The Museum of HP Calculators

HP Forum Archive 15

[ Return to Index | Top of Index ]

hp 41 or hp 49?
Message #1 Posted by e. young on 21 Jan 2006, 12:59 p.m.

which one of the 2 machines would be better for dayly use by an engineer? I know the 41 series has huge nostalgic appeal, and also that the 49 has had quality issues. Is one easier to program than the other? Is buying a 49 a risky move due to quality and bad design issues that have plagued the most recent calculators from HP (Kinpo)? Thanks.

      
Re: hp 41 or hp 49?
Message #2 Posted by Stephen Easterling on 21 Jan 2006, 2:12 p.m.,
in response to message #1 by e. young

I have never been able to adjust to the way the numbers are displayed on the 41 with the LCD segments. Any of the Pioneer displays are much better, IMO. A 42S, 32S, 32SII, or 33S are much better choices, IMO.

The 49G+ is better at almost anything that HP has ever made in a calculator. It's fast, inexpensive (compared to a nice vintage HP calc), has virtually unlimited memory (depending on SD card size), the display contrast is unbeatable, and the ROM can be upgraded. The downside: it's big and bulky and hard to fit in your pocket. Regarding quality, the new models now,... and using at least ROM 2.00, are great. The quality is more than good enough for me and still better than the competition.

I think a 15C is great if you like it's keypad layout and don't mind its speed. The 15C will go easy on batteries, plus I've never heard of a LR44 / 357 battery leaking as easily as others. The 15C is also made strong like a little brick! There's a reason why so many here want to "bring back the 15C." There aren't too many that say "bring back the 41." Again, this is just my opinion.

Edited: 21 Jan 2006, 2:16 p.m.

            
Re: hp 41 or hp 49?
Message #3 Posted by Massimo Gnerucci (Italy) on 21 Jan 2006, 3:52 p.m.,
in response to message #2 by Stephen Easterling

Quote:
There's a reason why so many here want to "bring back the 15C." There aren't too many that say "bring back the 41." Again, this is just my opinion.

Of course this is just your opinion, do you allow me to express mine?

While so many (?) shout that they want back the 15C, you can actually buy (right here, right now) brand new enhacements for the 41C (Clonix, NoVram, MLDL2000).

In 2006, for a calculator out of production since 15 years.
Doesn't this ring a bell?

And why do you find the numbers in the display of a 15C much better than those in the 41C?

Greetings,
Massimo

            
Re: hp 41 or hp 49?
Message #4 Posted by Massimo Gnerucci (Italy) on 21 Jan 2006, 5:08 p.m.,
in response to message #2 by Stephen Easterling

My apologies to Stephen for the tone used in my previous reply: it could easily sound offensive.
That was not my intention.

When someone doesn't like my lady the blood boils up...
Sorry Stephen!

Massimo

      
Re: hp 41 or hp 49?
Message #5 Posted by Howard Owen on 21 Jan 2006, 2:21 p.m.,
in response to message #1 by e. young

From a casual perusal of comp.sys.hp48, it appears to me that the latest 49G+ ROMs, combined with tweaking of parameters (dealing with keyboard debounce, it seems,) look like they have addressed keyboard issues with recent vintage 49G+ machines.

Otherwise, the two machines truly are like apples and oranges. Both have learning curves, but the 41C is a much simpler machine. This can be an advantage or not depending on whether you need some bell or whistle available on the 49G+ that's not there on the 41. Since the feature list of the 49 is so long, I can't suggest what that might be, except to note that the screen graphing capabilities and symbolic manipulations are missing on the 41. Many people here think those are overkill on a calculator anyway, given that PCs do them so much faster and better, and that numeric calculation is the proper domain of a handheld calculating machine.

Here are some other differences I can think of.

