HP 35s missing keystrokes, some more
Joel Koltner
enter the sequence...
1 enter 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +
...and obtain the correct answer of 45 if I enter the sequence as rapidly as I
can move my fingers -- sooner or later one of the keystrokes is missed so the
computation becomes corrupted. (I'm using one thumb for the numbers here, and
the other thumb for the "+" button.)
Anyone have a 35s where this consistently works?
I can do this on my 50g without problems (after having adjusted keytime...).
I also have no problems with a $20 Sharp EL-W516B that does this without
incidence.
While this is obviously a contrived example, I encounter similar errors often
enough in real-world usage that I'm planning on ditching my 35s machines and
just sticking with the 50g.
---Joel
Eric Rechlin
>I have two HP 35s calculators to play with, and on neither one can I
>usually enter the sequence...
>
> 1 enter 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +
>
> ...and obtain the correct answer of 45 if I enter the sequence as rapidly
> as I can move my fingers
Same problem here. It drives me nuts, and I run into this problem
regularly. I leave my 35s on my desk at work and frequently use it for
simple math while programming.
I have to use my 50g when I care about having a reliable keyboard. When I
use the 35s, I have to remember to type the keys slowly, or I end up with
the same problems with inaccurate results that you are setting.
It seems to me that the CPU in the 35s sometimes starts "thinking" when you
type a bunch of keys in succession and misses any keystrokes during that
period. It doesn't seem to be a result of the keys not being pressed hard
enough, and it also isn't because I am pressing two keys at the exact same
time (while that also would cause the keystrokes to be lost, that isn't the
case here because I can reproduce this while using only a single finger).
It's sad that this occurs, because the 35s has the best-feeling keyboard (to
my fingers) of any current HP calculator, but it's too unreliable to be
trusted.
Regards,
Eric Rechlin
Joel Koltner
"Eric Rechlin" <er...@hpcalc.org> wrote in message
news:78j696F...@mid.individual.net...
> It's sad that this occurs, because the 35s has the best-feeling keyboard (to
> my fingers) of any current HP calculator, but it's too unreliable to be
> trusted.
Agreed -- the mechanical design of the 35s keyboard is very good. I think
it's sad that HP largely perfected keyboards over 35 years ago now (I can't
imagine the original HP 35 had this problem!) and somehow they've regressed to
being worse than cheap $20 calculators in this particular area. :-(
---Joel
IMQ
> Thanks for the confirmation, Eric.
>
The missing keystrokes also occur in the silver 17bii+.
Curious, I tried this exercise on both the older 17bii and the 42s.
Both work perfectly every time!!!
The 42s instantly showed the result of each keystroke.
The 17bii, however, appeared to be slower than the 42s and or my fast
finger. It quickly flashed each number, then the result of each
addition before showing the final result, after I already finished
entering the last number. In short, it took a second or so after I
pressed number 9 for the result to show.
Maybe HP needs to update the ROM to account for keystrokes buffer.
BTW, how do you like the Sharp EL-W516B?
Michael & Maria Lopez
but no problems with it so far. We can press the keys as quickly as our
abilities will allow without missing a number.
Whilst I normally use a 50G, the 35S is a great feeling calculator & really
reminds me of the 2 HP41's I have.
Maybe they have improved the latest batch or just lucky so far.
Michael
"IMQ" <gte...@gmail.com> wrote in message
news:dd56f255-db90-4f87...@37g2000yqp.googlegroups.com...
Tom Lake
"Joel Koltner" <zapwireD...@yahoo.com> wrote in message
news:KdXUl.72015$6p1....@en-nntp-02.dc1.easynews.com...
> I have two HP 35s calculators to play with, and on neither one can I
> usually enter the sequence...
>
> 1 enter 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 +
>
> ...and obtain the correct answer of 45 if I enter the sequence as rapidly
> as I can move my fingers -- sooner or later one of the keystrokes is
> missed so the computation becomes corrupted. (I'm using one thumb for the
> numbers here, and the other thumb for the "+" button.)
I get an answer of 35 consistently on a very lightly used calculator
(no discernable key wear)
Maybe it's the New Math - HP Style.
Tom Lake
sc_u...@hotmail.com
>
> I get an answer of 35 consistently on a very lightly used calculator
> (no discernable key wear)
>
> Maybe it's the New Math - HP Style.
>
> Tom Lake
I wonder which two numbers the calculator is missing...only four
possibilities
S.C.
IMQ
wrote:
> I purchased a HP35S for my daughter about a month ago so it is quite new,
> but no problems with it so far. We can press the keys as quickly as our
> abilities will allow without missing a number.
>
Maybe HP listens to the complaints about the keyboards and quietly
updates or correct the problem in the new production?
Glad to hear that you have no problems. I am hoping that is the sign
that HP continues making improvements based on feedback here and,
perhaps, the MoHPC or HP business forums.
The HP35s is still quite useable provided that the users are not too
quick with keystrokes. It certainly has the look and feel of a tool
made for professionals. TI/Casio/Sharp are no match in that
department.
BTW, what grade is your daughter in?
One of my nieces will start 9th grade in the Fall. I don't know what
kind of calculator she will be required/recommended to have, but I
kinda suspect that it will be TI of some sort.
Just curious.
sc_u...@hotmail.com
> Maybe HP listens to the complaints about the keyboards and quietly
> updates or correct the problem in the new production?
>
> Glad to hear that you have no problems. I am hoping that is the sign
> that HP continues making improvements based on feedback here and,
> perhaps, the MoHPC or HP business forums.
>
I'm still waiting for a ROM revision to fix some 2-year-old bugs:
http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/articles.cgi?read=735
The HP 35s overview on HP's website contains the following quote:
"Scientific projects require accuracy, functionality and dependability
for success. The HP 35s scientific programmable calculator delivers
them—and more—"
I think it fails in the accuracy department (even as the 48/49 series
is amazingly accurate).
S.C.
Joel Koltner
I like it a lot. For $20, it's quite impressive!
Amazon has a bunch of reviews that do a good job comparing it to the main
competition, which is the Casio FX-115ES; I generally agree with them:
Generally the Sharp is better (as it should be, as it's several years newer
than the Casio), although the Casio has a couple of significant features that
the Sharp doesn't. Compared to an HP 35s:
-- In generally it's nowhere near as powerful owing to not being programmable,
of course.
-- It has solvers built in for simultaneous equations up to 4x4, quadratic and
cubic equations, and something like 8 different regression types. The 35s has
solvers for 2x2 and 3x3 built-in and a few regressions, but everything else
you have to program yourself. (In fact, one of the first programs I wrote for
the 35s was a quadratic solver.)
-- The 35s computes integrals better; I was a bit incredulous that the Sharp
literally just picks a number of evenly spaced points (you can specify how
many -- 100 is the default) and multiplies f(x)*dx and sums -- the simplest
possible approximation of an integral. Truth be told, though, I've never used
integrals in a real program on the 35s... although I have a couple of times on
the 50g.
-- The solver on the 35s is much more flexible: The Sharp solves an
expression, f(x), to be equal to zero. The unknown must be the variable 'x',
and it's up to you to change an equation in the form of f(x)=k (or whatever)
to the expressoin f(x)-k. Even the Casio is more flexible here...
-- There are standalone buttons (D1-D4) that you can assign any function you
feel like to. This nicely solves the problem of, "I use the pi key a lot and
it sucks that it's a shifted key!" -- on the Sharp you'd just assign pi to,
e.g., D1 and now it's only one keystroke rather than two.
-- The 4 directional arrows double as four "function" memories where you can
save and recall algebraic expressions. Very similar to the 35s's equation
list, except that you're limited to 4 and the Sharp actually just inserts
whatever's in the memory into the current expression, so if you have a snippet
of an equation you use again and again, it can save some typing. I stuck the
parallel impedance equation into one memory and the equations for converting
to/from dB into two others. (You can use the "ANS" variable to have the
equation refer to the last computed result.)
-- Using complex numbers on the Sharp is a different operating "mode" (instead
of complex numbers being just a different object type as they are on the
35s/50g), and the operations you can perform on them are quite limited: Pretty
much just addition, subtraction, multiplication, and division, and keys such
as x^2 and x^3 that effectively do the same thing. But none of the regular
"scientific" functions like square root or the transcendental functions
operate with complex arguments. You also loses the main six memories (A-F) in
complex mode... indeed, you're left with only one regular memory (M), although
you could use the four "function" memories in a pinch.
-- Base-n conversion is done much better than on the 35s... pressing
hex/bin/oct/pent (base 5!) really just changes the display mode of the
calculator, at which point the keyboard is redefined so that memory keys A-F
enter the digits A-F and the various scientific function keys become
AND/OR/XOR/etc. In other words, no menu selection of logical operators (which
even the Casio FX115ES requires); it all works quickly and intuitively, and as
soon as you hit "dec" the keyboard reverts back to all the regular key
definitions.
-- Sharp paid attention to how the display is used; you never have a situation
where you need to scroll just to see a numeric result, as the 35s requires at
times. (Complex results are displayed on two lines...) They also have a key
that rotates through up to three display represenations: Regular floating
point, a "mixed" mode with radicals and proper fractions (as appropriate), and
a mixed mode where the fractions are improper. This is quite handy in that it
encourages you to keep the machine in "write view" mode, where the mini-CAS
system is active and fractions and radicals are kept as such when possible.
(On the Casio it was a shifted function to get from mixed mode to a decimal,
which caused most people to just use "line" mode and forget about the
mini-CAS.)
-- As the above might imply, fractions are handled nicely, although there
isn't the "/c" functionality that the 35s has; the Sharp always reduces to the
least common denominator.
