LCD - Slightly off topic (brief) - Contrast issue - SOLVED

[Gabriel]

I was testing out some LCDs last night using PCM's Flex driver _which worked flawlessly as soon as i hit compile_ (thanks PCM!) but getting the contrast right was imposible...

I tried:
resistor to VCC
resistor to VDD
5, 10, & 20k pots
5, 10, & 20k pots with buffer amp (LM358 Unity gain)
direct to VCC
direct to VDD
Fixed resistor voltage divider

Power from Pickit 3 or power from FTDI RS-232 converter from Sparkfun
(approx 5V)

Nothing made rich dark characters....
Same problem 2 different screens

all i got was extremely faint (but correct) letters which i could only see by tilting the screen at the right light reflection angle and prayer to multiple gods...

anyone have any idea what might be happening?

EDIT: i should point out that the closest i got to any kind of contrast was when the contrast voltage was near ground.

G.
_________________
CCS PCM 5.078 & CCS PCH 5.093

[Ttelmah]

Some LCD's, _require_ a -ve contrast voltage.

Best Wishes

[asmboy]

my generic test bed for LCD's uses a 20K - 270 degree pot with a range of +2.5 to -2.5 volts

i do this with a 10k fixed resistor on both ends of said 20K pot - and this provides a quick eval of what bias voltage a given model of LCD requires, given source voltage of +5 and - 5V .

[Ttelmah]

Yes.
Historically most of the early LCD's needed a -ve bias. Then there was a fairly long period where the commonest designs all used +ve bias values close to ground. A lot of the newer high contast displays have gone back to needing a -ve supply, and also units with extended temperature range often require this.
Model number of the LCD?. Data sheet should specify the Vee needed.
Remember the current on Vee is small. If you have a MAX232 on the board, the -ve supply on this can be used.

Best Wishes

[Gabriel]

one is a Sanyo LC7985NA/ND chip set - I didnt find any reference to -ve on the datasheet....

the other one is a COB...and i dont have the nscreen on me right now...

anyways... ill give the -ve voltage a try...

Thank you!

G.
_________________
CCS PCM 5.078 & CCS PCH 5.093

[asmboy]

if you have a usable low I , -v at the ready- i have a cute circuit that uses a rail to rail opamp, some R C parts and gives you smooth contrast control using a PIC PWM output . have used it in numerous customer designs .

here is the link to said LCD bias circuit

http://64.128.110.53/lcdbias.pdf

NOTE: it is NOT a TL072 -- ( which is NOT rail to rail)
use ANY true R/R OA - slew rate of 2uv/us is fine with typical PIC pwm freqs

if i have to explain how this works - you should not be using it

[Gabriel]

hey! thanks for the Ckt... but i do not have a -v supply...

i think i will, for now just use a MAX232 or a simple ch-pump to generate the -ve...

ill let you guys know if i find blue smoke inside my LCDs...

(i harvested them from a lucrative dumpster dive containing old office phones)

G.
_________________
CCS PCM 5.078 & CCS PCH 5.093

[bkamen]

You can build a charge pump fairly easily with a diode and a capacitor off a PWM line of the PIC.

I've done this before for LCD's. Easy to search with Google.

So if you have an extra PIC line that can be a PWM output.. that'd work too.

-Ben
_________________
Dazed and confused? I don't think so. Just "plain lost" will do. :D

[asmboy]

As bkamen says - you might be able to use the SAME PWM drive as the input to the circuit IF your PWM values are within +/- 25% of 50% duty -
BTW 50% duty = 0 vdc output.

And the 10K values can be scaled to 20K if current is really an issue.

I routinely use an LM828 to get my -5v for this purpose and running a couple CMOS R/R op amps in a design that is fully powered by USB 5V.

The LM828 is VERY simple to use and a very nice part overall.

[bkamen]

[Gabriel]

hey guys, thanks for your help!

indeed the negative voltage solved the issue!

i used 2 diodes, 3 caps and Resistor to build a simple chg pump (@500Hz)
(i added an extra RC filter on the output for a cleaner -ve)

worked perfectly...

Thank you All!

G.
_________________
CCS PCM 5.078 & CCS PCH 5.093

[Ttelmah]

Just as a comment, worth understanding, that it is the LCD that needs the -ve supply, not the control chip. Hence you can get a dozen modules using exactly the same PCB, and chipset, with different LCD glasses on the front, with some needing -ve voltages, and others not needing this. Hence it is the 'module' part number that is needed, not the chipset.

Glad we were right. At this point several of the older posters will be making some comment like "in my young day all text LCD's needed -ve supplies".....

Best Wishes

[Gabriel]

yeah that makes sense...

if i remember correctly its related to the Backplane signal having to be of "reverse polarity" as the active(on) segements.... something like that...

i thought that the voltage was fed to the glass through the chipset... i remember using a plain 7 Segment LCD(no controller) wich required the backplane to be modulated... so i just assumed the chipset would be involved...

thanks for the clarification though!

G.
_________________
CCS PCM 5.078 & CCS PCH 5.093

[Ttelmah]

They are more complex than that.
Simple seven segment displays have just two phases like this, but multiplexed displays have something like six voltages used, Usually named V1....V5, and the logic supplies. The intermediate voltages are done by division between Vss to Vee, but none are directly connected to Vee.

Best Wishes

[bkamen]