CASIO fx-991CW

[lrdheat]

A note on calculation speed…while the CW is faster in general than the EX, due to it using 17 to 23 internal digits, some calculations that take significant time to converge on a result that will be accurate to full display precision will take longer on the CW on occasion.

[lrdheat]

The extra internal digits allow for solutions on the CW that are not found on the EX. When integrating (x^4)/sqrt(x^10 -2) from 1.07178 to 3 (I used 1.07178 as 1.071773463^10=~2) comes up with 1.1663633, same as on fx-CG 50 (which, interestingly cannot do area under curve for that domain in graphing app). The EX throws a math error. The EX throws an error raising the lower endpoint to 1.0718, even throws error using 1.08!

[lrdheat]

Accidentally had the -2 outside of the sqrt in the denominator. The EX and CG-50 in graphing analysis handled the problem.

[BINUBALL]

(04-04-2023 09:04 PM)pier4r Wrote:  if someone has willing and has spare time, one could try the summation benchmark here: https://www.hpmuseum.org/forum/thread-9750.html

I've tested fx-570CW several months ago.

(01-24-2023 01:12 PM)BINUBALL Wrote:  I have bought newer casio model, fx-570CW several months ago and reminded this thread. Test results below:

max = 10000
~ 581s - casio fx-570CW, 13955.8579 (sum function)
max = 1000
~ 59s - casio fx-570CW, 1395.346288 (sum function)
max = 100
~ 7s - casio fx-570CW, 139.297187 (sum function)
max = 10
~ 1s - casio fx-570CW, 13.71183502 (sum function)

fx-570CW is significantly faster than fx-570EX. Let's compare them.

1s / 2s (max = 10)
7s / 12s (max = 100)
59s / 109s (max = 1000)
581s / 1073s (max = 10000)

[lrdheat]

Was wondering why the U.S. version does not have the product operation, but then realized that I could use summation of logs from x to y, and then take the anti log of the answer. This can also be used to find factorials greater than 69! by using integer number part of the logarithm as power of 10 of the answer (if log was used), and raising 10 to the decimal part of the logarithm to find the part of the answer that is the number times the power of 10.

[Eddie W. Shore]

(04-02-2023 04:29 PM)Csaba Tizedes Wrote:  

(04-02-2023 01:48 PM)Eddie W. Shore Wrote:  When the Casios had the multistate command, storing calculations immediately terminated the entry ... Ans could be used ... For the fx-115ES Plus (1st or 2nd Edition), we could use both Ans and PreAns.

Of course you can use the following structure:

Code:

var1=expression1:var2=expression2: ... etc

Then press CALC (which is removed from 991CW )

This is same as STO but without entry cancellation.

Check this video (How to emulate IF...THEN structure on CASIO fx-991EX):

Now I get why you are upset about the CALC function being gone. Casio sapped it's algorithmic power when the fx-991CW was released.

I wish we could store mutli-expressions in memory on Casio. There are some Sharp calculators (unfortunately not all), such as the EL-506TS, that has formula memory (F1-F4), I will see if I can store multi-expressions there.

[Csaba Tizedes]

(04-11-2023 01:30 PM)Eddie W. Shore Wrote:  Casio sapped it's algorithmic power when the fx-991CW was released.

Unfortunately yes. I have some videos which based on this feature, eg. calculating average values with trapezoidal formula (with multistatement easy to sum the area, changing variables with minimal user data input (updating left boundary to old right boundary automatically in each turn, summation, etc...) )

I hope they won't cancel the 991EX or we must to buy until available. This is really big problem in education, in Hungary the 991CEX is preferred in schools and all educational materials based on this model. If they downgrade some features, it will be hard to update without them.

Cs.

[Eddie W. Shore]

Reposted review of the fx-991CW, which addresses the lack of multi-statements, and the lack of the integer part, fractional part, and sign function:

http://edspi31415.blogspot.com/2023/04/r...multi.html

The blog entry now includes the link to Csaba's video as well, in the Addendum section.

