SR-50 vs SR-52/56 Logic

[Matt Agajanian]

Hi all.

I wonder... Although it wouldn’t be until the SR-52 and SR-56 that a full AOS logic would be the norm, could the SR-50/50A and 51/51A have been created with the full version of AOS?

[Gene]

I suspect probably not. The memory requirements to store pending operations and to process them likely weren't ready for the SR-50 / SR-51 timeframe.

There is a discussion in the SR-51 owners manual about "Register level processing" which tells more about the internal workings of the machine, but it was by the time of the SR-52 some 8 months later to do a full AOS... and remember that was on a $395 machine compared to the much lower priced and older SR-50/51 architecture.

My 2 cents.

[Matt Agajanian]

(04-05-2019 01:28 AM)Gene Wrote:  I suspect probably not. The memory requirements to store pending operations and to process them likely weren't ready for the SR-50 / SR-51 timeframe.

There is a discussion in the SR-51 owners manual about "Register level processing" which tells more about the internal workings of the machine, but it was by the time of the SR-52 some 8 months later to do a full AOS... and remember that was on a $395 machine compared to the much lower priced and older SR-50/51 architecture.

My 2 cents.

Thanks! That’s worth a lot more than two cents.

[Gene]

Well it probably was the internal memory shortage that hurt the most. The SR-50 had 1 and the SR-51 had 3 addressable memories. From the initial marketing standpoint, TI probably thought if they **had** any extra available memories that it would be best to give them to the user, not use them internally...plus you have to have parentheses on the keyboard and internal processing to evaluate the deferred algebraic stack... all tough when being designed in early 1974 (9-12 months before being sold for the first time?).

When the SR-52 came out, memory was cheaper. It had 20 (ok 22 when you use STO 98 and STO 99) memories plus the AOS stack.