Calculator Community

I absolutely love this calculator, with one important caveat: I will never do integrations with this machine except as a test. Its integration algorithm tips over at ridiculously low speeds when it comes to edge cases. I find that it adds to the Sharp's charm, but it definitely affects its usefulness. Aside from that, this nifty little guy has a few features that you can't find on other calculators, and is arguably the best one out there for statistics and tables.

First of all, those buttons labeled D1 through D3 on the top row? Those are this calculator's super power. Programmable Buttons! Let's say that you're in Matrix Mode, and you are doing a whole bunch of stuff with matrices A and B. Typically, to use matrix A it's a three button journey: MATH -> 0 MATRIX -> 0 matA. Instead, you can hit STO D1 and it will let you assign matA to the single button press of D1. Any function from any menu can be mapped to one of those three buttons, saving tons of button presses if you get into the habit of using them.

Most scientifics don't make it too difficult to do base-n calculations, but Sharps make it the easiest out of any manufacturer that I've tried. They are also the only one that supports pental in addition to the standard hex, decimal, octal, and binary. Honestly, I've never had any reason to use pental, but if you're doing something that requires it, this is the only way to go!

It doesn't have any stats features that the Casio fx-991EX or CW don't have, but it does have persistent memory. Anything that you pop into the statistics lists or tables are retained after the calculator is shut down. This is a huge win over the Casios which clear everything when you switch modes or power them off. If you're working with stable sets of data over a longer period of time, this is a lifesaver!

For tables, the Sharp is also a clear winner. When making a table from a function, both the Casios and the TIs have a static value for where x starts. The Casios will only compute 15 steps, while the TIs will continue to calculate as you scroll down through the values. The Sharps not only do rolling calculations going forward, but backwards as well. It's a nice feature to have if you use tables quite often.

Then there's the not so rosy stuff. The aforementioned integrations being the most egregious. It handles easy integrations as well as any other, but when it comes to edge cases it falls down in a light breeze. Take int(e^-x^, 0, A), for instance. Every non-CAS calculator will crap out as A increases in value, but some are far better than others. The new Casio fx-991CW can handle A all the way up to 2.55x10^9^. Even the Catiga CS-121 can get up to 1500 before it gets inaccurate. The Sharp EL-W516T? 16. It dies at 16. Says that the integral is greater than one. It's embarrassing!

The other place that it distinctly craps out in comparison to other flagship scientific calculators is its SHIFT+SOLVE function. It will only solve algebraic equations with a 0 on one side. It won't solve integrations or derivatives like the Casio, and even the TI 36X Pro can handle equations with x on either side of the equals sign. With the Sharp, you need to perform the algebra to get one side equal to 0 or it errors out.

Overall, it's a great device that gets a lot of use on my desk. You just need to be aware of where it is is weak, and be prepared for those inaccuracies.

I got it in 1999 or so. TI-basic was my first-ever intro to programming.

I found this guy used on shopgoodwill.com for about $20 USD a few months ago. It was on my list of calculators to procure, but the going rate of $130 USD for a new one was way out of my range. I was ecstatic to find one in such great shape for a low price!
This is an amazing calculator with some glaring issues that make it more of a mixed bag than it could have been.

This is the last calculator produced by TI with the OS that was originally designed for the TI-92, TI's first calculator with a built-in CAS. The fact that it was written for a device with a full QWERTY keyboard is glaringly obvious. The UI has been crammed into a standard graphing calculator form factor in such a way that some key combinations can't be silk-screened on the keypad. Base 10 log is Green Diamond + 7, for example.

The other huge problem with it is the speed reflects when it was designed. This calculator was released in 2004, and has had no upgrades since. As such, it is ungodly slow to perform a lot of calculations. In most cases, it performs on par with the Casio fx-991EX, a calculator which costs 1/7 as much. Graphing speed is pretty close to the Casio CFX-9850G from the mid 90s. This thing is unbelievably slow compared to most currently available calculators.

On the other hand, with the software being as mature as it is, it is amazing. There are program packages for the TI-89 Titanium that have no rival on any other platform. EEPro and MEPro are both freely downloadable from TI and provide a suite of activities, functions, and references customized for electrical engineering and mechanical engineering, respectively.

I'm very happy with this machine, and it does get quite a bit of use in spite of its shortcomings.

I don't know much (or anything) about calculators, but I wanted to share a select few that I thought looked cool. Here's the museum's website: https://www.kennettclassic.com/

I've spent a couple of months with this guy, buying it just after it was released in the US. I've long been a fan of the 991EX, and the CW is quite a different beast in terms of UI and capabilities.

