Customer Reviews

Hewlett Packard 49G+ Graphing Calculator

5 star:		(19)
4 star:		(22)
3 star:		(12)
2 star:		(13)
1 star:		(23)

 

 

I purchased the HP 49g+ in August of 2004 as a replacement for its predecessor, the HP 49G. In my initial review of the 49g+, I downgraded it from 5 to 3 stars because of problems with keys not registering, and because of the amateurish user documentation. HP made a new firmware release available in May 2005, and it appears that the problems with key presses not registering were largely due to flaws in the way that key events were handled in the software, which evidently have now been fixed, with the possible exception of the function keys along the top row. That would have motivated me to increase my rating, were it not for the problem with the tendency for the keypad to wear out prematurely, which I have recently discovered. The Delete key on my 49g+ broke recently, and HP replaced it with a new unit. When I took the broken unit apart, I wasn't pleased with what I saw. Unless changes have been made to the keyboard design, I can confidently predict that the keyboards on all 49g+ calculators will break long before any external sign of wear, and when this happens, the calculator will become an expensive paper weight. Because I have no reason to expect the keypad design has been changed, I have no choice but to withhold any recommendation for this calculator. This is especially lamentable because if you consider only the remarkable capabilities of the software that runs on this calculator, it would easily garner a 5-star rating.

The user documentation that HP provides is woefully inadequate in numerous areas such as the important Equation Writer application and calculator programmability. The two CAS modes that each influence whether the calculator returns exact, symbolic results or approximate, numeric results, are fundamentally important. Given that these two mode settings are independent binary flags that differ in their esoteric effect, exact descriptions of the effect of each of these modes is fundamentally important. Yet, the descriptions given in the manuals are incorrect and useless.

In order for you to be able to anticipate what the calculator will do in any given scenario, you need to have a meaningful understanding of the rules that govern recursive expression evaluation, which is at the heart of calculator operation. Yet, you cannot find a meaningful discussion of expression evaluation anywhere in the user documentation. Prior to the introduction of the 49g+, you could acquire a meaningful understanding of expression evaluation by reading the manual page for the EVAL command in the 49G reference manual. But HP has since removed that manual from their web site. That not only removed the only accessible explanation of expression evaluation, it removed the only meaningfully complete descriptions of the calculator's 700+ commands. The only mention of EVAL in any of the documentation that is currently available is in the description of the keypad, which tells you that when you press the EVAL key, the EVAL command is performed.

If you write a program that has the essential characteristics of a mathematical function and you want to use the function plotter to plot its output, then you will have an opportunity to explore the nuances of the function plotter. When an algebraic expression is plotted, it is evaluated both before and after the x-coordinate value is substituted in place of the independent variable name. If you use a simple algebraic expression to invoke your program, your program will be passed the literal name that you have designated as the independent plotting variable, instead of the x-coordinate value. If you have written your program so that it will return a numeric result no matter the setting of those two important CAS modes, the static value of the actual global variable that you have named as the independent plotting variable will be used instead of the x-coordinate generated by the function plotter. If you are lucky, you will see a straight horizontal line in the plot window, which will give you a clue that the y-coordinate values are all generated using the same value for the independent variable. More likely, you will have trusted the function plotter to automatically set the vertical range of the plot window, and it will have chosen a vertical range appropriate for the static value stored in the actual variable and then will have changed its value so that when plotted, the constant function no longer falls within the vertical plot range. It is easy enough to steer clear of this lovely business if you can control the urge to smash the calculator long enough to discover that this nonsensical behavior only occurs when there is an actual variable having the name that you have designated as the independent plotting variable.

If you play with many of the plotters, you'll eventually get an error that simply says "Interrupted", and you won't find anything anywhere that gives you a clue why this happens. If you call HP's help desk, they will advise you to delete the reserved variable `PPAR', which is a nested list that holds the plotting parameters. Almost all of the plotters use this variable, but the structure and interpretation of the nested list varies from one plotter to the next. The various plotters don't bother to insert a descriptive label in the list, so when you open a plotter with `PPAR' left over from a different plotter, the plotter that you are currently trying to use will be confused by the unfamiliar structure, and may attempt an illegal operation that will lead to the useless error message.

The PC connectivity works great once you have installed it correctly and have learned to disconnect and reconnect the cable as needed. The install utility requires you to select what you want to install from a list and then it copies the files that make up the manuals, connectivity kit, USB driver, etc. It gives you no clue that you will need to install the special USB driver, but if you guessed that you need it, you got lucky, unless you tried to follow the driver installation instructions included on the CD, which don't work.

