hp 49 for $450????


Am I missing something, Or is this man insane?



Am I missing something, Or is this man insane?

Anyone who buys it at that price will deserve a different adjective: idiotic. The HP-49G was one of the worst calculators ever made by HP. And there's no justification for buying one now that the 49G+ is out with a much better keyboard and features -- for 1/3 of the price.

Long ago I bought a 49G, hated it, and traded it for a 41CX. I haven't missed the 49G for a single minute.



The 49G sounds like a good candidate, but I've never owned one, so I can't speak from experience.

My nomination: the 6s/6s Solar. Truly worthless. (But then, one could argue that H-P didn't actually "make" it . . . )


I second Paul's nomination. The 6s solar is a piece of junk. I originally bought one for shirt pocket use. I had hardly used the damn thing before lcd segments started failing. It started off with one segment, then another, and another... When it got to the point that I couldn't even guess the answer, I threw the calculator away.

The thing is I bought this calculator because I used to have a tiny casio scientific wallet type calc, more or less credit card sized. It was so light and convenient, and useful. Unfortunately it fell out my shirt pocket,fell 30 meters and broke. I was never able to find another one to buy (but come to think about it, I haven't searched ebay). I thought the 6s would be an ideal replacement, but it's actually too big.


That tiny Casio must have been a fx98 (see http://www.hknebel.org/Museum/Museum/Rechner/Taschenrechner/Casio_FX-98/casio_fx-98.htm).
A pure jewel !

On the worst HP ever, my vote goes to the HP-10BII, because of the horrible keyboard. Keyboards MUST be reliable.


Hi Neil. You may want to look at the Casio FX-260. It is not a beloved high quality HP but for $10 it is a decent buy. It is solar powered, fits nicely in the shirt pocket, is quick, fairly accurate and has many useful functions. Also, if it is dropped or lost you won't be out much money.




Tough call, Agree that the Hp6s was junk (rubber keyboard), but the early Hp10Bii had failed LCD segments shortly after I acquired (can't say about present build). Poor keyboard from day one, as some keys had no click while others felt fine. AND all that for a calculator that never left a pampered shelf. My Hp6 was used by my daughter in a bookbag, took punishment and held up well. Some keys were gouged out??? but worked when it was put away after a school year of abuse.

My vote: Hp10Bii


Isn't nominating one of the post 1990, outsourced models kind of like shooting fish in a barrel? How about a nomination from the time when they were supposed to be good? I'm thinking of a series that was basically ruined if you just happened to plug it into its charger without a battery pack, or maybe even if you plugged it in with a pack that was not making good contact. Don’t get me wrong, I like the woodstock calculators. I had a 29C until it got lost between the time I got a 15C to replace it and realized that I HAD to have one of every model ever made. But HP should never have let such a design out of the door.


I'm going to go against the likely grain of this thread and nominate the HP-55. Although it was novel in having an accurate timer and a limited program memory, every one that I've run across has LED segment problems to some extent and I really don't like the volatility of the program memory. You could use this argument against the HP-25 too, but it was half the price and was offered for sale during the same time period.


The HP55 has the same display devices (5 digits in a package) and display driver chips as the other classics (35, 45, 65, etc). There's no reason why it should have any more problems with missing segments than the other machines -- and in fact my 55 was perfect last time I turned it on, and has never been repaired.


I realize that the 55 uses the same components as the other "classics", but for whatever reason all 4 of the ones that I've had had 1 or 2 segment problems and I've corresponded with other collectors who have seen the same thing. I don't generally find these problems on other classics. I wonder if the problem is due to the calc being used in timer mode for extended periods of time which might cause certain segments to fail.


I've experienced the same sort of thing as Katie, both of the HP55s I have had bad segments when I got them. The LED modules are the same part number as the rest of the classics, but the inductors that provide the charge dumped into the segments are different (they may be the same value, though, I've never measured them).




Most HP55s have those 4-inductor modules in them (Note, the 2 modules are not the same, one contains 4 inductor of the same value, the other has an odd one for the (smaller) decimal point (period) segment). They're also used in the HP65. AFAIK they are the same value as the inductors used in the 35 and 45.


