0.92.3 Ultimaker-Smartcontroller Combo not Functional

Wildseyed

I did a sanity check on a GT2560 (Ultimaker 1.5.7 Compatible) and Smartcontroller LCD 2004 using the Arduino Example LCD Display sketch, using the following pins, and it worked fine.

LiquidCrystal lcd(LCD_RS, LCD_EN, LCD_D4, LCD_D5, LCD_D6, LCD_D7);

LiquidCrystal lcd(20, 17, 16, 21, 5, 6);

The 0.92.3 Firmware configuration tool produces source with different values for these pins, but even after changing to the correct values, there is no display. The buzzer and Click Encoder pins are also not matched, but I cannot test that without display.

My guess is that some defines are overriding since I have not selected RAMPS or RUMBA. Is it the case that these two boards are the only ones supported?

Repetier

What display did you select? I don't think any display has preset lcd settings for ultimaker, so you need to set them as manual configuration instead. Or modify in ui.h one similar display to use the right pin numbers and select that one.

Config tool does not set any of these pins. If you use a predefined display that is known for that board it has preset pins.

Wildseyed

Yes, I came to the same conclusion after examining the source. The thing is, I did modify uiconfig.h and pins.h, using Marlin pin assignments, and information found on the Ultimaker Shield wiki, but I got only a screen full of a single character (Japanese). After this is when I loaded up the LCD demo code from Arduino IDE, just to double check. It did work there.

If I understand you correctly, there might be additional things to change in ui.h

Repetier

Not regarding LCD except charset and timing. You can also use the liquid crystal library instead of the faster build in lcd library. Just make sure for the buildin to use the 4 bit parallel version and it normally works well. And most important set FEATURE_CONTROLLER to use uiconfig.h

Wildseyed

Added the following in ui.h.

It workds, mostly. SD card is always detected, even when removed. The display is a bit glitchy here and there. Any ideas on the glitch? I don't see this in Marlin, even with longer cables.

#elif MOTHERBOARD == 37 // ULTIMAKER_157, GT2560

#define BEEPER_PIN 18

#define UI_DISPLAY_RS_PIN 20

#define UI_DISPLAY_RW_PIN -1

#define UI_DISPLAY_ENABLE_PIN 17

#define UI_DISPLAY_D0_PIN 16

#define UI_DISPLAY_D1_PIN 21

#define UI_DISPLAY_D2_PIN 5

#define UI_DISPLAY_D3_PIN 6

#define UI_DISPLAY_D4_PIN 16

#define UI_DISPLAY_D5_PIN 21

#define UI_DISPLAY_D6_PIN 5

#define UI_DISPLAY_D7_PIN 6

#define UI_ENCODER_A 40

#define UI_ENCODER_B 42

#define UI_ENCODER_CLICK 19

#define UI_RESET_PIN 41

#undef SDCARDDETECT

#define SDCARDDETECT 38

#undef SDCARDDETECTINVERTED

#define SDCARDDETECTINVERTED 0

#undef SDSUPPORT

#define SDSUPPORT 1

Repetier

You could check in ui.cpp

void lcdWriteByte(uint8_t c,uint8_t rs)

{

#if false && UI_DISPLAY_RW_PIN >= 0 // not really needed

SET_INPUT(UI_DISPLAY_D4_PIN);

SET_INPUT(UI_DISPLAY_D5_PIN);

SET_INPUT(UI_DISPLAY_D6_PIN);

SET_INPUT(UI_DISPLAY_D7_PIN);

WRITE(UI_DISPLAY_RW_PIN, HIGH);

WRITE(UI_DISPLAY_RS_PIN, LOW);

uint8_t busy;

do

{

WRITE(UI_DISPLAY_ENABLE_PIN, HIGH);

DELAY1MICROSECOND;

busy = READ(UI_DISPLAY_D7_PIN);

WRITE(UI_DISPLAY_ENABLE_PIN, LOW);

DELAY2MICROSECOND;

WRITE(UI_DISPLAY_ENABLE_PIN, HIGH);

DELAY2MICROSECOND;

WRITE(UI_DISPLAY_ENABLE_PIN, LOW);

DELAY2MICROSECOND;

}

while (busy);

SET_OUTPUT(UI_DISPLAY_D4_PIN);

SET_OUTPUT(UI_DISPLAY_D5_PIN);

SET_OUTPUT(UI_DISPLAY_D6_PIN);

SET_OUTPUT(UI_DISPLAY_D7_PIN);

WRITE(UI_DISPLAY_RW_PIN, LOW);

#endif

WRITE(UI_DISPLAY_RS_PIN, rs);

WRITE(UI_DISPLAY_D4_PIN, c & 0x10);

WRITE(UI_DISPLAY_D5_PIN, c & 0x20);

WRITE(UI_DISPLAY_D6_PIN, c & 0x40);

WRITE(UI_DISPLAY_D7_PIN, c & 0x80);

#if FEATURE_CONTROLLER == CONTROLLER_RADDS

HAL::delayMicroseconds(10);

#else

HAL::delayMicroseconds(2);

#endif

WRITE(UI_DISPLAY_ENABLE_PIN, HIGH); // enable pulse must be >450ns

#if FEATURE_CONTROLLER == CONTROLLER_RADDS

HAL::delayMicroseconds(10);

#else

HAL::delayMicroseconds(2);

#endif

WRITE(UI_DISPLAY_ENABLE_PIN, LOW);

HAL::delayMicroseconds(100); // add this line

WRITE(UI_DISPLAY_D4_PIN, c & 0x01);

WRITE(UI_DISPLAY_D5_PIN, c & 0x02);

WRITE(UI_DISPLAY_D6_PIN, c & 0x04);

WRITE(UI_DISPLAY_D7_PIN, c & 0x08);

HAL::delayMicroseconds(2);

WRITE(UI_DISPLAY_ENABLE_PIN, HIGH); // enable pulse must be >450ns

HAL::delayMicroseconds(2);

WRITE(UI_DISPLAY_ENABLE_PIN, LOW);

HAL::delayMicroseconds(100);

}

and add

HAL::delayMicroseconds(100); // add this line

as indicate din source. This is how liquid crystal would do it also that extra paus eis normally not needed. But this routine is it which sends data to lcd so if there is a problem it will be in this function and it can only be timing.