[Dave Britten]

Just got my fx-991CW today, and I've found/confirmed a few rather important improvements:

1. You can now change the auto power off time to either 10 minutes (default) or 60 minutes. This is big - I've seen plenty of complaints about losing work due to the older fx-991EX powering off automatically. Now you can change it to 60 minutes and not have to worry about this happening nearly as much.
2. Spreadsheet data now survives switching modes. There may be exceptions to this, but I was able to go to stat mode, enter some two-variable data, and switch back and forth between the two apps without data in either app being deleted. (The ON key will still clear the spreadsheet.)
3. As mentioned, stat mode also retains its data when switching modes, but more significantly, data is also retained when powering off. Please correct me if I'm wrong, but I don't think this was the case on the ES or EX families.
EDIT: 4. And I just noticed that the spreadsheet has 40% more memory available compared to the EX: 2380 bytes free vs. 1700. But it's also storing higher-precision values, with each numeric cell entry consuming 14 bytes instead of 10. But since formulas are still seemingly stored as character strings with one byte per character, you should theoretically be able to fit more into memory if your sheet is using a lot of formulas. Both the EX and CW store both the formula definition and the result value for each cell, but the formula definition portion at least should consume a smaller portion of total RAM now.

Alas, the CALC button and multi-statement (the colon) appear to be well and truly gone. But given that you can now define two functions f(x) and g(x), and the spreadsheet's data is now more durable, I suspect many of the use cases of CALC and multi-statement can be satisfied by using these two other functions instead, perhaps to even better effect than the old functions. Set your auto-power-off time to 60 minutes, put a few formulas into the spreadsheet, enter variable values to get results, switch to other modes as needed, come back to the spreadsheet to tweak the variables a bit more, etc.

Those who extensively use engineering display or engineering symbols should probably steer clear of the CW. Entering engineering symbols requires pressing Catalog, using the page/arrow keys to scroll to "Engineer Symbol", then scrolling through the list of symbols to the one you want. And they're spelled out with full names, (e.g. "Femto", "Milli"), all capitalized, which is rather disorienting for the ones that are typically written with a lowercase abbreviation. If you want to shift the mantissa and exponent of a result up and down - a task you would accomplish with a pair of "ENG" keys on older Casios - you have to press "Format", select "ENG Notation", then use the left and right arrow keys to adjust the exponent. All in all, it's wildly unwieldy compared to dealing with engineering display on almost every other Casio I've used. If Casio had at least designed this to open the Catalog and Format menus with the cursor defaulting to the last-used item, that would have gone a long way to improving things (and not solely in this situation).

In general, that seems to be the theme: if you hate the TI-36X Pro because of its heavy reliance on menus, the fx-991CW is definitely not going to win you over. But if you don't mind menus, or if you find the very busy keyboard of the fx-991EX somewhat confusing, you'll find that the CW has good discoverability, and a keyboard that isn't overly cluttered. (e.g. the single "Format" key instead of two ENG keys, a fraction/decimal key, and a degrees-minutes-seconds key)

[lrdheat]

Thanks for the review. I had incorrectly believed that stats data was lost upon turning the calculator off. I had tested it, and was happy to see that the data was retained when switching through apps, but refrained from using the stats for uses that required new entries through the course of days (such as daily high/low temperatures as a month progresses).

There are some difficulties with the calculator finding roots of an equation when the root is at a discontinuity or when the slope at the root location is infinite, but I am quite happy with this gem. With the extra internal significant digits, results that appear in decimal form on other calculators (even the 991 EX) that are clearly recognizable as a fractional result are often presented in fractional form.

[lrdheat]

Upon completion of data entry, if you wish to obtain the statistical analysis, press “OK”. This will bring up the screen with choices for what you wish to do.

[Dave Britten]

(04-30-2023 04:48 PM)lrdheat Wrote:  Upon completion of data entry, if you wish to obtain the statistical analysis, press “OK”. This will bring up the screen with choices for what you wish to do.

Yeah, that's kind of a strange design. I would have expected that to be in the Tools menu.

And slight correction to my previous post: the fx-991EX will also retain stat data when powered off, but it clears the data if you switch modes. Meaning you have to power it off while still in stat mode if you want to keep the data. Now on the CW, the stat data stays until you clear it.

[Csaba Tizedes]

... QR of Spreadsheet formulas pass only values or formulas also?!? (I'll check, but I throwed my CW into the deepest shoebox after 2-3 days of crying)

BTW: it is possible to use integrating in form of INTEGRATE(function, lower bound, upper bound, tolerance). What if, if you use f(x) as function AND g(x) as tolerance?

Maybe possible to make adaptive stepsize for integration on these machines?!?