Yes, for some strange reason the HP55 is a rather failure prone machine. The HP55/65 introduced the silver lead LED packages. Earlier LEDs had gold leads and a slightly bigger LED character.

Besides the bad LED problems (also seen in a lot of HP65 machines), there are many with CPU problems. I don't know why they would fail more than HP35/45 machines. It is also interesting that the HP21 has a much lower ACT failure rate than the HP25, even though both machines share the same icky power supply design.

Also the green plastic HP55 case does not seem to be as stable as the original black cases. In particular, the latches on the battery door like to disintegrate.

The silkscreening on the keyboard is not very stable. Many machines have worn off labels... particularly the blue ones.

The keyboard key click tends to go mushy. (the HP65 key contacts tend to split vertically along one edge).

Finally the silver trim below the display window is almost always worn completely off on the HP55.

I think most of the problems are related to "cost reduced" manufacturing techniques over the earlier machines. I suspect the case problems are due to some minor change in the plastic formulation to make it green.


AFAIK the 2 types of LED package are interchageable, although the 3 packages in one machine should all be the same type. Mind you, it works fine if they're not, it just looks odd.

My HP45 has a mix of display types!. I kludged that machine together from a box of bits given to me by a fellow HPCC member. It's worthless to a collector (silkscreen rubbed off, battery contacts home-made, odd display packages, etc), but I love it because I got it working the hard way. The logic board was built from bits of 2 -- I had to work out what was dead on one board and hope it was OK on the other one. Fortunately it was!

The same displays are used in the later (09810-66542) display board for the 9810 desktop calculator. The older board (09810-66541) has individual digit modules -- a total of 45 of them!. Both boards have yellow and brown handles (to match the card guides in the chassis), AFAIK this is the only time where the handles in a 98x0 machine don't give the last 2 digits of the PCB part number using the obvious code.


All the HP display modules seem to have the same part number even though they are not exactly interchangeable due to the digit size. The leads in the 9810 (and the HP46/81) are trimmed differently becuase the modules are tilted down in the hand helds and up in the desk tops.

One should also check the display brightness code letter on each package. Usually in the range J..M, a difference of a single letter is not really noticeable, two letters and the casual user will not know, three or more can be noticeable.

I have an HP55 with chips and displays salvaged from HP35 and HP45 machines. About the only thing original in it is the passives and the ROM chip. The battery door and contacts are from a '35.


My 55 (ser # 1704s) is the only one i have ever used so i don't have a lot to compare it to. It has no problems with the display or anything. It is "well used" and has numerous dents and a big scratch but no worn off lettering. It's good that it functions well 'cause i paid a whole dollar for it and it would be a drag if it didn't work well.

BTW: i have noticed that a battery with just the right level of almost-drained-out-ness will run my 35 fine, run my 45 but it has a bunch of decimals, give ghost a display that drains in seconds on the 55, and won't even turn on my 65. The 67 runs but has the low battery light. I'm not complaining; just mentioning it.



You said that the 9810 display board handles (yellow -brown) are the only cases of not compliance with the obvious part number code. As Yellow-Brown means "41", I would like to suggest that a secret message for true RPN believers may have been hidden there for almost a decade !!

Fortunately enough for my "theory", the 9820 (not RPN) used a different display.

Also, I would never accept any Red-Gray (28), Yellow-Gray (48) or Yellow-White (49) combinations as a valid message for RPL justification!!

But ... I wondered why the 28 series used a red shift key and gray keyboard plate; instead of keeping with the usual yellow over black ... May here be a clue for shifted key color changes over time? If so, the 48G series should have been the "57" (green-violet)!!


I don't have a 9820 (YET!!), but I suspect the display board is a 09820-66541. HP subassemblies have part numbers consisting of 2 groups of 5 digits. The first group gives the model number where it was first used. For the 98x0 machines, -6654x numbers seem to be display PCBs. The
09810-66541 and 09810-66542 are the 2 versions of the 9810 display, the 09830-66541 is the 9830 display and the
09830-66542 is the 9830 display driver.



how does the extreme color scheme of the HP-70 fit into your theory? Of the classic type calcs, it's by far the ugliest combination (IMHO), although key groups on the HP-70 should be visible even to color-blind people;-)



I pretended to make some fun about the colors and model numbers (really part numbers, as pointed by Tony). But, just to let it make an extra turn, the color scheme of the HP70 may be decoded as:

Orange Keys => 3

Pale Green Body => 5

I suppose it means that the "secret wish" of the HP70 was to be as important as an HP35 !!

