HP 35S keyboard problem



#28

I'm interested to know what the final consensus is on the HP 35S keyboard. Is it prone to mechanical defects that causes missed keystrokes or is the problem caused by keys being pressed too quickly, or off centre?


#29

Personally I have never had a problem with the 35S keyboard. From what I have read it appears to me that the problems seen are a software rather than hardware issue caused by rapid input. I have never *tried* to break it with rapid keystroke sequences and in normal use I have not encountered any issues. People that are actually having problems can tell you more.

Cheers

#30

Cursor-up key is more or less gone on my sample, but it seems I'm the only user in this forum to have such a problem.

I guess care needs to be taken when pressing the 35s' cursor keys as they are not hinged. Very bad design decision, imo.

#31

I think it is a problem with the keys being pressed too quickly. The 50g has a similar problem that can be fixed by changing the keytime setting to a smaller value.

If I press a key as fast as possible, the 35s will only register 8 of the 10 presses.

#32

I don'tthink it can be a software problem. For me it only happens with the RCL key. I have to press it very firmly to make sure it registers.


#33

I believe there is defintely a software issue with the keys. At first when I was learning the 35s I did not have a problem, but as I got more used to it I began getting missed key strokes. This becomes partcularly apparent when typing in programmes or working through a formula I am familiar with. Quite often the 'busy' annunciator will come on too. It was particularly annoying when I was typing in Pauli's game (see article 1035).

-B

#34

I bought my 35s in 2009 (CNA 72701623) and use it daily, although not heavily. Missed keystrokes due to quick typing is not new for me, but 2 months ago the "+" key started to fail. When I press it on the inferior left side, it fails 5 to 6 times in 10. I feel the click, but it does not conclude the operation. There is another mechanical problem with the "SIGMA+" key: it is unbalanced and inclined to the left, though it continues working with no fail.

Silvio


#35

It sounds like there's definitely doubt as to the durability of the keys on the 35S. That's enough to put me off buying one at the retail price. I'd thought of getting one to sit next to my vintage HP35 as a comparison, but that would be like parking a Lada next to a Ferrari.
It seems that in the last few years the reduction in the MSRP of certain makes of calculator has been matched by a reduction in MTBF. I'd call this the China Syndrome. If the much-anticipated HP15C LE does actually materialize, I doubt that the construction quality will prove to be much better, and perhaps LE may end up standing for Lousy Engineering. I hope I'm wrong.

Nick


#36

Quote:
If the much-anticipated HP15C LE does actually materialize, I doubt that the construction quality will prove to be much better, and perhaps LE may end up standing for Lousy Engineering.

I doubt it. It should be physically identical to the HP-12C+, which AFAIK is quite good and solid, and nothing like the junky and flimsy construction of the HP-35s. I would expect it's quality to be closer to that of the HP-50g, which is quite good and much better than the HP-35s.


#37

Michael,

Your bias is showing <g>. While the 35S does have faults I don't find the construction to be one of them.

The problem with this approach to inquiring about quality is you have no idea whether those NOT responding are having problems or not. The number of responses to the original query does not really indicate a hardware problem since you WILL get failures--that's simply the nature of the beast--the question becomes what is the failure rate. Anecdotal evidence of the kind being gathered here does not provide an answer. You need a far larger sample to unequivocally answer that question. I even have a 41 that after years and years of dutifully registering every keystroke, every time, suddenly began missing the '5' key. Does that make the 41 "junk"? Hardly.

I am not saying that there is NOT a hardware problem, just that it is very hard to be sure there is one with the answers received so far. And yes, I have seen other complaints about missed keys, but, again, is that hardware or software, I don't know. Even Pauli (I believe it was) said that they would have used the 35S as a basis for their work had it been flashable.

Cheers,

Marwan


#38

The issue here is one of durability and consistent quality. If a calculator like your HP-41C works for decades before finally experiencing a failure, it is much better made than a calculator like the HP-35s that fails after only a few years of use. I have a heavily used HP-32SII whose keyboard buttons work as well as the day it was removed from it's box. Not to mention an even older and more heavily used HP-15C. When was the last time you ever heard anyone complain about the quality and durability of a Pioneer or Voyager ? Also, just because you don't hear a lot of complaints, doesn't mean there aren't a lot of people that simply haven't reported them, because most users don't even frequent this forum.