Item HP41C HP49G+ HP48GX PC ========================================================== Speed Slow Faster Intermediate Fastest Ease(Programming) High Very Low Very Low+ Varies Ease (Use) Medium Low Low Varies Software Lots Lots Lots Lots! Ease of loading Hard Easy Intermediate Too easy. 8) "Power" Low Medium Medium-Low High

The 49G+ is a 75Mhz ARM that emulates the Saturn CPU found in earlier 48 series machines. So some of its functions are reasonably fast, but others poke along due to the emulation. the 49G and 48G+ both ran a 4Mhz Saturn. In terms of speed, they were OK for their day, but have been eclipsed by 68K TI machines and the ARM based HP models. The 41C CPU ran at 750 Hz, but it was optimized for BCD math, and so performed a bit better than you would expect at that speed. Whether it's too slow for you depends on what exactly you need to calculate, and in what timeframe. If speed is really important, no calculator can beat a modern PC, running any recent processor I know about.

Ease of programming is somewhat subjective. A couple of months ago we engaged in extensive discussion about the relative nerits of RPN keystroke programming, which the 41C implements, and RPL, which is what the 48 and 49 series use. The RPL afficianados claimed that RPN seemed counter-intuitive and difficult for them, whereas most of the RPN enthusiasts said the same about RPL. Obviously, familiarity with one or the other plays a role in perceptions of this quality. But I don't think anyone would seriously argue that RPL wasn't a lot more complicated - and more powerful - than RPN. Having just completed implementing a game on the 48GX, 49G and 49G+, I can say that there was light at the other side of the first inflection point on the learning curve for me. I'm now pretty comfortable with User RPL, but it took a good six months of occasional effort to get there. By way of contrast, it took me something like a month to learn to program the 41C at fairly high proficiency back in the 80s, and that was the first time I'd programmed anything!

Ease of use depends on how easy it is to get to common functions, and how easy it is to get them to work together. Both the 49G+ and the 41C had a lot of though devoted to their respective user interfaces. So the ensemble works pretty well in both cases. However, there are fewer functions on the 41, so there are fewer that you need to spell out in order to get to. The 49G+ has menus, but customizing those is more trouble than assigning a key on the 41C. I give both high points for trying to make themselves usable, but the 41C wins on simplicity, in my opinion.

There's lots of software available for both platforms. Some of it overlaps, but as you might expect, the 41C's lean heavily toward numeric calculation, whereas the 48 amd 49 series have a load of symbolic and graphical apps too. Mostly though, the software for the 41 is as good for what it does as anything available for the 49. And of course neither can hold a candle to the number of high quality apps available for PCs.

There's an issue regarding loading software onto the 41C. If you don't want to key in a program by hand, you need extra hardware. There are lots of choices for this, but they are complicated, and invole hardware as old as the 41C itself.

Partially balancing this is the fact that 41C programs can only get so large before they fill the limited memory on the machine. This, combined with the slow speed of the CPU is why I give the machine a "low" power rating. The 48GX has more memory and a faster CPU. Also, RPL is more complicated, but also more expressive than RPN. I think that the 49G+ is more powerful than the 48GX by virtue of its faster CPU and larger usable memory. But there's also the fact that an assembler and other developer tools are built in to the machine. Once again the PC wins by a long mile over any of these machines. Clusters of Intel (and AMD) architecture machines are among the fastest supercomputers on Earth at the moment. And even desktop PCs run rings around the fastest calc I'm aware of. Combine that with the better software choice, and it's easy to see why the calculator has been supplanted for many functions.

But even a laptop is more trouble to carry than a calculator. That's one other thing the 41C has going for it - portability. It has better battery life and a smaller form factor than the HP49G+. Plus it has a special place in my heart. I use my 41C for everything up to the point where I exceed its capacity, mainly due to this nostalgic love for the machine. What works for you may vary.

Regards
Howard

Edited: 21 Jan 2006, 2:23 p.m.

            
Re: hp 41 or hp 49?
Message #6 Posted by e. young on 21 Jan 2006, 2:40 p.m.,
in response to message #5 by Howard Owen

Thanks for the excellent replies.

