Posts: 4,027
Threads: 172
Joined: Aug 2005
Hi, Geir;
I have the AECROM image loaded in the NoVRAM (programmed as low-power Clonix) and I ran the example at page #49 in the manual. After entering each data I VIEWed each of the registes' contents from R55 to R43 and they had nothing but 0.0000. After pressing [D] (find the best curve fit) I found the following contents: (FIX 9)
R Contents
55 -6.748806730
54 10,408.00000
53 830.0000000
52 2.000000000
51 3,212.000000
50 33.85502276
49 30,406.00000
48 -4.580173552
47 6.574829560
46 -22.96950673
45 -50.00000000
44 6.420402010
43 370.0000000
I did not try to relate these values with the input data, but I guess that with a smaller sample (maybe three points only) we could `see` the relation faster. In any case, if the manual states that 56 reguisters are needed and CURVE has one single program line to make sure it's been set so (line #05: STO 55), then I'd guess they are some sort of 'scratch pad'. The HP15C uses some registers as scratch area (up to 24) when numerically computing SOLVE and Integrate, so...
If you need any particular test, just let us know.
Cheers.
Luiz (Brazil)