SSD1963 EVK1100 interfacing

1 post / 0 new
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ok, it turns out that there is probably an input voltage error which causes the resets. I have put a capacitor on each panels voltage inputs (2200uF each) and it now looks good.

I guess, if everything will be put on a self made demo board, then it will work as supposed.


I am working on a project interfacing a SSD1963 display controller with the EVK1100. I use this SSD1963 eval board from Techtoys:

I have created a break out PCB to connect the two eval boards (see attached picture).

The problem which occurs is the following: Sending certain information from the AVR32 to the SSD1963's picture buffer causes the SSD1963 to reset. This can be reproduced.

When I don't power up the Techtoy eval board on it's 5V input, then however it functions correctly. Doing this it draws its power from the EVK1100's I/O connections. The SSD1963 1.2V domain has 1.2V, whereas the 3.3V domain is in about 1.8V, which according to the data sheet is above the recommended min. 1.65V. Using this setup causes the backlight to be very dim, as the charge pump does not get it's required 5V.

This made me think, maybe there is a Voltage tolerance problem between the AT32UC3A and the SSD1963.

According to the data sheets the AT32UC3A requires the following inputs:
Low level: max. 0.8V
High level: min 2V
The SSD1963 outputs the following levels:
Low level: max. 0.2xVDDio = 0.2x3.3V = max. 0.66V
High level: min 0.8xVDDio = 0.8x3.3V = min. 2.64

The SSD1963 requires the following inputs:
Low level: max. 0.2xVDDio = max 0.66V
High level: min 0.8xVDDio = min 2.64V
The AT32UC3A outputs the following levels:
Low level: max. 0.4V
High level: min 0.4xVDDio = 0.4x3.3V = min. 1.32 (edit: incorrect)
High level: min. VDDio-0.4 = 3.3V-0.4V = min. 2.9V.

So you can see, that the min. high level output of the AT32UC3A is lower than required by the SSD1963. This, and the fact that when the SSD1963 VDDio is running on ~1.8V then the controller works fine made me think that there is probably a voltage problem.

My idea was to put a ~100k pull up resistor on every line between the AT32UC3A and the SSD1963. But I am not sure if this would work, so I decided to ask you guys, if you maybe have some similar experience in any way.

By the way, I use the EBI module to interface the display controller. The attached pdf contains, how they are interfaced. The 10k pull up resistor for PA25 (EBI NCS0) was removed from the EVK1100, but the 10k pull up for PA29 (Reset) is still on its place.

Any idea or help would be greatly appreciated.

Edit: Ok, I did see it wrong, the High level min for the AT32UC3 is VDDio-0.4 = 3.3V-0.4V = min. 2.9V. So the voltage tolerances are correct. There still remains the question, why it does work correct when powered by the parasitic voltage from the I/O connections of the AT32UC3.