TFT LCD Projects/Ideas/Applications Open Discussion.

[Ian Rogers]

I have built the prototype..

I'm getting 16.66Mhz SPI with no problems.... I have had to move to MODE1 ( low idle and clock on trailing edge )..

As I have a 240 x 320 screen Update seems a tad slower than yours Eric.. I have the chip at 64Mhz on the internal OSC and PLL... so I can't go any faster!! I'm taking it home for Christmas so I can play and get a video done..

 

[ericgibbs]

hi Ian,
That's a fair SPI rate.
I found that the Arduino routines used Unsigned WORD's for the pixel settings, checking shows that the High Byte of the address Word is always 0x00.
Passing a BYTE for X and Y Addressing [ a WORD for colour is required] saves a little time and memory space..
Also IMO the additional external Calls within the Set Pixel procedure are time wasting, I have added the Code in these external Calls to to the Set Pixel procedure.

If you are using Oshonsoft Basic.? check the operation of the RS232 serial HW UART, I had to Define the Clock as 60MHz, even though the actual internal clock was 80MZ, in order to get the correct division rate for the HW UART.

E

 

[Ian Rogers]

so I can't go any faster!!

I Lied!! I have re-witten the LCD clear function and is now the same speed as yours...

If you are using Oshonsoft Basic.?

Nope!! Gave that up ages ago... MPLABC18!! I haven't started on the USART!! The pic18f26k80 has a second USART on the programming pins, So they are already exposed...

 

[ericgibbs]

hi,
Using the PIC's onboard ADC, set for Byte output, ie: 0 thru 255 counts, using SetPixel, the 9225 LCD displays a very good 5Vppk 100Hz waveform.
Lower ppk sine and triangular waveforms upto ~200Hz look OK.

As the waveform pixel is set, the 'y' component [theADC output count] is also stored in a 250 long buffer.
When the next waveform sweep occurs, the buffer is Read and the pixel set as Black just before the current pixel is set.
This erases the 'old' plot one pixel at a time just before the new pixel is set.
By using alternate colours for each sweep, it makes it easy to view the TFT/signal waveform.

The X axis timing is set by a delay between plots and increments by one for each ADC reading.

E

 

[camerart]

Hi All,
I was just looking through this post with interest, but it is something for me in the future.

These are the screens I bought : **broken link removed**
Would they work with the Programs here?

On a different point, regarding Crystals. When choosing crystals for a PIC, is it best to go for the fastest shown in the data sheet, or one that divides better for the project, 'say' 16MHz?
C.

 

[atferrari]

and BYTE6 is a random Seed.

Beau Schwabe

Hola Beau,

What is that random part for? Intrigued.

 

[Nigel Goodwin]

Hi All,
I was just looking through this post with interest, but it is something for me in the future.

These are the screens I bought : **broken link removed**
Would they work with the Programs here?

On a different point, regarding Crystals. When choosing crystals for a PIC, is it best to go for the fastest shown in the data sheet, or one that divides better for the project, 'say' 16MHz?
C.

I bought a couple of those, unfortunately they need a LOT of I/O pins to use, as it's a parallel interface.

Clock speed depends entirely on what you're doing, mostly I choose PIC's with internal oscillators and don't use external crystals, it wastes two I/O pins on any remotely modern PIC.

 

[camerart]

I bought a couple of those, unfortunately they need a LOT of I/O pins to use, as it's a parallel interface.

Clock speed depends entirely on what you're doing, mostly I choose PIC's with internal oscillators and don't use external crystals, it wastes two I/O pins on any remotely modern PIC.

Hi N,
Ok, thanks.
I did say sometime in the future, so it looks like a distant future if at all.
C.