HP Forum Archive 12

[ Return to Index | Top of Index ]

Calculator Design (A few questions) Message #1 Posted by Ben Salinas on 1 May 2003, 11:51 p.m. I almost have my design for the contest complete (http://www.hp-calculators.com/contest/index.phtml), but I have a few questions. I have designed the calculator using autodesk inventor, and have mapped the keyboard. I need help in deciding a few things. First, the basics on this calculator. It is slightly taller than a HP32sii, a bit thicker, and the same width. It has a 2 line LCD screen, which has a backlight, and many functions. I have mapped the keyboard, and written up a brief description (totals about 6 pages:) )(I will have it posted online tonight, and will follow-up with the address) of the functions. I need help in a few major areas. I need input to know whether i have left any vital functions out, or if i need to change any of the existing (like if they aren't useful). I also need info on how previous calculators have performed matrix operations. I also need to determine what flags should be included on the calculator. (I was looking at the 41's flags, and will probably include most of those)Finally, i need to decide on the cosmetics of the calculator. I need to decide on the color of the backlight, and the color of the calculator itself. Originally I was thinking black, but several of my friends have suggested a blue backlight with a silver-aluminum case. This would appeal to students, but I don't know how many engineers would want to walk around with a glowing blue calculator. What do you think? BTW, I have decided on a name. It is HP-51, because it is one of the only numbers not already used, but also because of some crazy "Ben logic." I decided to incorporate HP, which is PH backwards. PH makes an F sound in english, and F in Hexadecimal is 15. So, 15 (F or PH) is HP backwards, so HP would be 51, and hence HP-51.

