> The HP-35 had numerical algorithms that exceeded the precision of most mainframe computers at the time. [...] This forced time-consuming manual comparisons of results to mathematical tables. A few bugs got through this process. For example: 2.02 ln e^x resulted in 2 rather than 2.02. When the bug was discovered, HP had already sold 25,000 units which was a huge volume for the company. In a meeting, Dave Packard asked what they were going to do about the units already in the field and someone in the crowd said "Don't tell?" At this Packard's pencil snapped and he said: "Who said that? We're going to tell everyone and offer them, a replacement. It would be better to never make a dime of profit than to have a product out there with a problem". It turns out that less than a quarter of the units were returned. Most people preferred to keep their buggy calculator and the notice from HP offering the replacement.
That letter should be required reading and the anecdote further cements the 'fish rots from the head' proverb as accurate. After all, without ethics at the top there is no corrective action against such relatively minor nudges in the wrong direction and over time that will accumulate until you get a completely different version of the company. One that will never ever have the kind of stature that HP enjoyed.
Oh, how things have changed. In fact reading the whole article made me wish for a return to a time where people took this much pride in what they made.
Unfortunately, consumer electronics are not the same as the days of Braun. 70’s was the golden era.
I was extremely pleased with my DM42 from SwissMicros. All of their models either have a Titanium or Steel case. The heft and quality is just absolutely beautiful.
I spoke to Michael Steinmann and he seems to be a great guy passionate about what he does. I enjoyed our conversation about old school electronics.
Talking about Braun, I've been disassembling all kind of devices for a few monthes, and you can appreciate differences betweens brands a bit. Braun and a few brands, do have better materials, better internals. I guess 90% of consumers can't give a damn about that but it registered in my head. To the point that I would even start a website about that, some sort of ifixit for everything but phones.
Please...please start a blog, instagram, Youtube, whatever and post your findings. I've been meaning to purchase old Braun and B&O devices to learn how they're made but the expense is putting me off.
This would be so amazing if you could document your tear downs.
Well, my devices were those people leave on the curb monthly so finding Braun things there is possible (although not regular), B&O on the other hand .. you gotta be reaaaallly pressed to discard anything from them outside.
INTERESTING, thanks for sharing. Please, how is the keyboard? To me, more than anything, what made the HP calc fantastic was the great keyboard - which really matters when you start using them in anger.
My HP-15C survived all but the last of many drops. The HP-11C that replaced it still powers on on its original batteries!
Back in the '80s I had a dog named Ben. A lab-afghan-shepherd mix, truly a great dog. Whenever he took a nap, my cat Petunia would climb on top of him and take a nap with him. They were best friends.
Ben had only one bad habit. He liked to find my wallet or checkbook and bury it in the back yard!
One day he found my HP-16C in its "leatherette" case. Yes, it was just like a checkbook or wallet. So he buried it for me. Thanks, Ben.
And then he dug it back up and when I found it, it had major tooth marks in it!
I was pretty upset. This was not a cheap calculator. So I swatted him on the nose with it. [1] And then I was so angry I threw the calculator across the street!
Well, I felt pretty stupid after that. So I went across the street, picked up the calculator, and it still worked! And it still works to this day. (Lucky for me, Ben's tooth marks just missed the display.)
[1] Animal rights lovers, please do not get too upset. Ben and I had our disagreements from time to time, but we were best of friends, just like all the dogs and cats I've had the privilege of sharing a home with.
On a related note, in high school, my dog got hold of my nearly new HP-48GX; I wasn't so fortunate as you, as she bit right through the display, shattering it.
Figuring dog bites weren't covered under warranty, I sent the calculator — which had obvious "puncture wounds" through the display — off to HP Corvallis, with a note explaining the cause of the damage, and requesting a repair estimate for the damage.
A week or so later, I received a response — in the form of a brand new 48GX in retail packaging, provided at no charge.
Which, as expected, still works today (though these days, I rarely use any of my [embarrassingly many, including two revisions of HP-35] HP calculators other than my pair of 15Cs; though I do also have a 16C at my desk which I use on a semi-regular basis).
Casio calculator watch—the black plastic kind—went all the way through a black Lab named Brando. Came out the other end of the dog minus half of its wristband, with teeth marks in the keyboard and a cracked display but it still worked. Cleaned it thoroughly, replaced the band and wore that watch for many more years.
Greetings, fellow 32Sii lover. I also have enough spares to outlast me, and it's a good feeling. That said, perhaps I'm being overcautious as my first 32Sii (bought in the early 1990s) is still in great shape.
