Post by thread:PIC and LCD direct drive

Hi PICLISTERS ! I'm involved in a PIC project (a 2-digit timer, basically), using LCD display. All previous project I designed were using LED display, and I don't know much on LCD. 3 questions : 1/ Is 2-digit (+ dp) direct drive cost-effective rather than multiplexed one (anyway, for now, I couldn't find any 2-digit muxed one) ? 2/ Any pointers or links on harware design using this kind of LCD display ? 3/ Any pointers or links on software using this kind of LCD display ? My wish is to avoid an IC driver between PIC and display. Thanks for reading and best regards Laurent Haas ------------------------------------------------------------------------------- In a world without walls or fences, what use do we have for windows or gates? (Unknown author)

On Sat, 4 Jul 1998, Laurent HAAS - F6FVY wrote: > Hi PICLISTERS ! > > I'm involved in a PIC project (a 2-digit timer, basically), using LCD > display. All previous project I designed were using LED display, and I > don't know much on LCD. > > 3 questions : > > 1/ Is 2-digit (+ dp) direct drive cost-effective rather than multiplexed > one (anyway, for now, I couldn't find any 2-digit muxed one) ? You are not likely to find muxed LCDs with less than 3 1/2 digits. It just makes no sense to add that complexity to reduce the number of wires by 1 (or 3, depending on which kind of mux, 2-level or 4-level). > 2/ Any pointers or links on harware design using this kind of LCD display ? Is 'this kind' the un-multiplexed one ? If yes, then take a display data sheet, for example from eio.com's website and study it. It contains the control waveforms and all that. It is usually a good idea to look out for a display that can work well with the 5 Vpp that can be obtained from a PIC. Un-muxed ones will do that but muxed ones need more volts to achieve decent contrast. This means that you want to stay with a PIC for more than one reason (such as, supply voltage(s) charge pumps and power requirements). > 3/ Any pointers or links on software using this kind of LCD display ? > > My wish is to avoid an IC driver between PIC and display. You do not need any chip in-between but it is good to start out thinking that any deviation from 50% duty cycle on the backplane of a standard LCD display will slowly murder the LCD, reducing its lifetime to under a few months in some cases, and keep thinking about that all the time while you write the code. 2% off (i.e. 0.1 Volts DC across the LCD) is already VERY bad. Consult the LCD data sheet for exact numbers. Peter

Peter L. Peres wrote: > You do not need any chip in-between but it is good to start out > thinking that any deviation from 50% duty cycle on the backplane of a > standard LCD display will slowly murder the LCD, reducing its lifetime > to under a few months in some cases, and keep thinking about that all > the time while you write the code. 2% off (i.e. 0.1 Volts DC across > the LCD) is already VERY bad. Or just cheat, and put 100nF (SMD) chip ceramics in series with all the segment leads just in case! Much cheaper than replacing LCDs! -- Cheers, Paul B.

Status Distribution July 04, 1998 12:03:18 The message regarding "Re: your visit" sent on July 04, 1998 12:03:18 was sent by Status Recipient Type To Native Name spam_OUTHartTakeThisOuTquicklogic.com Foreign Native Name .....HartKILLspam@spam@quicklogic.com\n\n\nINTERNET Recipients Status Reporters Type From Name Domain NOTES Native Name CN=Roberto Zavatti/OU=Italy/O=IBM@IBMIT Foreign Native Name CN=Roberto Zavatti/OU=Italy/O=IBM\nIBMIT\n\n Organization IBM Org Unit 1 Italy Last Name Zavatti First Name Roberto Status 769 Explanation Invalid recipient X.400 Status 769 Explanation Router: Unable to open mailbox file D14HUBM1/14/H/IBM mail.box: Server not responding

>chip capacitors Hmm it just occurs to me that this solution may be a pain in the rear side of the human body. If the chip capacitors are good enough at insulating the LCD, any static potential built up in the case or on the LCD (f. ex. by wiping it) will light up segments for several seconds at least. We all know the effect from handling raw LCDs w/o static protection. Peter

