Best calculator for the working engineer

[Vtile]

(02-27-2016 06:21 PM)walter b Wrote:  A good engineer can use a lot, though some tools suit him better than others.

d;-)

I would still rather buy a good hammer from a smith than making one decent myself. (different thing ofcourse if the smithing is also a hobby)

Best calculator for engineer is still not produced, it would be a pocket size 50g with keyboard stencil ability (like in 41 models).

[Tugdual]

Rush on the 50g while it is still on the market...

[Logan]

This discussion has been fascinating. I'm in much the same position as the parent poster.

I've always enjoyed calculators and for the last 10 years or so have been using the TI-89 regularly. I knew one student who had an HP-50g and I'd always been curious to try it.

Recently I decided to take the PE exam and after looking at the approved calculators and comparing them, decided to buy the TI-36+ Pro. It's a good calculator, it has a lot of functionality. The problem for me was that the screen refresh was very sluggish when entering long equations, and that everything is so graphically based that it takes a long time to do anything.

I was familiar with the concept of RPN and decided to try the HP 35s, figuring I could use algebraic mode if I wanted. I liked the calculator, read through the entire manual, and decided I'd give RPN a go. It makes sense to me but I can also see why it's not widely adopted, especially with things like the TI-89 around.

Anyway, having spent a good bit of time with the 35s, I recognize it has a lot of very nice quick-access features, but it also has a few shortcomings. I've entered Stefan's matrix multi-tool, written a few programs of my own, added summation and differentiation routines, expanded the quadratic solver (the default one seems a bit lacking). It's been good but I've now been hunting for the ultimate scientific calculator.

I like the idea of the WP 34s, it just seems to suffer from low-quality base material and I'm dubious that the stickers will hold up. Smart phone apps are handy but just don't do it for me. I'm a EE so the 42s looks great but prices of $200+ for used ones seems extravagant. I was very interested in the Swiss Micros 42L, either the landscape or portrait version, and would definitely consider one, even if that ends up being $200.

So yes, I like the 35s quite a bit, it's replaced my TI-89 for most calculations (I love the quick conversions, for example) but it just seems to be a bit lacking in some areas.

This is a very interesting market, it seems like there would be more people like Swiss Micros who were working on the "ultimate scientific calculator" but then again, I guess there isn't too much of a market for that, at least not RPN ones

[Ron Ross]

Well, for the PE exam, you are limited to only a few select calculators and you already have the two best calculators available for that exam.
.
I suggest you buy an Hp 50G to try, they are fairly cheap to buy via Amazon, new even. I am kind of partial to the Hp 48G line and they are also very cheap.
.
For an EE, the Hp 42s is indeed a treasure. However, your spending $200 to find out that you think RPN is dated and unnecessary would certainly be a big waste of cash. An Hp 48G would give you a similar experience with a calculator of similar capability for about $50. If you feel that the Hp 48G is GREAT!, then consider splurging (or hunting) for the much smaller and portable Hp 42s (and knowing that you will like it, or simply "meh, I have a 48G and its okay").
.
The Hp 48G has the exact same lower portion keyboard (and keyboard construction) as the Hp 42s, aside from RPL vs RPN (but you would get that comparison from the Hp 48G vs the Hp 35s anyway.

[rprosperi]

(07-05-2016 10:00 PM)Ron Ross Wrote:  The Hp 48G has the exact same lower portion keyboard (and keyboard construction) as the Hp 42s, aside from RPL vs RPN (but you would get that comparison from the Hp 48G vs the Hp 35s anyway.

The 48G and 42S keyboard layouts are similar, but absolutely *not* exactly the same. Perhaps Ron is referring to the double-size ENTER key being in the same classic-HP location, but many of the functions are in different places and the 42S does not include many of the 48G functions. Keyboard construction is also similar, though again not exactly the same, but both do have that quality HP feel.

