Post by thread:[EE] Simple Clocklike LCD Usage

'[EE] Simple Clocklike LCD Usage'

If I were to buy a module like this one from Mouser:

696-LCD-S3X1C50TRA,

how would I drive it?

It specifies 3-5 V A.C., static driving mode.  Could I

hook the two COM pins to complementary output

PWM pins on a PIC at, say, 50 Hz, then drive the data

pins high/low to on/off them, or something else?  Maybe

High-Z for segment off and 50 Hz for on?

Microchip's AN563 seems to be along these lines, but

I'm not sure that all applies to this module.

Or maybe it does.  Maybe the thing to do is shell out for

a few and burn them up figuring it out.  The "datasheet"

on Mouser's website is less than informative, at least, for

me.

I guess I'm spoiled by the 44780 driven displays, but for

this project, one of those would be overkill, plus the current

draw wouldn't be appropriate.

There isn't a similar module for such displays, is there?

Mike H.

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

> If I were to buy a module like this one from Mouser:

> 696-LCD-S3X1C50TRA,

> how would I drive it?

> There isn't a similar module for such displays, is there?

>

I'm using a Microchip AY0438 32-segment LCD driver for a very similar panel. 

The AY0438 is a generic static-mode driver that takes serial data input, and 

you can wire the segments however you prefer.  You'll only need 3 output 

pins to operate it, and a 4th if you desire to provide switched power to the 

display.  The datasheet has an excellent example of using a 16C71 to drive a 

4 x 7 segment display, similar to yours.

I've found the AY0438 to perform fantastically.  Latchup time is fast enough 

that I don't have to place any NOPs in the output routine.  I don't have to 

worry about any contrast adjustments and the driving frequency is adjustable 

up to about 150Hz, which is nice for flicker-free display.  Speed is 

important in my application.

The only 2 drawbacks I have found is that the pins are not conveniently 

arranged so that the traces directly fan out to the LCD pins.  This results 

in a hefty ratsnest of traces between the 2 devices.  There's a workaround 

but it involves 4 times the number of lookup tables. The second drawback is 

the size of the chip.  It only comes in a 40-pin dip x .600, or in a 44 

PLCC.  As of yet I've only found one manufacturer who makes a smaller size, 

but its specs aren't nearly as desirable.

I'm happy with the AY0438 but would sure love to hear about a similar device 

where you can wire the traces directly to the LCD without having to provide 

so many vias.

Maxim supplies a similar device (ICM7211 IIRC) and takes BCD data from 4 pic 

pins.  It works okay, but you won't be able to utilize some of the extra 

characters on that particular LCD.

Rick 

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

Driving a static LCD isn't complex, but does require a little more 

effort than what you suggest.

Key points:

* Segments with a certian voltage potential across them are "on"

* DC potentials destroy the properties of the LC fluid.

Therefore they must be driven in an AC fashion.

The common pins flip occasionally - 20Hz-80Hz.  The segment drivers flip 

in sync - they match the common to turn a segment off, and are opposite 

to turn a segment on.  There are no other special devices needed - LCD 

segments can be modelled as capacitors.

So the segment drivers would somehow have to run in sync with your 

common.  Perhaps you can drive an interrupt every time the common flips 

and then flip all the outputs of your segments.  Alternately you can use 

external XOR gates - the PWM drives one side of XOR, and the PIC port 

drives the other side.  When the port is low the output of the XOR 

matches the common, and thus the segment is off.  When the output is 

high the XOR inverts the common and the segment turns on.  Lots of extra 

gates, though - one for each segment.

I would be tempted to use a timed interrupt and control the common and 

segment lines directly. 

Use some memory to hold the display data - this is where the program 

writes to change the display

Each interrupt

1) Invert the common line

2) If common is 0 then copy display data to segment output, if 1 then 

invert during copy

Since it's so slow I wouldn't tie up a PWM for it.

Also you don't have to connect both commons - they are essentially the 

same wire.

Driving a multiplexed (non static) display is much harder and not 

practical without a driver chip.

-Adam

Mike Hord wrote:

{Quote hidden}>If I were to buy a module like this one from Mouser:

>696-LCD-S3X1C50TRA,

>how would I drive it?

>

>It specifies 3-5 V A.C., static driving mode.  Could I

>hook the two COM pins to complementary output

>PWM pins on a PIC at, say, 50 Hz, then drive the data

>pins high/low to on/off them, or something else?  Maybe

>High-Z for segment off and 50 Hz for on?

>

>Microchip's AN563 seems to be along these lines, but

>I'm not sure that all applies to this module.

>

>Or maybe it does.  Maybe the thing to do is shell out for

>a few and burn them up figuring it out.  The "datasheet"

>on Mouser's website is less than informative, at least, for

>me.

>

>I guess I'm spoiled by the 44780 driven displays, but for

>this project, one of those would be overkill, plus the current

>draw wouldn't be appropriate.

>

>There isn't a similar module for such displays, is there?

>

>Mike H.

>_______________________________________________

>http://www.piclist.com

>View/change your membership options at

>http://mailman.mit.edu/mailman/listinfo/piclist

>

>

>

>  

>

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

I see.  So basically, I can just hook up my 24 or 25 control lines,

plus the common, and then call an interrupt every, say, 7 ms,

during which time I update the state, and then every 14 ms, 

complement the ports?

Nice.

Mike H.

On Wed, 06 Oct 2004 10:27:32 -0400, M. Adam Davis <spam_OUTadampicTakeThisOuTubasics.com> wrote:

{Quote hidden}> Driving a static LCD isn't complex, but does require a little more

> effort than what you suggest.