#39

Quote:
Also, just because you don't hear a lot of complaints, doesn't mean there aren't a lot of people that simply haven't reported them, because most users don't even frequent this forum.

And there's the rub. Until we have a nice large sample we can't tell how bad (or good) the situation is.

And yes, I do agree with your statements about the quality of the machines you mention. The point I was getting at was that sample size is important even when discussing durability.

My first 41C had a memory problem (failure) very early on in it's life. It was fixed under warranty and never had a problem after that. However, I could have assumed that because it failed soon after purchase that the durability of the machine was questionable. That would have been a bad assumption based on too small a sample size.

I am not arguing that the 35S has great hardware--although I have not had a problem with mine. What I am saying is that given the sample size I simply don't know. Remember, the squeaky wheel gets the grease, in this case that might mean that those most dissatisfied with their machines are the most vocal.

Another example is my 50g. I have two and one of them has a funny sound to the click on one of the keys. Still responds fine but it does not sound right to me. I could jump to the conclusion that quality control is poor since 1/2 does not meet my version of spec. I might even be right but again the sample size does not warrant a conclusion one way or another.

I would love to see a true random sample of say 1000 units tested to verify hardware functionality. Now if the sample set was truly random that would probably be a valid basis on which to make a conclusion as to the quality of the machine's hardware. Unfortunately that isn't going to happen on this forum.

I would add that I probably agree that the quality of the 35S does not match that of earlier machines such as the 32Sii but I would argue that that does not make it "junk" in it's own right.

I would also say that my one really big complaint about the 35S is that they never bothered to fix anything. Is this the only mid-range/high-end HP that has never had a ROM update?


Edited: 10 Aug 2011, 9:49 p.m.

#40

I was worried after reading all this type of bosh in the forum, but found it all to be not true and greatly exaggerated.

#41

I really don't know how fast you guys can type, but personally, the fastest I can manage is about 12 digits per second on my 35s (I can enter 123123123123 in just under a second) and they all register consistently and reliably.
I don't know if I have a faster than average calculator, or if I just type too slow for you, but does anybody *really* need the capability to type more than 12 digits per second?

Cristian


#42

You have to include operations too. I suggest you to try a chain of one-digit calculations in RPN mode. The longest, the best. And try to use two fingers (both thumbs, for example.


#43

Typing with two fingers may be a special problem if the keyboard does not have two key roll-over. This is a software thing to implement or not. Without it, only one key is detected and a second key pressed at the same time is ignored. With two key roll-over properly implemented the second key is accepted even while the first one is still down. This is what I've done in WP 34S. (At least I tried to implement it correctly.) To get multi key roll-over (three or more) you need diodes in the key matrix which would increase the production costs and is only really useful on large keyboards, not on a calculator.


#44

I made a few quick tests, and a second key pressed while the first one is still down is correctly detected in every case I've tried.
It also works for many three-key combination - i.e. press 1 and hold, press 2 and hold, press 3 - the display shows 123 and it doesn't change regardless of how you release the keys. It also works for 789, 147, 258, and many more. Two I've found that don't work are 456 and 753 - for these, only the first two keys pressed register.
I suppose it has to do with the row/column key encoding?

I've also noticed that digits are entered as soon as the key is *pressed*, while operations are performed when the key is *released*. Do you know why?

Cristian


#45

Cristian, thanks for testing. It looks like the keyboard driver is not too bad.

Executing a function on key release may be associated with SST/BST showing the step and executing it when the key is released or something to that avail. Maybe it's a provision for NULL (like on the 41C) without ever being fully implemented.

#46

I just tried pressing 2+2+ etc as fast as I could 66 times and nothing was missed.

My serial is CNA 03701109 if that helps with production dates.


#47

[nonsense deleted]

I can't beleive production went on in 2010 w/o correcting bugs ...


Edited: 11 Aug 2011, 7:03 a.m.


#48

Its probably the same production run.

Then again, mine doesnt have any bugs so maybe they fixed them?