The difference between RPL and RPN is a far bigger difference - Suggest you read some of the many articles here which compares and contrasts the 2 styles. Fundamentally, RPN is keystroke programming, where a program is a collection of the exact keystrokes you would use to manually do a calculation, whereas RPL is a full programming language. RPL has a much steeper learning curve, though the trade-off is it is also far more powerful and flexible. In addition, the 48G has a much larger set of built-in functions, a flexible solver, unit management, matrices, complex number support, etc.

[Logan]

(07-05-2016 10:00 PM)Ron Ross Wrote:  Well, for the PE exam, you are limited to only a few select calculators and you already have the two best calculators available for that exam.
.
I suggest you buy an Hp 50G to try, they are fairly cheap to buy via Amazon, new even. I am kind of partial to the Hp 48G line and they are also very cheap.
.
For an EE, the Hp 42s is indeed a treasure. However, your spending $200 to find out that you think RPN is dated and unnecessary would certainly be a big waste of cash.

I'm very aware there are only a few select calculators available for the PE, I mulled over the list for probably far too long

I do have an HP 50G which I am looking forward to learning to use, but it's not my priority right now.

As for RPN, I should have been more clear. I am dedicated to using the 35S for this coming exam, in RPN mode. I don't mind it at all, I enjoy it and see why it is appealing (I also taught myself to use the slide rule and I think that chaining calculations translates naturally from the one to the other). My grandfather used a 15C, which I inherited a while back. So RPN is fine, I just understand why people give funny looks when one talks about it. But I'm thinking of after the exam that I would like something even better than the 35S for general use, and the RPN calculators seem to be most designed with the engineer in mind, the "power user", so to speak.

So the 42S is very intriguing. I played around with Free42 a bit but really need to read the manual before I feel comfortable with it. I'm really hoping the build quality of the Swiss Micros 42 is good, and I wouldn't mind the landscape version if the function keys are intuitive. The WP 34s also looks good functionally but I can't get around the hardware

[Marcio]

If one is patient enough, knows how to get around (which is easy) the 22 bugs (http://www.hpmuseum.org/cgi-sys/cgiwrap/...i?read=735) and is willing to program the 35s to compensate for what it is missing, it can become quite the calculator.

Also, if you like the 35s, I recommend reading this thread.

[Thomas Radtke]

(07-06-2016 02:06 AM)Marcio Wrote:  [...] knows how to get around (which is easy) the 22 bugs [...]

Some bugs depend on the state the machine is in. I call this instability. This instability reveals serious problems not only in the implementation of certain functions, but also in a lower operating level where things like e.g. memory allocation happens. I'm under the impression the firmware contains quite serious problems in pointer arithmetic.

I'm using the 35s since 2007, but came to the conclusion that you cannot trust it.

Yes, the 22 bugs, if that's the current known number, aren't necessarily a problem.

[Logan]

(07-06-2016 02:06 AM)Marcio Wrote:  If one is patient enough, knows how to get around (which is easy) the 22 bugs (http://www.hpmuseum.org/cgi-sys/cgiwrap/...i?read=735) and is willing to program the 35s to compensate for what it is missing, it can become quite the calculator.

Also, if you like the 35s, I recommend reading this thread.

Thanks, I already did read that thread, and looked at all the 35s programs in the program board and on the site. I've entered or created every program I thought I needed or thought was lacking and read the User's Guide cover to cover. I feel like I'm pretty familiar with the 35S's capabilities

Writing a 400+ line program to handle matrices (I modified Stefan's) isn't the best solution in my opinion. It's cumbersome to do and could easily be erased. Ditto for summation functions, derivatives, cross-products and better quadratic solving. Plus, single-letter labels are pretty restrictive. It is a very useful calculator, I just can't help wishing it was better.

[Marcio]

(07-06-2016 05:43 AM)Thomas Radtke Wrote:  I'm using the 35s since 2007, but came to the conclusion that you cannot trust it.

I know this is a bit off-topic but would you be kind enough as to elaborate on this?

Thanks

[Thomas Radtke]

(07-07-2016 11:21 AM)Marcio Wrote:  I know this is a bit off-topic but would you be kind enough as to elaborate on this?