Re: Calculator Design (A few questions) Message #2 Posted by Ben Salinas on 2 May 2003, 12:02 a.m., in response to message #1 by Ben Salinas Well geocities wasn't working, but let's try this anyway: (Sorry its so long) Features And Functions When I started designing my entry, I knew the basic functions I wanted for my calculator. It needed to be scientific, and have all the functions I always wished my 32sii would have. Most importantly, it needed to be programmable. After designing my calculator’s basic appearance, I started on the daunting task of mapping the keyboard. Each key needed the perfect location, depending on how much it would be used. Starting from the top left going right and then down, the functions are as follows. Keys 01-06 (User keys) These keys have no predefined functions, but are used to select menus on the screen. They also have memory registers A-F, and are used for searching through the catalog (when in the catalog, pressing one of the buttons will take you to the catalog functions starting with the small letter above the key). The A key also doubles as the Alpha key and the B as the Alpha Lock. The F key is the = sign for the equation solver, and the D and E key are the < and > signs respectively. Keys 07-08 These keys, when using their primary functions, are the store and recall buttons. They recall numbers from the English/Latin alphabet registers, and strings or letters from the Greek character registers. They are also the G and H registers. Their other functions are % and pi keys. These are buttons that are important enough to be on the keyboard, but not vital enough to be next to the keypad, or as a primary functions. Keys 09-11 These keys perform the basic statistics of the calculator. Numbers are added to the stack with the Sigma+ buttons, and are removed with the Sigma- button. The previous entry is displayed in the top part of the LCD screen. The Up and Down arrow keys can be enabled in the SUM button. With this enabled, the arrows will scroll through the entries. The SUM button brings up a menu with the following choices: n, x, y, x2, y^2, x^y. The s,? button brings up a menu with the following choices: sx, sy, ?x, ?y. The xBAR,yBAR button brings up a menu with the following choices: xBAR(the symbol), yBAR, and xBARw. The L.R. button brings up a menu with choices: x (with ^ above), y (with ^ above), r, m, and b. These keys also have memory registers indicated by their letter to the right of them. Key 12 Key 12 is the LBL and RTN key, which is used for programming. It is also the L register. Key 13&14 Key 13 and 14 perform logarithmic functions. The primary functions use base e, while the secondary functions use base 10, or whatever base the calculator is in. They also have memory registers assigned to them. These are very useful functions, so they are directly above the enter key. Key 15-17 Key 15 performs squares and square roots. Key 16 raises numbers to powers, and roots. Key 16 is also the up arrow key, when the arrow keys are enabled (used for programs). Key 17 has the inverse and factorial button. These are more mathematical functions, which would be used while performing quick math, so they are closer to the number pad. Key 18 Key 18 is the up arrow, which will scroll through programs by default, but can be configured to scroll through Statistic data. It is also the flag button. The menu allows to check if a flag is on, clear the flag, or set the flag. Flags 0-4 are displayed on the screen. Key 19 Key 19 is the most important, most used key on the keyboard: the Enter key. This key could not be too close to the number pad, so the fingers have room, but can’t be too far. It is also the largest key, mainly because if you need to find one key on keyboard, it is the enter key. The secondary function is the Show key, which displays the checksum for equations, and programs, and shows all the digits, whether in decimal or fraction mode. Key 20-22 Keys 20-22 are the trig functions. The primary functions are sine, cosine, and tangent, and the secondary keys are arcsine, arccosine, and arctangent. The Hyperbolic functions can be found in the catalog. Key 20, 21, and 22 are also the left, down, and right arrow keys respectively. Key 23 Key 23 is the down arrow (like key 18). The secondary function is the input/view function, used mainly for programming. These two functions have been consolidated onto one button. To use the input function, you would press: Input/View STO Variable to store in The view function is used by pressing: Input/View RCL Variable to view Key 24 Key 24’s primary function is the MODE button. This pops up a menu allowing you to select: Complex, Standard, .(decimal) or , (comma). The C key doubles as the i key. So, to input a problem like (5i+3)(2), you would first put the calculator into complex mode, by pressing: MODE Complex Then, the problem would be inputted by: 5 ENTER C (i) x 3 + 2 x The constant will be displayed on the screen (6), and the imaginary part will above on the screen (6i). The secondary function of this key toggles between Degrees and Radians. The mode will only be displayed on the screen if it is in Radians mode. To convert from Radians to Degrees, the calculator must be in radians mode. The keystrokes to convert 6.29 radians to degrees would be: 6.29 ENTER Second 0 (Convert) Second MODE ?(degrees,RAD) The answer, 370 degrees will be displayed, and the calculator will still be in radians mode. This allows users who prefer to work in one mode to input data from the other standard without switching modes. Key 25 Key 25 is the R/S key (run/stop) and has the secondary function of PSE (pause). When used in a program, the R/S key will stop a program. When used in calculation mode, it will start running a program from whatever line it is one. The PSE key is used to pause a program or a message for a given amount of time (PSE 10 would be to pause for 10 milliseconds) Keys 26-29 Keys 26-29 are commonly used functions. Key 26 is one of my most used functions, the “switch key.” It is used to, as the name implies, switch the number on the screen with the number previous to it in the stack. The secondary function to this is the “switch with variable” key, which switches the number on the screen with the number in a memory register. In programming mode, this can also be used to switch strings of letters. Key 26 is the +/- key, which multiplies the number highest in the stack by -1. The secondary function, R ^(up), moves up in the stack. Key 28 is the E key, used for scientific notation (1.235E10 is the same as 1.235x10^10). The secondary key is R(down), which moves down in the stack. Key 29 Key 29 is also very important for anyone who makes mistakes. It is the backspace key. The secondary function, the CLR function, pops up a menu with the options to clear: x (the number in the LCD screen), stack, eqn (equations), pgm (programs), vars (variables), or all (everything). The calculator can be reset by pressing the ON, A, and F key. Key 30 Key 30 has a unique feature, the BASE function. This function allows you to operate in any base between 2 and 37. To change the base, enter the number of the new base, and press BASE. The base will be displayed on the LCD screen. To convert a number between bases, is very similar to converting from degrees to radians. The number on the screen is the number to be converted. Then press: Second 0 (Convert) Base to be converted TO BASE The new base will be displayed on the screen. To input letters, the variable keys are used. A-F are keys 1-6, but, for the higher letters, the alpha key must be used. To use this, press second A, or second B for the alpha lock. This will allow any letters to be inputted. There will be no conflict with trying to input numbers and letters at the same time because none of the letters fall on the same keys as a number. The secondary function on this button is the MEM function. This displays the amount of memory available, and allows you to clear specific equations, programs, and variables (including Stats variables). Keys 31, 32, 33, 34, 36, 37, 38, 39, 41, 42, 43, 44, 46, 47, and 49 all have primary functions as a number or basic math function. Only the secondary function will be described. Key 31 Key 31, along with 36, 37, 41, 42, and 47 all “work together.” They will be discussed at the end of this section. Key 32 Key 32’s secondary function is the equation solver. This opens up the “library” of equations. Equations are inputted algebraically. This solver, which is different than the 32sii’s solver, will solve for all of the possible answers. More about this solver is to come later Key 33 Key 33’s secondary function is the PROBability menu. This menu includes Cnr (combinations), Pnr, (permutations), RandomINT (chooses a random integer nearer to zero than the number entered), RandomDEC (produces a pseudorandom decimal), and SEED (for adjusting the RandomDEC chosen). Key 34 The secondary function of key 34 is the Parts menu. This menu allows you to take the INTeger part, the DECimal part, the ABSolute value, and the MOD of a number. Key 35 Key 35’s primary function is the XEQ command. By pressing this button, followed by a memory register a program in the memory register will run. While programming, this is the command which runs a subroutine. The secondary function is the display key. This pops up a menu allowing you to select SCIentific notation (including the number of decimal places), ENGineering (including the number of decimal places), FIX (fixing the number of decimal places), ALL (all digits), and SigFigs (more on this in the new function section). Key 36-37 The secondary function on key 36 is part of the unit conversion. Key 38 The secondary function of this key is the Pseudo-Graphing. More is available on this in the new functions section. Key 39 The secondary function of this key toggles between the programming mode and calculating mode. Key 40 This key is the second most important button on this calculator. This button is the second key, which effectively doubles the number of buttons on the calculator. Key 41 The secondary function of Key 41 is part of the unit conversion functions. Key 42 The secondary function of Key 42 is part of the unit conversion functions. Key 43 The secondary function on Key 43 is the GOTO command. When used in programming, this acts like the goto command in c++. It permanently sends the program to the chosen destination. GOTO variable will take the program to line 1 of the given variable’s program. GOTO . variable number takes the program to the line given by the number of the program given by the variable. In calculation mode, this “loads” a program, so it can be ran by pressing the R/S key. Key 44 The secondary function of Key 44 is the ) which is only used for the equation solver. Key 45 Key 45, is probably the most important, but most taken for granted button. This button turns on and off the calculator, and clears whatever is on the LCD screen. Key 46 The secondary function of this button is a unit conversion button. Key 47 The secondary function of this key is the FDISPlay, which displays numbers in fraction mode. Fractions are inputted like on the 32sii, using the .. as a / sign. Key 48 This is a new feature which, to my knowledge, has never been implemented on any “full time calculator,” only on PDA’s. This calculator features a backlight on the LCD screen, which allows avid users, who have memorized their keyboards, to work in the dark and be able to see the LCD screen for their answer. By turning the backlight off, the battery will be conserved. The secondary function is the catalog button, which displays all the functions that wouldn’t fit on the keyboard. A complete list is available at the end of this document. Key 49 The secondary function on key 49 is the ( which is only used for the equation solver. NEW FUNCTIONS There are several new functions on this calculator, which I have designed, or implemented in a way I have not seen before. The first, and most impressive, is the Pseudo-Graphing function. This function allows the scientific calculator to graph in a way more accurate than a graphing calculator. This function was put in this calculator to increase the market to high school students, most of who depend on their graphing calculator. This function creates a table of value of only important points. This will greatly decrease graphing time and allow students to be weaned from their graphing calculator. Important points are those which are directly adjacent to curves in a graph, or at a vertex, as well as a point ˝ to 1/3 of the way through a curve (which will make sure the line drawn on paper is accurate). Important points also include the x and y-intercept, and any other number which happen to be “nice” whole numbers. The table will display 2 lines at a time, and be scrolled through using the Up and Down “Program Navigation” keys (keys 18 and 23). Another function that is new to me is the BASE conversion, and operation function. Converting bases is a fairly simple program to write, and it is about time it is included on a calculator. This will help computer scientists, and electrical engineers, as well as some advanced math students. One of the most helpful functions on this calculator is the unit conversion. This unit conversion is similar to the unit conversion on the TI-86 graphing calculator, but never has been implemented on a scientific calculator. There are 5 buttons representing different types of units. They are TEMPerature, AREA, VOLume, LENGTH, and MASS. When one of these buttons is pressed, a menu system of units is displayed on the screen. This is to select the current unit that a number is in. The Convert button is then pressed, followed by another unit category (generally the same one). A second unit is then selected like the first. The number on the screen now reflects the new unit. The units for each category follow. Temperature: degrees Celsius degrees Fahrenheit Kelvin Area: Cm^2 M^2 Km^2 In^2 Ft^2 Yd^2 Mi^2 acre Volume: Liters Milliliters Cm^3 M^3 In^3 Ft^3 Yd^3 Gallons Quarts Pints Cups Fluid Ounces Length: Cm M Km In Ft Yd Mi Mass: Milligrams Grams Kilograms Ounces Pounds Catalog Functions Arrows Toggle- turn arrow keys in the center of the keyboard on or off. Left- Checks if left is pressed (works like a conditional test) Right- Checks if right is pressed (works like a conditional test) Down- Checks if down is pressed (works like a conditional test) Up- Checks if up is pressed (works like a conditional test) Power On With- turns this program on as the calculator turns on Power Off- turns off the calculator Matrices- toggles a menu to use matrices (NEED HELP WITH THIS) Sound (0-9) plays a sound (10 choices) Receive Infrared Signal- receives an infrared signal Send Infrared Signal- sends an infrared signal Connect with computer- “docks” with a computer via USB Send USB Signal- sends a USB signal Receive USB Signal- receives a USB signal Connect Infrared- makes a connection via infrared Connect USB- makes a connection via USB DSE ISG Start Timer – starts timer; stores time in TIMER register (on decimal button) Stop Timer- stops timer Reset Timer- resets timer to 0 HR --> HMS HMS-->HR Set Time- Set current time (Stores in TIME register, on 0 button) Hyperbolic- used with existing keys X?0- conditional tests X?Y- conditional tests RPN Mode ALGEBRAIC Mode NEED ANYMORE GOOD FUNCTIONS!!!! Thanks -Ben