First off, this thing is an absolute beast! It is the most accurate and precise calculator that I own by far. I knew when I put the fx-115ES Plus 2nd Edition through its paces and watched it soundly beat the 991EX that the CW was going to be a leap in capability, but I had no idea how large that leap would be. Take a look at arcsin(arccos(arctan(tan(cos(sin(9)))))-9 in degree mode, which should equal 0 but never will if your calculator is solving it arithmetically. most calculators will give an answer in the range of 1x10^-6^ to 1x10^-9^. The CW's answer is 7.5528x10^-18^, doubling the number of accurate digits!

Another mind blowing one is it's integration abilities. One of my favorite tests is int(e^-x, 0, A) and finding the value for A where the integration breaks. The value should approach 1 as A increases, but because the calculator's algorithm for integrating is not exact it will reach a value for A where the answer is wrong because it starts ignoring smaller values of x. Most of my calculators can make it to A being four digits long before they decide that the answer is either 0 or they declare an error. The CW gets all the way to 2.55x10^9^ before it craps out. I was astounded!

As for the UI, limiting each button to a single shifted function is a double-edged sword. It irks me that I need to dig through a menu to find some pretty basic things like integrations and derivatives. At the same time, I'm approaching 50 and my eyesight is not what it used to be. I don't need to take off my glasses and squint at the keys to see what they do. I wasn't sure what I'd think of the round keys, but they feel great in use. I'm also disappointed that they removed the ability to perform simultaneous equations with the : symbol, but it was a feature that not many people even knew about, let alone used regularly.

The only improvement that the CW let me down with is it still lacks persistent memory. It has a lovely spreadsheet function that is made useless by forgetting all of the entered data as soon as you change functions or turn it off. The same thing goes for data in lists or the statistical apps. This is not a calculator to take out into the field and use while gathering data. It really is the only thing that has kept the fx-991 series from being the perfect scientific calculator. It's still close enough to be really impressive!

I've seen these guys floating around on various sites, retailing for around $40 USD, but that was always too expensive for something that I figured would be sub-par. About a month ago, I saw one on eBay for $8, new and in the box. So here we are!

This little guy is delightfully crappy! It is bad in ways that make me giggle and add a little something to my smile. I've seen a lot of people assume that these are clones of the fx-6300G, but they just use the same screen. If it is Casio firmware, it has been modified beyond recognition.

The first thing I found that made me laugh was how much this calculator loves the number 4. It will randomly decide that any button that you pressed was actually a mistake and you meant to press 4. Arrow key? 4. 9? 4. I feel like a two year-old with a jack-in-the-box using this thing. Oh! And don't go typing too quickly, either. Sometimes it will decide not to register key presses that are too close together.

Its speed, accuracy, and precision are pretty close to the Casio fx-7700G from the early 90s. Even though it is graphing a very limited number of pixels, it still takes a while to output a simple sine wave. It can do integrations that are vastly more accurate than the busted ones that Sharps output, but they take forever.

I'm very happy to own this calculator, but I'm also happy that I spent less than $10 on it.

This lovely guy is the first HP calculator in my collection. I got it used from a member of the calculator subreddit a week ago, and am still getting used to it.

The speed of the HP Prime when graphing is nothing short of impressive. I was amazed by the 10 frames per second 3D plot rotations on the TI-nspire CX II CAS, but the instantaneous rotations and zooms with the touch screen are seamless and smooth. I'm used to seeing graphs get drawn to the screen from left to right since I've been playing with them since the early 90s. Even the TI-nspire has a bit of a hiccup when showing a simple graph. The Prime, though? Immediately shows the graph without hesitation for simple things. It does bog down when you start to push it, but it is still the fastest graphing calculator I've ever seen by far.

Given that it graphs so quickly, I was surprised at how slow it is when it comes to some things. sum((e^sin(atan(x)))^(1/3), 1, 1000) took three times longer on the HP Prime than it does on the Casio CG-50, which just blew my mind.

The other thing that I've found interesting is it's lack of precision and accuracy in non-CAS mode. Pop it into degrees mode and run arcsin(arccos(arctan(tan(cos(sin(9)))))-9, which should equal 0. Unless your calculator evaluates the function symbolically, there will always be some residual number due to the way that it performs the calculations. The smaller the number, the more digits your machine keeps track of internally. The best of my current collection is the Casio fx-991CW, which gives an astoundingly small 7.5528x10^-18. The only calculator that I know of that can beat it is the SwissMicros DM-42. The Prime delivered an answer of -1.35733x10^-6, which is about the same margin of error as my Catiga CS-121, or the Casio fx-115W from the mid-90s.

I haven't had much of a chance to really play with the CAS yet, but I have found that it will actually handle symbolic sums. That's something that neither the TI-nspire CX, nor the Casio fx-CG500 can handle at all.

I'm still figuring out where this beast will fit into my day-to-day, but I am quite pleased with it so far!

Do you collect calculators, have something cool to show off, or have a question about how to do something with your lovely chunk of plastic that plays with numbers? Feel free to post it here. Welcome, everyone!