Graphing Calculator With CAS, HP or TI?

[labarowski]

I'm a third year ECE student and I'm planning on upgrading my TI-84. I've been throwing around the possibility of buying a CAS calculator for a while, but I'm taking a semiconductor physics course right now which covers quantum mechanics. My TI-84 often rounds some of the ridiculously small numbers to 0, while my teacher's basic HP scientific does just fine, so this seems like a good time to pull the trigger and upgrade. I'm leaning towards the HP 50G. I've used RPN with xcalc on my pc and this is definitely a great feature. I would like to know how the CAS system on the 50G stakes up with the similarly priced TI-89 Titanium or a member of the Nspire family. A friend with an Nspire was saying that there is an option to show step-by-step solutions of calculus functions in the CAS. He also said he had never personally used the feature, so maybe someone can confirm this? I've also heard that the screen resolution is somewhat low on the 50G. All other things being equal, I would go with the 50G for RPN, but I wanted to get some opinions here first. I downloaded an emulator for the 50G, but I'm not sure if such a thing exists for the TI calcs. Anyone know if there is a TI-89 Titanium or Nspire emulator? Would be nice to try them all out first.

[T4P]

TI-89 Titanium yes, TI Nspire not. But that's on Android Not PC

[IanB]

The only purpose for calculators these days is to help you take exams and tests. Either college exams, or later on tests like the EIT or PE exams. So mainly, you'd want to check what calculators are permitted in those exams (hint: not the HP 50G).

If you want to buy a calculator like the HP 50G, you can choose based on its value to you as an entertaining toy. It doesn't have an actual professional use. In the real world, you will use Excel for basic calculations and Matlab or Mathematica, or dedicated simulation tools like LTspice, for advanced calculations and algebra. The only kind of pocket calculator you need in 2012 is a basic $20 scientific for those odd occasions when you don't have a computer handy--and even then your smart phone will have a better calculator built in.

[labarowski]

Thanks for the honesty guys. That's definitely not the answer I was expecting based on what I've read elsewhere. Still, I feel like the $100-$150 that I would spend on a calculator with a CAS is a fairly small investment in the long run. That's around 5-6 months of a mandatory smartphone dataplan. Although I suppose I could get an itouch or android mp3 play for around twice the price, which would be easier to justify than a smartphone/contract. I'm not really even sure what a Mathematica/MatLab license will cost. I can supposedly get both free from my university now. I've actually been using MatLab to do my physics of semiconductors homework to avoid the round to 0 issue. On tests I just keep track of decimal places manually and add them at the end. Both are a bit more tedious than just plugging it into the calculator IMO, but I'm not complaining. Mathematica looks promising but I haven't figured out how to get a license from my university yet. Anyway, I just spent over $500 for books for this semester (keep in mind I got all but one of them used), so another $150 doesn't seem too bad if it will be useful later on. Having said that, I'm not set on buying a CAS calculator. I've got the 50G emulator running and I'll see how I like it. For the sake of argument though, I would like to know which of these calculators people think are best. I'm sure there are other people who will stumble across this thread as well. I certainly have looked at quite a few 50G vs 89 threads - I'm just trying to target EEs by posting here.

It looks like I can't even use the TI-84 on the EIT/PE exams. These are really the only standardized tests that I have left, although some teachers have a no CAS policy.

[robrenz]

A student license of Maple is about $125.00  Way better than any calculator.

[IanB]

Ah, when I said Mathematica I probably meant Maple. Thanks for mentioning that.

The truth is that when you are at work you won't have to pay for those licenses, the company will provide the software tools you need.

Computer algebra systems can sometimes help with tedious work, but as a student isn't it better to learn how to do it on paper? It acts as a workout for your brain, like going to the gym.

I'm honestly serious when I tell you that as a professional engineer I haven't used a pocket calculator in decades. Every calculation I need to do is more conveniently performed on a computer screen. It allows me to edit and save and re-use the calculations, to print them out if needed, and to email them around when communication and collaboration is needed. A pocket calculator is hopeless for that, even if the most advanced ones should happen to have a USB interface.

