TI SR-50/51 AOS Possible?

[Matt Agajanian]

Hi all.

Although TI engineered the SR-50 as the competition to the HP-35, Arithmetic (Sum of Products hierarchy) is kinda cumbersome. But at least three levels of parentheses and pending operation would be something better at problem solving.

Although full AOS was a later reality with the SR-52 & SR-56 would TI have been able to create the SR-50 & 51 maybe with three parentheses/pending operations at the time of the SR-50?

[Steve Simpkin]

I agree that these early TI models without parenthesis keys were not easy to use on complex calculations. The mental gyrations needed for calculating parenthesis heavy formulas could be daunting if you didn't break down the equation into simpler subsets first. Even then, this may have required intermediate results to be written down and re-entered in order to solve some problems.

I have always thought that the reason they didn't implement parenthesis keys on these early models was because of transistor count. It takes approximately 4-6 MOS transistors to provide a bit of storage in these IC designs. If a register occupies about 56 bits (as in the early HP models) then that translates into approximately 200-300 transistors per register. The early MOS LSI IC designs in this era were often limited to around of 2000-6000 transistors per IC. The "cost" of adding 5-7 extra registers needed for a good parenthesis implementation on top of the existing working registers was probably just too great for these early TI models.

Conversly, HP models, with their RPN logic and a four level stack, were able to provide a workable entry system that allowed most problems to be solved without writing down intermittent results. The methods used to break down a problem and to work it from the inside out also had the added advantage of mimicking the methods you would have used to solve the problem with a pencil and paper (or a slide rule).

[Gene]

Essentially, no.

The SR-50 was introduced on January 15, 1974 and the SR-51 in January of 1975.

The first ** AOS ** model, the SR-52, was not introduced until September 1975.

Those time gaps were immense in terms of continued progress in the calculator game.

Absolutely no way AOS could have been on the SR-50 or SR-51 when introduced. They were already pretty expensive (although cheaper than the HP's).

[Steve Simpkin]

To expand on the above, incorporating the rules of precedence required for AOS entry would have also required more code space as well as registers for intermediate results. The fact that TI didn’t introduce a model with AOS until September 1975, a full 3-2/3 years after the HP-35, really drives home how difficult it was to implement at the time.

[Maximilian Hohmann]

Hello!

Apart from the question "was AOS a technical possibility in 1974?" another question comes to mind: "was the concept of AOS already around in 1974?". Or did AOS evolve in parallel with the hardware, possibly as a request of calculator testers and users. I do not know if there were desktop calculators before the SR-52 that implemented AOS. A bit like RPL which was invented together with the first RPL calculator.

Regards
Max

[ijabbott]

(04-15-2022 03:55 AM)Steve Simpkin Wrote:  To expand on the above, incorporating the rules of precedence required for AOS entry would have also required more code space as well as registers for intermediate results. The fact that TI didn’t introduce a model with AOS until September 1975, a full 3-2/3 years after the HP-35, really drives home how difficult it was to implement at the time.

And Casio didn't implement the equivalent until the introduction of the fx-39 in 1978!

[Gene]

Well, one minor clarification (IMO) :-)

The implementation of AOS was likely ongoing, but the ** idea ** of AOS was to evaluate entered expressions using (fairly) standard hierarchy.

The SR-50/51 implemented the sum of products (M/D before A/S), which was a big improvement for algebraic models at the time, but it was the rather large AOS stack of the SR-52 and its descendants that allowed AOS to evaluate more than simple sum of products.

Now keeping track of what was stored in that AOS stack was not simple at all...but that's a different story and a big reason I much prefer an RPN stack of some sort.

[Stevetuc]

The Commodore SR36 incorporated parenthesis. The first version was introduced in 1973.
https://sliderulemuseum.com/Calculators/...Manual.pdf

https://oldcrap.org/2019/12/12/commodore...tor-sr-36/

I remember in '74 sitting in class with my SR50 and looking enviously at my classmate's SR36 with [ ] keys!

[Dave Britten]

(04-15-2022 02:48 PM)Gene Wrote:  Well, one minor clarification (IMO) :-)

The implementation of AOS was likely ongoing, but the ** idea ** of AOS was to evaluate entered expressions using (fairly) standard hierarchy.

The SR-50/51 implemented the sum of products (M/D before A/S), which was a big improvement for algebraic models at the time, but it was the rather large AOS stack of the SR-52 and its descendants that allowed AOS to evaluate more than simple sum of products.

Now keeping track of what was stored in that AOS stack was not simple at all...but that's a different story and a big reason I much prefer an RPN stack of some sort.

What was the largest they ever made their AOS stack? I know the 58/59 have something like 8 pending values, but I don't recall what the 95 has off the top of my head.

My order of preference tends to be expression entry (with playback of at least the last entry, if not more), RPN, AOS. Though there are occasionally use cases where that order might change slightly.

[Steve Simpkin]

Gene, for reference the TI-52, TI-58/59 & TI-66 supported 9 levels of parentheses. The TI-95 Procalc supported 15 levels. The original TI-30 supported 9 levels of parentheses but only 4 pending operations!

The TI-36 Pro is limited to 23 levels of parentheses with up to 8 pending operations. TI’s graphing models appear have parentheses levels that are only limited by memory.

The Commodore SR-36 mentioned above supported 2 levels.