Re: Calculator Design (A few questions) Message #3 Posted by Christof on 2 May 2003, 2:01 a.m., in response to message #2 by Ben Salinas I haven't mapped this out yet on paper, but a quick reponse would read: you want keystroke programming, some menus, full alphanumerics. other than that, 'just' (hahaha) munge together the 32SII and 42S and you should be fine. A note on clocks: probably one of the most useful things on a calculator for me is a programmable count down timer, stopwatch, and appointment maker. with a loud enough alarm for me to hear when some other guy is blowing out his beakers with compressed air :p the system you use for units is pretty decent, I have used it before. it cna be helpful to integrate a set of units flags (6 flags should give you enough for common and SI) with some capability to display and convert in the solver. this would make a lot of chemistry work a bit easier for students who could figure out how to use it ;)

Re: one less keypress Message #4 Posted by glynn on 2 May 2003, 2:03 a.m., in response to message #2 by Ben Salinas In Conversions, you state: "...When one of these buttons is pressed, a menu system of units is displayed on the screen. This is to select the current unit that a number is in. The Convert button is then pressed, followed by another unit category (generally the same one). A second unit is then selected like the first. The number on the screen now reflects the new unit" The number in the X register or "focus" has been TOLD already what category of measurement it is. Pressing Convert tells it you want to perform that action on the unit in focus (very RPN-esque). At THIS POINT, the calculator should tell YOU what conversions are valid/possible-- you select from the menu right then. Eliminate that second "category selection" and save a keypress.

Re: Calculator Design (A few questions) Message #5 Posted by glynn on 2 May 2003, 2:46 a.m., in response to message #2 by Ben Salinas Ben-- Just looking at your description and trying to piece it together mentally, I am quite impressed. You are obviously aiming at a premium calculator, one that meets the needs of practically everyone, in a package small enough for someone to keep around. (I am frankly embarrassed to be seen with some of the larger "hip-flask" sized graphing calcs. Hey, at least McCoy had a holster for his Tri-corder.) I fear you want so much in this design, that you are going to scare the bejeesus out of a user trying to learn it. But I cannot argue that it is not well-enough thought out... it is very logical and for those who would study it for a while, it would certainly be a worthy tool for any task you put it to. Um, as far as a backlight-- ick-- but if ya wanna get crazy/novel, run a rail on each side of the keyboard, with little focused lenses incorporated in them. A set of light-pipes for two bright white LEDs inside the calc. The lenses could be aimed at various angles and spreads so that, like stage lighting, they spotlight the legends above/below the keys. There is also that spooky whitish paint now that they use on tv remotes so that when the lights are out, it glows for a while. Hahaha, Alternately: incorporate a mic and voice-recognition. "One-Fiver-Two-Point-Thureee-Thureee-Sevin... ENTER...." Talk about SPOOKY... ;0