For those not familiar with it, the electronics color code, used mostly for resistors but also for other components is:

0 Black
1 Brown
2 Red
3 Orange
4 Yellow
5 Green
6 Blue
7 Violet
8 Gray
9 White


Ok, say HP-48G & earlier. ("True" H-P.)

I then nominate the 28c. A dramatic departure, but hobbled by that legendary battery compartment door and WAY too ambitious an operating system (RPL) for the available memory.

The 28s fixed the memory problem, but then begged for serial I/O. (That and a programmed "wake up/alarm" function would have given them a PDA before the "PDA" acronym had even been coined -- though the battery compartment door remained a perennial issue.)


I second the 28C for the same reasons if we are going to restrict ourselves to true HP calcs.

Actually the 18C was every bit as poor, but its intended audience probably did not complain about its shortcomings because they wouldn't use the memory except for note taking.


I have a USA manufactured HP-28C (purchased new in 1988) and I like it very much. While I don't like the size N batteries, I've never had a problem with the battery compartment door.

The main reasons I like it are its display layout and its support for number base conversions and logical operations using an adjustable (1-64 bit) word size (a la HP-16C). As an embedded systems programmer, these features are much more important to me than graphics, complex numbers or matrix operations.



I agree with you Britt. I have a 28C from new 1988 USA build, and have never had a problem



I'm not going to nominate any for worst calculator overall, but I do think they have made some surprisingly bad design decisions at times. Most of the ones on my list are relatively recent, but even back in the good ol' days they didn't get everything right:

  1. Using the battery pack to filter and regulate the power from the AC adapter. Some models, most notably the Woodstock series, can be damaged by operation from the AC adapter without a good battery pack installed. Or if the battery contacts get corroded.

  2. Press-fit solderless assembly used on the Spice series (HP-31E etc.). US patent 4,197,586 describes this, and gives as the rationale that it avoids the possibility of damaging the chips due to soldering temperature. Sorry, I don't buy it. Most electronic products made since the 1960s have had the components soldered in, and the failure rate due to that is very low. On the other hand, the failure rate of the press-fit solderless calculators was extremely high, and HP did finally give up on the idea and retool to use conventional soldered assembly.

  3. The AC adapter connector on the Spice series. Junk. The earlier and later ones are much better.

  4. Eliminating double-shot injection molding of keyboard legends. Now the legends wear off over time.

  5. Eliminating the snap-disc keyboards. The snap-disc has the property that the tactile response is guaranteed to occur if and only if the contact will be made. Other inexpensive keyboard technologies such as membrane keyboards offer poor tactile feedback, and the tactile feedback can occur without electrical contact or vice versa.

  6. Replacing the traditional HP color scheme of gold and blue for the shifted functions with green and purple on the HP 48G/GX, and various other combinations since. Being partially color blind, I could barely tell the colors apart, while I'd had no trouble with gold and blue. Reportedly a focus group of high school students picked the colors. Since it's a calculator for engineers, how about letting engineers pick a reasonable color scheme that's functional, rather than attempting to look "cool". Looking cool is OK when it doesn't make the calculator harder to use. Let high school students pick colors for calculators targetted to the high school market. I doubt that they sold a single extra 48GX or 48G due to the colors, and in fact I bought more used 48S calculators rather than buying a new 48GX. They do seem to have gotten better about this on the latest models.

  7. Replacing the large, prominent ENTER^ key with a small key in the far corner. C'mon, guys, it's the most important key on the whole calculator.

  8. Putting the HP name on commodity calculators from China. (I don't mind Chinese manufacture of HP designs, but HP shouldn't have branded calculators they didn't design.)

