Strange QuickStart board and LCD (Wintk WDC2401P - a 24 x 1 line display)

HShanko

I assumed this LCD had a CMOS i/f, so connected the Data lines to P0..P7; same lines as the touchpad 'switches' use.

However after spending too much time trying to get the touchpads to work with the QuickStart with the LCD connected I find for some reason the LCD pulls up those Data lines and pull the Prop P0..P7 lines to about 3.96v Fortunately I anticipated using the Read mode to sense when Busy occurs, so have a current limiting resistor in series with P0..P7. Seven resistors are 1K and one a 4.7K; ran out of enough same values.

Scoping, I find the Prop lines are high and note there are those touchpad 'charging' pulses pulling the line down from the 3.96v level to about 3.3v. Thus the 100K resistors in series with the touchpads cannot ever pull towards to ground sufficiently to present a low level.

The LCD is never put into the Read mode. Presently I use the Write mode only. Thus the LCD Data lines should only appear as inputs. The LCD uses the USB generated '5v' power which runs about 4.88v.

I surely did NOT anticipate such a result; apparently the input levels draw more current than I'd expect. Anyone recognize why this effect is occuring? I don't have a schematic of the mezzanine board for the QuickStart module. I tried Jonny Mac's QuickStart touchpad demo and when the mezzanine board is installed, it appears to draw the Prop P0..P7 lines to clamping levels. But if I remove the mezzanine board, the demo program runs OK.

By the way, the LCD appears to otherwise function properly. I can display characters representing $00..$FF 16 characters at a time, along with their value beginning to end of the 16 characters.

[FONT=Parallax][SIZE=2]

LCD Control      (Reset, RS, RW, E)             ┌──────┐
   P8..P11        ───/───────────────│                 │
                                                                            │ 24x1       │──5v
LCD Data bus                                   R*           │  LCD       │
    P0..P7            ──/──┳───────── │                 │
                                         │                                └──────┘
                                         100K                                                     
                                         │                
   * 1K S0..S6                 │
     4.7K D7                  touchpad
[/SIZE] [/FONT] 

edit: For some reason, even though I finally got Parallax font on the iMac, the forum 'font selector' does not show it.

HShanko

Measuring across resistors between Prop and LCD, get 150 mv across a 1K for 160 µA and 330 mv across a 4.7K on the data buss. This neglects what the DMM draws. Seems excess leakage.

Chris Savage

The interesting thing about this LCD is that the pinout matches most standard panels except the Vo (contrast) pin, which on this one is a /RESET pin. Strange. Anyway, it's been so long since I worked with a Parallel LCD (one without a serial to parallel backpack) that I can't remember if the data lines are internally pulled high on the LCD side. It seems like we assumed they were when we (Parallax) first created some LCD test code and each data line had a 3.9K (4.7K would work) resistor in series. Of course, we were reading and writing data. In order to verify if the LCD has internal pull-ups you would have to disconnect lines form the LCD data and measure them with the LCD powered to see if they measure 5V.

HShanko

Thanks for the comments, Chris.

What documentation I could find on the LCD chip didn't include anything on the input current draw. I suppose I could just rewire to LCD data bus to a different 8 I/Os so as to let the touchpads be isolated from THIS LCD. Might make the LEDs flash a bit, but that isn't as drastic as the 'pull-up' that the LCD causes. Those touchpads have to be capable of being discharged towards ground.