pellepl wrote:That's good news! Would you care to enlighten a poor soul like myself, I'll put a picture of you on a piedestal by my desk I keep banging my head on this, close to coming to a nervous breakdown.
Did you make your own PCB - if so, I presume you used the QFP package? Did you utilize the PLL, and if so, did you make use of all the guidelines concerning PLL power supply? Also, what oscillator did you use?
Basically, I'd like to see your design (and probably steal it right off), but where's the sport in that?
Lol. I know how you feel. My team and I spent a ton of time getting it to work. And since it was such a pain and there isn't a lot of clear information, I am happy to share.
I used the QFP package, the board I built I wirewrapped. I used an adapter to convert the QFP package into something wirewrappable. But that doesn't really effect the operation of the design.
I used a 1MHz square wave input (from a can oscillator) and then using the PLL I upscaled it to 45MHz internally and then divided that down by 5 to generate the 9MHz PCLK required by the screen.
As for the power supply guidelines, I tried to get some more information out of some FAEs at Epson without too much luck. But basically I ended up not using it. However my power supply rails only had 50-90mV variances. But the data sheet makes it sound mega dangerous to have any variance.
Also it is critically important to put the device into sleep mode before you configure the PLL registers otherwise it won't work. There is a register to control the sleep mode.
I used a PIC18 microcontroller to run my system and it talked to the Epson chip.
I am planning to write up how exactly my team and I did it and I can even provide a register dump of the Epson chip. And some things we learned about the screen in the process. I'm hoping to write it up this weekend, when I do I'll post back.