.ps.gz)

** T. E. Hull, Thomas F. Fairgrieve, Ping Tak Peter Tang, “Implementing the Complex Arcsine and Arccosine Functions Using Exception Handling”, ACM Transactions on Mathematical Software, Volume 23 (1997) pp 299–335

Now, that's nice...

Edited: 5 Sept 2012, 9:25 a.m.

[ Return to Index | Top of Index ]

Go back to the main exhibit hall