Puzzling behavior from IMU, Arduino, serLCD

I am nearly stumped with a problem I’ve encountered. Here is my setup: I have a 6dof atomic IMU running through a 3.3v to 5v signal converter, so that it can communicate with an arduino. The arduino simply checks to see if there is data in the serial buffer, and then outputs either “no data,” or “data available” to a serLCD. My custom (perf board) power supply takes 8.5v unregulated, and regulates it to power both sides of the logic level converter, and also the IMU, Arduino, and LCD. The voltage regulators have caps on the input and output to smooth out fluctuations, but I didn’t put them in my crudely drawn schematic. All the grounds are connected via breadboard and then to chassis. Also, there is only one ground wire coming from my power supply board. Ok, so here is the strange behavior I’m getting: When I power up the board, it initializes the IMU (which responds and begins sampling) and as long as I leave the 3.3v switch open the Arduino outputs “No data” to the LCD, as it should. When I close the switch, and the 3.3v side of the logic converter gets juice, the buffer on the Arduino fills, and it will output “Data available” to the LCD, but with random/missing characters (i.e. D#ta avail *&) like something is interfering with the signal to the LCD. If I switch the 3.3v open again, then the LCD will read “Data available” like it should. So somehow, when the 3.3v circuit is closed, it affects what is put out on the LCD. I’m pretty much stumped so any advice will be greatly appreciated.

Below is the url of a crummy drawing of my power supply schematic.

http://www.diydrones.com/photo/power-supply

Ok so maybe to simplify, is it a problem that I’ve got all the grounds connected to the chassis of my project? The signal wire going to the LCD is isolated, so there isn’t any direct interference. I have the 8.5v from the wall, 5v and 3.3v from the power supply, and the grounds from the IMU, Arduino, and LCD all connected to a single rail, so is it possible that spikes to the ground could be causing garbage characters to appear on the LCD? It only occurs when the 3.3v circuit is closed tho, so that seems odd.

I’m pretty new at designing circuits, so if anyone has any ideas about what I’m doing wrong please help me out. Thanks.

Ok, well just in case anyone else has had a similar issue or was following this, I have narrowed the problem down and it seems to be a sofware issue whereby the process of data filling the serial buffer is interfering with the Software Serial ‘print’ function. In restrospect this was the wrong place to post this, but I thought that the problem was with my circuit design… Peace.