  9. Poor documentation on HP 49G. The HP 49G+ documentation is much better, but still not up to traditional HP calculator documentation standards.

  10. Insufficient QA before release. Especially noticable on the 12C Platinum.


The winner is the HP-48GII - even HP pulled it out.

When (if) they bring it back it still will not have Flash for bug fixes,

it still will have only 128K RAM leaving 80K for the user
AND THERE IS NO Equation Library build-in EVEN though the "48" is used as a model name.

The display is small.

The Serial port is a special mini-B clone, which needs a HP cable (unlike the 49g+ model, which takes any (video)camera cable.

The RS-port can NOT be used for controlling devices, since it pulls the power from the lines where only a PC (or other similar controller) offers it.

All the bad things in the 49g+ will also apply to the 48gII

Like: the IR is not 48 compatible

[ENTER] is small and in wrong place

Colors are horrible - as in ALL the new HP models

etc..ad infinitum



A few more..

The connectors in the HP41 fullnuts that are held in contact by the case screws. I must have fixed 50 or 60 of those darn machines.

The flexible PCB in the HP41 that connects the batteries, I/O ports, and keyboard PCB. That thing suffers terribly if the batteries leak.

[Not strictly a calculator..] The battery pack/charger circuit for the 9114 disk drive. The charger can't keep up with the current taken by the drive, so the battery goes flat when the drive's in use, even with the charger plugged in. You just can't run it continuously using HP-provided stuff (OK, hackers connect the battery connector in the drive (pack removed) to a bench supply).

Not putting an RS232 port on the 49G+. Have USB as well if you insist, but RS232 is useful for connecting directly to peripherals (printer, data logger, etc). I can't be the only person who uses a calculator without having my PC nearby.



nearly full ack,
and your point 10 also dramatically refers to the 49g (NOT +) which is still beta after all those years.

For the 48G series: I admit that the color schema isn't as good as on the S series, but since I used the G series very much, I know where the menues are;-)
And the overall keyboard layout is similar to the S series, with only a few exceptions.


Edited: 31 Mar 2004, 7:11 a.m.


I generally concur with all of Eric's comments about various poor HP design decisions over time.

I would add these items:

- the excellent Woodstock-style case should have been retained for HP41C;

- use of size-"N" batteries for the 41C;

- the "zebra" connector connecting a fullnut HP41C PCB to the keyboard PCB can be flaky;

- HP41C should have had sturdier support columns w/metal inserts - these can crack on overtightening, or be too loose and cause contact loss when the 41 "sandwich" spreads open a bit;

- "heat staking" keyboard PCB to front panel even in old calcs which were designed to be repaired (these were high-cost items to begin with and should be fully repairable);

- machines w/ *any* I/O (HP41, 71B, etc. and later) should have had standard UART serial I/O and associated support firmware. An external level converter to swing btwn +12/-12vdc for true RS232 levels would be OK.

- not having a "true RPN mode" (4-level XYZT+L stack) available on the 48 series;

- not using a true emulation layer of prior calcs' firmware for 33S and 12C-Platinum.

Bill Wiese

San Jose CA USA


I have to disagree about retaining an RPN stack for the 48 series. While it does make it more complex, it is more also more powerful. The 28/48 were supposed to expand the envalope and the 4 level RPN stack constricted that.

The 48 was to be much more than a 41. That said, there was still a real market for a 41 type replacement that was never filled and actually just plain abandoned. The 42s didn't have I/O and the 48G was way to BIG! and cumbersome because of its size and the complexity of RPL.

Just my $0.02.


I feel that which of {4 level RPN} or {'infinite' level RPN/RPL} you prefer depends on which you learnt first.

Although I had an HP41CV when it was current, I didn't use it that much (I was doing serious hardware hacking on the Sharp PC1500...). I really started using HP machines when the 28C came out.

I find that I actually prefer the RPL stack, named variables, different object types, etc. I love my 41s (don't get me wrong...) -- heck I like the 3-level RPN stack of the 9100/9810 as well. But I have no problems with RPL either.


Hi Tony,

>I feel that which of {4 level RPN} or {'infinite' level RPN/RPL} you prefer depends on which you learnt first.