It isn't, but I don't know how to be more clear than I have been except for the more or less trivial conclusion that if so many bugs have been found already (most soon after the release of the 35s), you'd expect many more waiting.

Still, it's a nice toy and represents a good concept for a true 32SII successor (which the 33s wasn't).

Ok, Marcio?

[Guenter Schink]

(07-06-2016 01:47 AM)Logan Wrote:  So the 42S is very intriguing. I played around with Free42 a bit but really need to read the manual before I feel comfortable with it. I'm really hoping the build quality of the Swiss Micros 42 is good, and I wouldn't mind the landscape version if the function keys are intuitive. The WP 34s also looks good functionally but I can't get around the hardware

The DM42 will be portrait version. Almost the same layout as the original 42S, plus a few additional keys to improve on the U/I. Mock up

Günter

[Logan]

(07-07-2016 12:32 PM)Thomas Radtke Wrote:  the more or less trivial conclusion that if so many bugs have been found already (most soon after the release of the 35s), you'd expect many more waiting.

Here are two examples just from last night:

First, the vector syntax bug popped up for me, where any vector I entered just gave a "syntax error" response. I found a sort-of workaround but no idea why it works or what causes the bug.
https://youtu.be/Amkf4pEzMoE

Second, I also was trying to solve an equation last night with the result "Bad Guess" and it wouldn't solve. After trying a bunch of things (all variables were cleared, stack was cleared, X-register input was close to the correct value etc.), it still wouldn't work. Entered the exact same equation on the line below and it solved it right away. Tried it with the original again (not changing anything) and it solved it too.

Things like this don't inspire confidence, especially when you intend on using these functions for a timed exam. I don't have time to troubleshoot during an exam.

[Pekis]

(07-08-2016 07:21 PM)Guenter Schink Wrote:  

(07-06-2016 01:47 AM)Logan Wrote:  So the 42S is very intriguing. I played around with Free42 a bit but really need to read the manual before I feel comfortable with it. I'm really hoping the build quality of the Swiss Micros 42 is good, and I wouldn't mind the landscape version if the function keys are intuitive. The WP 34s also looks good functionally but I can't get around the hardware

The DM42 will be portrait version. Almost the same layout as the original 42S, plus a few additional keys to improve on the U/I. Mock up

Günter

It begins to look like a Casio fx 5800 P ... ?

[Guenter Schink]

(07-09-2016 05:53 PM)Pekis Wrote:  It begins to look like a Casio fx 5800 P ... ?

If you think a Ford Escape looks like a Chevy Tahoe, then you're right.
Günter

[Dwight Sturrock]

(07-08-2016 07:21 PM)Guenter Schink Wrote:  The DM42 will be portrait version. Almost the same layout as the original 42S, plus a few additional keys to improve on the U/I. Mock up

Günter

Based on Swiss Micros offerings and promises, maybe the best road for the "working engineer" would be any 15C, plus the Free42, followed by the Soon to be offered DM42. Those 2 models have always seemed to me the best suited for work in electronics/ electrical engineering & technology.

The only problem I have had with the HP42S was that they only lasted so long, and by then they were out of production.

[Pekis]

(07-09-2016 09:21 PM)Guenter Schink Wrote:  

(07-09-2016 05:53 PM)Pekis Wrote:  It begins to look like a Casio fx 5800 P ... ?

If you think a Ford Escape looks like a Chevy Tahoe, then you're right.
Günter

It will be worth comparing features of both, and I don't think there will be so many differences, apart from better complex numbers handling and usb transfer... in favor of DM 42.

[Vtile]

(07-10-2016 12:42 AM)Pekis Wrote:  

(07-09-2016 09:21 PM)Guenter Schink Wrote:  If you think a Ford Escape looks like a Chevy Tahoe, then you're right.
Günter

It will be worth comparing features of both, and I don't think there will be so many differences, apart from better complex numbers handling and usb transfer... in favor of DM 42.