In May of 2005, when my 49g+ was not quite ten months old, the delete key broke, so HP sent me a replacement. HP didn't ask me to return the broken unit, so I took it apart. Three different techniques are used to attach the keys to the underneath of the faceplate. The four keys that are used to move the cursor (the arrow keys) are held in place by a single rubber sheet that is attached to the faceplate. Each of the three colored keys is a separate part, with each individual key rocking on its lower edge. There are two shades of grey keys on the keypad, and a block of black keys. There is one molded plastic part for each of these, i.e., one molded part for the top row of keys, another molded part for the block of grey keys, and another molded part of the block of black keys. These parts are fastened to the underneath of the faceplate, and each consists of a frame that holds the individual keys, all one molded part, and with the individual keys separated from the frame except by two thin strips at their lower edges. If you have ever put together a plastic car model and have seen how the individual parts need to be cut away from a frame of sorts, you have a good idea of what these three molded parts look like.

All three techniques for holding the keys to the faceplate permit the individual keys to press against a little rigid bump on a sheet of stiff plastic that covers the circuit board. Each bump is roughly 5 mm in diameter, and there is a thin conductive coating on the underside of each bump. When force is applied to the center of the bump, it inverts with a pop, pressing the conductive coating against the circuit board and closing the circuit across the contacts that are exposed on the surface of the circuit board. Except for the bumps, the underneath of the stiff plastic sheet has an adhesive that holds the sheet to the surface of the circuit board.

The problem with the top row of function keys is that the bumps are located against the upper edge of the plastic sheet. The effect is that the bumps on the top row aren't rigid enough to invert correctly with a pop that will cause the conductive coating underneath to make solid contact with the circuit board. The replacement unit that HP sent me does not seem to have the same problem. This may be because the physical design was changed slightly, or it may be because the de-bounce algorithm in the earlier firmware exacerbated a problem with a weak contact, but it is more likely because the plastic sheet on the new keypad hasn't broken in yet. But even if the problem with the keys on that row not consistently registering returns, soon thereafter it will be rendered moot when the keys break. My Delete key broke because the two thin strips that hold it to its frame, broke in two under repeated use. I observed that the same failure mode was ready to happen with the Enter key and the F3 key as well. When those two thin strips break, the key will move around and will become unusable because it has a short protrusion underneath that must be positioned directly at the center of the little bump in the plastic sheet underneath. Having inspected this first hand, I cannot avoid the conclusion that all 49g+ calculators are doomed to premature death because of this ridiculously cheap design.

If Bill Hewlett were alive, he would surely have something to say to the person responsible for allowing his legacy to be tarnished in connection with this junk. HP urgently needs to make a design change to the 49g+, and then needs to send a new calculator to all of the present owners of these calculators. All that is likely needed is to use the design that is used with the arrow keys, for the other keys that are attached to their frames with the two thin strips. With that change, these calculators could easily last for one or two decades, but if HP doesn't do something about this real soon, these calculators are all going to be converted to expensive paper weights very soon, and the damage to HP's reputation as a calculator manufacturer may be irreparable.

I've had a variety of hp calculators over the years and I've been very happy with them. From the first time I used one 15 years ago I recognized how superior they were to other calculators so I've used nothing but hp for the past 15 years. My most recent hp was the 48gx. I love the "feel" of the way its keys work, just like other hp calculators I've used. There's something special about the way the keyboard feel's. A good positive response. When you hit a number you could be sure it registered by the way it felt. Positive and firm, but not too firm (like the 49g+). Since I've received my 49g+ I no longer feel that way about hp.

HP made a MISTAKE in redesigning the "feel" of their keyboard. The 49g+ is NOT like other hp calculators I've used. If you're used to the hp "feel" and are considering the 49g+, you'll be very disappointed with this calculator. I made a mistake in buying it online without trying it first and "feel" like returning it. Next time I'll shop around more.

The 49g+ could also use some help in the display area. Although usable, the graphics are a little too crude. It's legible, for the most part (larger numbers are fine), but various menu's not so easy to read sometimes. I can live with the graphics, but the keyboard has got to go.

The advice from this review is from a long time HP user, with

very enjoyable "heavy user" programming experience with many

HP past "legends" like the HP-25, HP-67, HP-41C, HP-15C, HP-34C

and HP-48GX models.

