AVR Freaks

Do you really own a SPFD5408 controller?
Where did you buy it?
Is it possible for me to buy one?

Note that tft.readID() will return 0x5408 for a genuine SPFD5408.

David.

You are living in a world of Make-Believe !!!

Whether it is Ebay, DX, BangGood,  ...  they all reproduce the same rubbish.   e.g. http://www.dx.com/p/diy-2-4-tft-lcd-touch-screen-shield-expansion-board-for-arduino-uno-350121#.VkGxSr9n0ac

says:

- 4-wire resistive touchscreen - 240 x 320 resolution - spfd5408 controller with built in video RAM buffer - 8 bit digital interface, plus 4 control lines - Uses digital pins 5-13 and analog 0-3. That means you can use digital pins 2, 3 and analog 4 and 5. Pin 12 is available if not using the micro SD - 5V compatible, use with 3.3V or 5V logic - On-board 3.3V 300mA LDO regulator - Support 2GB micro SD TF card

Now look at the legends printed by each pin on the pcb.   The TFT use digital#2-9 for the data bus and analog#0-4 for the control signals.  The microSD uses digital#10-13.

Yes,   I am sure that several years ago,  there existed shields with the Orise Technology SPFD5408A controller chip.    I have never seen one.    If anyone has got one or knows a source,   I would happily buy it from them.

Incidentally,  you can try my library attached to https://forum.arduino.cc/index.php?topic=337473.msg2373456#msg2373456

It handles most of these "red" shields.    If you have one that it does not recognise,  please run the examples/LCD_ID_readreg.ino sketch.

If you are prepared to test sketches on a controller that you have,   I am happy to add support for it.

David.

It really does not matter what you have done to an Arduino during development.
Just use the Arduino IDE to Burn Bootloader. This will get you back to normal.
Solder RESET-EN after debugWIRE.

David.

david, modified the program a bit to see what whould happen...

   lcdInit();
    lcdReset();      //ensures that controller is in default state
    readReg(0x00, 16, "ILI9320, ILI9325, ILI9335, ...");
    readReg(0x10, 16, "Manufacturer ID");
    readReg(0x20, 16, "ILI9481");
    readReg(0x30, 16, "HX8357");
    readReg(0x40, 16, "ILI9341, ILI9488");
    readReg(0x50, 16, "ILI9320, ILI9325, ILI9335, ...");
    readReg(0x60, 16, "Manufacturer ID");
    readReg(0x70, 16, "ILI9481");
    readReg(0x80, 16, "HX8357");
    readReg(0x90, 16, "ILI9341, ILI9488");    
    readReg(0xA0, 16, "ILI9320, ILI9325, ILI9335, ...");
    readReg(0xB0, 16, "Manufacturer ID");
    readReg(0xC0, 16, "ILI9481");
    readReg(0xD0, 16, "HX8357");
    readReg(0xE0, 16, "ILI9341, ILI9488");
    readReg(0xF0, 16, "ILI9327");

to dump the entire structure.

the output now is:

Read Registers on MCUFRIEND UNO shield
controllers either read as single 16-bit
e.g. the ID is at readReg(0)
or as a sequence of 8-bit values
in special locations (first is dummy)

reg(0x0000) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     ILI9320, ILI9325, ILI9335, ...
reg(0x0010) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     Manufacturer ID
reg(0x0020) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     ILI9481
reg(0x0030) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     HX8357
reg(0x0040) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     ILI9341, ILI9488
reg(0x0050) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     ILI9320, ILI9325, ILI9335, ...
reg(0x0060) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     Manufacturer ID
reg(0x0070) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     ILI9481
reg(0x0080) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     HX8357
reg(0x0090) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     ILI9341, ILI9488
reg(0x00A0) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     ILI9320, ILI9325, ILI9335, ...
reg(0x00B0) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     Manufacturer ID
reg(0x00C0) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     ILI9481
reg(0x00D0) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     HX8357
reg(0x00E0) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     ILI9341, ILI9488
reg(0x00F0) 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00     ILI9327

so Something is really fishy on what is going on here

David,

I had bought 2 'uno' clones and 2 displays.

I just took out the second board and second display that behaves the same.

also on the second uno the first display behaves identical.

I know they tested the touch part as I can see that they wiped with a 'pen' over the protective foil and left a mark on it.

I would suspect that then they also did a quick display test.

I have not attached a scope yet, but that slowly seems to be the way to go.

unfortunately I will not be ad home tonight, but might have a look tomorrow.

I have lifted one of the displays from the shield and it gives me the follwoing numbers:

S241GF31 d09j06010 usp5280371

also on the flex there is ck241_rev0.1 2009.03.07 BH0923

IS9F01COP1P

if the reset line was stuck I would not be able to get garbage on the display, and I have managed that. might still be that the rd line is not ok.

my guess is that the first number is interesting

S241GF31

that seems to relate to it being a samsung display.

I have to go now, but perhaps this does ring a bell. I did see more people complainging about this although google initial search was only a couple of results

Hi David,

I have taken the swtft-shield stuff and teh adafruit-gxf_library master

All I have changed is:

void setup(void) {
  Serial.begin(9600);
  Serial.println(F("TFT LCD test"));

 

  tft.reset();

  uint16_t identifier = tft.readID();

 
    Serial.print(F("LCD driver chip: "));
    Serial.println(identifier, HEX);
    
//identifier = 0x7783;
  identifier = 0x9341;
 
  tft.begin(identifier);

Both 0x9341 and 0x7783 as identifier have the same result.

I see the display flickering, that suggests that atleast it is responding to something of the data written to it.

My next step now is that I will de a re-solder of the shield PCb and teh arduino uno clone pcb I have to make sure that therre is no short circuit on there somewere.