Industrial design in the post-Imac era. Message #6 Posted by glynn on 2 May 2003, 6:17 a.m., in response to message #5 by glynn Monsieur Ben; There is always the temptation to be trendy-- and that is not NECESSARILY a sin, except when it gets in the way of the basic functionality of the device, or misdirects the customer base about the nature of the device. Plenty of companies (HP included) have been making mistakes on both counts when their marketing departments seek counsel from focus groups, attempting to uncover what the target audience would most like. Shiny chrome is in, and blue or violet lights, and translucency, and curves that make one think of a swedish kitchen appliance (or an alien marital aid) more than a professional computing tool. A lot of this is the fickle nature of the public taste, and their sometimes odd perceptions and associations. Back years ago, TDK was trying to market their premium cassette formulation against Maxell, and getting roundly whipped in the marketplace. A long market study ensued. A/B tests established that the audio and durability aspects of the products were essentially the same, and in some ways TDK was superior, but consumers were buying lots more of the Maxell at slightly higher prices. The study concluded that, in the U.S. market, audiophile consumers didn't like all the cheap low-grade crummy tapes, and all the cheap low-grade crummy tapes were molded of black plastic and had a black-shell box--- and, uh-oh, so did TDK. Maxell had a grey plastic cassette and a clear-shell box, so to the consumer, it was a higher-quality tape, one that would take your music seriously. *(Ben, you are young, so if you don't remember a time when people taped music onto cassette, don't worry. It was mainly for listening in the chariot.)* TDK did a very dumb thing as a result. They let the marketing department pick the color of their cassette. The marketing department asked people to pick which color cassette and box they liked best. They asked the GENERAL PUBLIC. Bad mistake, because the general public was not worried about picking the cassette with the best audio quality or anything like that-- they were picking what coordinated with their decors, or matched their suits. This was the early eighties. They picked "Midnight Blue"... Needless to say, TDK now had a product crippled by popular aesthetics. Yes, they recognized the mistake after only a year or so, and went slate-grey-- which is what audiophiles had been looking to buy-- and been buying for years from Maxell. Played catch-up ever since. Right now, Ben, ANYTHING that has a blue or purple LED on it is "IN", because, quite frankly, up to a couple of years ago, we had all the red and green and yellow LEDs and NO BLUES. They were too expensive and inefficient to incorporate into products. Just recently, some manufacturers figured out how to make the blues, violets and whites cheaply and brightly. Consumers still think they are "new", and that they are "cool". But in a couple of years, it'll change. Public opinion always does. Similarly, these coke-bottle or yam shaped calculator devices are going to be giggly props on "That 00's Show" whenever they get around to making that. Yes, it fits your hand. So does a baseball bat. Ever punch buttons on your baseball bat? Designers often like to do wacky things with engineered materials because they CAN. If they had tried some something with last year's styrenes or polyethylenes, the product wouldna worked-- "ahh, but NOW, with these Lexans and Urethanes, I can do THIS..." and they show off. But the stuff that WILL BE AROUND year after year doesn't often have a "look at me" quality. They tend to have an "I know what you are" quality. This is not to say they are not stylish or distinguished... just that they are rarely flashy in any way. A great watch, for instance, never looks like you are supposed to control Sonic the Hedgehog with it. It will enhance the look of the wearer, without anyone yelling "Hey, check out that watch!!!" from across the room. The statement a LASTING watch often makes is one that implies tasteful elegance and discrimination, or care and precision, or some aspect of the wearer (his masculinity, sensitivity, all-business attitude, etc.). Because the job of personal possessions MOMENTARILY may be to make a "statement" or "express individuality", but LASTING personal possessions tire neither the user nor annoy his friends. So, just as analogy, you may have a hopped-up Camry with TRD decals and neon underneath and it suits you fine NOW-- and gets the girl-- but fifteen years out you'll look at the pictures your friends took and realize, gee, that was a fun car I wouldn't be caught Dead driving anymore (you may even say the same about the girl); the Lexus built on exactly the same chassis and rather understated looks (BTW, and it has a truly kickin' engine!) won't get your friends to look twice, but fifteen years out, you'll still be polishing it on weekends. It won't wear on you, you'll wear IT, and look good the whole time. What should a calc say? Well, is it a "Student's" model, or a "Professional Tool" or a "Handy Friend"? You determine, as a designer, the personality the product exudes; but it MUST MATCH the nature of the device, and it MUST MATE with the perceptions of the target user-- else it will fail. When looking at the nature of the device, some things just naturally fall behind, and others come forward. A 2-line calc is NOT going to be a game machine, for instance. And that fact alone GREATLY reduces the odds that it will *ever* be manipulated in the dark!!! So worry about glare, contrast, and such, so that it is useful indoors or out, but do NOT try to make it glow like Chernobyl. The perceptions of the user are trickier to guess, but if you have a CLEAR picture in your mind of who the user should be, it will be easier to draw your conclusions. Black has always been a so-called "safe" color (notwithstanding cassette tapes) because there is nothing really to object to in it. It can't be accused of looking dirty or of getting dirty too quickly; it doesn't look either masculine nor feminine; it doesn't look "kiddy-toy" nor "high-falutin"; and so on. Problem is, it doesn't communicate much of ANYTHING to the user or intended user. Note the 12c (best selling model HP has ever done). In its current incarnation, it has a gold toned bezel and dark chocolate looks. Like an expensive box of Neiman-Marcus truffles. In fact, it just looks like it is RICH and VALUABLE. What else would you sell a financial type? So; is your user a Physicist, Doctor, Architect, Building Contractor, a Biologist, Oceanographer, an Anthropologist? Get the one or two in mind and stick with him/her, study what THAT PERSON values. The aesthetics will extend from what he/she already is comfortable trusting for qualities you are trying to impart in YOUR design. If I chose a Physicist, say, I'd look at a Titanium finish (that slightly dull darkish silver with the grainy feel). Think of the focus ring on a damn nice camera. He'd identify with that metal, certain qualities. I'd choose to do my legends and such as engravure with enamel fill-- shiny black, shiny white, shiny blue, shiny red, shiny green. The look of them against the Titanium would not be the Ultimate in contrast, no, but would impart a look not unlike the tick-marks on graduated calipers or on a vernier dial-- not bad associations. I WOULD do a blue-tint on the LCD CHARACTERS if manufacturing processes allowed it. If I were backlighting, flourescent white. And so on. I'd want the Physicist to immediately feel like this was one of the highest quality precision instruments ever delivered to the Lab. (That was an example, Ben, not a specific recommendation of a design scheme.) And now I'll say a word about Imac. For the audience Apple hoped for, it was a success. And then EVERYBODY started copying elements of its design for the "cool" factor. But Apple had made Imac deliberately plasticky and fruit-flavored and Volkswagen-Beetly for the VERY GOOD REASON that they wanted a product that the Common Man and Woman would immediately buddy up to. It was unintimidating, it was unbreakable-looking, it was not going to require a degree to run it and you knew this because it had the aesthetic of a kid's learning toy. And, believe it or not, it took incredible engineering to pull off all that. Notice how the imitators said, Hey! Translucent is IN! Bright ("Bondi") Blue is Cool! GrapeJuice and Tangerine and Lime are HOT! And so every PC maker proceeded to create mice and keyboards and cases that looked cheezy and like you got it out of a gumball machine. Didn't last long, did it? That's because they were using Apple's words strung into nonsense sentences-- misusing the elements of communication by not trying to communicate (evoke) anything in particular except ape-ing of the language. And so the fad died. And Apple could not use it anymore, since it had been sullied by the misusers. So they moved on, to the embedded silver skin... All design is communication. Some designers say "Form follows from Function", and that is right, partly. But it also must *communicate* its function, and its qualities that aren't as easily guessed, and it's creators vision of its personality and place in the life of its user. Know you have your design mostly figured out, but just thought I'd respond to those of your friends with the Nextel/Sprint fixations. ;-)