[amvakar]

I learned to program efficiently by hacking various functionality into the TI-84, including arbitrary-precision arithmetic without the rounding errors. But I just can't justify treating it as anything more than a programming challenge anymore. If I'm too lazy to do basic maths, my phone has Wolfram Alpha. If I'm doing anything serious, there's no reason to leave the multi-gigahertz machine with multiple cores to playing background music and turn to the microcontroller for help. The only times I use a calculator are during exams and when I'm working on something that has to be above napkin standards but isn't worth the time of grabbing a laptop.

It should also be noted that there is nothing more unpleasant than hunting down AAA batteries five minutes before an exam. That experience was enough to get me to go for a simpler design; the eventuality of the EIT/PE helped decide which one.

[bingo600]

If you have the 50G emu , just install the TI-89 emu and compare them

http://forums.afterdawn.com/thread_view.cfm/358142
http://lpg.ticalc.org/prj_tiemu/

I just installed the TI-89 on Linux.

Btw: I have bought the HP-50G , the TI-89I , and lastly the TI Nspire CX CAS.

The TI-89I was for the daugthers highschool (i bought it 2 years ago), but the school changed track this year to the Nspire , so i had to get one of those instead. 
TI (teacher programmes) apparently affects the contents of the math lessons , and the teacher insisted that the students buys the 1'year PC-Nspire licence (each year). As the lessons contains examples made for that calculator.
But looking at TI's homepage i found out that buying the Physical Nspire included a non-timelimited PC-Nspire also , at +25% of 3 x 1'years licences.
So basically i got the PC-Program and the Physical calc for 25% extra.

I keep the 50G for my self , and i like it better than the TI89.
Even though i have it set for Non-RPN.

I'm an old TI fan , and have always had TI's before , since i got my TI-59 back in 79'
But i do like the HP50G better than the TI-89I , the TI-89 seems slower and a bit out of date.

I haven't tried the Nspire much , but it might be able to convert me back to TI.
The Nspire is a nice calculator , It's fast and i like the color display , and you can program it via the LUA language   

But i'd check it's capabilities , for engineering before getting a Nspire.
And make sure it can do the same or more than the TI-89 , i can't really decide if i think the Nspire is more of a Highschool calculator with a "Geek touch .. The color screen"

Edit: Ohh and i think i saw the Nspire CX CAS , was also forbidden in some US tests/exams , the non CAS was allowed i seem to remember.ams


/Bingo

[h2o]

I'd get the hp.  I haven't used the 50, but the 48 was a wonderful tool.  Obviously no one should ever use a calculator.  Do it in your head.  However, if you happen to need to use a calc, HPs are great.  RPN is a real time saver.  Do be swayed by the haters.  I get $$ to program excel, but I love an HP.  The calculator can fit on a desk or in a drawer and just works.  Less distractions and I don't think a phone will ever replace the nice tactile feed back for me.  Now in 2112, maybe there will be a better neural interface.  Computer programs have their place, and are superior for many things.  They aren't substitutes.  Once you go RPN, you won't go back.

[labarowski]

Thanks for the input guys.

IanB, for the record, the CAS is just a selling point. Because these calcs are so closely matched though, it is a deciding factor. I agree that in general it is much better to do math by hand. I have friends who graph functions and find the intercepts rather than do basic algebra by hand. It's ridiculous and good luck finding imaginary answers like that. There are definitely some instances on exams that allow a CAS where I might use it to check answers or save time though. I definitely feel like a CAS calculator would promote laziness less than Mathematica or Maple though.

amvakar, any chance that you released the arbitrary precision arithmetic function? That is, after all, one of the reasons that I am considering ditching the 84. You bring up a good point with the EIT/PE exam too. My teachers HP scientific doesn't have rounding errors, so maybe I should go that route. I could probably borrow a calculator for those exams though.

h2o, all other things being equal, RPN is the deciding factor. I definitely like RPN based on previous experience. I am reading that you can add RPN to the 89 though.

bingo600, exactly the models I'm looking at. I will definitely check out the 89 emu. Guess the only way to get an NSpire emu is to buy TI's software. The new NSpire definitely packs a lot of features - it looks a bit awkward to carry around and use for day to day stuff though.

