Weird Keyboard behavior: I got "L" if I press any key long enough

[manterola]

I think the title is a good description of the problem. This is a 600XL with Stackpole membrane type keyboard. For  example, if you press "r" you get:

RRRRRRRRRRRRRRRRR;LLLLLLLLLLL

There is always some random character just before the Ls. At first, I thought it was only the delete key that was having this problem, then the arrow down key, while selecting something in the fujinet configurator. 

Then I realized it happens with any key. if you stop holding the key, it goes back to normal/correct behavior, so I can type, select using the arrows, etc.

My first instinct is to replace the 4051s and see what happen.... but first before ordering anything I would like to know your opinion.

I am also intrigued by the "L"... why the L?  I checked the keyboard matrix picture and the atascii and internal code for some clues, I really have no clue.

 

 

Edited December 9, 2021 by manterola

[+bf2k+]

You said 600XL... is this happening in BASIC?

 

Does it stop if you unplug the Fujinet?

 

[manterola]

9 minutes ago, bf2k+ said:

You said 600XL... is this happening in BASIC?

 

Does it stop if you unplug the Fujinet?

 

It happens in any program, so far tested on Basic, self test and DOS, , with and without Fujinet installed.

Edited December 9, 2021 by manterola

[Rybags]

Swap the multiplexors first and see if the behaviour changes.

 

L is "special" as it's code is 0.  Default no key pressed code is FF.

[manterola]

I swapped the multiplexers (the 4051s) and I still got same results. I remember @Nezgar mentioned something regarding resistors to fix keyboard errors.. maybe it has something to do with them. You said that L is zero. So maybe there is a capacitor not able to sustain the voltage drop produced by pressing and holding a button for several seconds. I actually believe that I observed a shortened time to get the "L"s if I keep pressing and holding keys.

I guess I need to check the schematics to understand how the keyboard works.

 

 

 

[_The Doctor__]

resistors for keyboard are put in parallel to the other resistors for the keyboard.... that actually lowers resistance and is done for the console keys not the main keyboard matrix... sounds like a voltage or pokey issue if something isn't floating around in the keyboard itself, sometimes swapping the 4051's doesn't help... one may be defective in a way that zero as described... so if all else fails you might revisit them... pokey, plexors though  sometimes weird GTIA/PIA issues can muck things up... it doesn't sound like the latter two are in play though... no harm in trying the resistor mod or replacing the membrane if it continues... though it does sound like something warms up and fails so I do not think there is such an issue with resistance.

more current usually equals more heat when things are failing in the chip world.

Edited December 10, 2021 by _The Doctor__

[+Nezgar]

42 minutes ago, manterola said:

I remember @Nezgar mentioned something regarding resistors to fix keyboard errors..

Heh, yeah that mod/hack was specific to the console keys as @_The Doctor__ mentioned... they are tied directly to pins on the GTIA. The extra resistors let a little more voltage through to compensate for increased resistance that XE membranes often develop over time.

 

I like the steps considered though: swap 4051's, swap POKEY.

 

Somewhere in these threads is a matrix diagram that might hint at what interconnects result in "L" ...

 

Edit - here's one: 

 

[DavidMil]

When you hold down a key, would you mind counting the characters that appear before the odd character and L's start showing up?  I'm

curious to know if it is always the same amount of character repeats (no matter which key is held down) before the odd character and L's

star showing up.

 

DavidMil  

[djglish]

I found that if I tightened the screws to the keyboard too tightly that I would start to get weird multiple key errors like that.  

If you loosened them and retightened them that could be the cause.  I finally figured it out by trial and error.

 

[manterola]

On 12/11/2021 at 6:10 PM, DavidMil said:

When you hold down a key, would you mind counting the characters that appear before the odd character and L's start showing up?  I'm

curious to know if it is always the same amount of character repeats (no matter which key is held down) before the odd character and L's

star showing up.

 

DavidMil  

The number of characters gets shorter and shorter. Typically, the fullest time you can get a full line of 40 characters or do before the Ls.

 

[manterola]

On 12/11/2021 at 7:32 PM, djglish said:

I found that if I tightened the screws to the keyboard too tightly that I would start to get weird multiple key errors like that.  

If you loosened them and retightened them that could be the cause.  I finally figured it out by trial and error.

 

Thank you, I haven't tried that yet. 

The bad news is that I exchanged the keyboard and multiplexers and I still got the same behavior. But when exchanged the POKEY the problem was solved. That is very bad  news given the difficulty to get POKEYs and the $$$ involved.

I still think that the problem is pretty odd.... And maybe changing a capacitor or trying a type2 keyboard might solve the problem.... Maybe just wishful thinking...

Edited December 15, 2021 by manterola

[Rybags]

Of the main ICs, Pokey should be the easiest to get (except maybe PIA) - there were big stockpiles plus plenty of surplus 7800 Ballblazer cartridges as a source.

 

I've got over a dozen in my posession, one of these days aiming to do stereo in most of my machines.

[+DrVenkman]

11 hours ago, Rybags said:

Of the main ICs, Pokey should be the easiest to get (except maybe PIA) - there were big stockpiles plus plenty of surplus 7800 Ballblazer cartridges as a source.

 

I've got over a dozen in my posession, one of these days aiming to do stereo in most of my machines.

You are by far the exception. I think I currently have 1 spare POKEY, and that’s sitting in a spare non-working 800XL given to me by a friend. Since I haven’t yet investigated the machine, it could be that POKEY is the fault. 

 

But generally, POKEYs are pricier by the month, which why there have been so many efforts to replace them with FPGA solutions in recent years. Arcade cab owners, 7800 homebrew buyers and 5200/A8 repairs have dried up the ready availability of cheap replacemements. 

[mimo]

£60+ for a ballblazer cart now in the UK, so nope to using them as a cheap source .

Luckily I have 2 spare pokey chips, but very much looking forward to hokey being available to put one in my concerto cart to free up another spare 

[+bf2k+]

My spare pokey is bad and it sits in my "chip test" 800XL.  Some of the keys don't work on it but it is good enough for the test machine for testing the other chips.

 

At $20 ea, Brad@Best will only sell me one.. it is on the way here now.

 

[manterola]

Yep, Brad Kenobi, you are our only hope..