I have gotten by with the "low" end hp's (11c, 32sii, 33s) since I first got turned on to RPN. The 33s has gotten me into programming because I can now have many more in it as compared to the 32sii. I do a lot of numeric calcs, I don't often get involved with the higher end stuff, and when I do, I use pc. So far I find RPN keystroke programming to be very and intuitive.

      
Re: hp 41 or hp 49?
Message #7 Posted by Garth Wilson on 21 Jan 2006, 3:06 p.m.,
in response to message #1 by e. young

Although most users used the 41 as a calculator, it definitely qualifies as more of a hand-held computer than the 49 "supercalculator" (all one word, if I can coin it) and similar products. I'm not referring to execution speed, or quantity of memory, or graphics; but the 41 was, as far as I know, the first one that could be interfaced to a lot of professional lab equipment, even a whole rack of instruments simultaneously, while its contemporaries only pretended, being limited to maybe a tiny printer and a microcassette storage device. This interfaceability was important in my work in the 1980's; and in an engineering lab where I woked, almost every engineer had his own 41 that took almost no workbench space, went with him to his desk, went home with him every night in a corner of his attache case, and could be programmed anywhere including on a train ride to work. At least at the time, it was much quicker to whip up a control program on the 41 than it was on PCs. Unfortunately, HP dropped the HPIL which was the key to this interfaceability. An IR link doesn't cut it. For more, see my post at http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/memories.cgi?read=224

My work has evolved over the years; and although I do sometimes still use the 41 and 71 for their interfaceability, more of my workbench-interfacing use of computers involves a home-made workbench computer that interfaces directly to circuits I'm working on, without the bus protocols of HPIL, HPIB, USB, etc..

            
Re: hp 41 or hp 49?
Message #8 Posted by ECL on 21 Jan 2006, 5:07 p.m.,
in response to message #7 by Garth Wilson

In regards to the 33s: It is a good tool. My only wish is for any update to handle small matrices. I've said 10x10 in the past (any more and I'd probably input to a pc utility).

The 49g+ is obviously much more capable than the 41. The human/machine interface is the only fault. My copy is very old (obtained in '03) Perhaps the new ones really are better. If the keys were a bit softer and the bounce/skip issue is resolved, then I'd like to purchase one.

Is there a running consensus on the CURRENT batches? ECL

      
Re: hp 41 or hp 49?
Message #9 Posted by Vieira, L.C., (Brazil) on 21 Jan 2006, 7:24 p.m.,
in response to message #1 by e. young

Hi e.young, all;

although many others have already posted many good stuff about this issue, and their opinions/thoughts/considerations are mainly regarding to the functionality and/or use, all Id add is that I'd rather taking an HP49 instead of an HP41 when leaving home. Main reason: I have no idea of what kind of people will ask to borrow my calculator... (I hope I wrote it correctly)

I have five working HP41s and three (working?) HP49s.

Cheers.

Luiz (Brazil)

            
Re: hp 41 or hp 49?
Message #10 Posted by Howard Owen on 21 Jan 2006, 9:25 p.m.,
in response to message #9 by Vieira, L.C., (Brazil)

You reminded me of another point, Luiz. If you have a 49G or 48GX, you can run Hrast's HP41X emulator, and have the best of both worlds. The 49g+ version exists too, but it isn't quite as functional. Perhaps functional enough for many purposes, though?

Regards,
Howard.

                  
Re: hp 41 or hp 49?
Message #11 Posted by Vieira, Luiz C. (Brazil) on 21 Jan 2006, 9:44 p.m.,
in response to message #10 by Howard Owen

Hi, Howard;

you're more than right, very well noted! To be honest, only one of my HP49G's is 'as is' and ready to use (RPN, though). The HP49G+ and the other HP49G are actually running HP41X. Their best use, indeed!

I had the chance to purchase Hrast's HP41 + Voyager's emulators; believe me, that's what I call 'money with a purpose'!

Cheers.

Luiz (Brazil)

                        
Re: hp 41 or hp 49?
Message #12 Posted by Howard Owen on 21 Jan 2006, 9:51 p.m.,
in response to message #11 by Vieira, Luiz C. (Brazil)

Hi, Luiz,

Yes, I remember you said you had HP41X. That's what reminded me. 8)

I have HP41X and HP71X. I'm planning to acquire the others, too. Great stuff!

Regards,
Howard

Edited: 21 Jan 2006, 9:53 p.m.

                        
Re: hp 41 or hp 49?
Message #13 Posted by Massimo Gnerucci (Italy) on 21 Jan 2006, 9:55 p.m.,
in response to message #11 by Vieira, Luiz C. (Brazil)

Hello Luiz,
nice move!

In one of my 48s (a double-speed G internally expanded to 1280Kb) I have Hrast's 41X, 42X, 71X, 11E, 12E, 15E and 16E.

Aaah, what a nice machine!

Greetings,
Massimo

                              
Re: hp 41 or hp 49?
Message #14 Posted by Vieira, Luiz C. (Brazil) on 21 Jan 2006, 10:10 p.m.,
in response to message #13 by Massimo Gnerucci (Italy)

You lucky b... Oops! (LOL... d8^D)

I'm happy for you, Massimo!

Cheers!

Luiz (Brazil)

Edited: 21 Jan 2006, 10:12 p.m.

                              
Re: hp 41 or hp 49?
Message #15 Posted by HrastProgrammer on 22 Jan 2006, 1:28 a.m.,
in response to message #13 by Massimo Gnerucci (Italy)

I have Hrast's 41X, 42X, 71X, 11E, 12E, 15E and 16E.

What?!? You don't have Hrast's 57E?!? Shame on you ;-)

                                    
Re: hp 41 or hp 49?
Message #16 Posted by Howard Owen on 22 Jan 2006, 2:15 a.m.,
in response to message #15 by HrastProgrammer

Oh yeah. I've got that one too. 8)

                                    
Re: hp 41 or hp 49?
Message #17 Posted by Massimo Gnerucci (Italy) on 22 Jan 2006, 4:50 a.m.,
in response to message #15 by HrastProgrammer

Hi Hrast!
of course yes, I have it; but on my HD, not on the calc... not yet? :-D

All the best,
Massimo

                                          
Re: hp 41 or hp 49?
Message #18 Posted by HrastProgrammer on 22 Jan 2006, 4:53 a.m.,
in response to message #17 by Massimo Gnerucci (Italy)

http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/forum.cgi?read=85688 :-)

      
Re: hp 41 or hp 49?
Message #19 Posted by Donald Ingram on 22 Jan 2006, 7:40 p.m.,
in response to message #1 by e. young

In the mid 80's, I had reached a state where my the HP41CX was becoming a bottleneck and I was very excited at the launch of the HP48sx.

I used complex numbers and matrices a lot for microwave circuit analysis - the HP41 stack could not really cope and my fully loaded 41CX never had enough memory. So as a poor recent graduate I had to sell it. Occasionally I read a nostalgic article on the HP41 and write a few lines of code on the emulator, but find the limitations of the 41 too great and switch to the 49.

I now have the HP48sx at home and a horrible blue HP49 at work - both still get used a lot for simple conversions e.g. VSWR - RL, Vpkpk - dBm, Xc etc. but all the large calculations e.g. microstrip impedance, filter design are now on the PC.

However on that PC runs the HP41 emulator for when I loose the HP49 under a pile of paper.

The Hp48/49 series is easier to program and archive than the HP41. ( If only HP had make a RS232 module instead of HPIL - the HP41 series could still be in production as a Instrument terminal ) .

Today if you need a calculator as a tool - the HP49 series would give you most run for your money. Check http://www.hpcalc.org and you might find code that fits your needs. Also check out the old HP application pacs - I still often refer to HP41 and HP67 codes examples ( it's easier to rewrite them using the formulas than decoding old keystrokes ).

