3x4 Matrix Keypad Decoder

[lewiseason@googlemail.com]

I know this has been discussed before but the solutions provided are
not what is needed for this project, so here it is.

What I have is a matrix keypad (3x4).

What I need is when a key is pressed _one pin_ of the decoder is
connected to ground. The ideal solution would be a series of logic
gates or similar basic circuitry.

Thanks for your time,
Lewis

 

[Gary Tait]

"lewiseason@googlemail.com" <lewiseason@googlemail.com> wrote in
news:1186398597.576186.175670@19g2000hsx.googlegroups.com:

I know this has been discussed before but the solutions provided are
not what is needed for this project, so here it is.

What I have is a matrix keypad (3x4).

What I need is when a key is pressed _one pin_ of the decoder is
connected to ground. The ideal solution would be a series of logic
gates or similar basic circuitry.

Thanks for your time,
Lewis

Microcontroller would be electroniclly simple, but coding is required.

If there is a common on the keypad, it could be easier.

In logic, conceptually:

You have an oscillator (1) feeding a two bit binary generator (2).
That feeds a binary decoder (3), going to the rows of the keypad (4),
The columns (pulled high) go to a 4 to 2 encoder (5), and a 4 input NAND
gate (6). Two bits from the binary gnerator (2) and the 4 to 2 encoder
(4) got to a 4 to 16 decoder. (7) Its output is enabled by the NAND
output (6) which also inhibits the clock (1).

 

[petrus bitbyter]

<lewiseason@googlemail.com> schreef in bericht
news:1186398597.576186.175670@19g2000hsx.googlegroups.com...

I know this has been discussed before but the solutions provided are
not what is needed for this project, so here it is.

What I have is a matrix keypad (3x4).

What I need is when a key is pressed _one pin_ of the decoder is
connected to ground. The ideal solution would be a series of logic
gates or similar basic circuitry.

Thanks for your time,
Lewis

So you have that 3x4 matrix (7 wires) and that decoder (12 wires + GND). A
solution for a conversion is not half as easy as it looks like, especially
when using basic building blocks like gates. Problem is the matrix keypad
which cannot be read out using combinatorial logic only. You will need some
clock and a shift register or similar to put signals on the rows (or
columns). When correctly designed, the signals on all seven lines determine
which key was pressed (if any). This code is not continuously available
however as other keys also needs to be scanned. So you will need to build in
some timing for validation of the key pressed. Another method is to stop
scanning as long as a key is presed. Conversion from the 7 bits code to the
12 bits one is straightforward using combinatorial logic, although you will
need a lot of gates.

This type of circuits has been build and I still can build them if I really
need to. Otherwise I won't. Using a simple, cheap, 28 pins micro or similar,
will be far easier and faster to design and requires much less components
and so board space.

petrus bitbyter

 

[Lostgallifreyan]

"lewiseason@googlemail.com" <lewiseason@googlemail.com> wrote in
news:1186398597.576186.175670@19g2000hsx.googlegroups.com:

I know this has been discussed before but the solutions provided are
not what is needed for this project, so here it is.

What I have is a matrix keypad (3x4).

What I need is when a key is pressed _one pin_ of the decoder is
connected to ground. The ideal solution would be a series of logic
gates or similar basic circuitry.

Must it reject multiple key-presses? (Not all keypad decoders do that
anyway). If not, use series resistors. 0,1,2 K on one buss, 0,3,6,9 K on
the other. Using a constant current source, the voltage across the
resistance can be fed to a linear dot/bar driver IC like the LM3914 which
drives 10 outputs but can be chained to get more. Use dot mode to get
single output at once. No timing circuits needed, no oscillators, just a
scale and offset adjust which can be fixed resistors once you know what
they must be.

Siemens UAA180 has 12 outputs and is linear. Getting one might be awkward.

 

[clifto]

Lostgallifreyan wrote:

Siemens UAA180 has 12 outputs and is linear. Getting one might be awkward.

A statement like that suddenly makes me want several, and I don't even know
what I'd do with them.

--
Spammer gets 30 years in the slammer
Suddenly wishes Viagra was harder to come by
<http://www.theregister.com/2007/08/02/spammer_gets_30_years/>

 

[Lostgallifreyan]

clifto <clifto@gmail.com> wrote in
newsrako4-ou3.ln1@remote.clifto.com:

Lostgallifreyan wrote:
Siemens UAA180 has 12 outputs and is linear. Getting one might be
awkward.

A statement like that suddenly makes me want several, and I don't even
know what I'd do with them.

Me too. I never really thought of them much until I wanted to solve that
question just posted. The ability to take a voltage and put out an
appropriate signal on a unique pin makes them way useful, for much more
than the original intent. That might be why they stayed around for 20
years. The standard ten segment ones, anyway. What could be nice is a
single IC with 16 outs.