accuracy of integration and solve routines HP 15C LE

[Jan]

Hi,

Has anybody tested the integration- and/or solve-function of the 15C LE? And if so, do they seem work properly or not? Doing a lot of algebraic work I primarely use my calculators for checking results. With an increase in speed of 100 times the speed of the original 15C (of which I got 2) and therefore even exceeding the speed of my 32SII and 48G, the 15C LE seems a perfect tool for the job. That is, when I can rely on it. Thanks for your respons.

[Marcus von Cube, Germany]

The accuracy of the 15C LE is exactly the same as that of the original 15C because it runs the same software. All problems are related to the operation environment of the emulation (display, battery, keyboard).

[Jan]

Thanks for the info. I had the idea though that the original firmware had to be reverse-engineered since the original documentation was lost at HP. Or is that just a rumour? If not, was the reverse-engineering done in an appropriate way?

[Marcus von Cube, Germany]

I'm not the most qualified to answer your question but I doubt that any reverse engineering was really necessary. The original code is run in a virtual environment as if the nut processor were still there. It's not, the instructions are interpreted by some software running on the ARM chip. This piece of software not only has to implement the processor but all of the peripherals, too, and map them to the real hardware. Here the trouble has its origins.