I've been eyeing one for a long time. I use an original HP-42S I stole from my father that's older than I am, but I almost spilled coffee on it the other day, which made me realize it's probably time to invest in the SwissMicro just based on the value of the 42.
Looks like the DM42 also fixes the only complaint I ever had about the 42S, namely, the display contrast. While I've had a 42S for many years, and have spent many pleasant hours fiddling with it, it never replaced my pair of 15Cs as a daily driver on account of the display.
I've had good experiences with their EliteBook lines. Solid machines, if you are closer to the rugged notebook vs thin trophy book on the scale of laptop thickness.
OTOH, if you are engineer faced with the problem of picking some number between the respective economical constraints (ie. pick a number not much higher than 2.25 or not much higher than 3, both of which are somewhat well known estimates of manufacturing overhead) you are somewhat likely to come up with these two transcendentals.
At most hardware companies, this is the starting point. As in, a flat multiple is the baseline, before S&M and other overheard costs are re-integrated by Finance into product cost.
Indeed. Same thing when I was a hardware engineer. However, I think the GP loves it due to the use of Pi and 'e' as the constants. We used 3 I seem to remember, but we included some amount for assembly and test, which I bet would come close to Pi.
It really is hard to over state the impact that scientific calculators had on industry. My great uncle who was an engineer at the Panama Canal company had log tables, these were tables of logarithmic values given a number, and the inverse. He needed the heavy tome because he couldn't get enough digits of precision out of his slide rule. He was one of those people who immediately bought an HP-35 when it came out and was thrilled.
My parents bought me the TI SR-52 when it came out and it was the coolest thing I could imagine. I wore the numbers off the keycaps on that thing.
I have my dad's slide rule, and his HP-35. He's a retired industrial research scientist.
At my dad's workplace, every scientist had a Friden 55. The managers compared the price of the HP to the annual service contract on the Friden, and bought an HP to try out. They were going to let each scientist have it for a week, then pass it on to the next person. Instead, after just a couple weeks, the Fridens were all in the dumpster.
It's fully mechanical and can do addition, subtraction, multiplication, and division. I saw one of these things work, and it was quite an experience. There's a YouTube here:
And a few people have published images of the service manual.
My dad said that while the Friden was accurate, it was not portable or quick, and he usually used his slide rule, notebook, and graph paper. For a chemist, slide rule accuracy was usually sufficient.
As someone who switched away from RPN (well RPL) I disagree. I was an HP48/50 series fanboy since about 1995 and before that RPN on a HP41. I was the weird kid at school who didn’t have a normal calculator and the teachers couldn’t help me.
However I bought a Casio FX991EX recently after using my daughter’s unit which has a modern entry system, solver, engineering units, conversions and dual power. It’s pretty much perfect. Anything else more complicated, I roll out python, scipy, sympy etc.
Yea, it's much easier to enter a matrix on a modern Casio than the HP calcs and requires fewer keystrokes I believe. I guess navigation of the indices takes up time on the Casio, but it is so intuitive.
I highly recommend anyone to convert to RPN but it does make you feel like a idiot when you can't use a normal calculator for the minute it takes to switch back over to normal. I find it doubly frustrating though to use a normal one because I know now the forbidden fruit of how calculators can be improved and college sometimes prevents me from using it.
I bought an HP-28C in when I graduated from high school using my own money. It was $235 which was a lot of money for me. My parents thought I was crazy. I later upgraded to the 28S followed by the 48sx.
At college, I would enter and win the "Calculator Olympics". They would give you these super complicated algebraic expressions to evaluate. The TI people would lose track of all the parentheses and the older HPs would exceed the 4-level stack.
I was amused a couple years ago to discover many of my old programs are still preserved on the internet: https://www.hpcalc.org/authors/166
POLY was hugely popular at the time. I just wrote it to simplify my classwork.
My father worked at HP in the 70s in Tool&Die, and made some of the molds for plastic pieces of the HP 35. He bought an HP 21 in '76. It's now my son's.
It still works perfectly, except the battery is lost so we run a wire into the battery compartment.
It's the most intuitive calculator I've ever used.
While I am definitely fond of the HP-35 in a nostalgic sense, my real-life working calculator was the HP-48GX when I became a land surveyor. I added a COGO card that would calculate pavement design, storm water pipe modelling, and the area of a polygon after the coordinates of each point were given. And away I went. There wasn't much I couldn't do with that, graphing or calc-wise. Made my job so much easier in the field.
When they killed the HP-48GX, and replaced it with inferior models, it was a dark day in land surveying.