Not always.

The calc I learnt the phantastic RPN with was a 33E which a friend lent me for a few days. After that experience, I bought my first HP-41C and HP-11C, and used and programmed them (especially the 41) every day, mainly in ML.

Writing RPN programs with optimal stack usage was a nice game, at least at the user code level. The automatic T duplication or seamless stack roll feature were (and are) very handy sometimes.
But it can lead to a very dirty program style, because you never have to clean up the stack mess your program left.

With the event of the first RPL machine (28C) I (and many other people) had to learn the new stack concept. First it seemed a bit difficult, but after all it's as easy to handle as the traditional RPN stack.

When programming in RPL, you have to watch your stack parameters, and take care of other programs' stack parameters.

However, both worlds have their advantages. For programming I much prefer the HP-48 (RPL and ML), for daily adding of a few values the HP-41 is more handy, but I sometimes ask myself how I could be satisfied by a single line display for all these years;-)




I have particularly bad memories of my relatively expensive HP33E routinely going crazy in university examinations. HP didn't seem to be able to fix it so I went to the dark side of the force (TI58) for the next three years.

The lack of protection of the screen on the HP48 also seems poor to me. How many people have broken that ? Even Casio and TI give you a protective cover for graphing calculators.

On the other hand, my HP41CV is still going after 20 years and my HP11C has lasted 15 and I wouldn't change them. It's just that we have such high expectations of HP's engineering.


Bill Weese wrote:

"- machines w/ *any* I/O (HP41, 71B, etc. and later) should have had standard UART serial I/O and associated support firmware. An external level converter to swing btwn +12/-12vdc for true RS232 levels would be OK."

That would have really been nice. The whole thing could be interanl, but there should be a way to shut the RS-232 drivers down so you don't have to keep disconnecting the cable every moment you're not using it, since RS-232's power requirements are a bit heavy for small batteries.

I do have the FSI164A HPIL-to-RS232 interface converter though, which is similar to an HP82164A with an extra RS-232 channel (and optionally up to 8 channels). I've used it a lot with both the 41 and the 71.


the excellent Woodstock-style case should have been retained for HP41C;

The Woodstock series clearly had the best industrial design of a calculator ever. Given how atrocious the industrial design of the Spice series was, I can only guess that the industrial designers didn't want to move to Corvallis, Oregon when the calculator division moved there in 1976.

In my opinion, the 41C industrial design is fairly good. The real problems are the internal connections, as Tony Duell has pointed out.

And the Voyager industrial design is very good, provided that you don't mind the horizontal layout. Some people love that, some people hate it.

The one that surprised me was the 38G. They put a lot of effort into the industrial design of the flip cover, but the calculator itself is just a rectangular box with keys. I thought this model was supposed to appeal to high school students, but I can't imagine that it particularly did unless there was some fad for boxy products that I haven't heard of.

Edited: 1 Apr 2004, 1:03 a.m.


Battery doors on the clamshells.


HP-48 G user interface in calculator mode [error messages (DROP: Too few arguments) instead of a NULL keystroke nullifier] and RPL fundamentalism (everything in reverse notation: 1 10 FOR...)

The HP48 is not a bad product at all but, in my humble opinion, it was a lost opportunity, with that much power...

I feel someone's compass failed while approaching high latitudes...

A Classic RPN emulated mode, would have eased some pains.


and RPL fundamentalism (everything in reverse notation: 1 10 FOR...)

For me, it is not fundamental enough:
<< 1 10 FOR I >> instead of << 1 10 I FOR >>
and << IF test THEN...>> when you have << test IFT ...>>



Why not << 10 1 I FOR >> or (even better) << 01 1 I ROF >> ?




view your support I


weiv ruoy troppus I

RPN was good not because it being backwards, but for being intuitive, practical and understandable. But I don't want to start this old discussion again; I understand there are people who like RPL as much as I like RPN, and who think of RPN as something as sad as parenthesis in a calculator keyboard. I respect such positions, but I liked just to give my humble opinion.

