HP Forum Archive 14

[ Return to Index | Top of Index ]

Deep PI - On a 42S Message #1 Posted by Bill (Smithville, NJ) on 7 May 2004, 8:47 a.m. Started playing around with the Deep PI routine that Hugh had posted. Entered it into two HP-42S, one of which is the earlier ROM which supports "Fast" mode. I've been curious about how fast is fast mode and thought the Deep PI routine would made a good comparison. I set up a infrared printer with a HP-41CX and the two HP-42S pointed at it. The 41 is running a short endless loop program that prints the date and time every 5 minutes so that I have a time trail on the printout. One 42S is running normal mode and one is running in fast mode. Both running the PI loop (1, 7, 13, ...) and then print the results. Ran it for an hour last night and the Fast 42S was calculating about 2.5 times faster. I plan to load new batteries into both 42S tonight and then I'll just let them run until the batteries die. Be interesting to see how many decimals they calculate and how long the batteries last. The printer I can run on an AC adapter. The 41CX has a rechargable battery pack. If I run the 41 with the AC recharger pluged in, will it basically run for ever?? Or should I load up the 41CX with fresh regular batteries to get maximum constant on time? I don't want the 41 to run down before the 42S batteries die. I'll post the results next week for anyone who's courious.

Re: Deep PI - not on a 32SII :-( Message #2 Posted by Bram on 7 May 2004, 9:30 a.m., in response to message #1 by Bill (Smithville, NJ) I had already converted the program into 32sii format (on paper). It has just enough LBLs and STOs to fit. I started keying, but not even half way I got the memory full message. pity. curious about the 42s results.

Re: Deep PI - not on a 32SII :-( Message #3 Posted by Veli-Pekka Nousiainen on 7 May 2004, 11:16 a.m., in response to message #2 by Bram "I had already converted the program into 32sii format (on paper). It has just enough LBLs and STOs to fit. I started keying, but not even half way I got the memory full message. pity. curious about the 42s results." You need a hp 33s !!! If you have the program ready, please send it here so we can test it! Thank You! {VPN} AXL

Re: Deep PI - not on a 32SII :-( Message #4 Posted by Bram(nl) on 9 May 2004, 7:36 a.m., in response to message #3 by Veli-Pekka Nousiainen No objection to publish my conversion, but as mentioned, I´ve done it on paper. I would have to scan it. Conversion is fairly straight forward, though. I downloaded www.voidware.com/calcs/pi.txt, printed it and replaced every label and storage number with the `matching` letter. 01 becomes A, 02 becomes B aso. 00 became Z and 05 became i, because RCL IND 5 translates to RCL (i). I wrote a little routine for MOD (32sii hasn´t) and began typing, but as the machine complained so soon, I judged it to be too ´simple´ to cope.

Re: It is not working (input is ex:2x+3=6) Message #5 Posted by Veli-Pekka Nousiainen on 9 May 2004, 8:16 a.m., in response to message #4 by Bram(nl) 33s has enough memory to store the program {VPN} AXL

Re: Deep PI - not on a 32SII :-( Message #6 Posted by Bill (Smithville, NJ) on 7 May 2004, 1:46 p.m., in response to message #2 by Bram Bram, I wonder if the code could be optimized to fit into a 32SII. Not necessarily optimized for speed but for space. I don't have a 32SII any more - I tried one for a while, but just couldn't get used to it. I did love the large display!! Probally could be adapted to the new 33S.

Re: Deep PI - not on a 32SII :-( Message #7 Posted by hugh steers on 7 May 2004, 2:27 p.m., in response to message #6 by Bill (Smithville, NJ) i can make it smaller and, in fact, faster if you dont mind not having the first 76 decimal places!! i was tempted to do this on the original version. but i wanted the whole set of digits. you could, of course, store the first 76 places in the program manually. hey bill, you are trying the experiment i want to try soon. basically i have a 41 and a mains adapter but no printer. my plan was to acquire a printer and leave it for two weeks or so. im keen to see how it compares on speed. i'd like to think it will give the first 1000 places in 12 hours, but i could be completely wrong here. if it is really really slow, i might try to speed it further. i have some ideas but they require considerable work. good luck with your tests, and please post your findings.

Re: Deep PI - not on a 32SII :-( Message #8 Posted by bram(nl) on 9 May 2004, 7:42 a.m., in response to message #7 by hugh steers >i can make it smaller Even then I think it won´t fit in an HP-32SII. The machine already complained at about a third of the program. Each number takes 9.5 bytes and there are a lot of numbers and too few registers left to hold them all or even some. Better to follow Veli-Pekka´s advice (or accept a current offer for an HP-42S for 60 euros. Thinking about it).

Re: Deep PI - not on a 32SII :-( Message #9 Posted by Veli-Pekka Nousiainen on 9 May 2004, 8:15 a.m., in response to message #8 by bram(nl) 33s has more memory

Re: Deep PI - On a 42S Message #10 Posted by David Smith on 7 May 2004, 12:52 p.m., in response to message #1 by Bill (Smithville, NJ) The 41CX will run forever on a rechargeable pack that is plugged in (or at least until the battery physically cells go bad).

Re: Deep PI - On a 42S Message #11 Posted by Bill (Smithville, NJ) on 7 May 2004, 1:51 p.m., in response to message #10 by David Smith Thanks David, That's great news. I won't have to worry about the 41 running down before the 42S. I'm really curious at how long a 42 can run on a set of batteries. Also, will be able to see what difference Fast mode makes on battery consumption. Of course, having the printer IR port turned on will reduce battery life. Question: On the 42S, does having the IR port turned on reduce battery life IF one never prints with it? Does it consume power even if not printing?

Re: Deep PI - On a 42S Message #12 Posted by Richard Garner on 7 May 2004, 3:20 p.m., in response to message #11 by Bill (Smithville, NJ) Yes,if the IR is in trace mode it is sending information even though there is no reception by the printer. I would say no, if trace mode is off and you are not using any commands that would activate the IR to send to the printer.

. . . how about rigging an external supply to the 42s? Message #13 Posted by Paul Brogger on 7 May 2004, 6:44 p.m., in response to message #11 by Bill (Smithville, NJ) Someone here could suggest the correct voltage -- something like 4.5 V or less. Connecting the right leads from, say, a Radio Shack AC-to-DC converter (or maybe just a few D cells?) should enable you to test for significantly longer . . .

Re: . . . how about rigging an external supply to the 42s? Message #14 Posted by Tony Duell on 7 May 2004, 7:09 p.m., in response to message #13 by Paul Brogger The 42S runs off 3 cells at about 1.5V each, so 4.5V should be fine. However, those little wall-wart mains adapters are often not regulated and the off load voltage can be as much as double what you expect. For that reason, and because I've had them catch fire, I won;t allow them in my workshop! Use a decent, regulated, bench power supply and you'll have no problems. 3 D cells in series should be fine as well.

Mastech HY 1500 Message #15 Posted by Wayne Stephens on 8 May 2004, 11:33 a.m., in response to message #14 by Tony Duell Yesterday the UPS man brought me the Mastech HY1500 regulated power supply which I purchased off ebay this week ($85 including shipping). I have been playing with it quite a bit. It has analog V and A meters, and provides adjustable 0-15V and 0-3A, offering restricted current protection. Thus far I have used it to power a 32E with broken battery terminals (2.5 v); a 11c (4.5v); and an old 9V Novus Mathemetician. It works GREAT. I highly recommend it. One of these units will power virtually any calculator indefinitely. Take care. Wayne.

[ Return to Index | Top of Index ]

Go back to the main exhibit hall