Unfortunately, sharing a consense with many other reviews on

the HP-49G+, my experience after just 3 months of use of the

actual Top of Line HP product, is simply the Worst possible.

I will not readress many of the blatantly evident designing

flaws, already mentioned on previous reviews, but stress on

some not so common problems, which might have a direct impact

on a potential new buyer decision (towards avoing this

extremely risky HP model).

A Serious Hardware Flaw:

Aside from the delayed keybounce problem, with some keys

sometimes not being recognized, even though firmly pressed,

(which in my opinion may not be just a hardware problem at

all, but simply some flaw on the 49G+ operating system ROM

program, persisting on the most actual version 1.23, which

sometimes "looses" the key pressed due to some bad pooling

of the keyboard scan rate), there is a Real, Worse and

Serious hardware flaw, which HP tries not to "disclose"

to its customers, and such project flaw is that some of

the plastic keycaps, do not have enough resistence to hold

up with the stress of even a regular (casual user) usage.

This fact is noticed on another review, which indicates that

after 3 months of regular use, 3 of the keys where left on

a "floating" state, like hanging on the air...

That "floating" state of a key, is just the first stage of a

Total and Complete keycap failure, leading on a couple of

hundred keystrokes later, to a Completely broken key, which

although entirely loose from its plastic sustentation, does

still function if pressed with extreme care so as not to

definetively damage the key sensor mechanism.

The described problem happened to my HP49G+ after just 3 months

of regular use, with an initial audible "crack" of one of the

plastic sustentations of the backspace key, and a hundred strokes

later, on another definitive crack of the second and last plastic

sustentation, leading the key to a completely "loose" state

(even though still working if pressed with extreme care).

Contacting an official HP representative, I have found

that the maintenance cost of a keyboard replacement,

is about 3/4 the price of a new calculator, turning

such procedure in an extremely non attractive option.

My opinion, endorsed by other comments, is that there are

evident keyboard project flaws, that HP insist in not

public revealing, since that would imply on a complete

re-engineering of the actual Chinese production line of

the 49G+ (and probably also 48G II) keyboards.

A Non Economic project flaw:

Even though mentioned that the battery life problem could

be somehow fixed by the new ROM 1.23 version, my experience

with the most up to date ROM, is that the 49G+ "eats" Top of

line (Ultra) alkaline batteries, in a faster rate than any other

portable electronic device of the same kind.

If you careful notice, the HP manual does not even provide a

crude estimation of any battery life expectation...

My experience hanged from an initial 2 months life with very

low usage (not demanding excessive computer algebra processing),

to an unadmissible less than 1 month life, with some usual

computer algebra routine calculations, on a regular basis.

A Not so serious Hardware flaw:

Sometimes the calculator simply turn Off while pressing some

key. Sometimes this problem happened near the end of battery

life, but before the low battery sensor indicated the need

of battery replacement.

Of the noticed hardware flaws, this seems to be the less

harmfull one, since after pressing the On button, the calculator

returns to its previous unaltered state.

Notice however that some other users have experienced complete

hardware "freezes", meaning that some unavoidable loss of data

might end up, if one eventually succeeds in some "awekening"

procedure.

Some Unavoidable software flaws:

The CAS simplification routines, simply try to factor

everything it is faced up with.

Simply enter for example:

[square_root]( 5 - [square_root](2) )

with [square_root] denoting the square root key.

The result one ends up is just:

(5-sqrt(2))*sqrt(115+23*sqrt(2))/23

which clearly is Less simple than the original

expression (even tough, expressed on a factored

form...).

Such kind of nuisance happens only on some expressions

of the form: sqrt( a - sqrt(b) ), while expressions

of the form: sqrt( a + sqrt(b) ) are left intact.

Another software flaw is that Taylor series expansions

are limited to results of 20th order terms at most,

even tough an expert user can program his own Taylor

series routines to further develop a function series

with any desirable number of terms.

"Unrestricted algebraic manipulations is what

one expects from a Top of Line CAS calculator with

plenty processor and memory resources to do Exact and

Complete arithmetic and algebric calculations". That is

not the case of the exposed Taylor series expansions

treshold, since one would normally expect the system to

impose no limitations on the number of terms (which a

competitor like the TI-89, have no problem in dealing

up with).