What about the "Drop" error message


Hi Andrés,

mine were not meant as flames to start another fire: I agree that each one has his/her preferences; mine are not RPL oriented but this is my fault since neither do I have the time nor the will to master another language.
As far as calculators are concerned I started with RPN and do not need more but YMMV, of course...

The looping excerpt got my attention since last week I took my 49G+ for a quick and dirty benchmark against some of its ancestors and, well, I found that I forgot the right syntax!

And yes, I got that "drop" error too many times ;-)

Take care,


How about

 << 1 10 I ORFAY . . . EXTNAY >> ? 

I remember long ago some wag suggesting that the then-current fashion in programming languages of bracketing constructs with mirror-image verbs such as

   if ... fi
do ... od
should be taken to its logical extreme with
   comment ... tnemmoc

(I've always loved that one.)


It is impossible for a simple theoretical reason to have "1 10 I FOR" instead of "1 10 FOR I".

The reason is that if you take the first 3 steps of "1 10 I FOR" you have a perfectly valid program which pushes 3 elements on the stack, while for the third this is not what you want : you want to declare a local variable.



One implementation of a "proper" behavior for << 1 10 I FOR ... >> might be if the "I" pusheS an identifier on the stack, and then FOR uses the identifier in declaring the local variable for iteration.

However, the interpreter writers out there may have a more informed opinion to share . . .


You are right, what I really meant was:
<< 1 10 'I' FOR ... >>.
As far as RPL purity is concerned I really prefer sysRPL but it is slightly more troublesome than UsrRPL so I stick to the latest on for a quick little program on the side.



I have never owned a bad HP calculator. Of course, I haven't owned all of them either. I have had (and still have) the 11c, 42s, 32sii, 48G, 48GX, 33S, 12c and a newly obtained 41c. Not a loser in the bunch as far as I am concerned.

Take care.



Missing something? Yes, it's not your everday, ordinary run of the mill HP49G - it's preloaded with "Directional Drilling" software... Just in case you have an oil field in the back yard... Of course if you do, then you don't care about the $450... Chump change.

In that case, I've got an oil dowser for $10,000 should you need to find the proper place to drill... IMO about as useful as a 49G.


Just a minor clarification Randy; the directional drilling software does not tell you "where to drill", rather it tells you "where you are drilling to". Some interesting trigonometric calculations based upon depth, deviation from vertical and azimuth - but something easily duplicatable in a 41 with reasonable accuracy. Mind you, a 41 might bring that price in as well - and it is arguably more robust!

Your dowser would be better served in the seismic industry;)

My .02 as a pet. eng.


Possibly Related Threads...
Thread Author Replies Views Last Post
  HP Prime... NOT meant to replace HP48,49,50 ? Chris Pem10 21 4,212 11-18-2013, 03:30 PM
Last Post: Chris Smith
  ANN: A MacOSX Kermit program for HP48/49/50 calculators Paul Onions 2 1,221 09-08-2013, 04:55 AM
Last Post: Paul Onions
  Nibble reverse (HP-48,49,50g) Gerson W. Barbosa 44 8,305 07-28-2013, 10:10 PM
Last Post: Gerson W. Barbosa
  (OT) Pandigital expression (HP-48,49,50g) Gerson W. Barbosa 19 4,022 07-19-2013, 05:59 AM
Last Post: Gilles Carpentier
  downloading HO 229 v. 2,0 from HP 49/50 Astronomy Programs list Al 10 2,083 11-26-2012, 05:38 PM
Last Post: Al
  Best Keyboard of the HP 28/48/49/50 Series Eddie W. Shore 9 2,332 07-12-2012, 07:21 AM
Last Post: Dave Britten
  49/50 PC Development Tools Under Linux John R. Graham 1 758 04-24-2012, 03:06 PM
Last Post: Han
  49? What about 99 (or even 98)? Matt Agajanian 11 2,140 03-21-2012, 11:41 PM
Last Post: Eddie W. Shore
  HP-48/49/50 software sites? x34 3 1,043 04-07-2011, 05:44 PM
Last Post: David Hayden
  gamma approximation in 48/49/50? Eric Smith 19 3,433 09-25-2009, 08:06 PM
Last Post: hugh steers

Forum Jump: