Got a 32s, musing about the 33s


I managed to snag a nice 32s off of "the auction site." It's got a couple of scratches on the bezel that kept the collectors from bidding it up too high, and everything works. So far I'm very pleased. I *love* the Pioneer form factor and keys.

I don't have a manual yet, so I downloaded HPs PDFs of the 32sii and 33s manuals to figure things out. I was fooling around with the RPN solver and the TVM equation Jeff Kearns posted last month (thanks, Jeff!). And something dawned on me.

The 33s has 32K of memory. People have balked because it has the same limited number of variables and labels as its ancestors. But then I realized what the memory is for--equations.

It takes a lot of equations to make much of a dent in that 32K, right? You can easily store variations of your most-used equations to solve for the variables you need, and run them directly, without ever having to use the Solver. Meanwhile, you've still got plenty of program space when you need it.

Now, if only we could retrofit a 32K memory chip in 32s and 32sii. And all the best calculators could benefit from a a USB port to plug in an SD card reader/writer for backup and restore. I know the latter will never fly on the current models due to the exam certification requirements. But hey,I can dream, can't I?


You're invited to a little roaming through the museum archives to see where dreaming may take you to.

Ceterum censeo: HP, launch a 43s.



Haha, there was an enormous amount of unfounded griping about the lack of variables in the 33s (like, go get a 48GII or a 50G!). Like, the 32sii was so limited in memory, you could never use all the variables and labels, so with increased memory everyone's griping the other way!

Yes, the real bonus is the equations and simply the inexhaustible space for equations or programs. Even better, you can use equations as labels (that is also true in the 32sii which you don't have on the 32s). With essentially unlimited memory, you can put in long descriptive equations which actually make it possible to keep infrequently used equations or programs in memory with some chance to actually know what they are and how they work!

All of this also applies to the 35s except that it has a huge number of indirect memory addresses.

Edited: 23 Feb 2009, 11:31 a.m.


Let me guess. It's the maximum deflection at the center of a simply supported beam with a uniformly distributed load, where W=wL = total load and there should be a 384 in the denominator to the right of the ExI. Just asking out of idle curiosity; I'm a retired civil engineer.

As far as memory is concerned, the HP-32SII has a puny 384 bytes of memory, but the HP-42S has a reasonable 7K of memory, which has always been plenty for me. All my long programs are loaded onto my HP-48SX, which has 160K memory with an expansion card and serial communication with my DOS PC. I just bought a used HP-50g on "that auction site", and am looking forward to experimenting with it.


I find the solver in the 33S ideal in that it makes direct solutions where possible. It is sad that someone found it necessary to "style" it so it looked like a rocket ship and made it hard to repair due to the rubber side strips which rules out a slipcase like former calculators and led to the bulky horrid case. Sam


As far as memory is concerned, the HP-32SII has a puny 384 bytes of memory, but the HP-42S has a reasonable 7K of memory,

The HP-32S had an already-miserly 390 bytes, then its successor HP-32SII lost six bytes even as the capability to store equations was ported from the HP-22S. This made the RAM even more inadequate.

The programming paradigm was good enough on both models to make it worthwhile to retain crafted user programs in permanent residence. Unfortunately, that can leave little room for other things -- very annoying.

I'd say that 2 kB would have been just right for the HP-32S/SII -- including 216 bytes reserved for 26 letter variables and the indirect register "i", so as not to illogically declare a variable nonexistent if its value happened to be zero.

Why couldn't the RAM from the HP-28C have been used?

The HP-32SII does offer a byte-saving trick: Non-negative integers not exceeding 254 can be stored in a program or equation using only 1.5 bytes, instead of the 9.5 bytes that all other numbers would require.

Here's an equation-based example for calculating balanced 3-phase apparent power [in MVA] from phase-to-phase rms voltage [in kV] and rms current [in A]:

M=K*A*SQRT(3)/1000      26 bytes

M=K*A*SQRT(3)/100/10 21 bytes

The latter runs a bit slower, but saves five bytes of RAM because "100/10" consists of three 1.5-byte objects, while "1000" is a 9.5-byte object.

Users really ought not to have to resort to tricks like that, though...

-- KS

Edited: 24 Feb 2009, 3:16 a.m.


All of this also applies to the 35s except that it has a huge number of indirect memory addresses.

Also the HP35S has the direct GTO line number feature that can save scarce label resources.

Possibly Related Threads…
Thread Author Replies Views Last Post
  HP 32s LCD part Trisnadi Sutrisno 2 1,401 08-27-2013, 02:12 PM
Last Post: Han
  HP-32S II Expansion Matt Agajanian 14 3,693 07-21-2013, 10:19 PM
Last Post: Matt Agajanian
  HP 32S-II Vertical Curve Program Ron Cardwell 2 1,368 05-20-2013, 07:54 AM
Last Post: Thomas Klemm
  HP-32S/33S/35S emulator with Stack Overflow sensing/Stack display x34 3 1,577 10-26-2012, 04:56 PM
Last Post: x34
  HP-41's ARCL on the 32S-II Matt Agajanian 2 1,173 05-04-2012, 04:45 PM
Last Post: Matt Agajanian
  Tips & Programs for the 32S Matt Agajanian 0 845 03-16-2012, 02:24 AM
Last Post: Matt Agajanian
  Just how tough is a 32S-II ? Matt Agajanian 12 3,297 03-11-2012, 05:36 AM
Last Post: Thomas Radtke
  HP 32S style ridgid case Anoop Sahal 0 828 05-10-2011, 04:08 PM
Last Post: Anoop Sahal
  HP 32S Successor ka6uup 10 2,510 07-30-2010, 09:20 PM
Last Post: db (martinez, ca.)
  optimized prime factor finder for HP-32s Don Shepherd 6 2,015 07-07-2010, 06:23 PM
Last Post: Don Shepherd

Forum Jump: