8 Jan
2014
8 Jan
'14
11:34 a.m.
On Jan 8, 2014, at 12:00 PM, ARD wrote:
(Quote slightly modified)
Point taken; the TI gets very close to 2.00 for my test and should show 2.00, but for the
near-zero result, it should show the scientific-notation value it gets, not 0.00. Now I?m
curious; I?ll have to try to get my TI running long enough to try this. Hopefully Pi and
Sin are two of the keys with fewer bounce problems than the others?.and I really don?t
know what I?ll do for a battery pack. Sigh. At least the cells look like standard
Ni-Cads.
And no, although I did think about trying to write an arbitrary-precision set of
subroutines for the TI using multiple registers to represent a single number, I never did
it and 100 digits of Pi, or 100 digits of Sin(x) function, would have been beyond my
math/programming skills anyway.
- Mark
THe HP41 was a much mroe expandable ssytem than the
TI59. Or at least,
IU've never seen an TI59 controlling a benchtop of HPIB instruments,
savign the readiungs to floppy disk or tape and then trasnmittign the
logged data over an RS232 link. I do that with my HP41 all the time.
I don;t think the TI ever had a realtime clock, did it?
And of course the HP had an alphanumeric dispaly.
All true to the best of my knowledge; expandability and extensibility (sic?) was one area
where the HP was leaps and bounds ahead of the TI. I was insufficiently precise; I was
looking at total storage capacity and program execution speed as my main ?performance?
criteria at the time.
FWIW, the TI would print alpha characters on its PC-100 printer/cradle, if you had one of
those. But it was pretty kludgey, and I don?t think there?s any way to get alpha out of
the handheld part of the system. Well, OK, get the result 07734 and hold the calculator
upside down. :-)
2) I quibble
with Tony?s recommendation to reject a machine that says
Sin(Pi) = 0. I?m pretty sure the TI says that; the quicker but
essentially equivalent test I always used to taunt my HP-41-equipped
friend was (Sqrt(2))^2. The TI said 2, the HP said 1.99999? I claim
Actually, that is something rather different.
If you take a decimal approximation of SQRT(2), round it after, say, 13
digits, then square it and round the reult to 10 digits (or whatever),
you will get 2. But if you take any finite number of dgits of pi,
calculate the sin, the answer is not zero. It's of the order of
10^-(number of digits). So even if you take 13 digits of pi, the SIN
should not display as zero. I suppose you could argue that if you took
over 100 digits, the result would be zero (since a number of the order of
10^-100 will underflow to 0), but I doubt that the TI uses 100 digits
anywhere. 
9 Jan
9 Jan
11:46 a.m.
All true to the best of my knowledge; expandability and extensibility
(sic?) was one area where the HP was leaps and bounds ahead of the TI. I
HP and TI kept on leaping past each other here. There wa a printer cradle
for TI machines long before HP had any expandability at all. The official
way to get a printout of an HP67 program was to buy an HP97 calculator,
which had a built-in printer.
The TI59 had the ability to use the printer, of course,and also plug-in
CROMs (the'Solid State Software'. But you could not add a card reader to
a TI58 to make it into a TI59 (and I always thought it was a pity there
was no TI59-C). But HP then came out wit hthe hP41, which had very little
built in, but which could tke a printer, card reeader, ROM modules, RAM
modules, etc. Later they added the time module (real time clock and
stopwatch) and HPIL.
was insufficiently precise; I was looking at total
storage capacity and
program execution speed as my main ?p rformance? criteria at the te.
HP were always a bit mean with RAM...
Of coure the HP41 could _later_ be expanded with the extended memory
modules, but those were not avaialble when the machine first came out. So
you would probably not have considered those.
FWIW, the TI would print alpha characters on its
PC-100
printer/cradle, if you had one of those. But it was pretty kludgey, and
Sure.
I don?t think there?s any way to get alpha out of the
handhelpart of
the system. Well, OK, get the result 07734 and hold the calculator
upside down. :-)
Indeed. I am not sure what non-normalised numberes displayed as on
the TI59. On the HP67 (also a machine with a 7 segment numerical dispaly)
you could get a few alpha characters that way. Incidentalyl, never try to
print non-normalised numbers on an HP97. The print routine gets confused
ans leaves the pritnhead turend on for too long. The result is a
burnt-out head.
But yes, these machines were numerical only.
Point taken; the TI gets very close to 2.00 for my
test and should
show 2.00, but for the near-zero result, it should show the
YEs, 1.4142...^2 is close to 2 (wherever you truncate/round the decimal
for sqrt(2) so roundign it to 2 is nto a major error). In much the same
way I do not complain that COS(3.1415926..) (inradians mode) returns -1.
The cosine of that decimal, truncated or rounded to a given number of
digits, is essentially -1 to machine accuracy.
But SIN(3.14159...) is close to 0, but not 0. It is a non-zro number htat
hte machine could display. But it doesn;t
scientific-notation value it gets, not 0.00. Now I?m
curious; I?ll have
to try to get my TI running long enough to try this. Hopefully Pi and
Every non-HP machine I've tried (TI, Sharp, Casio,...) gives 0.
Sin are two of the keys with fewer bounce problems
than the others?.and
I really don?t know what I?ll do for a battery pack. Sigh. At least the
cells look like standard Ni-Cads.
The TI59 uses a BP1A battery pack which is 3 AA NiCd's in a plastic
housing. No other internal electronics.
If you have a PC100 printer cradle, it will run from that without a good
NiCd pack anywhere. Or you can runn it from a 3.75V bench supply. There
is a TI59 service manual on the web (Google foudn it for me this morning)
which gives the polarity, etc.
-tony
4005
days inactive
4006
days old
1 comments
2 participants
participants (2)
-
ard＠p850ug1.demon.co.uk -
mtapley＠swri.edu