Csaba

[Dave Britten]

(05-02-2023 07:34 AM)Csaba Tizedes Wrote:  ... QR of Spreadsheet formulas pass only values or formulas also?!? (I'll check, but I throwed my CW into the deepest shoebox after 2-3 days of crying)

BTW: it is possible to use integrating in form of INTEGRATE(function, lower bound, upper bound, tolerance). What if, if you use f(x) as function AND g(x) as tolerance?

Maybe possible to make adaptive stepsize for integration on these machines?!?

Csaba

Not sure, I made a simple sheet with some formulas, and navigated using the QR code, and I ended up on the ClassPad web site with a view of all the data that was in my spreadsheet. I didn't see any way to view the formulas, or edit the sheet, but it looked like there were some options to do statistical calculations.

[ThomasK]

Got a Casio fx-991DE CW today, for 10.99€ (German "Mediamarkt" store)!
Not sure if that price was an error, "sold out" a few hours later.

Tried summation benchmark first:
max = 10000
~ 598.5s - casio fx-991DE CW, 13955.8579 (sum function)
max = 1000
~ 60.8s - casio fx-991DE CW, 1395.346288 (sum function)
max = 100
~ 6.8s - casio fx-991DE CW, 139.297187 (sum function)

A little slower than fx-991CW or may be not all calculators have exactly the same speed.
This calculator is light as a feather - is there anything inside? ;-)

[Steve Simpkin]

(05-04-2023 05:40 PM)ThomasK Wrote:  Got a Casio fx-991DE CW today, for 10.99€ (German "Mediamarkt" store)!
...
A little slower than fx-991CW or may be not all calculators have exactly the same speed.
This calculator is light as a feather - is there anything inside? ;-)

Here is an internal photo of a Casio fx-991DE-Plus posted by Stefan. It seems to mostly just be a blog of epoxy. Amazing! Epoxy could not be used to do anything as complex as a calculator when I was a kid. of course I'm not sure epoxy was all that common when I was a kid...

https://www.hpmuseum.org/forum/thread-67...ml#pid5507

[Dands]

I am an electrical engineer and the 991EX was my go to calculator during school. Passed both FE and PE exams using it as well. Here's why I think the 991CW was a HUGE fail for Casio given that they are still marketing this calculators to engineers:

- They got rid of the ANGLE key(shift option): if you are an electrical engineer and rely on phasor notation, Casio screwed you. The easily accessible angle key is gone. To enter a phasor you will need to go through a series of menu options. It is also hard to convert between rectangular and polar forms, more menu options. Huge miss.

- They got rid of the ENG key. The 991EX made it easy to toggle between 10ˆx powers (both positive and negative!) using the ENG key. It's GONE.

- Forget the amazing S<=>D key. Now you have to use this terrible FORMAT key and go through a series of options to switch between standard and fraction notation. You will have to do this EVERY TIME. What were they thinking?!

- Plus all the weird key placements - ANS is not next to EXE (or =). STORE and M+ keys gone....total miss.

I returned mine and will buy an extra 991EX as a backup during the coming zombie apocalypse....if you're studying to be an engineer, or are already practicing the profession, stay away of this CW model. Casio, hope you read this and release a Version 2 as soon as possible. A super clean keyboard is NOT always the best option.

Thanks

[jonmoore]

The unfortunate truth regarding Casio calculators such as the 991 series is that bad design decisions in large part happen because of pressures from eduction institutions. You witness this in significant differences is functionality that's available across the regions of the world.

As a prime example, in the last generation of calculators, this effected the UK with GCD and LCM being removed, whilst it was included across most of continental Europe and Australasia (not totally certain regarding the US market, but I believe it was the same as the UK).

I strongly agree with the central tenet of pre-university mathematics education that symbolic manipulation is a problem-solving skill that can only be learnt through lots of practice, but to include mechanistic calculation functions such as GCD and LCM within this tenet is wrong as it proves nothing regarding the mathematical abilities of the student. But the UK teaching pressure groups were pedantically opposed to such functions, so they had to go.

As a design strategist, I hated the industrial design of the 991EX which tried far to hard to appeal to the fashionable young student with a faux carbon fibre fascia that obscured the labelling on said fascia (particularly under artificial light). I believe this was one of the primary reasons for the simplification of the design of the current generation of 991 calculators. One of the clichés of design is that rounded design elements are more 'friendly'. I can almost hear the equivalent of me saying this when pitching the design to Casio executives. The biggest irony here is that Casio already had a great design for the previous generation of the 991, and that was the second generation Casio fx-100AU PLUS, the equivalent model for the Australasia market (apologies for the giant image below, I lazily linked to an online image without first resizing it, but on balance at least everything is legible).

