Hi Jean Claude Just received my 2 graphics boards and very, very nice they are too. However, I've tested both displays with 2 jeenodes and observed the following with both glcd_demo and glcdClock programs: 1. on the right hand side, there is a vertical column 3 pixels wide of gibberish ie non-sensical collection of active pixels. 2. in your programs, you've set brightness to 0x15. This makes the whole display so dark it can only be read from a very acute angle. setting it to 0x5 gives an excellent display. 3. with glcdclock, it gets real interesting! the bottom half of the digits appear in the top half of the display and vice-versa - looks like a b&w telly where the picture has rolled up half a screen - that's long before your time 8=) maybe there is a difference between your latest batch of displays and the original ones? regards, Bill
graphics board - weird problem(30 posts) (6 voices)
thanks lennart but jeefiles.equi4.com/st7565.zip with file dates 22/11 and 20/09 are the ones I'm using. with the simple glcd demo, the display shows, starting at the very top line "graphics board", "jeelabs", "arduino", "st7565" ie the display shows lines in the sequence 4,5,6,7,0,1,2,3 Bill
The brightness command is actually a misnomer, it's actually the twist of the crystals in the LCD and usually called contrast. It does change the viewing angle. So changing it varies the viewing angle.
The "rolled" screen effect is actually quite interesting. I've been doing a lot of work on the GLCD library, and there is provision in some mention in the specs for the ST7565 chip to have the horizontal banks in a different order to the default 0,1,2,3,4,5,6,7.
Can I ask which panel you have, is it the early one, the blue/white one (the one I have), or is it the newer (and clearer) more traditional silver/black one?
I hope JC won't mind me leaking this a little before it's officially made public next week (although some others have already found it on SVN), but would you mind trying the all new library from http://jeelabs.net/projects/cafe/repository/revisions/6397/show/GLCDlib
That is the library we're currently working on. If you could give that a try and confirm if the problem persists. Make sure you get the new demo from the examples library as the class name has changed to keep it separate from the original ST7565 library.
If so then I can make some changes to the library to support the alternate line bank mapping.
Hi tankslappa, it's the new display, the batch which came in circa 10dec with darkgrey on silver. i've just tried the new library with a clean restart etc and the result is unfortunately the same. many thanks, Bill
Right, I'll have to code a test program which you can run which will let me work out exactly the order of the horizontal banks, and also sort out the odd characters you're getting on the right hand edge. The later is probably caused by the ST7565 chip supporting a horizontal resolution of 132 pixels, whilst the panel is 128 pixels. The original panel had the 4 "missing" pixels positioned on the left hand edge, so as far as the graphics library low level code is concerned, the top left pixel is actually 4,0 (this is all taken care of for you by the library).
Can you just confirm the "jibberish" you are seeing on the right edge is still there with the library I pointed you to just now, and that it is 3 pixels wide (I would expect/hope it was 4).
tankslappa, I'm using the new GLCDLib. I found it last night after I assembled the new glcd and realized the ST7565 files weren't working for me. I realize now that (in the old st7565) the black/white are reversed. I also tried glcdclock and the displayed graphics were split horizontally in half.
When I'm using the glcdlib, I'm getting random text on the left, about 5-8 characters from the left. Sometimes numbers or smiley faces. It's usually only a single character and only once in a while.
I'm currently trying to get it to display the time and scroll other text. Mainly incrementing an int and writing it and scrolling. It seems to reset the jeenode after a short while. It doesn't always reset at the same place.. It generally lasts longer and longer until the next reset. Weird.
update: It seems to be working better now. Not sure what I did wrong before. I do notice that handleInput(Serial.read()) from RF12demo stops the glcd from working. No text or backlight.
Hi Jim, thanks for your info, it certainly sounds like there is a difference in the horizontal bank mapping, that's easy enough to fix, but a pain as it means there is going to have to be a switch or setting that will need to be changed depending on which LCD panel people have. I shall have a chat with JC about which way route he prefers. It could be that all the silver/black panels have the alternative mapping, and as they are the ones he is now supplying I will have that the default. The early adopters like myself with the blue/white will have a setting to change.
Regarding the corruption, is there any way you could take a picture of it for me? Same goes for you bgreene, it really would help. You can either send it to me on flickr (I'm the admin of the Jeelabs group) or email it to me at tankslappa dot com, anybody at that domain will come to me.
I'll take a picture if I see it again. Maybe it was my improper use of pointers and itoa. I haven't done C since school.
I'm currently just building up on the rf12demo sketch to use the glcd along with the rtc plug. "handleInput(Serial.read());" seems to cause the glcd to not work. No text or backight.
No backlight? How do you have the LED connected, via the jumper to power or the jumper to the interrupt line?
Have you called display()? All the function calls "paint" the pixels in the 1K ram buffer on the ATmega. They are only copied to the screen when you call the display() function.Posted 2010-12-19 ago #