Well one thing is for DM42 favour, it is open platform in many ways..

---

On the other hand I feel this conversation (as many others a like) is inrelevant in way that they are many ways looking behind. Even DM42 as a replacement of HP42S.

Also I find somewhat amusing the conversations in the past that said that the handheld calculators are dead in all aspects except for education market and the future is for PDAs. The reason I feel it is amusing is that the radio were the victim of similar conversation when tele-vision system were invented and yet now 80 or so years later the cars (which use 100+ years old power sources for a good reason) are delivered with a radio! Same story have been with the paper books, it is invention that in current form have been around for last hmm 550 years.. and still going strong and it will not go away any time soon it is optimal set of features in many aspects for many situations over the ePaper / digital data. (Ie. micropower, compatibility and self-life) I have one book that is 300+ years old and still in working condition, by just turning it on. Same goes for the classic physical device called handheld pocketcalculator, it does have very good set of features and it is purposedly build device (should be precision instrument) a tool for certain tasks. While PDAs even in current form of "NonSmartPhones" and "TableTopTablets" are merely a swissarmy knifes.

I said that many of the conversations are looking behind, I will also look behind, since I will be talking about 50g. On the other hand I will not talk about big enter key, which for some seems to be "the feature" of the calculator, while arguably it is the "trademark" feature of old HPs.
We should take the 50g for a baseline, as it is extremely versatile system, but... It is based on 25 year old technology (Saturn emulation). ..But I don't need the versatily if I have matlab/GNUOctave/mable/mathematica etc.. that extra feature you do not need is needed a collague in different department his add-hoc / mock-up / prototyping and even you might see it handy in some opscure situation you can not see beforehand. Speed, speed, speed can be seen in many topics discussing the usability or the value of the calculator (ie. 8-queens benchmarks etc.), I would rather say battery life, battery life, battery life. That is something more valuable to consider for something like pocketcalculator than speed comparable or exceeding of modern portable phones and tablets or even PCs. Even the 50g do have plenty if it wouldn't run on 25 year old technology (forementioned Saturn emulation), in fact it is in pretty good position in some speed benchmark listing made a member of MoHPC forums, when running at ARM ASM. I went through my highschool and lower uni.degree with Ti83+ (which were pretty lame compared pretty much everyone elses Ti89tits and a few HP49geesh), then when I two years ago did jump back to school bench again I got one of those Casios FX-CP400s which is an hybrid of the PDA and calculator, It were all except a nice user experience, it weren't as versatile as PDA (touchscreenphone) and weren't as purposedly build tool as ie. 50g or even Ti83+ calculator. On the other hand it is purposedly build teaching device, not serious calculator. From there I did jump to 50g (to find out it is 25 years old device inside) and also have 35S that is sub par device of 21st century (except it does have a big ENTER, big EXIT could be more usefull though).

I could continue this ramble to infinity so it is better to stop.

All I want to say that 50g even in its discontinued state is the (last) calculator for the working engineer. Unfortunately its formfactor and mentality in many ways is from the beginning of 1990s, not 2010s. The calculator for a working engineer in my eyes is $4 scientific or 50g-level of calculator, there is nothing between these days (of course with exceptions). The grey mass between is for education and certificate tests.

Why? Well 50g style because versatily, memory, real programmability (and fast scripting with UserRPL), CAS and screen estate for ie. more than 2 lines of stack and "equation handbooks with pictures/description" to customize the tool to suit your needs. You can't get them on these par minimum programmables with 16x2 display or other 1900s legacy issues anyway so you can better stick with matlab, reference books et. al. Why RPL? for many reasons, one is the debugging on-fly while doing work with it, compared to algebraic (textbook / pretty print etc.) where you catch (if you notice them) the errors of calculation after you have the final answer. Why not RPN as programming method? because it belongs to history box with its 4 accumulators else it could be fine ASM like language.

