The Museum of HP Calculators

HP Forum Archive 19

[ Return to Index | Top of Index ]

Missing keystrokes
Message #1 Posted by Quan on 13 June 2009, 3:18 p.m.

Had it ever been determined the problem with missing keystrokes was hardware or firmware related in the current versions of 17bii+ silver and 35s? Maybe both?

I tried to read through the numerous posts here and other places, and what I got from these were that some have problem and some don't.

I did the simple test suggested in other posted by adding up single digit numbers consecutively, keying them in fast enough to cause the problem to occur on both my 17bii+ silver (CNA751...) and the 35s (CNA833...).

At first I thought the keyboards were the cause since I read so many complaints about their older siblings, but then I tried to punch the number keys only to see if they are registered. Neither calcs fail to record the keystrokes.

I performed these tests with both hands holding the calcs and using my thumbs for quickly entering the numbers and the add key.

The test seemed to fail only when doing (+ - * /). My guess is that they would fail for other function keys as well but since their locations are further away from the number keys, I didn't run the test on them. Somewhere in the sequence of number-add-number, one of these keys was missed or not registered.

I know at the end, as far as the users are concern, the problem is still summarized as annoying missing keystrokes, but I wonder if they could be fixed with a firmware update for the next production run. Or the problem is not that simple.

BTW, the tactile is quite good on the 35s, though not the same as the older 32sii/48sx I have. Also, none of the older brothers have missing keystrokes problem, no matter how quickly I hit the keys. They worked every time.

      
Re: Missing keystrokes
Message #2 Posted by Mark Edmonds on 13 June 2009, 7:39 p.m.,
in response to message #1 by Quan

Key weight is just about OK on the 35s but I don't like the key profile in relation to the pivot point, nor the feel of the matt surface. Given how absolutely perfect the keys on the Pioneers and 48 series are, it is such a shame that such an important part of the design has degraded so much.

I haven't read much about the verdict on the new 20b keys - are they similar to the 35s?

Mark

            
Re: Missing keystrokes
Message #3 Posted by Thomas Chrapkiewicz on 14 June 2009, 12:38 p.m.,
in response to message #2 by Mark Edmonds

I've recently received a 20b and am VERY let down. It is not even close to the 35s. Further, keypresses result in a scraping/scuffing sound - not certain if the key is rubbing on the edge of the hole or what exactly is causing this. Pressure and throw are also unacceptable. I was expecting the feel to be closer to the 35s.

My 'point of reference' by the way goes back to the 45/65 through the 48 series - all of which I have found to be ideal. I use a 35s and 48 daily.

I certainly hope that newer machines are closer to the 48 or at the very least the 35s.

Regards, TomC

      
Re: Missing keystrokes
Message #4 Posted by John B. Smitherman on 14 June 2009, 1:17 p.m.,
in response to message #1 by Quan

Hi Quan. I assume that you are referring to the test in Joel Koltner's post at link

Using Joel's test I'm unable to get the 35s to calculate the correct answer consistently regardless of whether I'm using +, -, x or /. However, if I only enter numbers as fast as I can it makes no mistakes.

I've noticed that when I use the +, -, x or / in a sequence that the "B" symbol comes on in the display which means that the calculator is "Busy". I'm guessing that our missed keystrokes are due to the calculator being busy after certain function keys are used.

My conclusion is that HP should be able to fix this problem with software improvements.

Regards,

John

            
Re: Missing keystrokes
Message #5 Posted by Don Shepherd on 14 June 2009, 1:40 p.m.,
in response to message #4 by John B. Smitherman

Quote:
the "B" symbol comes on in the display which means that the calculator is "Busy".

John, it has always been obvious that there is a substantial speed difference between the 35s and the 33s; the 33s being much faster. While I have never seen a definitive reason for this speed difference, we do know that the 35s allows several different formats of data in its registers, such as vectors, complex numbers, etc. Since it takes some processing time for it to figure out what exactly it has in the registers, that always seemed to me to be a possible reason for its slowness relative to the 33s. And this could be a reason for the "B" annunciator.

                  
Re: Missing keystrokes
Message #6 Posted by Mark Edmonds on 14 June 2009, 3:12 p.m.,
in response to message #5 by Don Shepherd

It is almost as if the 35s doesn't implement a keyboard buffer or an interupt driven handler so it can only detect keystrokes when it is in the main-prompt idle state. Having said that, you can break out of programs quickly but maybe the On key is a special exception? Lots of armchair guessing here!

One thing for certain:

In the Pioneers and 48s, you can pound away on the keyboard and all the keystrokes get buffered so you can watch them being processed as the calculator catches up.

On the 35S, fast keystrokes whilst the "B" (broken?) indicator is on get missed as well as the extra problem of physical contact failure causing intermittent keystroke loss. Not a happy combination for not a happy calculator.

Mark

                        
Re: Missing keystrokes
Message #7 Posted by Quan on 14 June 2009, 10:36 p.m.,
in response to message #6 by Mark Edmonds

Quote:
One thing for certain:

In the Pioneers and 48s, you can pound away on the keyboard and all the keystrokes get buffered so you can watch them being processed as the calculator catches up.

. . .

Mark


Hi Mark!

This was exactly what happen when I ran the test on the old 17bii. On the 42s the results were displayed as soon as the keys were pressed.

The 17bii and the 35s are still very usable. In their case, slow is better.

            
Re: Missing keystrokes
Message #8 Posted by Quan on 14 June 2009, 10:29 p.m.,
in response to message #4 by John B. Smitherman

Quote:
Hi Quan. I assume that you are referring to the test in Joel Koltner's post at link


Hi John!

Yes, that was the test I ran because it was so simple.

Quote:

Using Joel's test I'm unable to get the 35s to calculate the correct answer consistently regardless of whether I'm using +, -, x or /. However, if I only enter numbers as fast as I can it makes no mistakes.


Exactly my experience.

Quote:

I've noticed that when I use the +, -, x or / in a sequence that the "B" symbol comes on in the display which means that the calculator is "Busy". I'm guessing that our missed keystrokes are due to the calculator being busy after certain function keys are used.


You said it better than I did.

Quote:

My conclusion is that HP should be able to fix this problem with software improvements.

Regards,

John


That's what I hope as well. But all we can really do is hurry up and wait. :)

Regards,

Quan


[ Return to Index | Top of Index ]

Go back to the main exhibit hall