The Museum of HP Calculators

HP Forum Archive 17

[ Return to Index | Top of Index ]

HP-42S I/O
Message #1 Posted by Gerry Schultz on 7 June 2007, 1:33 p.m.

Okay, I just got my 42 and it works great. Now, I have some questions. I've read previous discussions here about the 42 and limited I/O but what's the deal? After looking at memory management in chapter B, it looks like the maximum available memory in the 42 is about 6977 bytes, about the same as a 41CV or CX. With my 41, I had magnetic data cards, tape drives and even 3 1/2 disk drives for mass storage. There's no way to load or store 42s programs!?

I find this unbelievable. Part of the success of the 41 line was its great selection of peripherals. I don't get it; what happened? This must be a doozy of a story given the 42 was suppose to be the replacement for the 41. I would think that HP would have wanted 41's peripherals to work with a 42. Oh, don't tell me, to save money. Boy, I hope not.

I do have an HP 82240B printer that I bought for my 48GX and it should work fine with the 42 but I'll have to test it. The 42 is a fascinating machine in that it seems to bridge the gap between how the classic HP calculators and the newer RPL calculators are designed. Even with the I/O limitations, I'm interested in using this machine.

As I mentioned in a previous post, I got an HP-97 and I just finished reading the owner's manual. Yes, I do read the manual. My boss thinks I'm nutty and he prefers the "poke around until it works or breaks" method.

Anyway, as an old PPC member, I have all the back copies of the PPC manual on paper and since the PPC was originally founded on the 65 and 67/97 calculators, its pages should be full of programs for me to try on the 97. Do any of you have a favorite to share from the old PPC? This is where I learned about NNNs. Thanks, Valentin Albillo for your warning.

Since I lost contact with the PPC or the HPPC after the 41, did these user groups ever publish 42 programs? If they did, where can I get copies of these issues? Are they on CDs or DVDs in PDF like the files here at HP museum?

Thanks for the feedback, guys.

Gerry

      
Re: HP-42S I/O
Message #2 Posted by Dave Johnson on 7 June 2007, 2:14 p.m.,
in response to message #1 by Gerry Schultz

Your right - there is no input for the 42 and only printer output (though with some tricks I believe you can get memory dumps via IR transfer but not quick or easy...) I am sure it was hotly debated within HP whether the 42S should ever be released and it is not really a follow up for the 41 series. The debate was won by supporters of limited calculators such as the 32 series with limited memory, functionality and no menus. I'm sure it was disconcerting for true calculator supporters within HP at the time.

      
Re: HP-42S I/O
Message #3 Posted by Thomas Okken on 7 June 2007, 4:27 p.m.,
in response to message #1 by Gerry Schultz

The HP-42S is not the successor of the HP-41. It is the successor of the HP-15C: small but powerful with a rich set of functions -- and it just happens to be designed around the HP-41 programming model.
Rumor has it that the 42S was initially supposed to have bidirectional I/O, but that that was scrapped in order to prevent it competing with the HP-48 series -- and in terms of being a hand-held computer that can interface with the outside world, the 48 series is the true successor of the 41.

- Thomas

            
Re: HP-42S I/O
Message #4 Posted by Ron Allen (Fairhope) on 13 June 2007, 2:01 p.m.,
in response to message #3 by Thomas Okken

Sorry I had to reach you off topic, but I wanted to thank you for putting me on to the lectures by Weynman, especially v1 15. He seemed to anticipate all my questions and answer them before I could formulate them. Instincts from teaching so many freshmen. The full development of those equations are gratifying to see.

Amazon is running a special right now - I ordered my own set. The "Lectures" was certainly a great idea by those students at Cal Tech. What a teacher!

Thanks again,

Ron

                  
Feynman lectures
Message #5 Posted by Thomas Okken on 13 June 2007, 9:37 p.m.,
in response to message #4 by Ron Allen (Fairhope)

Quote:
Sorry I had to reach you off topic, but I wanted to thank you for putting me on to the lectures by Weynman, especially v1 15. He seemed to anticipate all my questions and answer them before I could formulate them. Instincts from teaching so many freshmen. The full development of those equations are gratifying to see.

Amazon is running a special right now - I ordered my own set. The "Lectures" was certainly a great idea by those students at Cal Tech. What a teacher!


You're welcome!
Feynman was not just a great scientist and teacher, but also an all-around interesting person. For some really wonderful light reading I recommend "Surely you're joking, Mr. Feynman" and "What do YOU care what other people think?". The latter has fascinating first-hand accounts of his work on the Manhattan Project and the Challenger inquiry.

- Thomas

                        
Re: Feynman lectures
Message #6 Posted by bill platt on 14 June 2007, 8:21 a.m.,
in response to message #5 by Thomas Okken

One of my favorite (auto)biographies.

                              
Re: Feynman lectures
Message #7 Posted by Prabhu Bhooplapur on 22 June 2007, 1:30 a.m.,
in response to message #6 by bill platt

More than two decades ago, just after I had started work after my graduation, I read "Surely, You must be joking Feynmann" and was immensely fascinated by his personality and subsequently read his other works including some of his lectures. Three weeks ago I gave the book (Surely...) to my son who is pursuing his research interests in Physics related electrical engineering at an American university. I am sure he will quite appreciate it. Feynmann is a must read for people interested in science and technology.

      
Re: HP-42S I/O
Message #8 Posted by Karl Schneider on 8 June 2007, 2:29 a.m.,
in response to message #1 by Gerry Schultz

Hi, Gerry --

You stated:

Quote:
... the maximum available memory in the 42 is about 6977 bytes, about the same as a 41CV or CX. With my 41, I had magnetic data cards, tape drives and even 3 1/2 disk drives for mass storage. There's no way to load or store 42s programs!?

Part of the success of the 41 line was its great selection of peripherals. I don't get it; what happened? ... I would think that HP would have wanted 41's peripherals to work with a 42.


Thomas Okken stated:

Quote:
Rumor has it that the 42S was initially supposed to have bidirectional I/O, but that that was scrapped in order to prevent it competing with the HP-48 series.

The other "high-end" Pioneer-series models (namely, the HP-17B/BII and HP-27S) also have 7 kB of RAM, but no I/O except the IR output to printer.

I believe that the primary reasons for lack of full I/O are engineering-based physical and practical considerations. Bidirectional IR would be convenient, but those users who would exchange programs and data via I/O would certainly demand reliability and low error rate. This dictates hardwire I/O and inexpensive batteries with adequate energy storage for the demands of data transmission.

Remove the translucent red cover on a Pioneer model, and what do you see? Three 13/44/76/357 1.5V "button" cells in a row, a place for the IR transmitter (present or not), barely enough room (given a minor redesign) for the IR receiver, but insufficient room for a robustly-constructed 4-pin hardwire plug, such as that used on the original HP-48S/G series.

The HP-48 had three bigger and higher-capacity "AAA" batteries placed in a different location within its much larger package. The Pioneers, however, had no other place for its button cells inside its slim package. The costly button cells -- designed for medical devices -- don't leak, but also don't hold a great deal of charge. Use of the HP 82240A/B printer will shorten battery life considerably, as will keeping the clock running on the HP-17B/BII or HP-27S.

Space could have been made available in the front by using thin "coin" cells placed side-by-side elsewhere (such as the two CR2032 cells in the HP-33S), if these were even available in the 1980's. However, the limited mAh capacity and cost per mAh would have been further exacerbated.

As for the HP-42S' lack of support for HP-41 peripherals, the obvious problem is the lack of space in the Pioneer package for the peripherals' plugs, even if a suitable adapter could be designed. Another engineering problem is power consumption, as some peripherals -- such as the card reader, wand, and HP-IL -- draw their power from the HP-41's batteries. The broader issue, however, is that PC's became available and increasingly affordable in the 1980's, enabling them to usurp the niche of peripheral control filled by the HP-41 and HP-71B.

-- KS

Edited: 10 June 2007, 6:07 a.m.

      
Re: HP-42S I/O
Message #9 Posted by Christoph Giesselink on 10 June 2007, 5:59 p.m.,
in response to message #1 by Gerry Schultz

Some aspects don't mention from the authors before.

Memory limitation to ~7000 bytes.

The 8KB and 32KB 8 bit static RAM chips have a quite similar pin layout. The PCB of the High-End Pioneers is prepared for the use of both chips. So it's possible to replace the RAM chip with the 32KB on all high end pioneers. 2nd part are the MMU mask fuses in the Lewis chips. These are programmed for a memory device at address #50000 with a size of 32KB on all High-End Pioneers.

But why only the HP42S can make use of this additional memory getting ~31500 bytes of free memory whereas on the HP17B or HP27S there's no difference?

The answer is very simple. Only the HP42S firmware contain code expecting more then the usual 8KB of RAM (history repeats). So we can say that HP thought about HP42S with more memory but never sold them.

Now to the limited I/O capabilities of the High-End Pioneers. Therefore we have to go back into Saturn history. The HP71B, the first calculator with the Saturn CPU, was build out of several modules, display driver, ROM, RAM, ... made the complete thing quite expensive. On the other hand we have had a powerful mobile CPU (Saturn) which was predestinated for further use. The HP18C used this CPU but now in connection with a more integrated customer chip (1LP2 Centipede) containing reset control, oscillator, RAM, display driver, timer, input/output, ... reducing the amount of necessary chips to build a calculator. But from history we can see that this was still too expensive too, because only the HP18C and HP28C used this hardware.

Now my short comment to "history repeats". The HP28C was sold with 2KB of RAM (from the two integrated Centipede chips). But Educalc sold versions with more than these 2KB of RAM. The HP28C hardware was prepared to add HP71B RAM module chips and the HP28C firmware was adding this additional found RAM to it's memory pool. I don't know if HP ever thought to sell memory upgraded HP28C by themselves, but HP added code to the firmware to exactly do this without firmware change.

The successors, the High-End Pioneers and Clamshell calculators, use the Lewis chip. In a very very simple view the Lewis chip is a combination of the Saturn CPU and the Centipede customer chip. From the I/O side the Lewis chip inherit the I/O capabilities of the Centipede chip, nothing more, nothing less.

This is in fact a single bidirectional Input/Output pin. On the HP42S and on all other calculators with the Lewis chip this hardware pin was used to drive the IR transmitter LED.

I verified this on a HP28S (also using two Lewis chips), where I removed the IR transmitter LED and connected a simple hardware to adjust the calculator logic level to RS232C level. The interface is mostly powered by the PC RS232C interface, the power consumption out of the calculator battery is < 4 uA. The rest for the RS232C communication between HP28S and PC is software.

Cheers

Christoph


[ Return to Index | Top of Index ]

Go back to the main exhibit hall