A Graphic calculator with a "Textbook" formula style

Equation Editor, without an equivalent Textbook

format previous Stack elements Viewer. That means that

when you try to revise past calculations that have left

out the screen, instead of seeing the pretty type Textbook

format of previously shown outputs, one gets only the

unfriendly algebraic form of the expression (originally

displayed in Textbook style at the first time the result was

produced).

With a TI-89, pressing the up-arrow key, to review

past results, one sees all the previous formulas in

amenable pretty print format, while the HP 49G+ simply

shows back all input and output expressions in algebraic

format only.

A calculator which can show Extremely Big Numbers, in

just a single line... (Even though having 7 to 9

available lines for splitting Big Numbers, the 49G+

insists on showing up numbers of any size, on just

a single line, leading one to view only a portion of

what could be showed at a time).

For those who seek alternatives at Texas Instruments:

Even though the TI's have a far superior craftsmanship

in comparison with the actual HP line of calculators,

(specially with regards to the keyboard quality), in many

other aspects, even the actual Top of Line Texas products

like the TI-89 Titanium, leave plenty to desire,

specially with respect to calculation speed, and some

CAS limitations like for example expanding factorials

or powers beyond a few hundred places.

Another serious CAS flaw of the TI's is their inabilities

to cope up with limits and Taylor expansions of the form:

( a + x ) ^ ( b / x ) leading to exponential developments.

In summary:

As the title of my review suggest, one is advised to try

an HP 49G+ at its own Risk !

Of the exposed project flaws and potential hardware failures,

one should seriously expect the possibility of keycap brokening

due not to excessive usage stress, but simply due to extreme

lack of durability of the plastic key components.

If a choice should be made beetween a HP49G+ or a competitor

like a TI89 titanium, I would still consider a 49G+ due to

the power of its extensive CAS commands library.

Another missing feature on both HP's and TI's Computer Algebra

Based actual calculators, is the ability to cope with floating

point numbers of arbitrary precision. (Both manufacturers

provide intrinsic Indefinite Integer and Rational precision

capabilities, while lacking arbitrary precsion on Real floating

point calculations).

A viable alternative instead of considering a Top of Line

Graphic calculator, is opting for a PDA with a Windows CE

version of some Top of Line Computer Algebra Software, like

Maple or Mathematica, which would end up costing a bit more,

but with plenty more unavailable hardware and software

resources, in comparison to a simple calculator.

I have owned several HP calculators: 15C, 28C, 28S, 48S, 48GX, and the 49G. I was dissapointed with the 49G (it was slow and had flimsy keys and a poor screen) so I purchased the 49G+ when it came out. Unfortunately, the keyboard on the 49G+ is soooo BAD it is nearly useless. Even though the keys "click" when pressed, at least 25% of the time the keypresses DON'T REGISTER. You have to look at the screen to verify EACH and EVERY keypress, otherwise you could make a mistake and not know it...

I recommend WAITING until the keyboard problem is resolved, if HP will fix it. At this time (12/03) they haven't admitted the problem...

My first HP calculator was an HP-45 in the mid 70's, then an HP-55, 25, 15, 11, 41CX, and finally I got a couple of 42s's. The 42's are the best I've used with a keyboard that is perfect for me. I bought the HP-49+ on the web as no one carries them locally. The calculator's calculation powers are very good -- those of us who went to college in the 60's and learned to do calculations with slide rules and all calculus problems by hand will shake our heads in amazement.

But HP missed the boat on this calculator by not demanding what was good about their previous calculators. Here are my major disappointments with the product:

1. The keyboard is atrocious. You can get a click as tactile feedback that you've entered a character, but that's not sufficient to see if it was actually entered. As one reviewer noted, you have to look at the display. This is a major, major design screw up. I can only guess that the hardware was not designed by HP (contracted out to some Taiwan firm more interested in cost reduction).

2. Another reviewer noted the design screw up with respect to complex numbers. I like to see my complex numbers displayed as either '1 + 2i' and occasionally as magnitude/phase. But, you're only given the choice of (re, im). This was a major oversight.

3. The manuals are far from the quality of earlier calculators. I'd desperately like to have a printed manual, but I guess such a thing doesn't exist. The manual was also written by a person to whom English wasn't a first language (although, to be fair, it's much better than some other documentation I've seen from the Far East). Finally, the manual should have been editted by an experienced editor -- there are just too many mistakes.