The eagled eyed among you will spy the GCD and LCM functions, missing from the UK 991EX. I ended up purchasing the Australasia model for my pre-university aged kids here in the UK as the school allowed them for tests and exams (the mathematics department agreed with my view - I sit on a PTA group responsible for science and mathematics teaching in the school and there was not a single objection when I raised the subject for discussion).

I could go on, but the key discussion point I wanted to describe is that young students (and professionals) are affected by the arbitrary viewpoints of regional education pressure groups. Plus I agree with the majority of views put forward by other reviewers, including those of Eddie and Dave. However, regarding performance, most UK students are recommended to supplement their 991-ex with the graphical Casio Model, the fx-CG-50, which is very fast and approved for those UK exam papers which allow calculators. Unfortunately, it's far from featherlight and pig ugly to boot! It's also a model where many of the key functions are hidden behind menus. In it's favour it has an exam mode which disables functions, programs and any text notes on the device. This is critical as e.g. the CAS from the HP Prime can be installed without any "hacks".

[jonmoore]

I forgot to mention that Calculus functions don't exist on the fx-100AU Plus as these are only encountered when students are older and at that point they are recommended the CG-50 (I believe this model is a universal design across world markets). Plus Calculus is one of those mathematics subjects that not all students are taught. It's only taught to those aged 16 and over in most education boards around the world, and by that point students are filtered to curriculums based upon their abilities.

A key advantage of removing Calculus functions from the fx-100AU is that it's a design which features far less menu diving for functions. This has a tangible advantage in the classroom, where teachers can spend more time on teaching mathematics and less teaching their students how to use their calculators.

[Dave Britten]

I agree with pretty much all the sentiments in this thread so far, but interestingly, my preference is slightly for the CW. There are a few very specific reasons and improvements that might drive you this way:

- The spreadsheet is much more viable now that you can set the auto-power-off to 60 minutes, and you can switch modes without losing your data (the spreadsheet workflow is a bit like working out a lengthy problem spatially on a piece of notebook paper, but you can go back and update your inputs or intermediate calculations without redoing all the subsequent steps by hand). It's also faster, and the page up/down keys make navigating the sheet quicker.
- Similarly, statistics mode doesn't throw away your data when you switch modes or power off. It's actually pretty difficult to inadvertently lose your data now, making the CW a realistic option for recording observations over a period of hours or even days.
- This is largely personal preference, but the feel of the CW keyboard seems slightly better to me. Not that there's anything particularly wrong with the EX, but I like the feel, spacing, and size of the keys on the CW just a little bit more.

There are some other niceties, like being able to define and use f(x) and g(x) not just in table mode (you can save even more space in spreadsheets by using these), but those three bullet points are the main differentiators in my mind. If those don't matter to you, stick with the EX. There are, unfortunately, a lot more reasons to not switch to the CW than there are to switch:

- Heavy users of the ENG keys will have to spend a LOT of time in the Format menu, and entering engineering symbols from the catalog menu is absolutely not worth the bother vs. just keying in numeric exponents. If you mainly only use engineering symbol output and don't rely on use of the symbols for input, or use of the ENG keys, you might still be okay here.
- DMS entry fares a little better, but you still have to press Shift to input the DMS separators, and displaying DMS is another trip to the Format menu.
- Converting results between fractions and decimals also requires the Format menu (I'm seeing a pattern here).
- Base-N mode is a lot less convenient, as you have to press Shift to enter A-F in hex mode, and you have to cycle through all the bases with the Format key (at least it's not a menu) instead of having four keys to directly choose a base. Fine for a one-off problem now and then, awful if you're a microcontroller firmware developer.
- Unit conversions, while not all that great on the EX to begin with, take more keystrokes here for digging into the menus.
- The CALC feature is totally gone, though the spreadsheet is potentially a much more powerful alternative, with the downside of taking slightly more effort to set up the problem. Personally, I never used CALC, but I know some swear by it.
- Multi-statement (the colon) is totally gone.
- No Shift-Solve from the main Calculation mode; you have to go to Equation mode and pick Solver, then type in your equation.

For the minority where the positives outweigh the negatives, the CW is worth considering if you want better statistics, and a $20 spreadsheet workflow in your pocket. For everyone else, I'm sure it will be decades before having to worry about the EX becoming difficult to find.