Loading ...
Sorry, an error occurred while loading the content.

38154Re: SR Re: RPN calculators (O.T.)

Expand Messages
  • Steve Rooke
    Nov 12, 2009
      The Sinclair Scientific was introduced in 1974 and used RPN on a custom
      programmed TI TMC0805NC. This chip was designed by TI as a normal 4 function
      calculator but they implemented it in a programmable ROM design. This
      allowed Sir Clive Sinclair to shoehorn in full scientific functions. I
      cannot find a datasheet for this chip but I understand it was very small and
      he had to redesign the algorithms to get it all in. In his own words:-

      �We did it by taking an existing TI chip which was a four function
      calculator chip�, recalls Sinclair, �TI had made it internally programmable
      � you could change the ROM � but it had only three registers.�

      �We re-programmed that, much to TI�s amazement, to create a full scientific
      calculator.�

      �A friend of mine, Nigel Searle, a computer scientist, did the programming,
      and I did the algorithms because no algorithms existed at that time which
      would have worked in just three registers.�

      �We went to Texas, stayed in a hotel room, and did the whole job in a few
      days. We took the programme into TI, then went back and collected the chip.�

      �TI were completely baffled by this�, remembers Sinclair, �there was a chap
      at London University, a professor who specialised in algorithms, and he
      couldn�t figure out how it could ever be done in just three registers. He
      thought it was technically impossible.�
      I attended a seminar at Sussex University in the UK by Sir Clive and he was
      an amazing man.

      Steve

      2009/11/12 David Smart <smartware.consulting@...>

      >
      >
      > >I read an article several years back claiming the original HP-35 was
      > > programmed internally with a less-than-2K ROM; I think the number was
      > > slightly over 1500.
      >
      > Could well be. I've worked with Forth implementations that were awfully
      > tiny. Making an RPN engine can be fairly compact (and also relatively
      > straightforward).
      >
      >
      > > But then, it doesn't claim to have an infinite stack ...
      >
      > But that's not in ROM anyway.
      >
      > Regards, Dave S
      >
      >
      > ----- Original Message -----
      > From: "Vince Tessier" <vlt@... <vlt%40glempire.org>>
      > To: <sliderule@yahoogroups.com <sliderule%40yahoogroups.com>>
      > Sent: Thursday, November 12, 2009 2:25 PM
      > Subject: Re: SR Re: RPN calculators (O.T.)
      >
      > >I read an article several years back claiming the original HP-35 was
      > > programmed internally with a less-than-2K ROM; I think the number was
      > > slightly over 1500.
      > >
      > > But then, it doesn't claim to have an infinite stack ...
      > >
      > > fwonghc wrote:
      > >>
      > >>
      > >> RPN calculators do have some advantages over *algebraic* calculators.
      > >> Nevetheless it is not the paneacea.
      > >>
      > >> The HP-35S (I also bought it 2 or 3 months age), when you do
      > >> "programming" you have to enter the equation (i.e. mathematical equation
      > >> ABC = xyz + pqr ...etc). Obviously there are limitations to the RPN.
      > >>
      > >> I tried the interscholatic Numbet Tie Test (for slide rules)
      > >> with the RPN calculator (cal98, previous free download, but no lonager
      > >> available) and the HP-35S. There are squares and square roots, cube and
      > >> cube roots. May be I am not familar, (to me) it is error prone.
      > >>
      > >> With the RPN notation, the HP-12C, there is no way you can enter
      > >> calcuate say 20 / 80 % (= 35), you must change 80 % to 0.8 mentally.
      > >>
      > >> ***
      > >> (Indeed this several years there are some calculators you must enter the
      > >> whole alegbraic equation (or mathematical equation), I am not used to
      > >> it, though I have several different models.
      > >>
      > >> I enjoy using RPN because there is not need to enter brackets, and is
      > >> the natural way to calculate when "manually". You must enter some extra
      > >> "hidden" brackets even the equations do not requrie.
      > >
      > >
      > > --
      > > take care,
      > > -- Vince
      > > Speed of light is 1 799 884 800 000 furlongs per fortnight.
      > >
      > >
      > > ------------------------------------
      > >
      > > Yahoo! Groups Links
      > >
      > >
      > >
      >
      > ----------------------------------------------------------
      >
      > No virus found in this incoming message.
      > Checked by AVG - www.avg.com
      > Version: 8.5.425 / Virus Database: 270.14.61/2497 - Release Date: 11/11/09
      > 19:41:00
      >
      >
      >



      --
      Steve Rooke - ZL3TUV & G8KVD
      A man with one clock knows what time it is;
      A man with two clocks is never quite sure.


      [Non-text portions of this message have been removed]
    • Show all 69 messages in this topic