mystical keyboard keys

[transkawa]

can someone help me solve this non-mystical so i want to believe
problem?
The keyboard on my laptop fails sometimes. It's a DeLL Latitude C600
series. Though the laptop is old but the keyboard is only two years or
so old. The keyboard fails in the sense that a particular set of keys
fail. The keys are T and Y keys. Then the Tab and Caps Lock keys. I make
htem work again but discharging them with a screwdriver and the tips of
my lips.
I do this several times once in a while but this mystical way of making
these work is way towards the alchemical. Can someone proffer a solution
to the problem?
xnt

 

[TVeblen]

can someone help me solve this non-mystical so i want to believe
problem?
The keyboard on my laptop fails sometimes. It's a DeLL Latitude C600
series. Though the laptop is old but the keyboard is only two years or
so old. The keyboard fails in the sense that a particular set of keys
fail. The keys are T and Y keys. Then the Tab and Caps Lock keys. I make
htem work again but discharging them with a screwdriver and the tips of
my lips.
I do this several times once in a while but this mystical way of making
these work is way towards the alchemical. Can someone proffer a solution
to the problem?
xnt

Paul recently posted a diagram of the wiring of a keyboard, and unlike
the ordered, logical, and neat wiring schematic you would expect, the
diagram looks like a plate of spaghetti. So while it seems random to you
and me, the keys are on the same circuit.

And the circuit is failing. Time for a new keyboard/laptop.

 

[Paul]

can someone help me solve this non-mystical so i want to believe
problem?
The keyboard on my laptop fails sometimes. It's a DeLL Latitude C600
series. Though the laptop is old but the keyboard is only two years or
so old. The keyboard fails in the sense that a particular set of keys
fail. The keys are T and Y keys. Then the Tab and Caps Lock keys. I make
htem work again but discharging them with a screwdriver and the tips of
my lips.
I do this several times once in a while but this mystical way of making
these work is way towards the alchemical. Can someone proffer a solution
to the problem?
xnt

If I was working on it, I'd want to start by reading the part number
off the keyboard encoder IC. This is an example of an encoder IC.
This datasheet is not as complete as I would like, but it should give some
idea how it works.

http://web.archive.org/web/20010613141313/www.semtech.com/pdf/sysmgmt/doc5-418-ds-109.pdf

The keys are arranged as a matrix of wires (row and column wires). At any
one time, most of the column lines are open circuit, except for one column
line driving to ground. That means the column lines have no "drive to logic 1"
capability. They either "drive to logic 0" or are open circuit.

The row lines are inputs.

There are some external resistors tied to the lines. If no keys on a column
are closed, then all the row signals will be pulled to logic 1 by the resistors.
If a single key is pressed, it joins a column line to a row input. If that column line
is currently being scanned (driven to logic 0), then the row input will see
a logic zero as well. So seeing a logic zero, means the key is active or closed.

So that describes a bit about how the keyboard matrix is scanned. The keyboard
encoder chip "debounces" the key, and only declares a key closure, if the
key remains closed for at least two scans of the keyboards (roughly).

Now, this article describes some additional components that can be added.
See "Getting Rid Of Ghosting and Masking" on page 8, where a switching diode
has been added for each and every keyswitch on the keyboard. The diodes are
added, when you want a "deluxe" keyboard supporting N-key rollover. N-key
rollover is preferred for gaming applications, where the user can be pressing
a number of keys at the same time. The diodes solve the "ghosting" problem,
although at the cost of adding $3-$5 worth of diodes per keyboard. The thresholding
on the CMOS circuit must be such, that it can handle the slightly higher input
voltage of a logic 0, caused by the diodes in the circuit.

http://www.dribin.org/dave/keyboard/keyboard.pdf

If you can find the datasheet for the main chip, you can get the table
of key mappings, which may help you explain why those particular
keys are affected. They could all be members of a group experiencing
"ghosting", meaning the keys are failing in the closed position,
and the keyboard controller has "locked them out" to prevent them
from interfering. That would not explain though, why you need to
"discharge" the circuit. Discharging a circuit to restore operation,
is something you might do if encountering CMOS latchup. But in this
particular circuit and application, I don't see a reason for that
to happen. And in any case, I don't see how you could add any components
to the circuit to actually fix it.

To test the PCB of the keyboard, I'd use a multimeter. I'd set it to
a low power ohms range (where the circuit will not apply enough
voltage to forward bias a diode). Using the meter, I'd test for
shorting between rows or columns or the like. The test is potentially
limited in sensitivity, by the pullup resistors.

I notice in the Dell C600 documentation, they mention the presence
of "diagnostic software". Perhaps you could run that software,
and check for any error codes related to the keyboard. The keyboard
has internal status information, and if the Dell diagnostic software
can extract that status information, perhaps it can detect a
"stuck key" or "ghost key" problem. And that might give you more
clues as to the source of the problem.