Re: Industrial design in the post-Imac era. Message #7 Posted by Ellis Easley on 2 May 2003, 11:31 a.m., in response to message #6 by glynn The latest 12C's are made of black plastic. I loved the rich chocolate brown of the Voyagers and the 71B, 75, HPIL peripherals, 28, some of the Pioneers, 82240 printer, etc. From what I have seen, some of the low-end pioneers started out in black plastic. Is it just a matter of the cost of a standard black plastic vs. a custom color? The one HP color I don't understand is the somewhat sickly looking (to me) greenish brown of the 65 and 67 (but they are still great calculators!)

Re: Industrial design in the post-Imac era. Message #8 Posted by glynn on 2 May 2003, 12:16 p.m., in response to message #7 by Ellis Easley Does have to do with cost-- for two reasons. One, if you have any "color" but black-- even whites and such-- you must provide all your subcontractors with samples and pointers to specific materials/recipes to get there, and then inspect from time to time to make sure you still have color match. If you don't, maybe the battery door will not exactly blend in with the back. But with black, you can get a lot less picky. Our eyes do not differentiate blacks apart very well. Two, those recipes mentioned above are done with ratios of specific dyes in "VIRGIN" plastic. If you can used "recycled" plastic (sprue and crunched-up pieces of stuff that came out mal-formed during the last molding run), you can reduce material costs a bit. The two cases which allow this are: A> if you don't care what color comes out (like an internal part); B> if it is black-- because remelted plastic (and the whole) can be made black easily with no fuss. Hmmm, apparently someone at HP has decided that either the subcontractor will mess up less by using black, or that the audience does not care enough about the chocolate look *(the gold accents being enough to impress)*. And the subcontractor no doubt wants to use remelt as a proportion of the plastic. Once again, black connotes: nothing. (One exception: black anodization on aluminum-- public connotes blackened metal with images of high technical sophistication. It must LOOK metallic though; extruded or brushed). Sorry to hear the 12c is not as "warm" as it once was. :-/

Re: Industrial design in the post-Imac era. Message #9 Posted by Ellis Easley on 3 May 2003, 5:14 a.m., in response to message #8 by glynn "(One exception: black anodization on aluminum-- public connotes blackened metal with images of high technical sophistication. It must LOOK metallic though; extruded or brushed)." - Have you ever seen a Western Digital Microengine? It came in a massive chassis with an extruded aluminum front panel with black anodization. I can't get to the machine right now but I could scan the picture from the cover of the manual. It's about the size of a 19" rack mount box, about 6" or 8" high. The front panel has broad and narrow extruded horizontal stripes. The recessed stripes are black anodized, the front surface is brushed - 11 transitions altogether, dominated by one wide black and one wide brushed stripe - and the name in stylized letters is on the bottom right corner in black. There are no controls or openings on the front. If only I could find an OS diskette for it!

Western Digital Microengine Message #10 Posted by Ellis Easley on 3 May 2003, 5:16 a.m., in response to message #9 by Ellis Easley I should have mentioned: ca. 1979

Re: Industrial design in the post-Imac era. Message #11 Posted by Bill Platt on 2 May 2003, 12:36 p.m., in response to message #6 by glynn <snip> 'All design is communication. Some designers say "Form follows from Function", and that is right, partly. But it also must *communicate* its function, and its qualities that aren't as easily guessed, and it's creators vision of its personality and place in the life of its user.' Very good point, and the whole post is a nicely conceived discussion of the problem of designing. It may even deserve to be posted somewhere else, as well, (is there an alt.design newsgroup??) And I really appreciate Ben salinas work on his new design. That is a comprehensive, and impressive document. Much more impressive than most anything I conceived of while in high school.. -Bill

Re: Industrial design in the post-Imac era. Message #12 Posted by Michael F. Coyle on 3 May 2003, 10:32 p.m., in response to message #6 by glynn Hi Glynn! Great essay, as usual. I'm saving it in case my company tries to do something stupid. :) (BTW, what do you think red LEDs communicate these days? Strictly retro, or maybe some positive qualities too?) - Michael

