HP15c CE questions

[DA74254]

I recently received my HP15c CE from The Calculator Store, including the cable and HP16 overlay.
Looking forward to be aquainted with it.

The questions:

Is there any benefit, apart from more storage, using the 15.2 over the 15 mode? Do I "lose" anything from switching to 15.2? I'm aware that using/reverting between 15(.2) and 16 mode will erase the settings/programs. With the cable, I assume I can store and restore states/memory from PC.

I ran the 9 sin cos tan atan acos asin "torture test". Result was 9,000417, which is "less" accurate than my other HP's (8,9999986). Is the precision still 10 digits as in the original?
(OK, my Soviet Elektronika MK-61 is 9,088 so probably 6 or 8 digit precision there).

I'm not "dissing" the calculator as I understand the logics and are happy that the calc doesn't give me "the answer I want". Just want to have confirmed the precision. I also have several SM calcs with 34 digit precision, which have an accuracy beyond all my ever possible needs. (Even NASA uses "only" 16 digits precision).

I'm sure that as soon as I get used to the landscape form factor, this one will be my daily driver.

[AnnoyedOne]

(09-28-2024 06:06 PM)DA74254 Wrote:  Is there any benefit, apart from more storage, using the 15.2 over the 15 mode?

Not that I know of. I keep mine in 15.2.

(09-28-2024 06:06 PM)DA74254 Wrote:  Do I "lose" anything from switching to 15.2?

No. You can also switch from 15.2 back to 15 unless you're using too much memory.

(09-28-2024 06:06 PM)DA74254 Wrote:  With the cable, I assume I can store and restore states/memory from PC.

Yes.

(09-28-2024 06:06 PM)DA74254 Wrote:  Is the precision still 10 digits as in the original?

The HP-15C CE emulates the original 15C firmware so it's the same.

A1

[naddy]

(09-28-2024 06:06 PM)DA74254 Wrote:  I ran the 9 sin cos tan atan acos asin "torture test". Result was 9,000417, which is "less" accurate than my other HP's (8,9999986).

9.000417403 on my 1985 vintage HP-15C, which is to be expected, since the CE runs the original firmware in emulation.

[DA74254]

Thank you for your replies.
I start to like this calculator more and more. Just have to get used to the programming and the one-line display, including the non-floatong-point decimals. (I guess that FIX 4 will be my new normal).

[naddy]

And you are encouraged to switch the display to whatever makes sense in your current working context. I typically have it at ENG 4, but for money things I switch to FIX 2. It's just three keystrokes.

I think it is important to understand that those old HP calculators were conceived as engineering tools. When dealing with actual technical-physical systems, the number of significant digits is always limited. You never have, say, a perfect number "3", you have, for instance, 3.000 +/- 0.005. Even if some mathematical operation spits out a perfect "3", e.g. if you take the derivative of x^3, in an engineering context you will always end up feeding values into your formula that have been measured to a limited number of significant digits.

This engineering approach is made explicit in a number of places, e.g. for the matrix operations, where an inverse can be calculated for any matrix, even one that can't be inverted in a strict mathematical sense. From page 150 of the Owner's Handbook: If the specified matrix is a singular matrix (that is, one that doesn't have an inverse), then the HP 15c modifies the LU form by an amount that is usually small compared to round-off error. For 1/x, the calculated inverse is the inverse of a matrix close to the original, singular matrix.

[Johnh]

I'm enjoying it too. I think with its simple, compact and accessible layout, and very clear though limited display its best role these days is for ad-hoc calcs, backed up with the extras such as Solve, Integrate, Matrices and Complex numbers. I find programming it is best used for a few very familiar routines that help with what I frequently do.

The main remaining limitation for me is the lack of Alphanumerics, to help remember the inputs and outputs of a program. The power of the maths and programming is plenty to set up some quite sophisticated and extensive routines, if you can remember how they work!

Now that Save to PC is available, I'm thinking to make a card to keep behind the calculator, with a summary of input and output order or memory references for my main programs.

The use of an Hp15c is also greatly enhanced by the availability of several very good emulators, to use on a mobile phone.

[DA74254]

(09-28-2024 09:41 PM)naddy Wrote:  I think it is important to understand that those old HP calculators were conceived as engineering tools.

Yes, that is why I like the HP scientific range of calculators.
I'm a mechanical/shis engineer and it is quite a few calculations in my work that is so much easier with RPN calculators. Programming routines where one can set a variable with STO x and run multiple calculations through routines with RCL x and save 5-10 keystrokes for each calculation both saves time and reduces the fail rate of mistyping a repetitive value.
I have small routine to convert cubic metres into metric tonnes based on (a variable) density for fuel when bunkering multiple fuel tanks and which must be specified in MT in the oil record book. Also, calculating fuel consumption for our voyage, in daily consumption and total consumption, is a breeze with preprogrammed routines.

[AnnoyedOne]

(09-28-2024 07:51 PM)DA74254 Wrote:  I start to like this calculator more and more.

Party like it's 1982. Or something

(I bought my HP-15C, brand new, in March 1982 and it still works!)

Sometimes newer isn't "better".

(09-28-2024 09:41 PM)naddy Wrote:  When dealing with actual technical-physical systems, the number of significant digits is always limited.

Exactly! Scientists use 8-9 significant digits. Mathematicians use as many as they can get. Engineers use 3-4. Components have tolerances and equations have assumptions. The trick is having enough knowledge/experience to judge.

A1

[naddy]

(09-28-2024 06:06 PM)DA74254 Wrote:  I ran the 9 sin cos tan atan acos asin "torture test". Result was 9,000417, which is "less" accurate than my other HP's (8,9999986).

BTW:

Code:

g RAD 9 f →RAD SIN COS TAN g TAN⁻¹ g COS⁻¹ g SIN⁻¹ g →DEG
8.999999955

[Thomas Klemm]

(09-28-2024 06:06 PM)DA74254 Wrote:  I ran the 9 sin cos tan atan acos asin "torture test".

We should rather call it Calculator Forensics.