-- An interesting/odd feature is that you can program two lines of text to be
displayed when you hit the "off" button... this is suggested to be, e.g., your
name and phone number in case the calculator is stolen, and that's what I
programmed, but it occurs to me that a sticker on the back is probably even
better for this functionality. :-)
-- The keyboard layout is pretty good -- STO and RCL get their own keys! I
would have ditched giving hyp (as in hyperbolic sin/cos/tan) its own key
(seriously, how many people use those functions on anything approaching a
regular basis?) and used it to make square root or 1/x unshifted.
-- It has a few more conversion functions and probably a few more constants
than the 35s, but I've never needed any the 35s didn't have. The Sharp does
have a nice feature in that if you press a letter in the list of constants or
conversions, it'll immediately move you to the items that start with that
letter.
-- The keyboard is a bit jigglier than the 35s's -- definitely a notch lower
in build quality -- but I found it perfectly usable.
-- And of course the Sharp doesn't have RPN, which is disappointing, but
obviously not unexpected when other than HP almost no machine does.
For a "throwaround" calculator, it's easily worth the $20. For the
educational market, it'd certainly be OK through middle school and much of
high school; kids who were thinking of engineering paths would probably
already want something a bit programmable in their later years, at which point
a 50g (or TI, if you insist) would be best.
Back in the late '80s, I had a Casio FX-4000p programmable calculator. While
the programming was quite limited (only 10 programs and some hundred of steps,
although it did have branching and all), if I had to choose between the 4000p
and the EL-W516B, I'd take the Sharp... although I'd whine a bit that the
programming functionality was absent!
---Joel
Joel Koltner
news:1befeb4c-a256-422c...@p4g2000vba.googlegroups.com...
"I'm still waiting for a ROM revision to fix some 2-year-old bugs:"
Same here -- in fact #15 on that list is one I discovered, and it was from a
real program I'd written for work too, not just an obscure programming
exercise looking to test "corner cases" or somesuch.
I'm disappointed that HP hasn't given any indication they intend to fix these
after so long. While some might take awhile to track down (e.g., the
"sometimes it works, sometimes it doesn't," like #17) and some need attention
from someone well-versed in numerical methods (e.g., #2/#3), the vast majority
of those bugs look like easy fixes that any programmer could crank out in no
more than a week or so worth of work. :-(
Truly scary bugs like #15 and #16 combined with an unreliable keyboard make
the 35s a calculator I can't recommend to most people. For "serious" work the
50g still takes the top spot, and for a "throwaround" machine, the $20 Sharp,
Casio, and TI "textbook display" calculators are a very good value.
---Joel
IMQ
That is an impressive feedback!
I, too, like both the Casio/Sharp. At least for their intended
purposes.
IMQ
> <sc_use...@hotmail.com> wrote in message
The 35s has been out more than long enough. HP should have fix these
issues. At least some of them.
For serious work, I don't want to wonder if I missed the keystrokes
because the result look strange.
--IMQ
Steve Keeley
"Joel Koltner" <zapwireD...@yahoo.com> wrote in message
news:KdXUl.72015$6p1....@en-nntp-02.dc1.easynews.com...
> (I'm using one thumb for the numbers here, and the other thumb for the "+"
> button.)
>
> Anyone have a 35s where this consistently works?
I tried this on my 35S using the "two thumbs" method 5 times and got 45 each
time. Perhaps my thumbs and/or reaction time aren't as fast as yours,
although I was pressing the keys a LOT faster than I normally do when
performing calculations.
I have yet to encounter a missing keystroke with the 35S in regular use.
--
________________
Steve Keeley
San Diego, CA, U.S.
John H Meyers
but that was a very old model, and had a lot of work to do
on every keystroke.
One of my early Casio LCD scientifics had an opposite problem --
it calculated so fast that sometimes one wasn't sure
whether or not a "function" button had actually been responded to.
Eventually, manufacturers (including HP) got the idea
that one needs to cause a definite "blink" or other indication
(e.g. an "hourglass" with a minimum "on" time,
or even a "beep" or "keyboard click"),
to make sure that the user is aware that a response has occurred.
Perhaps the "fashion" has now swung too far,
to calculators which once again can't even keep up with keypresses?
Or is it a failure to implement "key rollover,"
in which "fast pressing" may inadvertently cause rollover
(two or more keys down simultaneously), which then gets ignored?
If someone said that they could cause this even by rapidly pressing
the same key, then perhaps not -- unless "spongy keys"
also simply do not release fast enough :)
Or is it so much "debouncing" (a distinct issue with 49G+ ?)
that it considers even normal repeats to be "bounces"?
Whatever the case, no Casio or Sharp product of mine
has ever had any such problem, nor had any keyboard failure at all --
what trade secrets do they know?
[r->] [OFF]
Michael & Maria Lopez
science degree but has little experience with RPN calculators. She claims I
bought the HP 35S to indoctrinate her into the World of HP calculators but
off course I deny all ;).
Michael
"IMQ" <gte...@gmail.com> wrote in message
news:05bfafa2-b866-4a73...@h2g2000yqg.googlegroups.com...
Joel Koltner
perhaps I'll have to try to find a newer one to play with.
---Joel