HP Forum Archive 20

[ Return to Index | Top of Index ]

Chip-Level Question Message #1 Posted by Jeff on 18 Mar 2011, 11:18 a.m. I was wondering how HP calculators arrive at answers. With a PC, of course, it uses the ALU part of the CPU. But how about calculators? Is it software embedded in ROM? What about simple 4-function units? Thanks.

Re: Chip-Level Question Message #2 Posted by Eric Smith on 18 Mar 2011, 3:53 p.m., in response to message #1 by Jeff Most calculators have software in ROM, often called "firmware". In the early days HP and TI firmware was referred to as "microcode". Some of the early four-function calculators were hardwired rather than using a processor, but that went out of style quickly once it became relatively easy and inexpensive to use a processor. In the past most calculator processors were four-bit, with HPs use of a 56-bit word being a notable exception. AFAICT, most calculators now use 8-bit processors, with high-end calculators using 16-bit or 32-bit processors.

Re: Chip-Level Question Message #3 Posted by Marcus von Cube, Germany on 18 Mar 2011, 4:06 p.m., in response to message #2 by Eric Smith The newer ARM based HPs (12C, 20b, 30b) use a 32 bit core. The "processor" is in fact an all in one solution with all necessary functionality on one chip, including memory. TI introduced this technology in the early days of the TI-30, the LED model. A single chip, bonded directly to the board which carried the LED bar on its back, contained the complete calculator. The functionality has of course vastly increased over the years. We can assume that the chip contains a processor and its firmware but you can't tell from the outside.

Re: Chip-Level Question Message #4 Posted by Eric Smith on 18 Mar 2011, 4:25 p.m., in response to message #3 by Marcus von Cube, Germany Actually the single-chip calculator predates the TI-30 by almost five years (September 1971, vs. June 1976).

Re: Chip-Level Question Message #5 Posted by Marcus von Cube, Germany on 18 Mar 2011, 4:39 p.m., in response to message #4 by Eric Smith At the age of 12, my interest in calculator electronics hadn't sufficiently developed to recognize this. ;)

Re: Chip-Level Question Message #6 Posted by Mike Morrow on 19 Mar 2011, 1:48 a.m., in response to message #4 by Eric Smith Joerg Woerner's interesting TMS-0102 webpage gives some details of the early one-chip design (minus display drivers) that was in many four-function machines of 40 years ago. Including my first, the Bomar 901B that cost me the 2011 equivalent of $700.

Re: Chip-Level Question Message #7 Posted by Joerg Woerner on 19 Mar 2011, 9:43 a.m., in response to message #4 by Eric Smith How "single" is single-chip??? The TI-30 included the display drivers while the TMS1802 (1971) needed four additional integrated circuits to drive the LED's. The TMS08** (1974) needed one display driver and the TMS9** (1975) was the first "single-chip" solution from TI. Think TI-1200! Regards, Joerg

Re: Chip-Level Question Message #8 Posted by Eric Smith on 19 Mar 2011, 4:29 p.m., in response to message #7 by Joerg Woerner I seem to recall that there were single-chip calculators using the Mostek MK6010 by mid-1971. However, they did use some discrete transistors as well.

Re: Chip-Level Question Message #9 Posted by Frank Boehm / Germany on 18 Mar 2011, 7:00 p.m., in response to message #3 by Marcus von Cube, Germany The bonded die has been used by National Semiconductor as early as 1975 (Novus 650). Novus 650 (Even if the page states 1976, the date code 540 aka 7540 is clearly visible.)

Re: Chip-Level Question Message #10 Posted by Joerg Woerner on 19 Mar 2011, 9:53 a.m., in response to message #9 by Frank Boehm / Germany Two chips below the blob ;-)) Joerg

[ Return to Index | Top of Index ]

Go back to the main exhibit hall