Personally, I wouldn't have cared if the calculator was $150, $200, or whatever. I buy something like this for long term use and the initial cost is not terribly important. I want it to become a long term, dependable friend. I still have my two HP-42s's and my HP-11c and I treat those with kid gloves, because they're not easily replaceable. I do use the HP-49+, but somewhat grudgingly, as it's not quite the friend it should be.

I worked at HP for 25 years and I'm a bit ashamed of this product -- it's not up to HP's standards.

Finally, I'm not so sure of the place of this type of calculator. If I have a significant problem to work on, I'll use my own software or some tool like Mathematica. I can see how some of the symbolic calculations might be of use to a student, but I wonder how useful all those features are to the average user. I think HP would do its customers a favor by starting up production of some of the popular calculators from the past. The two I'd recommend are the 42s and the 15C. The programmer's calculator in the 15C form factor also still appears popular -- anyone I know who has one certainly won't sell it.

The screen is great; the contrast is the best on any calculator I have seen. The speed is very good, and there are many evolutionary improvements to the keyboard layout compared to the HP49G. There is no flicker; this problem was a software problem fixed a long time ago now. Apparently the alarms are not working reliably yet, so if this is important to you, be careful.

I have the version with the improved keyboard (bought in March 2004), and it is acceptable compared to previous HP standards (that is, it is not as good as the 28C for instance, seems roughly comparable to the 17BII, is much better than the rubber keys of the 49G) ... if you are not upgrading from an HP then all these comparisons won't mean much. The keyboard is ok, but purists will never be happy with it. It is considerably better than the first production runs (I had one of those but swapped for the new one under warranty), but it is not a great HP keyboard. The new version of the keyboard is much quieter (if the keys at the bottom sound the same as the keys at the top, you have the new keyboard; the original version was loud and hollow sounding).

Although famous for its sophistication, this calculator is good for every day calculation tasks because the screen, the stack and easy programmability make it very handy. The unit conversion feature, the equation writer and equation solving features are great. For lending across the table to non-HP users, swapping to "algebraic" (ie convential) mode is no bother (people who know the famous HP "RPN mode" will generally not use convential mode if they have a choice). What may not be obvious to newcomers is the very good selection of free programs from the user-community: advanced statistics packages, improved unit-conversion programs, sophisticated financial packages, and editable equation libraries for example. HP fans love to complain about this machine, but nearly all the best applications have been ported to the new machine, and nearly all posters have migrated to the 49G+. Thanks to the USB connection, it is really easy to transfer programs and data.

The manuals are excellent by normal standards. For programming they cover only the basics; the power and sophistication of this calculator is almost incredible; it inherits years and years of HP DNA going back to the HP48. These are things you don't notice initially.

Personally, I find this a tremendous calculator. I give it four stars because five stars would mean it is perfect, and it is not. I hardly ever give fours stars for a review.

The HP calculator community (check out the hp48 newsgroup) is a wonder of the internet, and thank goodness the HP49G+ proves that HP is still serious about calculators.

I will start by saying that I prefer doing all of my computaions in RPN. Currently, HP is the only company that I know of that still has calculators with this capability. I also previously owned the HP 49G, which was all but useless since it was too slow and seemed to get hung up on calculations that a TI 89 could whip through. Also, the HP 49 had abysmal documentation on how to actually operate any of the features or actual definitions of all the function calls built into the calculator. When I heard that HP was coming out with an upgraded version though, I was curious. I had even found an advanced release to the full instruction manual to the HP 49G+. At this point I found out what the HP calculators were able to do and found them to be more robust than the TI 89, assuming that you were willing to put the time into learning how to get the most out of this calcultor. I eventually shelled out the money for this calculator when it came out.

Ever since then, this calculator has been a bit frustrating. Being an electrical engineer, complex numbers appear almost in everything that I do. I found the way that this calculator handles complex numbers to be horrible. I think that the TI 89 does a better job at this by allowing the user to format complex numbers in either rectangular or in magnitude/phase format. Instead this calculator has a the format "(real,imaginary)". I guess that this is fine if you are doing everything in rectangular format, but usually it is usually more relevant to have it in magnitude/phase. I mean the obvious solution to put into this format requires an additional 4 keystrokes or just write a program, but the TI 89 allows for displaying complex numbers in either format. There are some other things that bother me about this calculator as far as the interface goes, but let's talk about the things about the hardware that bother me.