Oooh, red LEDs. (Mmmm, Forbidden Donut), Message #13 Posted by glynn on 4 May 2003, 1:51 a.m., in response to message #12 by Michael F. Coyle I think red LEDs are cool, myself... but I am a retro kind of guy. Yes, red LEDs DO impart a retro feel to anything they are incorporated into, but that is not necessarily negative. I know I had quite warm fuzzy remembrances of a TI LED watch I had in high-school. This led me to buy a Fossil LED watch, only to be rather disappointed in it. Unlike the TI, which, at the touch of its button, the time was displayed for 5 seconds-- the Fossil stupidly made you HOLD the button for as long as you needed to look at the time. Fossil, ya didn't do your homework... I don't think anyone has ever REALLY asked some of the serious questions that go along these lines, though-- if consumers could choose a calculator with long battery life and a reflective display, or the same calc with a somewhat hungrier emissive display, what would they choose? I know that Luiz V. here in the Forum has stated that he had that sort of choice in mind when he went with the more battery-frugal 41c over the 34c. He even paid a premium to do so. You know for me, when I first saw the 41c at my college's bookstore, the display was the deal-breaker. I could not imagine staring at that gray panel for my numbers. To pay THAT much for a calculator I could not read as well-- no way!! For that reason, I have been interested mainly in the older LED model calcs, and in non-HP models I have a couple of Canons and an Elektronika that are all vacuum-fluorescent. I am pretty sure that if a new calc came out that had a really nice display, it would sell. What technology? I don't know. Nowadays VF is way too hungry, and standard LED seems to be that way; I have been trying to find out if there was any gain in efficiency with the organic LEDs, but my preliminary look in that direction gave me the impression that efficiency was not its best feature. Which leaves a backlit LCD matrix as one option; the thought that maybe a "reversed" filter system would display white characters on a black background, well, I'd like to see it. And there is the question of ALPHA. Most "retro" LED calc lovers take pains to point out that a calc doesn't need it. No, it doesn't-- but if it HAD it, it could do what LCD-based alpha-capable calcs DO: have more advanced file-storage capabilities, descriptive names and the ability to share files with other devices. HP (and a few others) made advanced little dot-matrix displays that would have done this job quite well-- in fact, am I mistaken in thinking they were used on 9825 and similar? But they were a rather crude 5 x 7 cell, so nowadays a better LED display could/should be made with resolutions similar to that in the LCD world-- one up to the challenge of alpha, including attributes such as bolding and italic and underlines. As for red-- no, if I were designing myself one from scratch, I have to say I'd go AMBER on black. It's WAY easier to live with on a day-to-day basis. If I wanted to get the pilots on my side, though, I'd go green, as they find it easiest to focus upon without having to adjust from their other instrumentation and surroundings. Red was an LED color because they could make it very bright and efficient in that band of wavelengths. But my long stay on a Lear-Siegler ADM-11a has convinced me of the superior qualities of amber. Ya know, I was ALWAYS this picky. It's just less Damning if I let people assume I'm getting crankier with age. ;-)

Re: Oooh, red LEDs. (Mmmm, Forbidden Donut), Message #14 Posted by Ellis Easley on 4 May 2003, 3:04 a.m., in response to message #13 by glynn I bought a box of HP amber LEDs on Ebay a while back. Apparently, the model I got was already obsolete, I couldn't find the exact part number on the HP (or was it Agilent?) website, but I'm quite sure they are from a thriving family of devices made for use in outdoor highway signs. They are incredibly bright with normal current levels, like 20 mA (they are in the ordinary 5mm round package). I don't know if this scales sown - if smaller dice assembled into 7 segment or 5X7 displays would be very efficient, but it is a possibility

Re: Oooh, red LEDs. (Mmmm, Forbidden Donut), Message #15 Posted by David Smith on 4 May 2003, 5:46 p.m., in response to message #13 by glynn I bought my first HP41 specifically because of the LCD display... you can read it in sunlight. And it ain't too shabby in room light either. I also still use a Toshiba T1000 laptop quite a bit... the brighter the sun, the better the display.

Re: Oooh, red LEDs. (Mmmm, Forbidden Donut), Message #16 Posted by Michael F. Coyle on 4 May 2003, 10:30 p.m., in response to message #13 by glynn Hi Glynn! I don't know about the 9825 specifically, but alphanumeric displays used to be provided by "dumb" 5x7 dot matrix displays with all memory, refresh, timing, etc. done externally. Displays like that are still sold. Nowadays you can also get "intelligent" displays with 4-8 characters and controller all in one package. Even the newer displays are limited to 5x7 or 5x8, though; I'm not sure if this is due to technological limitations or no demand for higher resolution. As for color and efficiency, new red materials such as GaAlAs are very good -- same brightness as so-called "high-efficiency" red but with about 1/4 the current. (They can also be driven with more current to produce a sunlight-readable display, or so the data sheets claim.) Orange, yellow and green are available but not with super high efficiency. If you want to see what's currently available check out the Agilent Web site. The HCMS-29xx is a good place to start. Prices (qty. 1) are about $25 for 4 characters, $50 for 8; that is, about $6 per character. Cheaper in quantity of course. I'm considering picking up one or two of these to play with, to see what's really possible. Modern rechargable batteries should give a nice long time between charges. I'm also wondering if it would be possible to use one of those high-capacity lithium photo batteries (e.g. CR-123) as a long-life disposable. After all, if it can run camera motors and such for dozens of rolls of film, it should have no trouble with LEDs. - Michael

hp led source Message #17 Posted by db(martinez,california) on 5 May 2003, 3:06 a.m., in response to message #16 by Michael F. Coyle fwiw: there are cheap surplus hp leds here (just not OUR leds): http://www.bgmicro.com/prodinfo.asp?sid=855092592592593E-021669044200&prodid=ACS1076&page=1&cri=led+&stype=3

Re: Calculator Design (A few questions) Message #18 Posted by hugh on 2 May 2003, 11:31 a.m., in response to message #2 by Ben Salinas random comments that occurred to me, if you do want backlighting, shouldnt you light up the keys too (like a phone)? i have yet to see conversions work really well with a comprehensive list. i think units on the 28c/48g is overcomplicated. also, are you to cope with different versions of, for example, the inch (european inches are 2.54cm, US inches are 2.54+2e-6 cm, ie + 2ppm and correspondingly all other lengths). base 37! is there a good reason to stop there. alternatively, i dont think it would be a cheat if you supported only the usual 2,8,16,10 etc. from your description it looks like you only "goto" a number not a label? your new graph feature; what you describe is difficult technically. to find the truly important parts of a graph, extrema, asymptotes, inflexions etc. involves symbolic work (claim). otherwise you are always at the mercy of sampling aliasing.

