Four types of RPN

[Thomas Klemm]

(09-11-2014 10:19 AM)Marcus von Cube Wrote:  The HP 35 has limited memory and uses its T register for intermediate calculations in some transcendental functions. It makes sense to clear T after its "abuse" in order to avoid presenting garbage to the user. So as long as you don't use any transcendental functions, the behavior is "top copy", but top can be cleared on occasion.

There's an instruction stack -> a that drops the X register from the stack and moves it to register a:

X → a
Y → X
Z → Y
T → Z

There's another instruction c -> stack that pushes register c to the stack. After executing this instruction the previous value of the T register is lost.

Z → T
Y → Z
X → Y
c → X

Thus the intermediate result during the calculations of transcendental functions is saved in the X register and not in the T register.

Quote:Edit: I just read in another thread that only the trig functions are affected and that the HP 35 duplicates Z to T after such an operation.

With the instruction stack -> a the T register is duplicated to the Z register.

Cheers
Thomas