The HP-48G/GX was also my calculator of choice and I really disliked the later HP49. However, I then tried the HP50G, and was instantly hooked. Once you get over the differences in key layout, it's like a HP-48GX loaded with tons of useful software, in a slightly better form factor. I now have a HP50G and a HP-48G, but I never use the older model.
On iOS, I use the Emu50G app, which works, but could really use some improvements and optimizations. It's a fork of emu48.
Incidentally, I think the HP-50G is the last good engineering calculator. I'm puzzled as to how people get around without one these days: don't tell me about python, scipy and other similar shells: there is nothing else out there that can do so much, so efficiently.
Have you tried Calc? It's loosely modelled after the HP-calculators and it's my first choice of calculator when I have a full blown computer at my hands.
I bought a HP-42S in West Germany in 1989, my first trip to the West when I was 14 (I was Hungarian at the time, now that's fixed, I am Canadian). Oh the memories.
Later, 1995, I got a HP 48G via Usenet (eBay opened the month when it arrived from the USA so it's not like I could've bought it there). A friend of mine have soldered four 128K SRAMs in there, stacked, with most legs just ran together vertically with the few data legs separated. Dave Arnett himself told us it's possible to change a G into a GX... https://groups.google.com/d/msg/comp.sys.hp48/CS0lTpKkBHw/vn...
I still have it although I do not use it, it's one of the very few things I carried over when I immigrated.
Ps. Isn't the Internet fantastic? The easy communication with people you'd never be able to connect with otherwise and the astonishing archives allowing me to dig this 23 year old conversation up with ease.
I've a SKB case filled with HP calculators. It is your android FREE 42s I pack in the kit. Other than 3-D just-so HP's haptic double-injected keys, it's fantastic! Thank you Thomas for your phenomenal RPN gift to humankind!
A neighbor gave us one when he moved up to an HP-25 (lower number, but a later and better calculator). I used it in college. I remember running across campus on my way to a chemistry test, playing catch with my calculator (in the padded case) as I ran. I had confidence that, even if I dropped it, it would still work.
I accidentally sat on my HP-15C in college and broke the LCD. I wrote to HP to see where I could get the LCD replaced. I was a college student and the thing was like $125 1986 dollars, I wanted to fix it, not buy a new one.
They just sent me a new calculator for free. I was so amazed! Heh.
I repeated almost the same accident with an early kindle and Amazon did the same free replacement. I don't think they do today.
Amazing. I got a HP30s in 2003 as it was the only calculator approved by the university. Everybody I knew hated every single aspect of it. The arrow button was soft and imprecise, leading to long calculations being deleted by accident. The display was too short for longer expressions (compared to the competition). The display scratched easily. The buttons were round and didn’t give good tactical feedback. My jaw dropped when I started working in industry and people would guard jealously their ancient HP calculators with nice clicky buttons and crisp display.
I'm in the age group that spans the end of the slide rule with the beginnings of calculators and personal computers. As a high school senior, trigonometry was the highest math class and we learned to use the slide rule in it. I still have the one my dad bought me for my birthday that year. In college freshman calculus, I remember one guy brought his HP-35 to class and everyone else being awed and amazed by it, including the professor. It was too expensive for me but I was able to buy a TI-50 the next year. Functional but it wasn't an HP.
What an interesting story. My dad had an HP-35 that he used as a grad student at Stanford in the 70s, and he handed it down to me decades later for me to use in my high school math classes.
When I was growing up in the early 80s, my engineer father had an HP-35 that he adored. He ultimately added an HP-12c, too, and when I was in middle school I was gifted an HP-32sii. I was the only weirdo using RPN. :) I upgraded to an HP-48g in high school when everyone else was using Ti-82.
I can't really think of any other company or product line I've had such a longstanding relationship with. FWIW, I still have both of my HP calculators and they work great!
Oh, how I love old HP calculators. My HP 28s is 30 years old and still working great. - I learned programming in a Forth like language (stack based) with it. And it inspired me to learn Scheme and the Mathematica language. Love it. Although these days I rather use a version for my iPhone (HCalc)
wow... what an amazing article. Few interesting bits...
- HP was offering so many products that it took a four-pound, nearly 600-page catalog to describe them all.
- In what has to be one of the most famous design briefs in electronics history, Bill Hewlett asked Osborne and Cochran to shrink the 9100. “I want it to be a tenth of the volume, ten times as fast and cost a tenth as much.”
- At the time it seemed like an impossible request, but Hewlett didn’t let the idea go. Cochran, who for a time lived across the street from Hewlett, would occasionally give him a ride to work.