[amvakar]

In order for it to have been worth releasing, it would have been necessary to risk bricking the calculator with the improved math functions baked into the firmware. As acquainted with TI's code that I became during high school, I was not willing to go that far.

As far as borrowing calculators for exams, I've never liked doing so, but that's because I've grown to like RPN far too much after using an HP.

[nanofrog]

I own a 50g, and prefer it to the TI89 (not had access to an Nspire), and RPN to CAS. I like the tactile feel and dedicated function keys of a calculator when needed or is more convenient than a computer.

[Someone]

My TI-84 often rounds some of the ridiculously small numbers to 0, while my teacher's basic HP scientific does just fine, so this seems like a good time to pull the trigger and upgrade.

Do you have some examples of the computations that are rounding out? The legacy TI computational system is a floating point decimal based design with a silly huge range of exponent.

The truth is that when you are at work you won't have to pay for those licenses, the company will provide the software tools you need.

Its this sort of using a massively excessive tool for the job that those software companies live on, the costs of providing matlab seats is huge and many smaller companies do without it. While the free alternatives are good and largely replace matlab for anything other than physical interfacing they're still only numerical computational systems (more on that later).

All the jokes around about the cost of high end calculators is rather moot when you look at what equivalent software costs.

I'm honestly serious when I tell you that as a professional engineer I haven't used a pocket calculator in decades. Every calculation I need to do is more conveniently performed on a computer screen.

Calculators survive because of their optimised user interface, otherwise they'd all have turned into blocky touch only devices by now. Once you are well practised in a specific calculators use its much faster than computer based alternatives, and they dont have a UI refresh every second year that gets in the way of the workflow and you have to relearn.

One Calculator/Software package is not enough!
Trying to use a symbolic system for numeric analysis and vice versa ends up being much more complicated than needed. Just as anyone using software would have a good grasp of both techniques, the associated software, and when to apply them, trying to shoehorn a symbolic calculator into numerical engineering work is painful.

I have a TI-86 that is used almost daily in my professional work with a huge multinational engineering company, but its one tool among many. On the occasion that I need symbolic solving I have a Nspire CAS, its sole use for me is symbolic calculus and I've pushed a fair few problems into it that arent solvable. While we have matlab seats floating for use there is no symbolic tool licensed for us, even big business wont provide anything you want.

[robrenz]

One Calculator/Software package is not enough!
Trying to use a symbolic system for numeric analysis and vice versa ends up being much more complicated than needed.

Not with Maple.  That is the beauty of Maple compared to the others It work in both realms seamlessly

[Someone]

One Calculator/Software package is not enough!
Trying to use a symbolic system for numeric analysis and vice versa ends up being much more complicated than needed.

Not with Maple.  That is the beauty of Maple compared to the others It work in both realms seamlessly

As with most software packages its always growing, but maple as an example is missing a lot of classical computational things such as the meagre statistical analysis included (and really, when it gets big there are dedicated packages for that). Even a mid-high end handheld calculator offers easier to use and more powerful statistics than maple. A few other missing computational things without looking too hard are the lack of rounding modes and fixed point support. There is no one tool to rule them all.

[labarowski]

My TI-84 often rounds some of the ridiculously small numbers to 0, while my teacher's basic HP scientific does just fine, so this seems like a good time to pull the trigger and upgrade.

Do you have some examples of the computations that are rounding out? The legacy TI computational system is a floating point decimal based design with a silly huge range of exponent.

This is the equation that I'm referring to - although my textbook actually gives the equation with plank's constant^3 and has only (2m)^(3/2) <(this is why I like the idea of RPN.) This is a key difference. Using plank's in Joules (6.625x10^-34 J-s) to the third power, my TI-84 running 2.55MP (updated. not from factory) will round to 0. It will handle plank's to the power of 2 just fine, so it looks like I should just use wikipedia's equation. We also are dealing with a set volume too. Volume of a sphere I believe.