Paul

 

[jerry]

can someone help me solve this non-mystical so i want to believe
problem?
The keyboard on my laptop fails sometimes. It's a DeLL Latitude C600
series. Though the laptop is old but the keyboard is only two years or
so old. The keyboard fails in the sense that a particular set of keys
fail. The keys are T and Y keys. Then the Tab and Caps Lock keys. I make
htem work again but discharging them with a screwdriver and the tips of
my lips.
I do this several times once in a while but this mystical way of making
these work is way towards the alchemical. Can someone proffer a solution
to the problem?
xnt

I worked in an office once and I had some problems with the key board on
my desk. My desk was made of metal. The center drawer of the desk rode on
nylon guides, but there was a piece of metal spring at the back of the
drawer to cushion it it case someone slammed the drawer in. I sat in a
chair with hard rubber roller feet. And my chair was rolling on a plastic
chair mat on the floor.

When I rolled the chair across the mat I would pick up a static electric
charge. And if I then pushed the center drawer of the desk closed so that
it contacted the metal spring at the back of the drawer bay, there would
be a tiny spark emitted. And that spark produced EMI which the keyboard
was sensitive to, and it would stop working.

I finally figured it out and got a static discharge mat to go under the
keyboard which fixed the problem.

good luck!
Jerry

 

[emekadavid]

If I was working on it, I'd want to start by reading the part number
off thekeyboardencoder IC. This is an example of an encoder IC.
This datasheet is not as complete as I would like, but it should give some
idea how it works.

http://web.archive.org/web/20010613141313/www.semtech.com/pdf/sysmgmt...

Thekeysare arranged as a matrix of wires (row and column wires). At any
one time, most of the column lines are open circuit, except for one column
line driving to ground. That means the column lines have no "drive to logic 1"
capability. They either "drive to logic 0" or are open circuit.

The row lines are inputs.

There are some external resistors tied to the lines. If nokeyson a column
are closed, then all the row signals will be pulled to logic 1 by the resistors.
If a single key is pressed, it joins a column line to a row input. If that column line
is currently being scanned (driven to logic 0), then the row input will see
a logic zero as well. So seeing a logic zero, means the key is active or closed.

So that describes a bit about how thekeyboardmatrix is scanned. Thekeyboard
encoder chip "debounces" the key, and only declares a key closure, if the
key remains closed for at least two scans of the keyboards (roughly).

Now, this article describes some additional components that can be added.
See "Getting Rid Of Ghosting and Masking" on page 8, where a switching diode
has been added for each and every keyswitch on thekeyboard. The diodes are
added, when you want a "deluxe"keyboardsupporting N-key rollover. N-key
rollover is preferred for gaming applications, where the user can be pressing
a number ofkeysat the same time. The diodes solve the "ghosting" problem,
although at the cost of adding $3-$5 worth of diodes perkeyboard. The thresholding
on the CMOS circuit must be such, that it can handle the slightly higher input
voltage of a logic 0, caused by the diodes in the circuit.

http://www.dribin.org/dave/keyboard/keyboard.pdf

If you can find the datasheet for the main chip, you can get the table
of key mappings, which may help you explain why those particularkeysare affected. They could all be members of a group experiencing
"ghosting", meaning thekeysare failing in the closed position,
and thekeyboardcontroller has "locked them out" to prevent them
from interfering. That would not explain though, why you need to
"discharge" the circuit. Discharging a circuit to restore operation,
is something you might do if encountering CMOS latchup. But in this
particular circuit and application, I don't see a reason for that
to happen. And in any case, I don't see how you could add any components
to the circuit to actually fix it.

To test the PCB of thekeyboard, I'd use a multimeter. I'd set it to
a low power ohms range (where the circuit will not apply enough
voltage to forward bias a diode). Using the meter, I'd test for
shorting between rows or columns or the like. The test is potentially
limited in sensitivity, by the pullup resistors.

I notice in the Dell C600 documentation, they mention the presence
of "diagnostic software". Perhaps you could run that software,
and check for any error codes related to thekeyboard. Thekeyboard
has internal status information, and if the Dell diagnostic software
can extract that status information, perhaps it can detect a
"stuck key" or "ghost key" problem. And that might give you more
clues as to the source of the problem.

    Paul

diagnostic replies okay. it gives only one problem but that is non
keyboard related. not using the laptop right now and not at home but i
think the only failing on the diagnostic side is a dst error message;
haven't solved that problem yet also.

 

[emekadavid]

I worked in an office once and I had some problems with the key board on
my desk. My desk was made of metal. The center drawer of the desk rode on
nylon guides, but there was a piece of metal spring at the back of the
drawer to cushion it it case someone slammed the drawer in. I sat in a
chair with hard rubber roller feet. And my chair was rolling on a plastic
chair mat on the floor.

When I rolled the chair across the mat I would pick up a static electric
charge. And if I then pushed the center drawer of the desk closed so that
it contacted the metal spring at the back of the drawer bay, there would
be a tiny spark emitted. And that spark produced EMI which thekeyboard
was sensitive to, and it would stop working.

I finally figured it out and got a static discharge mat to go under thekeyboardwhich fixed the problem.

good luck!
Jerry

i think myself it's a static problem, though my desk is not made of
metal. i think i'll work on this premise.
thanks