            
Re: hp 41 or hp 49?
Message #20 Posted by Garth Wilson on 23 Jan 2006, 2:07 a.m.,
in response to message #19 by Donald Ingram

Quote:
If only HP had make a RS232 module instead of HPIL - the HP41 series could still be in production as a Instrument terminal
You should have gotten the HP82164A or FSI164A HPIL-to-RS232 converter. Better (or in addition), you should have gotten the HP82169A HPIL-to-HPIB converter, which would allow you to interface to a whole rack of IEEE-488 equipment all at once. RS-232 takes too much power to implement directly on the 41, you need a separate RS-232 port for every piece of equipment you want to interface to, and most pro lab equipment was available in IEEE-488, not RS-232. We did set up automated testing with the 41 as a controller. That was never possible with the 49.
                  
RS232 would have worked
Message #21 Posted by Donald Ingram on 23 Jan 2006, 8:19 a.m.,
in response to message #20 by Garth Wilson

I don't buy the RS232 power consumption - The HP41 can suply greater currents - e.g. the card reader and optical reader.

Especially if you limit the data rate to < 19200 baud the Maxim data sheets show less 10mA or 20mA at those rates for RS232 TX. As TX is intermittent and RX level conversion does not really reqiuire anything, the average current would be less than 5mA = lots of hours from a 1000mAh N battery. A HP41 with RS232 is entirely practical.

The HPIL solution, while elegent, was too expensive for many applications ( and certainly for me ). Controlling lots of expensive mains powered instruments, the cost of a HP85 would be less significant. I was wanting to control a <$100 worth of electronics.

The HP41 could have ended up as a single instrument RS232 controller/data logger/ terminal working as an industrial ( and hobby ) process controller. Imagine lego mindstorm type robots with a HP41 controller on board.

In the 90's before laptops really took off, I used a lot of 'craft terminals' single line dumb terminals to configure PDH/Sonet mux equipment. A RS232 HP41 would have done the job with extra functionality and lower cost.

                        
Re: RS232 would have worked
Message #22 Posted by Garth Wilson on 23 Jan 2006, 2:26 p.m.,
in response to message #21 by Donald Ingram

Ah yes, the cost of the HPIL and Extended I/O modules plus the interface converters was not trivial. In fact, the joke where I've worked (even among those of us who thought it was worth it) was that "HP" stood for "high-priced." A plug-in RS-232 port would have been nice for people in your position.

RS-232 is still a good workhorse and I still use it, although almost never without the hardware handshaking lines (which again add more current). Not everything supports software handshaking, and the 41 is not fast enough to go without some kind of handshaking. Hardware handshaking would just about be mandatory.

Then there's still the limitation that if you want to connect to five things at once, you need five ports, further multiplying the current requirement. HP would have had to derive the negative supply voltage in the module, and kick the positive supply up a little since 4 partly-used-up N cells don't deliver enough to get the minimum needed output voltage after the line driver drops a bit of it. Today there are quite a lot of line driver/receiver ICs that take care of the voltage conversion onboard for you, although increasing the voltage comes at the expense of supply current.

Making another module is something that could still be done if someone wanted to take up that project. (I'm not the person to do that though.) One nice thing I've seen some hand-held RS-232 devices do is to scavenge their power from the incoming lines, such that in some cases they didn't even need a battery. What the 41's module could do is to use it battery power only to fill in what was lacking from incoming lines. Then connecting to a bench instrument would not cut the 41's battery life at all, and yet the 41 could still be connected to another 41.

Another small connector would have to be used, but that should be no problem.

HPIL was a lot faster than the fastest (19,200bps) RS-232 standard speed in the early 80's, but even that doesn't matter since the 41 couldn't operate the HPIL at more than about 3% of its maximum speed.


[ Return to Index | Top of Index ]

Go back to the main exhibit hall