- It soon became clear that the entire project was going to cost around a million dollars.
- Stanford Research Institute did market survey and their conclusion was clear: “we don’t recommend that you go ahead with this.”
- Hewlett, despite the SRI report, decided he wanted one and thought his engineers should have one as well.
- For the log and exponential functions they used pseudo division and multiplication algorithms from Briggs’ 1624 Arithmetica Logarithmica. For the transcendental functions they used Volder’s CORDIC algorithm, originally developed for B-58 navigation.
- It took them more than a little finessing to fit everything into the 5140 bits (or 0.6 K).
- First industrial design featuring angled the display, a textured case, rubber feet serve as battery compartment latches.
- The entire project took 14 months, half of HP’s typical design cycle.
- Hewlett said the name would be the HP-35, after the device’s 35 keys. HP’s computerized inventory system only recognized four-digit names
- HP typically priced their equipment at the cost of the material list × π (or in an especially competitive market, list × e)
- “We [had to worry about] sales per square foot on the first floor of Macy’s, vs. the second floor.”
- HP-35 “[was] something only fictional heroes like James Bond, Walter Mitty or Dick Tracy are supposed to own,” a device that Captain Kirk of Star Trek was supposed to own.
- climbers carried to the top of Everest to do altitude calculations; Apollo astronauts used it in space to calculate re-entry coordinates
- In all, 100,000 HP-35’s (or more than 10× their estimate) were sold in the first year—accounting for more than half of the company’s total profits
It's amazing the similarity between HP and Apple's success stories. And then the fact that Steve Jobs bought the property that summer and the site is now part of Apple’s new Pentagon-sized spaceship headquarters.
I think pretty much every thing you need to know why companies succeed and fail is in this story.
It uses a pack of rechargable AAs, I have an HP-45 (the immediate successor, with the same battery pack) and replaced the original NiCD cells with NiMH. When I got it the batteries were clearly already a few decades old. AAs NiCDs were pretty common back then, I have an old TI and Soviet-made Electronika calculator that also use them.
Right. It mostly wasn't a problem so long as you remembered to charge them before exams, etc. In college, I'd still bring a slide rule to final exams as backup just in case something happened although I never had a problem.
I started college just as scientific calculators were appearing on the scene. I had a TO for a couple of years because the HPs were still so expensive but I was able to pickup an HP-55 after a bit for a relatively affordable price (still probably talking about high 3 digits in today's dollars). That's what I used until I bought an HP-41CV when I went back to school and needed a more programmable calculator.
To add some context: essentially all scientific calculators with LED or VFD displays had rechargable NiCd batteries and plug packs. Typical engineer mostly used them plugged into wall. (And then there was my dad who used his IIRC Sharp VFD scientific calculator as charger for AA NiCd cells well into late 90's)
100K units in the first year, vs your king of consumer electronics today: the iPhone, which pre-sells and ships 100K units in what, several minutes?
I understand not everyone buys calculators, but that’s sort of the point of the article that this was marketed as a ‘consumer’ electronics product rather than scientific equipment.
Also there’s much less friction and much more marketing today, but it’s wild to imagine how millions of iPhones ‘move’ in a single weekend, every year for nearly the last decade and what is required to keep supply in pace with demand.
That's Moore's law in action over the longer term. A device like the iPad back when the HP-35 was introduced would have been (1) impossible to make and (2) would have cost millions of $ given the number of transistors in the device (besides the fact that the internet is what makes an iPad useful).
You really can't compare those numbers, yes, the HP was a consumer device, but the manufacturer of any consumer device back then would have been happy to sell in those numbers.
The good news is now it can be yours for under $2 if you already have an iPhone:
> The HP-35 had numerical algorithms that exceeded the precision of most mainframe computers at the time. [...] This forced time-consuming manual comparisons of results to mathematical tables. A few bugs got through this process. For example: 2.02 ln e^x resulted in 2 rather than 2.02. When the bug was discovered, HP had already sold 25,000 units which was a huge volume for the company. In a meeting, Dave Packard asked what they were going to do about the units already in the field and someone in the crowd said "Don't tell?" At this Packard's pencil snapped and he said: "Who said that? We're going to tell everyone and offer them, a replacement. It would be better to never make a dime of profit than to have a product out there with a problem". It turns out that less than a quarter of the units were returned. Most people preferred to keep their buggy calculator and the notice from HP offering the replacement.
And here is a scan of that recall notice: https://imgur.com/K1k0cSQ (original source: http://www.hpmuseum.org/forum/thread-2821.html)