On Sun, 5 Jul 1998, Paul B. Webster VK2BZC wrote: > Peter L. Peres wrote: > > > You do not need any chip in-between but it is good to start out > > thinking that any deviation from 50% duty cycle on the backplane of a > > standard LCD display will slowly murder the LCD, reducing its lifetime > > to under a few months in some cases, and keep thinking about that all > > the time while you write the code. 2% off (i.e. 0.1 Volts DC across > > the LCD) is already VERY bad. > > Or just cheat, and put 100nF (SMD) chip ceramics in series with all > the segment leads just in case! Much cheaper than replacing LCDs! > -- As a quick hack it would work, but adding 12-odd chip capacitors to a design that has a part count of 5 or 6 without is not my idea of a production-ripe unit. Maybe your country is richer ;) Peter

>I'm involved in a PIC project (a 2-digit timer, basically), using LCD >display. All previous project I designed were using LED display, and I >don't know much on LCD. > >3 questions : > >1/ Is 2-digit (+ dp) direct drive cost-effective rather than multiplexed >one (anyway, for now, I couldn't find any 2-digit muxed one) ? That's correct. The only multiplexed LCDs are graphical dot matrix types. The simpler digit types are direct drive. This is because of the difficulty and complexity of LCD multiplexing. >My wish is to avoid an IC driver between PIC and display. You don't need one. You can connect the PIC directly to the LCD segments. The LCD is a voltage device (as opposed to LEDs which are current devices), and the segments will go dark with an applied voltage. Basically connect one PIC pin to one segment. Switching the PIC output on or off will switch the segment on or off. There's only one other consideration. If you drive the segments with a continuous DC signal, you'll eventually wreck it. Thus, you need to connect the backplane (the common) to an AC signal. Set up a 555 timer circuit for something like 50 kHz, 50% duty cycle, and connect it to the common. And that's it. A simple low power (virtually zero current consumption from the LCD) display. Stephen.

On Mon, 6 Jul 1998 08:25:09 +1200 Stephen Rothlisberger < {Quote hidden}> >You don't need one. You can connect the PIC directly to the LCD >segments. >The LCD is a voltage device (as opposed to LEDs which are current >devices), >and the segments will go dark with an applied voltage. Basically >connect >one PIC pin to one segment. Switching the PIC output on or off will >switch >the segment on or off. > >There's only one other consideration. If you drive the segments with a >continuous DC signal, you'll eventually wreck it. Thus, you need to >connect >the backplane (the common) to an AC signal. Set up a 555 timer circuit >for >something like 50 kHz, 50% duty cycle, and connect it to the common. > >And that's it. A simple low power (virtually zero current consumption >from >the LCD) display. > I think a DC component will "wreck" the LCD pretty quickly. Also, I think a backplane drive of 30 Hz is more common. The backplane is driven with the 30 Hz 5Vpp square wave. Segments are driven with a square wave in phase with the backplane to make the segments not visible. They are driven out of phase to make the segments visible. You can probably generate all the required signals using a timer interrupt on the PIC. Also, have a look at the Siliconix DF411 or the Intersil ICL7211. Both of these are LCD driver chips that accept 4 digits of BCD and 4 digit select lines. They output appropriate drive signals for nonmuxed LCDs. Harold _____________________________________________________________________ You don't need to buy Internet access to use free Internet e-mail. Get completely free e-mail from Juno at http://www.juno.com Or call Juno at (800) 654-JUNO [654-5866]

Harold M Hallikainen wrote: > On Mon, 6 Jul 1998 08:25:09 +1200 Stephen Rothlisberger < >> Thus, you need to connect the backplane (the common) to an AC signal. >> Set up a 555 timer circuit for something like 50 kHz, 50% duty cycle, >> and connect it to the common. > Also, I think a backplane drive of 30 Hz is more common. The > backplane is driven with the 30 Hz 5Vpp square wave. Something else you missed was the absurdity of using a 555 (all the more since it would have needed to be a 7555 anyway) to generate the backplane drive (where segment drive is supplied by a PIC - a major synchronisation nightmare). He was presumably going to suggest using EXOR gates for all the segments too? -- Cheers, Paul B.