> 

> Key points:

> * Segments with a certian voltage potential across them are "on"

> * DC potentials destroy the properties of the LC fluid.

> 

> Therefore they must be driven in an AC fashion.

> 

> The common pins flip occasionally - 20Hz-80Hz.  The segment drivers flip

> in sync - they match the common to turn a segment off, and are opposite

> to turn a segment on.  There are no other special devices needed - LCD

> segments can be modelled as capacitors.

> 

> So the segment drivers would somehow have to run in sync with your

> common.  Perhaps you can drive an interrupt every time the common flips

> and then flip all the outputs of your segments.  Alternately you can use

> external XOR gates - the PWM drives one side of XOR, and the PIC port

> drives the other side.  When the port is low the output of the XOR

> matches the common, and thus the segment is off.  When the output is

> high the XOR inverts the common and the segment turns on.  Lots of extra

> gates, though - one for each segment.

> 

> I would be tempted to use a timed interrupt and control the common and

> segment lines directly.

> Use some memory to hold the display data - this is where the program

> writes to change the display

> Each interrupt

> 1) Invert the common line

> 2) If common is 0 then copy display data to segment output, if 1 then

> invert during copy

> 

> Since it's so slow I wouldn't tie up a PWM for it.

> 

> Also you don't have to connect both commons - they are essentially the

> same wire.

> 

> Driving a multiplexed (non static) display is much harder and not

> practical without a driver chip.

> 

> -Adam

> 

> 

> 

> Mike Hord wrote:

> 

> >If I were to buy a module like this one from Mouser:

> >696-LCD-S3X1C50TRA,

> >how would I drive it?

> >

> >It specifies 3-5 V A.C., static driving mode.  Could I

> >hook the two COM pins to complementary output

> >PWM pins on a PIC at, say, 50 Hz, then drive the data

> >pins high/low to on/off them, or something else?  Maybe

> >High-Z for segment off and 50 Hz for on?

> >

> >Microchip's AN563 seems to be along these lines, but

> >I'm not sure that all applies to this module.

> >

> >Or maybe it does.  Maybe the thing to do is shell out for

> >a few and burn them up figuring it out.  The "datasheet"

> >on Mouser's website is less than informative, at least, for

> >me.

> >

> >I guess I'm spoiled by the 44780 driven displays, but for

> >this project, one of those would be overkill, plus the current

> >draw wouldn't be appropriate.

> >

> >There isn't a similar module for such displays, is there?

> >

> >Mike H.

> >_______________________________________________

> >http://www.piclist.com

> >View/change your membership options at

> >mailman.mit.edu/mailman/listinfo/piclist

> >

> >

> >

> >

> >

> 

> 

> _______________________________________________

> http://www.piclist.com

> View/change your membership options at

> http://mailman.mit.edu/mailman/listinfo/piclist

>

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

> >If I were to buy a module like this one from Mouser:

> >696-LCD-S3X1C50TRA,

> >how would I drive it?

I'd suggest the ICM7211. Use the segment drives for digits 2

3 and 4 as normal, and a small table to convert digit 1 data to

meaningful drive for digit 1 and those annunciator segments if

needed

http://home.clear.net.nz/pages/joecolquitt/0sp505.html

That's an old page, written around a prototype. These days I use

the VI402 display (Varitronix) and the 16F628A. Circuit is much

the same but swisher

As mentioned on the page, 4543 works fine, 1 per digit or assorted

segments. You need to add an oscillator for the backplane. I may

move back to the 4543 as 4 are cheaper than the 7211 (NZ$6.53+)

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

Mike Hord wrote:

>I see.  So basically, I can just hook up my 24 or 25 control lines,

>plus the common, and then call an interrupt every, say, 7 ms,

>during which time I update the state, and then every 14 ms, 

>complement the ports?

>

>Nice.

>

>Mike H.

>  

>

>  

>

The two interrupts seem to be able to conflict - you might never get a 

valid display with those two tasks.  Also, with the LCD you are using, 

IIRC, the common line probably needn't be run faster than 40Hz.  So:

Interrupt every 25mS

  Complement common line

  If common is high

    Copy display data to W, invert, send to output lines

  else

    Copy display data to output lines

It's essentially the same thing as setting common low, copying the data, 

then 1/2 cycle later complementing everything, then start over agin once 

the cycle completes which is what I suspect you were suggesting above.

The reality is that with your LCD you can even run it much less than 

30Hz (could change it 10 times a second if you wanted to) since the 

digits have a very slow attack and decay period.  You run it more 

quickly to avoid beating effects with flourescent lights and such.

Good luck!

-Adam

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist

> It's essentially the same thing as setting common low, copying the data,

> then 1/2 cycle later complementing everything, then start over agin once

> the cycle completes which is what I suspect you were suggesting above.

I kind of worded it poorly.  The 7/14 ms thing would have been the same

interrupt.  Maybe I should have said "on every other interrupt, the register

which tracks the active segments gets output in its inverted state".  

> The reality is that with your LCD you can even run it much less than

> 30Hz (could change it 10 times a second if you wanted to) since the

> digits have a very slow attack and decay period.  You run it more

> quickly to avoid beating effects with flourescent lights and such.

> 

> Good luck!

Thx.  Should be fun!

Mike H.

> -Adam

_______________________________________________

http://www.piclist.com

View/change your membership options at

http://mailman.mit.edu/mailman/listinfo/piclist