First, the whole reason that I am even writing this rant is because I needed to replace the batteries, which itself is a topic that I will get to shortly. Before I removed the AAA batteries, I was checking to see if I needed to move any programs around in memory and then turned it off. When I took the batteries out of the calculator and inserted new, out of the package batteries and then tried to turn it on, the calculator did not turn on. I then tried hitting the reset button on the back of the calculator but this still did not solve the problem. I tried looking for the troubleshooting section in the online manual for the calculator, but it was of no help. I them assumed that maybe the addition CR2032 battery for the memory was kaput, and replaced that, but guess what. You got it, the calculator did not turn on. I have also tried all of the various hard/soft restart keystrokes and it still has not come to life. I figured that I might be able to repair this thing if I were only able to open the calculator up, but the plastic case is held shut by what looks like plastic rivets, meaning that once I pop this thing open, there will be little chance of putting it back together again properly.

Second, I noticed that since I have owned this thing, it consumed batteries like nothing. When I first was using this calculator, I had to replace the batteries in the first 3 months. Tonight, in reading some other reviews, apparently the battery life can be extended by downloading a newer firmware revision, so this can be remedied, but it is still no excuse. It appears to me that HP's first solution to the super slow 49G was to increase the clock speed. This is great, but obviously someone forgot that faster clocks also increase power consumption. Duh!!! They didn't even attempt to imporve any of the internal algorithms at all.

Third, the keyboard is a bit ackward. This comment is noted by many others, and I eventually got used to it, but this is definitly not a calculator you would want to have while taking a test. You may not have a problem with it personnaly, but your neighbors might.

I give the calculator 2 stars beacuse it has RPN. It seems that overall HP did very little to actually improve the performance of this revision. They only seemed to modify the hardware and little in the software, and the job that they did with the hardware was quite poor. I'm really disappointed with HP. It seems that the quality of their products has taken a turn for the worse. If HP does decide to revise their calculators again, I would hope that they would do more than just giving their machines a face lift and do some actual improvements to everything and not take the cheapest way out.

I am-or was-a devoted user of HP scientific calculators. As an engineer I use HP's daily, and have developed several hundred programs over almost 30 years of use. I have purchased models from the HP-25 up to the 48GX and have developed hundreds of useful programs for various models over the past 25-or-so years. So what a major disappointment my purchase of the 49GX was. Yes, its power is awesome; head and shoulders above the 48's. But the tactile feel of the keyboard of earlier calculators has been replaced with keys so stiff even turning the calculator off sometimes requires three or four attempts. And the keyboard problems result in numerous computational errors: Triple-checking calculations often resulted in three different answers! The calculator is made in China and, believe me, it feels like it. After several weeks of trying very hard to get used to this model I have relegated it to the bottom drawer where it will probably remain forever. Regrettably, I have concluded that I must now abandon my love of RPN and start looking at Texas Instruments for a more powerful graphical calculator.

I recently purchased a HP 49g+. I also own a TI 89 Titanium. Both have stunning capabilities, but for a variety of reasons (familiarity being one of them I imagine), I prefer the HP. However, this calculator has the worst keyboard I have ever seen on a calculator. That is saying alot, I go back to the days of the MITS kit calculator! The keys on this thing often do not register. This caused me endless mistakes doing my taxes recently, especially before I fully realized the problem. This registration issue has been talked about at length by other reviewers, I had hoped (and thought) the problem was fixed by now. However, it most definitely is not. So while I would give this calculator the highest marks for capability and I personally find it easier to learn than the TI, I have to really fault HP for lousy engineering on the keyboard. If you buy this calculator, be sure to press the keys slowly and hard. Very frustrating.

I recently purchased a 49g+ to upgrade from 48gx. After 2 replacements from HP I gave up on getting one with a keypad that works properly and returned it for a refund. I had read where the keyboard problems were solved but it does not look that way to me.

The folks at HP tech support ensured me that the two replacements were from the latest batch (Jan 04) and that the alignment problem did not exist.

On the first calculator, the F1 key was intermittent. On the second, the problem was with the backspace and swap. On the third, the '1' key felt significantly different than the others and the swap key also worked intermittently.

The problem is that you sometimes feel the mechanical 'click' before the key press is actually registered. This means that you have to constantly watch the display to make sure your keystrokes were acknowleged.

Other than the keyboard problem I really liked the calculator. It was much faster than my 48GX and the contrast and size of the display was very nice. I may try again in a few months and see if they worked out the problems.