Four bangers and calculators like 35S, TI36abc etc. (of course with exceptions) replaces just ten to twenty first pages of mathematical reference pamflet (attachment), while calculators like 50g replaces whole handbooks, paper, pencil and so fort when done correctly. So does a computer or a phone, but those are jacks of all trades and generalist needs a specialists opinion at times.

What comes to Stack and RPL vs. Algebraic I don't see why they couldn't be working fluently together in the same machine. 50g did come pretty close and at times it just handy to jump to EQW for "pretty print / textbook" input method, but it does have a few too many "glued on top of RPL OS" quirks on ALG mode. While Prime did not get close with it's separated (with too high wall between) HOME/CAS applications, what I have read & seen. In general modern algebraic machines start to look a lot of RPL stack machines with reduced functionality on their history stack.

<Ramble End>

PS. I have money reserved for DM42.

[Logan]

I've been putting a lot of thought into this subject recently and have been solidifying my own feelings based on research and trying things out.

A calculator fills a niche position for the working engineer. Most computations, complex or otherwise, are documented and reproduceable in things like programs (for me, that's MATLAB) or Excel. In this sense, a graphing calculator is of limited or no use.

Things are a little different in a school situation because you are specifically limited to NOT having a computer for exam situations. Even so, I graphed only a few functions to double-check myself. Nearly all of my classwork I could have done more efficiently with a good scientific calculator.

So particularly for the "working engineer", a graphing calculator is overkill and tries to do too much and ends up doing most of it very inefficiently. Plotting is mediocre. Matrices are mediocre. Conversions are mediocre. Even most calculations are mediocre because the keyboard is not built as a scientific calculator: for quick computations, it's built for everything you could possibly want to do. I've been playing around with the HP-50g and while I'm impressed, I can't see myself using it for much now. I'd much rather use something efficient for a quick calculation (HP-35s or other), or something more powerful (computer). The best menu system out there is still going to be inefficient on something like this. I'm sure the Prime remedies this in part.

A smartphone app can only be so good too. It's good because you always have it with you, but it's not very fast, and depending on what kind of a calculation, more error-prone than a physical calculator. Not the best solution either.

So that brings me to the scientific calculator. This seems to be the niche market for the working engineer (at least this one). Either you're going to use a computer, or you're going to pick up a scientific calculator for some quick number crunching. I've been using the HP 35s a good bit and have appreciated it for just that. The main advantages that my TI-89 had were symbolic integration and differentiation, the larger screen (with computation history), and being able to solve simultaneous equations in any form. None of which I really use on a daily basis.

The HP 35s isn't perfect by any means, particularly because it seems to be designed around a few exams, not necessarily for work. Nevertheless, it does most things efficiently and pretty well. I'm even starting to think that matrices aren't really necessary for the "working engineer" anymore. Most of that really ought to be done on a computer.

Naturally, everyone's needs are going to vary, but the HP 50g just doesn't fill that niche use for me. I might evaluate the 48G since someone said it seemed more useable.

Candidates now are the 35s, the 42s, the 32sii and maybe the 48G. I really like the concept of the 15c, it just takes a big hit in programming for me, and I prefer the two-line (+) screen.

If the DM42 has a 4-line screen, I may have found my perfect scientific calculator.

[Csaba Tizedes]

(07-07-2016 11:21 AM)Marcio Wrote:  

(07-06-2016 05:43 AM)Thomas Radtke Wrote:  I'm using the 35s since 2007, but came to the conclusion that you cannot trust it.

I know this is a bit off-topic but would you be kind enough as to elaborate on this?

Thanks

What does it mean "do not trust in a calculator"? Something like this: I stored approx 50-70 equations in my 35S - not a big amount I know it. Mainly equations for my job and 4-5 programs to calculate something little complicated than a simple equation (for example settling velocity of particles, orifice sizing, fitting characteristics curves for blowers data from catalogs, etc...).

During a battery change all of them are lost.
That's all.

FYI: That was NOT MY FAULT. Since then my 35S sits in a shoe box - but it seems to me I'll never forgive it.


Csaba