Shop OBEX P1 Docs P2 Docs Learn Events
6 digit glass LCD panel, direct drive from Prop? — Parallax Forums

6 digit glass LCD panel, direct drive from Prop?

Heater.Heater. Posts: 21,230
edited 2011-03-08 11:47 in Propeller 1
A store near here as a couple of hundred 6 digit glass LCD panels on clearance for 1 Euro a piece. So of course I had to pick one up.

It has 50 pins DIP style rather like the 60D070 here http://www.femacorp.com/products/lcd/glass%20panels/glass%20panels%20index.htm

So how to drive this from our favourite MCU?

What I have gleaned so far:

1) One common pin and 49 segment pins.
2) Needs AC drive, no more than 100mv or so DC component.
3) Common pin driven from 50 to 200Hz square wave
4) Segments driven from in phase or anti phase signal makes them OFF or ON.
5) Contrast can be varied by changing "dead time" in signals. Patent held by STC.

So one Prop has enough pins to drive only half of this. Two Props would be a waste. Some kind of output expander is called for.

Anyone have any neat ideas for this? DIP solutions preferred, in keeping with the display and y Props.

Anyone know any more LCD secrets to keep in mind.

Comments

  • Heater.Heater. Posts: 21,230
    edited 2010-09-08 04:45
    What, no takers for my little LCD project?

    Here is an app. note from STC describing how to drive seven segment LCD panels in software from an MCU's general purpose I/O pins.
    www.st.com/stonline/books/pdf/docs/6292.pdf

    The Prop has nowhere enough pins for a 6 digit non multiplexed display with one common pin and 49 segment drive pins. So some external hardware is required.

    As I'm old fashioned so my first thought is to use six 8 bit serial in parallel out shift registers like 74HC595. In DIP packages at 50 cents a piece they match the style and cost of the display and I can actually bread board the thing.

    These can be chained together so now we only need 2 Prop pins to drive 48 signals out. Or up to 5 pins if we want control of output enable, output latch, and reset.

    Using the output latch is a good idea as it syncs the segment drive outputs with the common pin drive. Both square waves at 100Hz remember.

    48 is one output short so we can sacrifice a decimal point or I have to drive it from another Prop pin.

    Is HC the right technology for Prop interfacing?
  • LeonLeon Posts: 7,620
    edited 2010-09-08 04:48
    I remember CMOS chips for driving those sort of displays. I found one: the MC145003

    http://pdf1.alldatasheet.com/datasheet-pdf/view/166617/MOTOROLA/MC145003.html

    I think there were smaller ones.
  • Heater.Heater. Posts: 21,230
    edited 2010-09-08 05:24
    Leon,

    That MC145003 is not good here. It drives 128–segment, multiplexed by four LCD panels. This panel is not multiplexed so there would only be 32 segments usable from it. Perhaps there are similar devices but I have my heart set on DIPs.

    However you were right in suggesting 4000 series, which I noticed before you edited it away.

    Locally I can get CD4056 which is a BCD to 7 segment decoder + input latches + it does level shifting so the outputs can be bigger than 3v and can be adjusted for contrast.

    Then there is 4543 BCD to 7 seg with latches but no level shifter.

    There are a couple of other LCD drivers without latches which is no good if we want to save pins.

    These are still six chip solutions though. Looks like I have to scour the catalogue some more and do some experiments.
  • AleAle Posts: 2,363
    edited 2010-09-08 05:57
    heater:

    I used to use some BCD to 7 segment LCD drivers and some cmos xnor gates for a similar but only 4 digit display. the backplane was driven with 50 Hz.
  • Heater.Heater. Posts: 21,230
    edited 2010-09-08 06:16
    Ale, that's a lot of chips in this case.

    Even the BCD to seven segment LCD drivers would require 12 bits of BCD data to come from some where. That's 12 Prop pins or again it's serialized with three 8 bit shift registers.

    So to minimise Prop pins and chips and the amount of hand wiring to do it looks like the shift register idea is not so bad. It also means one can drive segment patterns that are not just "0" to "9" "A" to "F".

    Question remains is the 3v available enough to get the LCD to work?

    I'm thinking that HC series logic will work up to 6v. Driven in anti phase gives a 12v peak to peak across the segments which surely is enough.

    This would only possibly 2 or 3 pins on the Prop. The 6 volts could be derived from a charge pump and adjusted for contrast. Two or three transistors could take care of the level translation from Prop to the HC.
  • AleAle Posts: 2,363
    edited 2010-09-08 07:41
    you have to connect 49 wires anyways, the shortest should be the shift registers. The display I used was 5V based (driven by). remember that you are not multiplexing, so you do not need high voltage, even 3 V may do, I think.
  • Heater.Heater. Posts: 21,230
    edited 2010-09-08 09:19
    Yep, 49 wires. I start to balk when it gets to 100+ with extra layers of xor or multi bit inputs etc.

    That's a good point re: voltage and multiplexing.

    I have never lit up a naked LCD panel before (if that's a good expression for it) so this is all new to me. It was a surprise even to find out that I need AC drive yesterday.

    First experiment will be to drive common and a few segments from two anti-phase outputs from the Prop. Wish I had a signal generator around.
  • Heater.Heater. Posts: 21,230
    edited 2010-09-09 07:27
    You are right Ale.

    First experiment is a success. Two Prop pins driven in anti-phase at 100Hz lights up the LCD segments really dark, if you see what I mean:)

    Now to get some shift registers.

    This is going to be quite neat, the shift registers will sit underneath the LCD panel.
  • Dr_AculaDr_Acula Posts: 5,484
    edited 2010-09-09 16:10
    HC595 is what I first thought of too. I think you need three propeller pins to drive these. Just daisy chain them all together, and a single cog driver. It should be similar to a picaxe driver http://www.picaxeforum.co.uk/showthread.php?t=9116 There were some great suggestions on that thread that shortened the picaxe code - similar principles would apply to PASM code. This sounds a nifty little project.
  • Heater.Heater. Posts: 21,230
    edited 2010-09-09 23:51
    Dr_A,

    Actually there is a nice 595 driver object in OBEX which I will kick off with.

    It's kind of spooky being able to get segments displayed just by touching the 100Hz output from the Prop with one hand and some random display pins with the other.
  • AleAle Posts: 2,363
    edited 2010-09-10 00:02
    It is kind of spooky just to touch the pins and light some segments without touching any prop pins... 50 Hz are all around us :)
  • Heater.Heater. Posts: 21,230
    edited 2010-09-10 00:37
    Humanoid:

    "...50 Hz are all around us"

    Wow, you are right. By touching the displays common pin to a good ground, like the tap in the kitchen sink, and then holding the kettle lead with one hand I can "light up" the segments by touching the display pins with the other hand. Great experiment.
  • Erik FriesenErik Friesen Posts: 1,071
    edited 2011-03-08 11:38
    Not sure if I should start a new thread or not, but did you ever get yours working? I dredged this old display out of something, I think its a 5 volt display, anyway it works fine. Straight to common, 1K to the segment.
    1024 x 1326 - 130K
  • BeanBean Posts: 8,129
    edited 2011-03-08 11:46
    Basically hold the common at 1/2Vcc with two 100K resistors. Toggle the segment pin to make it dark, make the segment pin an input to turn it off.

    Not sure if 3.3V will be enough to make it very dark but give it a try.

    Bean
  • Erik FriesenErik Friesen Posts: 1,071
    edited 2011-03-08 11:47
    Will that work with a 5v display though? You don't get the same amount of drive. I am doing some experimenting here, I have another 5v rated display coming, I think it will drive from 3.3, not sure though yet.
Sign In or Register to comment.