Happy to try anything you can reproduce. Can also post youtube videos of the bugs not happening, as opposed to user error being repeated as heresay.

Daniel.


#49

Quote:
Then again, mine doesnt have any bugs so maybe they fixed them?
Missed that. You worked through the whole buglist and found all points to be fixed? That would be great.
#50

You did everything in the Bug List (article 735)? Yes, please post youtube videos of any of those that you can verify as fixed (e.g. no.1 would not be easy to verify). Perhaps HP did another "silent roll-out" (as with the 12C+, but this time so silent they did not even let the forum know?)


#51

Not everything, i.e I dont know what (1) is. I expect much is just user error and heresay.

Looked at the so called Cos and Tan "bugs". Its algorithm is less accurate than the Pioneers for values 89.99999 to 89.9999999, at 89.99999999 its the same. For values not near the extreme its more accurate than the 42s etc. So do we say the 42s is bugged? Of course not.

4 and 5 are irrelevant

8 Once again, minutia

9. Works fine on my 35s

10. Is purely opinion

11. Works fine on mine

15. Works fine on mine, though I used RPN as I dont know how they put an eqn one line. People were complaining that i had an ALG mode, so if kts an obscure ALG mode bug, who cares?

16. Is just people programming bad, a continuous loop will overflow or crash most computers.

After finding all this stuff was bs I coulnt be bothered with the rest. Dont believe everything written on the internet, try for yourself!


Edited: 11 Aug 2011, 10:01 a.m.


#52

Quote:
15. Works fine on mine, though I used RPN as I dont know how they put an eqn one line. People were complaining that i had an ALG mode, so if kts an obscure ALG mode bug, who cares?


To write an equation on a program line, just press [EQN] and copy the equation given in the example. Note that the first minus sign is a unary minus, thus has to be keyed in with [+/-]

My 35s reproduces this bug. That is really weird and gives me the creeps. Apart from that, I love this machine for everyday math!

#53

Quote:
16. Is just people programming bad, a continuous loop will overflow or crash most computers.

Here I disagree. The last computer I had that crashed if the user entered a continuous loop, was the Commodore 64, and that only if you programmed it in machine language.
With any modern programmable device, a continuous loop can crash the program, but not the whole operating system.
What would you say if in your computer, you wrote a small program with a loop, and that, besides locking the whole computer, also wiped all your personal data and programs? :) There should always be a way to stop execution of a user program, on a modern programmable machine.

#54

So, with all the bugs being irrelevant or b/s to you, the 35s is for you. Enjoy it :-).


Possibly Related Threads…
Thread Author Replies Views Last Post
  32Sii keyboard problem Peter Klein 4 1,828 07-25-2012, 01:53 AM
Last Post: Peter Klein
  HP-29c keyboard dead 2nd row problem Xavier A. (Brazil) 6 2,376 07-17-2012, 07:09 AM
Last Post: Xavier A. (Brazil)
  HP 21 keyboard problem Yvan Van Hoof 2 1,312 11-23-2010, 12:21 PM
Last Post: Jeff O.
  HP 35s - Choosing Initial Guesses for SOLVE problem Pablo P (Spain) 3 1,303 10-03-2010, 08:35 PM
Last Post: Karl Schneider
  HP 35s keyboard weirdness Hiawatha Wheelwright 6 2,066 01-12-2010, 05:02 AM
Last Post: Bart (UK)
  My HP 35s keyboard improvements kiril t 25 6,353 01-07-2010, 11:52 PM
Last Post: Mohammed Hadi
  HP 35s - Problem with conversion from polar to rectangular coordinates. Samir Kopas 1 864 08-31-2008, 10:58 PM
Last Post: V-PN
  An idea on HP-35S Keyboard Problems Palmer O. Hanson, Jr. 4 1,511 04-17-2008, 03:28 PM
Last Post: Tom Mathes
  hp 35s keyboard issue jdgc (spain) 3 1,368 03-28-2008, 04:18 PM
Last Post: Ivan Latorre
  HP 35S replaced (keyboard problems) Tom Mathes 15 4,478 03-18-2008, 11:05 AM
Last Post: Tom Mathes

Forum Jump: