Path: utzoo!mnetor!uunet!mcvax!ukc!its63b!hwcs!zen!frank From: frank@zen.UUCP (Frank Wales) Newsgroups: comp.sys.hp Subject: HP-28 Message-ID: <1169@zen.UUCP> Date: 4 Mar 88 10:48:58 GMT Organization: Zengrange Limited, Leeds, England Lines: 52 [This is posted on behalf of a friend who has no access to Usenet; please direct any replies to one of the addresses at the end. frank] Here are some more notes on the HP-28C, HP-28S, and a book about them. 1. Careful use of SYSEVAL has led to discoveries of useful routines in the HP-28C and HP-28S ROMs, but SYSEVAL can also be used to locate the RAM, and the (separate) display RAM. Users who have gone from an HP-28C to an HP-28S may like to know that the display RAM has moved from the address range #40000 - #47FFF to #FF800 - #FFFFF on the 28S. 2. The user RAM was at #4F000 - #4FFFF on the 28C. On the 28S it is at #C0000 - #CFFFF. Although the amount of RAM has gone up from 2K to 32K, about 400 bytes are reserved for the system, so in fact the HP-28S provides 17 times more RAM available to the user! Oddly enough, the HP-28S user RAM can also be accessed at addresses #D0000 - #DFFFF. This suggests HP designed the HP-28S with two "memory controllers", each able to control 32K of RAM, but in the end they provided only one 32K block, and therefore had to hardwire BOTH controllers to it, otherwise there would have been a controller claiming to provide 32K RAM, but unable to come up with the goods. 64K RAM in a pocket calculator would have been amazing, but with no input method (except the keyboard) it would hardly have been worth it. It seems to me that with the HP-27S now available as an alternative to the HP-28C at a fairly low (for HP) price, they should have produced a more expensive version of the HP-28C with more memory AND some useful means of interfacing with offline storage. 3. The PEEK and POKE programs in "Customize Your HP-28" work on the HP-28S as well as on the HP-28C (that's why they were written in machine language - so they would be independent of ROM code changes), but on the HP-28S the command HOME should be included at the beginning of the PEEK and POKE programs so the machine code will not go into a subdirectory. As the top of RAM is now at #CFFFF, all three occurrences of #4FF in PEEK and POKE should be replaced by #CFF. 4. I shall be away for 3 weeks from March 7 to March 28, so I shall not be able to answer questions about the book. I'm sure the folks at Zengrange will carry on answering questions, and with luck we might even get some interesting answers from inside HP too. Anyone who has missed the information posted by Jurjen N.E. Bos would be well advised to go back and read it - fascinating stuff for the HP hackers among us. Many thanks to Jurjen for posting it. W.Mier-Jedrzejowicz, Ph.D., EARN/BITNET: MIER@SPVA.PHYSICS.IMPERIAL.AC.UK JANET: MIER@UK.AC.IMPERIAL.PHYSICS.SPVA UUCP: uunet!mcvax!ukc!spva.physics.imperial.ac.uk!mier Smail: Physics Dept, Imperial College, London SW7 2AZ, England Disclaimer: Neither my employer nor HP are to blame. As for me, I accept no responsibility for what use anyone might make of my ravings. Just the same, have fun!