HP Forum Archive 18

[ Return to Index | Top of Index ]

35s revision? Message #1 Posted by Rick Brown on 23 Apr 2008, 6:53 p.m. I am in the market for a new scientific calculator and recently came across the 35s on HP's site. I did a little bit of research on it (mostly on this forum) and gather that there is a bug in the cos() calculations. Is there any information as to when an update that corrects this will be? or is there a way around it by using the sin() function and a trig identity? Sorry if this is a common / already answered question. Thanks for the help.

Re: 35s revision? Message #2 Posted by Paul Dale on 23 Apr 2008, 7:28 p.m., in response to message #1 by Rick Brown The cos(x) bug is more a loss of significant figures near x=90. I doubt it will be fixed any time soon since the problem is also in the 33s and has been know about for ages. I also doubt it is a significant problem for normal use of the calculator. Hence, it hasn't been fixed. - Pauli

Re: 35s revision? (trig bug) Message #3 Posted by Karl Schneider on 23 Apr 2008, 11:28 p.m., in response to message #1 by Rick Brown Rick -- Bugs stemming from calculation of sine and cosine of arguments far from zero -- but where the result is very near zero -- were introduced in the HP-33s and carried forward to the HP-35s. However, the bug was not discovered and reported in time to be corrected for either model. The following MoHPC Forum post of mine was in a thread provided in electronic form by me directly to HP's lead calculator R&D person at the 2007 HHC Conference. I would certainly expect that any future revised version of the HP-35s will include a remedy for this bug: HP-33s & 35s trigonometric bug tabulated It's true, though, that the bug is not very serious. Results are still correct to at least six significant digits. -- KS Edited: 23 Apr 2008, 11:38 p.m.

Re: 35s revision? Message #4 Posted by Ivan Nejgebauer on 24 Apr 2008, 8:25 a.m., in response to message #1 by Rick Brown Quote: Is there any information as to when an update that corrects this [cos() loss of significant digits] will be? Others have addressed the impact of the cos() glitch, but generally speaking, and judging by past experience, HP will probably quietly issue a bug-fixed 35s sometime in the next six months. It happened that way with the 33s -- only that one was more obvious since the updated calculator had a different display as well. (Actually, there was a production run with the new display, but old ROM; search the archives for the gory details.) There is an article which lists known and suspected first-generation 35s bugs and annoyances. It shouldn't be too difficult to check a new calculator against that list (well, most of it). i.

Re: 35s revision? Message #5 Posted by Rick Brown on 27 Apr 2008, 1:30 a.m., in response to message #4 by Ivan Nejgebauer Thanks for the replies. If HP does decide to issue an updated ROM is there any way to re-flash the one on board? Or would you have to purchase a new one?

Re: 35s revision? Message #6 Posted by Ivan Nejgebauer on 27 Apr 2008, 11:20 a.m., in response to message #5 by Rick Brown Quote: If HP does decide to issue an updated ROM is there any way to re-flash the one on board? No. The 35s really has ROM, not flash, EEPROM or anything reprogrammable. You would have to buy a new calculator to get it. i.

[ Return to Index | Top of Index ]

Go back to the main exhibit hall