All the jokes around about the cost of high end calculators is rather moot when you look at what equivalent software costs.{/quote]
Like I said earlier, the cost of a high end calculator is small even compared to what I paid for textbooks this quarter. It's a drop in the bucket in the long term. I just don't want to throw around money if it would go unused and I certainly don't want to buy more than one CAS unit, so I want to pick the right one the first time around. If I should wait for some other calculator to be released I would definitely be open to that also, but I get the impression this doesn't happen often. A lot of people seem to like the HP48G over the 50G so...

One Calculator/Software package is not enough!
I have a TI-86 that is used almost daily in my professional work with a huge multinational engineering company, but its one tool among many. On the occasion that I need symbolic solving I have a Nspire CAS, its sole use for me is symbolic calculus and I've pushed a fair few problems into it that arent solvable. While we have matlab seats floating for use there is no symbolic tool licensed for us, even big business wont provide anything you want.

Would you say that the HP50G is lacking in the CAS department then? I thought that it had a fair amount of calculus functionality...

[IanB]

All the jokes around about the cost of high end calculators is rather moot when you look at what equivalent software costs.{/quote]
Like I said earlier, the cost of a high end calculator is small even compared to what I paid for textbooks this quarter. It's a drop in the bucket in the long term. I just don't want to throw around money if it would go unused and I certainly don't want to buy more than one CAS unit, so I want to pick the right one the first time around. If I should wait for some other calculator to be released I would definitely be open to that also, but I get the impression this doesn't happen often. A lot of people seem to like the HP48G over the 50G so...

One Calculator/Software package is not enough!
I have a TI-86 that is used almost daily in my professional work with a huge multinational engineering company, but its one tool among many. On the occasion that I need symbolic solving I have a Nspire CAS, its sole use for me is symbolic calculus and I've pushed a fair few problems into it that arent solvable. While we have matlab seats floating for use there is no symbolic tool licensed for us, even big business wont provide anything you want.

Would you say that the HP50G is lacking in the CAS department then? I thought that it had a fair amount of calculus functionality...

I didn't make the quote above--perhaps you would correct the attribution?

[IanB]

This is the equation that I'm referring to - although my textbook actually gives the equation with plank's constant^3 and has only (2m)^(3/2) <(this is why I like the idea of RPN.) This is a key difference. Using plank's in Joules (6.625x10^-34 J-s) to the third power, my TI-84 running 2.55MP (updated. not from factory) will round to 0. It will handle plank's to the power of 2 just fine, so it looks like I should just use wikipedia's equation. We also are dealing with a set volume too. Volume of a sphere I believe.

I'm sorry to tell you, but that is not a problem with the calculator it is user error. What you describe is not "rounding error" but a problem with scaling. You need to evaluate the formula in a proper order to retain reasonable scaling of the intermediate results. If you were evaluating that equation using a slide rule in the 1950's you would have had no problem--the issue is that you are being misled by an over reliance on the calculator rather than gaining an understanding what it is you are evaluating.

[labarowski]

IanB, sorry about mis-crediting the quote. It is corrected now. And yes, you are somewhat right. I used the equation as it was given in my textboook while simply pulling the ^3 outside the ratio would solve the issue (as it is in wikipedia's equation). As I said before though, I've just been keeping track of the power manually and adding it in after the fact up until now - so its not like I'm incapable of doing the problem without an electronic mind, I'd just rather have the equations work as they are given on a midterm or final. Anyway, this class is not the reason I'm thinking about another calculator, its just got me considering it again.

EDIT: Actually, based on what little I know about slide rules, I don't think that they (in general) could handle a power like -34, which would leave me handling the powers separately and adding them in at the end - which is exactly what I've been doing. That being said, your point is by all means valid - I don't want to use a calculator as a crutch. But if I'm honest, I'm not above over-using my calculator on a test. Regardless of my skill level, I'll always be more confident in the calculator's ability to do basic arithmetic than my own.

