One thing to add regarding buttons is that we find using bitmap icons is often a good way where you have lots of buttons. If your icons are all the same size, using bitmap cells (where you create a large bitmap image which is one button wide and many buttons high) is a really handy way of doing this with minimal display list content plus you can give the icons your own desired appearance. You then place the icon by VERTEX2II which includes the handle and the specific cell number you want to place. Or by the CELL instruction followed by VERTEX2F.

There are several ways to implement this including:

• Having one image per icon and use conditional code to do one of the following when a variable indicates the button should be 'pressed'

- using the bitmap transform to re-size slightly
- add a slightly larger rectangle behind to give a glow round the edges
- set COLOR_RGB before the icon is placed and re-color to indicate pressed

• Having more than one image per icon which you select using the cell number using similar conditional code:

TAG_MASK(1);
BEGIN(BITMAPS)
COLOR_RGB(255,255,255)  // white displays following bitmaps in their original colours
TAG(Button_0);
If(Button0_Pressed == false)  // Button0_pressed is a local Boolean in your code
VERTEX2II(100,100,5,0)  // display cell 0 which is your button 0 unpressed
Else
VERTEX2II(100,100,5,1)  // display cell 1 which is your button 0 pressed
With most of these methods you can also keep the display list size the same regardless of button state.

As you mentioned, it is often quite efficient to re-send the entire list especially when using SPI burst writes when something on the screen changes if you have a lot of dynamic objects. For anyone looking to use the append in the original way (with RAM_DL content rather than commands) then you may also find using icons rather than buttons along with using multiple appended sections as smaller building blocks helps to minimise variation in length of the DL.

Thanks for your inputs Rudolph and Jesse, I'm sure your experiences here will help other users too,

Best Regards,
BRT Community

Hello,

Thanks for pointing that out, we will update the documentation.

Best Regards,
BRT Community

Hello,

I cannot see anything inherently wrong with your schematic.

I would suggest checking that all the power supplies are correct and ensuring there are no shorts on your board.

Best Regards,
BRT Community

Hello,

Theoretically you can set the REG_PCLK register to equal 1, but I would recommend against it.
I am going to make an assumption here, but surely the reduction in graphics intensity would negate the need for a higher frame rate anyway?

Yes, you would need re-initialize the IC to change the REG_PCLK settings.

Best Regards,
BRT Community

Hello,

It is not recommended to run REG_PCLK = 1, this will significantly short the computation time for each pixel.
Theoretically it is possible, but the likelihood of you running into issues when adding components on the screen and the GPU not having enough computational time to calculate for a given pixel is high. So yes, you can run the PCLK at 60MHz (if your chosen LCD supports this), but I suggest once you actually start developing your application you will need to lower this.

Again, EVE is designed primarily to be an HMI, and is optimized to give a good visual experience for such use cases, it is not designed to be for example to be a high frame rate video player or similar.

Best Regards,
BRT Community

Hello,

Please have a look at the following application note, this has an example of loading a custom font and an image concurrently:
https://brtchip.com/wp-content/uploads/Support/Documentation/Application_Notes/ICs/EVE/BRT_AN_014_FT81X_Simple_PIC_Library_Examples.pdf

Best Regards,
BRT Community

Hello,

There should be a min. 300ms delay followed by checks of the REG_ID and REG_CPURESET before you can expect FT81x to be ready.

    MCU_PDlow();                                                                // PD low                                                               
    MCU_Delay_20ms();
    MCU_PDhigh();                                                               // PD high
    MCU_Delay_20ms();
 
    Set ext clock here if required
    MCU_Delay_20ms();

    EVE_CmdWrite(FT81x_ACTIVE, 0x00);                                           // Sends 00 00 00 to wake FT8xx
   
    MCU_Delay_500ms();                                                          // 500ms delay (EVE requires at least 300ms here))
       
    // --------------- Check that FT81x ready and SPI comms OK -----------------
   
    while (EVE_MemRead8(REG_ID) != 0x7C)                                        // Read REG_ID register until reads 0x7C
    {
    }
     
    while (EVE_MemRead8(REG_CPURESET) != 0x00)                                  // Ensure CPUreset register reads 0 and so FT81x is ready   
    {
    }

Now continue to set display registers, load bitmaps/fonts etc.

Best Regards,
BRT Community

Hello

Please see table 4.11 in the BT81X datasheet (page 27), it is not possible to run the PCLK at the same rate as the system Clock (60Mhz in your calculation), the highest achievable would be 36Mhz:
https://brtchip.com/wp-content/uploads/Support/Documentation/Datasheets/ICs/EVE/DS_BT81X.pdf

This would give you an approximate FPS of (36,000,000/679,540) = ~ 53.

I would however advise that running the PCLK at the fastest possible rate reduces the computation time available for each pixel on the screen. This can cause issues if the GPU runs out of time to calculate what it should display for a given pixel if many graphics elements intersect this given point. In simple terms the faster you run the PCLK the less complex of a Display List you can run on the screen.

Best Regards,
BRT Community