Greetings

        Thanks to all that responded to my posting!



>>         I purchased an inexpensive (about $5) solar/battery dual power
desktop
> >calculator for exploring the possibilities of using the bold 0.5 inch 8
>> character display as the readout for my project. The characters are in
the
>> solid bar 7 segment format as opposed to the dot matrix type. I am
hoping
>> that I will be able to interface the micro controller to the display via
a
>> standard multi-character 7 segment LCD driver chip.
>>
>>         1) The display is connected to the calculator PCB via a 28 lead
ribbon
>> cable.
>>
>>         2) Battery positive supply is connected via a 1 K resistor to
lead 1.
>>
>>         3) There is a 120 Hz symmetrical square wave clock signal
applied to
>> lead
>> # 3.
>>
>>         4) There are 3 phases of sub harmonic 40 Hz 1/3 duty cycle clock
>> signals
>> applied to leads 2, 27 and 28.
>>
>>         5) A 40 Hz symmetrical square wave is applied to the remaining
leads.
>> The
>> relative phase of these signals vary with respect to the clock depending
on
>> the particular characters displayed.

>Interesting; from your description, the display is using a 1/3 multiplex
>design though I'd have to look at things to see exactly what's going on.
>A common way of doing 3:1 displays is to use nine elements per digit
(seven
>segments, a dot, and something else) and then have those in a 3x3 matris
(one
>side of the 3x3 is common throughout the entire display).

>The thing I find odd, though, is that you say most of the other pins have
>40Hz "square waves"; I'd expect them to look kinda wierd depending upon
what
>is being shown.

>Perhaps if you could clarify what sort of annunciators, punctuation, etc.
>the display has I could better guess what you've got.

        I wish to add the following to my original description:

        1) Each digit of the display has a right hand decimal point.

        2) The voltage applied to pin 1 is more or less equal to the battery
voltage of approximately + 1.5 Volts. This may be a reference baseline for
bipolar logic levels as the clock and signal pulses extend from 0 to + 3
Volts.

        3) All of the waveforms have a peak-to-peak amplitude of approximately 3
Volts and the step for the square waves is from 0 to +3 Volts with respect
to battery negative. As the 3 Volt level is twice that of the battery I
presume there must be a charge pump voltage doubler in the circuitry to
provide the higher voltage.

        4) The baselines for the 1/3 duty cycle, 1/3 clock frequency signals on
pins 2, 27 and 28 are at +1.5 Volts with negative pulses extending down to
0 Volts and positive pulses up to +3 Volts.

        5) The shape of the remaining 24 squarewave logic waveforms does not var
y
from a blank display (with only the right hand 0 showing) to a display
filled with 8 digit eight characters showing all segments. The relative
phases with respect to the clock signals do vary thus the algebraic sum of
the signal lines and clock signals would also vary and I presume this is
what is used to selectively illuminate the digit segments.

        Although I have been successfully utilizing LCD alpha/numeric display
modules in some of my projects which simply interface with ASCII code I
have not gotten around to studying the theory of the LCD element operation.
I do not for example know what voltage combinations it takes to illuminate
an LCD segment. I have the following questions:

        A) Is there a way that I could statically test this display to determine
how to illuminate a specific segment or must it be clocked with a tri-phase
bipolar source in order for it to work?

        B) Must LCD displays be driven with bipolar logic signals.

        Thanks again to all that responded to my posting!


Very Best Regards,
Thomas M. Alldread
e-mail: tmaldred@mail.island.net