Re: Calculator Design (A few questions) Message #19 Posted by Christof on 2 May 2003, 1:27 p.m., in response to message #18 by hugh Teh best use of backlighting for me is in 'odd' light conditions. Cafes, dusk/twilight, situations with high aount sof artificial light pointed in unpleasant directions. Backlighting is- in my real world use on devices- more for time when high contrast just isn't cutting it than it is for absolute darkness. So I say, go with the backlight, but I also say, keep the current mild blue (not the bright obnoxious blue) as it is a bit easier to read than the greens. I'd also lay out the keyboard in such a way that a certain set of buttons (up down enter? six function keys at top?) are bigger, higher (1.2 mm makes a huge difference when operating by touch) or something. that would make properly designed programs and menus easy to operate in lower light conditions. I'm also- I have to say it again- very very big on programmability. I find nothing wrong with the programming capacities of the 28/48/49 machines- the UI is a bit harsh, though. a logical extension of the 41/42 command set seems reasonable, but... well, to reduce the command set just doesn't. Wehn you get right down to it- the older TI85 I just pulle dout of a drawer is smaller and has better weigth distribution than the 48GXIt also has a farily well layed out keyboard (aside from crappy keys). The reaosn I don't use it is mostly the programming interface. A good, intuitive, learnable, USABLE programming language means a lot. And, not to quote Gene too much- games programming is going to be important. Everyone can look down their noses at it, but it is important. to learning programming, to becoming familiar and comfortable with the machine, and to the human psyche. Laugh at me all you want, If you don't consider game programming in the calc, you might as well use a 7seg display. Graphing is great, but graphing on a 2 line display (which I DO on my 42S)is just not quite enough by itself to justify not using a really capable machine that's a few ounces more. In fact.... I'd suggest further that you go full tilt and duplicate the pixel count of the 28S. it's a 4 line display type that will fit in your machine's physical limits. RPN is a lot more fun with XYZT labels on the lines. and 3 lines visible when using menus is a lot nicer, too. And this isn't just for games. Technically- I'd prefer the 4 lines in a display. Ask db down in martinez about how much easier surverying operations woudl be with the 2 extra lines. What about 2 var statistics? or coordinate pairs? if you have a menu running an donly one line left, you can't just look at the display and SEE the X coord you entered when the prompt is up for a Y coord Forget all about Z. One other minor note- any handheld with programming capabilities- or that cna run programs- must needs be fully and (as much as possible) easily programmable from within the device. I could spend hours discussing the reasons- and a lot of poeple will never use the programming features. *shrug* I'm never going to use hyberbolics in my real life, but I'm not arguing haivng them there. :) Programming is important. On many levels. Anyone who disagrees is welcome to trade me a 41CX for my 6G anytime ;)

Re: Calculator Design (A few questions) - Long Message #20 Posted by JimC on 2 May 2003, 1:17 p.m., in response to message #1 by Ben Salinas I have but a couple of comments on the "form follows function" vein. I really dislike the flat surface of the newer HP's (32 and 48's, etc). The larger volume region of my 41 with the slanted display allows for placing the calc. anywhere on my desk and still allows a view of the display. As laid out so very well in a previous post, colour plays a large role in product acceptance. Engineers, scientists, etc., are very conservative, so therefore I tend to think that a more understated approach might translate into more sales. (but then I'm in engineering, not marketting). Translation: I still like black. For me, RPN rather than RPL works. I can write a routine in the field, rather than sit down and program. I find it is usually a function of keystroke memorization rather than actual input/output window programming. After all, the idea is to save time-my 41 allows me to do that, my 48 doesn't; however I do understand that some persons like and want to use RPL. Therefore, might there be a way to record keystrokes so that non-programmers like me can still utilize simplistic programming techniques? Interconnectivity with PC's for backups. A flag to use AOS vs RPN Not as many menus! My 48 drives me crazy when I am trying to use a unit conversion. ("I'm sure I could assign that to a USER key") The ability to add cards/memory on, peripherals, etc. The same quality tactile feedback as my 41, not my 48. (bigger buttons for a clumsy old bugger like me) Last but not least - good, conclusive documentation. Like my 21, not like my 48.

Re: Calculator Design (A few questions) - Long Message #21 Posted by Joan on 2 May 2003, 1:43 p.m., in response to message #20 by JimC Well, I also have a couple of comments, Today I just receive a new brand 41cx and: 1. Keyboard feeling is good but softer than the 42s and 48sx. 2. Display in the 42s is bad but it is much better than the 41cx. 3. I have to say that I'm happy with the 41 but for me the 65 gives me better sensations. 4. After all, I know that nothing is going to be like the hp machines that we all know. 5. My ideal machine is a 42s size calc with the 49g capabilitys and the two times the 42's screen and exactly the same tactile feeling. 6. Fortunately I meet the hp's on time, because I'm not expecting anymore. Joan

Re: Calculator Design (A few questions) - Long Message #22 Posted by Ben Salinas on 3 May 2003, 12:46 a.m., in response to message #21 by Joan Thank you all very much for your input (and for reading the 6.5 page document:)) I agree that blue is very trendy, and it is true that 5 years ago backlights were not blue (mainly because blue LED's hadn't been invented, and when they were, cost 50 bucks a piece (maybe)) However, i feel that green is too BLAH (for the lack of a better word), and red is hard on the eyes. This narrows it down to either light blue, or white. As for colors, after reading these opinions, i realize that the 32sii does not look 10 years old, because it is black. That would probably be the best color (as the titanium is probably too faddy) I do like the idea about using glow in the dark paint for the keys though. That would allow low light use. The main reason I put the RPN/Algabraic function in the catalog, instead of in the flags is because the people who would take it out of RPN mode (Which it defaults to) are not the people who are avid HP users. It took me a few months to figure out what the flags did. While on the subject of flags, would you all suggest that most features also have a flag, so that settings can be changed, and checked in a program. And, are flags generally used for functions that are only on or off (as in RPN/Algabraic)? To the catalog, I have added some common symbols (!, "," , ?), and the highest denominator (the /c on the 32sii), the %chg (tenative), and set date (to be inputted as .mmddyyyy). As soon as I can find a viewer for Inventor (Autodesk's new Autocad), I will post my inventor files online. Everything will be more clear when you can see the calculator body itself. Once again I want to thank all of you for your input and praise. I hope to complete most of my design by next thursday. (due on the 15th of May) -Ben