There are several libraries similar to mine. My one probably supports more different controllers. For example, SWTFT only supports the ST7781. THe Adafruit Master supports several controllers.

It does sound as if there is something seriously wrong if you cannot read the chip.

You can just try every chip on the MCUFRIEND_kbv list in tft.begin(id)
but I suspect there is a wiring problem.

David.

As I say in the Read_ID sketch,   there are two styles of command.

1.  16-bit command + 16-bit data:    ILI9320/9325/9335, S6D0154, ST7781, R61509, R6809, SPFD5408, SSD1289, ... 

2.  8-bit command + variable number of 8-bit parameters:  ILI9327/9340/9341/9481/9486/9488, R61581, ST7735, ...

The first type generally return a 16-bit ID from a readReg(0)

The second type have their ID in different places and often require reading a different number of data bytes.

Quite honestly,   it seems crazy that the Manufacturers do not all follow the readReg(0) pattern.   You can always read extra registers after you have determined the "basic" controller model.    Most Manufacturers seem to produce controllers in both styles.

In answer to your OP.   The Mcufriend UNO style shields might contain any of the 10 different controllers that I support.    And it looks as if meslomp might have a "new" one.

At least the shield "should" be readable.    Which means you can identify the correct controller at runtime.    You can test every controller with the same binary running on a UNO.    The UTFT library requires you to "know" the exact details of each display.    You can rely on the documentation from ColdTears or EastRising.   When it comes to DX, BangGood, Ebay, AliExpress, ... you just have to guess.    Most of their information is completely untrue.

Davd.

David,

I feel very very very ashamed.

I think I have really F...d up somewere in the process of making the modifications that are written in the post #79 on the arduino forum:

http://forum.arduino.cc/index.ph...

I again downloaded:

https://github.com/adafruit/TFTL...

and

https://github.com/adafruit/Adaf...

did again the modifications to them as described in the arduino forum post and now everythings seems to be working just fine.

I have attached both libraries as I downloaded them.

Changed the "graphicstest" file and saved it under "graphicstest-new"

for me this has resulted in a working display on both my UNO boards + shield.

David,

I will next try your modifications to your readid program to see that that works on my boards too.

Simonetta,

I hope your board will also work with the zip file and example I attached.

regards

edit:

Simonetta, my display seems to be color inverted. during text rotation the background should be black and the text white, yellow, red and green.

in the "adafriut_TFTLCD.cpp file, there is a function called "void Adafruit_TFTLCD::begin"

inside that function there is a section that starts with "} else if (id == 0x9341) {"

there is a write to register "ILI9341_SLEEPOUT"

below that write I have added the line "    writeRegister8(ILI9341_INVERTON, 0x00);    // see registers.h"

now for me the display functions as the demo code was intended to.

hope this helps

I just replied and lost it.

Anyway, I would put money on you having an ILI9341 that has its extended commands disabled.
Just call tft.begin(0x9341) and it should work.

It is definitely a variable-sequence style controller. Of course it might be Himax, Sitronix, Samsung, Renesas, ... and not Ilitech.

In which case, it is worth using the power up defaults. Please let me know how you get on.

Tft.begin() should default to ILI9341. I am not at a PC. I will look to see what code you are actually using.

David.

Hi david,

did you try the code I attached in the zip file?

you should go to the "graphicaltest_new" it should be present in "libraries\tftlcd-library-master\examples" The original "graphicaltest" that is in there to indeed does not work on my LCD. I had to tweak it and then had to save it to a new location as the original was protected so I called it new.

Hi David,

I took the "graphictest_kbv"

changed

//    g_identifier = tft.readID(); // <- commented it out

added

    g_identifier = 0x9341; <- to make it hard coded the display controller we think it is.

the display seems to be doing something, but the background keeps changing from dark to light and there are lines over the display that change color.

So my guess is that indeeed there is a difference in the initialisation that makes the other code work.

Hop you find it adn can use it to your advantage and make your code better.

That probably has to do with how the manufacturer has wired the default settings.

I scanned through the ili spec and IIRC I have seen that the default colors and orientation could be done by setting pins to a certain state. So it might as well be one manufacturer choose one setup and the other the inverted.

Although it can also be that they wired it the complete opposite for one or the other reason.

I have no idea what you have.  But it is a modern "8-bit sequence" style.  Try a few more registers:

void setup()
{
    Serial.begin(9600);
    while (!Serial) ;
    Serial.println("Read Registers on MCUFRIEND UNO shield");
    Serial.println("controllers either read as single 16-bit");
    Serial.println("e.g. the ID is at readReg(0)");
    Serial.println("or as a sequence of 8-bit values");
    Serial.println("in special locations (first is dummy)");
    Serial.println("");
    lcdInit();
    lcdReset();      //ensures that controller is in default state
    readReg(0x00, 2, "ILI9320, ILI9325, ILI9335, ...");
    readReg(0x04, 4, "Manufacturer ID");
    readReg(0x09, 5, "Status Register");
    readReg(0xBF, 6, "ILI9481");
    readReg(0xD0, 2, "HX8357");
    readReg(0xD2, 3, "NVM Read");
    readReg(0xD3, 4, "ILI9341, ILI9488");
    readReg(0xDA, 3, "RDID1");
    readReg(0xDB, 3, "RDID2");
    readReg(0xDC, 3, "RDID3");
    readReg(0xEF, 6, "ILI9327");
}

I presume that you have tried tft.begin(0x9341) but it would be far better to identify the actual make and model properly.

David.

This display contains a Samsumg chip named S6D04H0.

It is roughly identical to well-known ILI9341 but with a different initialization sequence.

This tested source-code snippet works for that display:

http://www.tu-chemnitz.de/~heha/...

henni