Some assorted HP67/97 questions

[Guenter Schink]

(06-04-2016 11:14 AM)Dieter Wrote:  I am currently translating an HP41 a program for the 67/97. All I got is a simulator that seems to have several bugs, so I would like to ask users of "the real thing" a few questions. ;-)

1. Display formatting: when the display underflows in FIX mode, does it switch to SCI 9 or does it keep the number of digits set by DSP? In other words: if FIX DSP 2 is set, will 1/300 be displayed as 3,33 -03 or 3,333333333 -03?
2. When the program reaches a RTN statement and there is no pending subroutine call, does the program stop at the current line number so that R/S would continue there (like e.g. the 41) or does it return to step 000 (like e.g. the 34C)?
3. On the 34C and 41 I always tried to avoid the slow y^x function, e.g. by calculating cubes via [ENTER] [x²] [x]. I assume this is also a good idea for the 67/97. But what about x⁸ ? Is [x²] [x²] [x²] still faster than [8] [y^x]? I suppose it is, but maybe someone can check this. I also remember a table with timings of several functions in some HP-related magazine, but I cannot seem to find it. #-)
4. Can someome recommend a good 67/97 emulator for Windows that runs the original microcode? Nonpareil sounds perfect but it seems to be unavailable.
   

Thank you very much for your help.

Dieter

O.k. from the real 67:

1. DSP 2: 1/3=0.33 /10=0.03 /10=3.333333333-03

2. RTN: the Calc will stop at the following line. R/S will simply continue from there.

3. this short program with [x⁸] runs about 15 sec

Code:

LBL A
01
00
STO (i)
LBL 0
π  
8
y^x
DSZ(i)
GTO 0

Using [x²] runs in about 7 to 8 seconds

Code:

LBL A
01
00
STO (i)
LBL 0
π    
x^2
x^2
x^2
DSZ(i)
GTO 0

Günter

Edit: formatting