PDP8: Anybody with RALF assembler experience?

Hi folks, I recently added a LAB-8/e (PDP-8/e) to my collection. Most of the machine is just running fine (except that I still urgently need at least 8K of core!!). One of the two racks contains the FPP-12 floating point processor. So I can write programs in FORTRAN IV, compile them and run them on the FPP. An awsome device - with it, the 8/e becomes a real dual core machine as the FPP-12 fetches its own instructions from core by data breaks (getting at most half of the 8's memory bandwidth). Now I want to run my Mandelbrot set program written for the H316 on the 8/e with FPP as well. The FORTRAN IV code was quite compatible (had to adjust some of the FORMAT statements and IO channel numbers). But now I want to write a "putchar" routine needed by the program. On the H316 that was simple: One CPU, perfectly documented calling conventions etc. On the 8/e the FORTRAN IV compiler produces output for the RALF assembler. RALF is capable of producing machine code for PDP-8 *and* the FPP as well. The main part of FORTRAN IV programs (nearly everything except IO) runs on the FPP. Subroutine calling and argument passing are also done on the FPP. I now want to know if there are people out there who are able to help me to program a simple PDP-8 assembler subroutine that can be called from within a FORTRAN IV program. Sounds easy - but isn't. I have written a template function and then compiled to RALF assembly - I don't understand very much :-( If anyone has experience with the FPP, RALF and "trapping" (they really call it that way) to PDP-8 code, **please** get in touch with me! Many thanks and best wishes, Philipp :-) P.S.: In FORTRAN IV the desired subroutine would have a signature like "SUBROUTINE PCHAR(I)"