[danb35]

I'm still waiting for a few parts to be able to complete my system ($#&@* 7809s), but I have the WD v2.01 board assembled, the display (http://www.ebay.com/itm/261802361429) connected, and the firmware configured and loaded onto the Nano. When I power up the board using a USB phone charger to the Nano, I just get garbage on the screen. It isn't completely repeatable (i.e., the garbage doesn't always look the same), but one example is attached.

I've checked the solder joints and the wiring of the display, and I don't see any obvious problems, but I guess it's possible I'm missing something. I don't have the RX or TX boards running yet, but I expected I'd at least get an intelligible display on the WD board even if there was no data. Where should I start looking?

[werk_ag]

... but I expected I'd at least get an intelligible display on the WD board even if there was no data. Where should I start looking?

It should work even with no data.
Do you also get garbage when the WD unit is powered through the PC USB port?
How long are the cables between the board and the display?

[danb35]

Do you also get garbage when the WD unit is powered through the PC USB port?
How long are the cables between the board and the display?

Yes, but different looking garbage--see attachment. The wires are 10-12" long.

[werk_ag]

Can you please recheck the wires or try with shorter ones (5 to 6")?

[danb35]

Can you please recheck the wires or try with shorter ones (5 to 6")?

I'm pretty sure the wiring is right (picture attached, in case I've missed something after looking 3-4x). I can order some shorter wires, but it will take a while for them to arrive.

[werk_ag]

The wiring seems good.
The most common cause of garbage on the screen is due to interference in the SPI communication between the screen and the board, this is why I have suggested you to try with smaller cables, but maybe there is a problem with the onboard logic level translator, or even with the display.
If you could test the display with another arduino (with a logic level translator, those displays requires 3.3V logic) will be great, but I suppose you will not have the required parts at hand.

Just in case, you could also try to carefully reheat the weldings at the TXB0104 IS (this IC is sensible to ESD, don't touch it with hands without ESD protection).

[danb35]

Your mention of SPI communications trouble got my thinking, so I hooked up my scope to the SCK and MISO pins, first at the WD board, then on the display module. Unfortunately, I don't know that well what I'm looking at. I can tell the overshoot is quite a bit greater when plugged into the display module, but beyond that I'm not too sure.

Here are a few as measured on the WD board:
   
   
   

...and measured on the module itself:
   

[werk_ag]

Unfortunately those images also don't give me any clue.
Usually more simple approaches may give more quick conclusions. If you don't have shorter cables at hand, try with other set of cables, maybe there is a problem with one of the wires (sometimes happens with those dupond wires).

If you want to do some other tests with your oscilloscope, you can use channel A and B to check the same SPI signal (Arduino pin 13, 11, 10) and DC (Arduino pin 6) before and after the txb0104 IC. Both channels should display the same signal, just with smaller amplitude at the output of the txb01004 (display connector).

[strada916]

move the mosi to the miso see if that helps. :o) I could be wrong.

[danb35]

OK, here are more traces from the scope. Channel 1 (yellow) is on the nano; channel 2 (blue) is at the header for the display. All are at 2 volts/div vertical; horizontal is 1 uS/div except for the DC trace, which is 5 uS/div.
CS10 (pin D10 on the nano):
   

MOSI (pin D11 on the nano):
   

SCK (pin D13 on the nano):
   

DC (pin D6 on the nano):
   

What I'm noticing on these: MOSI, SCK, and DC show expected levels with good rise and fall. Ch2 (the output of the level shifter) is showing greater over- and under-shoot than Ch1. CS10, though, is showing much slower rise/fall, and the output of the level shifter is still 5 volts peak-to-peak. I'm not sure if this is what's supposed to happen, but it isn't what I'd expect.