Flags Message #23 Posted by Christof on 3 May 2003, 5:50 a.m., in response to message #22 by Ben Salinas Flags are often used either singly or in tuples. I like the idea of having an abundance of flags- with negative integer flag numbers being for 'OS' use and positive flags being for "user" application. I see no reason to hard code a limit in, but 256 total is a good number- once you go beyond that you are probably addressing a total of 65,536 (16 bits of addressing) but... well, why not? Memory constraints aren't what they used to be and I can think of reasons a computer scientist would use more than 256 flags. SIngle flags are things like pos/neg, printer presence, etc. Tuple flags are binary counting mechanisms. a two flag tuple would have 4 positions, a 3 flag tuple would have 8, and so on. Uses are varied- you *could* in theory (discounting menus) have 3 font settings on the display of a 42S - that would require a 2bit flag tuple. An IR I/O serial device would be handled the same way- generally you would have several bits on the flag tuple to help define various serial speeds- 300,600,1200,2400,4800,9600,etc. On another note- this brings to my mind som problem issues with USB. 1: an area of concern with some people is unneeded openings on a device. IR requires no 'de-hermeticizing' or extra opening in the device, where a USB port does. and IR to usb converter is not hards to oobtain, nor supply. this also applies to serial in (in potential) HPIL- or any other serial two 'wire' system (including modems!). 2: power. USB, if used to connect to something like a 'pen drive', has to *supply* power. Incredibly large amounts of it in calculator terms. To power a USB port is going to do truly nasty things to your battery life. 3: versatiliy- as mentioned above, it is easy to get IR to usb and serial converters (and an easy enough homebrew for any other two wire system). USB locks you out of whole ranges of controllable devices- from telescopes to data recorders to scopes to field hardware. A lot fo this stuff simply requires serial. I can, with small to huge amounts of programming, connect an HP48 to the control serial console port of a Sun E15k supercomputer, or a meade telescope, or a small lab data logger, or even a modem to talk to my HP75 (ongoing project) I can't do that with a usb only device. remember, a USB to serial converter is going to try to draw power from the *USB* side, and an IR convertor HAS to. (and lets not even get into the broken by design silliness of bluetooth. bluetooth would have been a good idea 15 years ago. right now, it's a giant step backwards.) Long, rambly thoughts over :)

Re: Calculator Design (A few questions) - Long Message #24 Posted by Gunnar Degnbol on 3 May 2003, 10:28 a.m., in response to message #22 by Ben Salinas The .mmddyyyy date format is a US-ism. It is a midlle endian mess, and is quite weird if you are not used to it. In most of Europe the order is ddmmyyyy, in Japan and many other places (and in the ISO standard) it is yyyymmdd. The latter is more logical (a strange number system for counting days), and would be understood by most people, including americans. If this format is used it might makes more sense putting the digits before the decimal.

Re: Calculator Design (A few questions) - Long Message #25 Posted by Christof on 3 May 2003, 2:06 p.m., in response to message #24 by Gunnar Degnbol While I se no reason not to allow different encoding systems for dates as a moality- I prefer the ISO standard. Might be the military experience :) I'd suggest yyyy.mmdd as it allows working with known dates prior to the common start date of the Common era, and allows you to work with rough dates a bit more handily through rounding. -C

Re: Calculator Design (A few questions) - Long Message #26 Posted by Gordon Dyer on 3 May 2003, 1:14 p.m., in response to message #22 by Ben Salinas Make sure you include a ddmmyyyy for us Europeans !

Re: Calculator Design (A few questions) - Long Message #27 Posted by unspellable on 5 May 2003, 12:32 p.m., in response to message #22 by Ben Salinas I'm a newbie here. Can anyone tell me why the 32SII seems to be a hot item on eBay comnpared to other models?

Re: Calculator Design (A few questions) Message #28 Posted by Ben Salinas on 3 May 2003, 1:14 p.m., in response to message #1 by Ben Salinas I have found the Inventor viewer, and have created some basic views of my calculator. This viewer only views in 2s, so bear with it. To view the files, you must download the viewer at http://usa.autodesk.com/adsk/servlet/index?siteID=123112&id=2952277&linkID=3016127 and the files at www.geocities.com/piguy31415926/calculator.zip (Disregard the Letters on the keys. Eventually those will be the function names) On USB, I had contemplated the power issue (I believe it supplies 4.5 V, which is quite a bit) Perhaps an unpowered USB port. The main reason I like USB is because it can be divided up, and is bidirectional, and has a small footprint. (Can IR be divided up?) On the calculator, there is also an IR port, for wireless communication. USB also provides higher speeds, though I dont know if that will make a difference. On the subject of dates, YYYYMMDD probably would be the best choice. This way the input could be YYYY.MMDD (Unless the year is like 0001). Thanks again -Ben

Re: Calculator Design (A few questions) Message #29 Posted by Gordon Dyer on 3 May 2003, 1:27 p.m., in response to message #28 by Ben Salinas I just had a look at your design and it is nice. However please round off all corners to make the calc comfortable to hold. I bought a 28S and the sharp top edges when it is open make it very uncomfortable. I also think that yyyy.mmdd is a good format for us all, but if you are displaying dates a set of the usual options would be good. Also if you have used a 28C or S then there are someee things to learn NOT to do! the ergomomics of the key layout were not good, but if you stick to the 42S type of key layout thats fine. I am looking forward to a full colour drawing with key legends and an emulator for Windows.......

screen grab? Message #30 Posted by Christof on 3 May 2003, 2:14 p.m., in response to message #28 by Ben Salinas Okay, I admit it. Not only do I not have Windows, I don't even have any intel x86 CPUs aside from the one in my 200LX in the house. This software doesn't seem to have binaries for sparc or ultrasparc architectures, so- cna anyone do a screen grab and save to jpeg or tiff or png, etc? thanks! (and I'll table USB, but it seems to me a bad idea to run around making non standard ports for a standard)

[ Return to Index | Top of Index ]

Go back to the main exhibit hall