EDIT2: Actually, it looks like it wasn't uncommon for for slide rules to have scales up to 1000 power - but I am now fully convinced that it was always easier and less time consuming to do basic arithmetic by hand than use one of these. Also, it sounds like accuracy was never guaranteed with large arbitrary powers. I am quickly diverging from the topic though.

[IanB]

What was saying about slide rules is that a floating point number like 6.625x10^-34 is really the number 6.625 with the decimal point moved 34 places to the left. When you worked with a slide rule you just took the number as 663 in your multiplication or division and worked out where the decimal point went afterwards--just like you were saying about keeping track of the exponent manually.

That said, the equation you referred to does have some pretty ugly scaling and getting decimal exponents over +99 or below -99 can certainly upset many calculators. It's nice if you can just let the calculator handle the arithmetic but apparently on occasions like this you may have to pre-order the calculation to avoid overflow or underflow. That's life in the world of physics I guess.

[labarowski]

Cool, well at least we are on the same page then and I got to learn a thing or two about the history of early calculators.

[Someone]

That said, the equation you referred to does have some pretty ugly scaling and getting decimal exponents over +99 or below -99 can certainly upset many calculators. It's nice if you can just let the calculator handle the arithmetic but apparently on occasions like this you may have to pre-order the calculation to avoid overflow or underflow. That's life in the world of physics I guess.

Looks like you have it spot on, TI-83 descendants have inherited its 8 bit signed exponent as this must have been deemed adequate for high school math problems, where as the high end TI-86 and TI-89 have a signed 3 decimal digit exponent.

One Calculator/Software package is not enough!

Would you say that the HP50G is lacking in the CAS department then? I thought that it had a fair amount of calculus functionality...

I cant comment on its peculiarities as I havent used one, they are held in high regard but the best judge of whats good for you is you. Try and test drive one before purchasing.

[h2o]

...
Regardless of my skill level, I'll always be more confident in the calculator's ability to do basic arithmetic than my own.
...

No! You can always do basic arithmetic better & faster than a calc, if you want/try.  Have faith.

[olsenn]

Trust me; TI makes FAR better calculators than HP! I own a TI-86 (which I've used heavily for over a decade) and a TI-nspire CX CAS (fairly new) by TI, and a HP-50g and a HP-12C by HP. I have used many other calculators by both companies as well.

The HP calculators are build cheaply and they will not last (the buttons have already begun to stick in my HP50g, and I've also lost a line on the display from one small drop). Texas Instruments calculators on the other hand, are build to last a lifetime! I've droped my TI-86 countless times and it still works just like new. I hardly ever use the HP-12C, so I can't really comment on the durability of that, but it looks and feels cheap also. As a side note though; I think the 48-series of HP calculators, prior to the 49 (and 50g) was released are MUCH better quality; but still not as good as TI.

The battery life of the HP-50g is abysmal. My TI-86 lasts years of intensive use, but my 50g only gets a few months. One point a will make though, is to go for the TI-89 or the regular nspire CAS model, as opposed to the nspire CX CAS -- the latter uses a rechargable battery; it's meant for high schools (which is where all their money comes from with graphing calculators), and if you're putting it in a charging cradle, it is great. However, give it a full charge and go to use it a month later, and it's dead! There's really no need for a colour display on a calculator anyways!

Lastly, the HP 50g is MUCH more difficult to learn than the TI-89, but you can work more quickly with it after you've spent a long time getting used to how to use it. However, the nspire series of calculators are not only even easier to learn than the TI-89, but you can do your work far more quickly than even the most skilled 50g user can! In short, RPN is better than standard type, but textbook style entry on the nspire beats the hell out of RPN any day.

[FenderBender]

My school loans us Ti-89s for our calculus course. It's a very nice calculator. Has all the functions you could ask for. As far as an EE calculator, it's relatively useless. If I'm actually designing something electronics-wise I use my Sharp EL-W16X. It has some actually useful features including a Decimal <> Binary/Hex etc. converter, an engineering mode that can use prefixes like nano, pico, kilo, etc and it's not too jumbled like the Ti-89. I believe it also has some sort of solver feature, but I've never used it.