I did this one in February of 2014.
I had this idea about being able to reduce the amount of wiring needed to hook up cameras, video switcher, OSD, transmitter, and get power to all the devices.
I'm going to back track a bit to show how the original proto board was made. First let's start with the parts. These first two shots are for size comparisons.
Switcher and 12volt step up regulator
Skylark Tiny OSD w/GPS and Current/volt sensor.
I'm going to make the parts list simple. This shot has all the parts for the project.
Going left to right on the back row:
Fat Shark CMOS and CMQ 1933X sitting on the box with another CMQ 1933X in it. The black square is another camera housings. Next are the PL1303 and PL1304 starlight cameras.
Going right still is the Fat Shark 5.8Ghz transmitter. Next to it the Skylark Tiny OSD with its add ons. Last on the towel is a pocket DVR.
Next row down on the left is header pins and breakout boards. Next is a 5volt step down regulator and below it is a 12vilot step up regulator. Next is assorted servo connector jumpers. Going right still is the 3 camera switcher. And last is heat shrink tubing.
Bottom row is my small tools needed for this sort of project. On the far right is a AR8000 receiver.
Off the towel on the right is my Spektrum DX7s transmitter, The Fat Shark Predator V2 goggles, and behind the is a 22" monitor.
The two starlight cameras were a great disappointment as the method used to pump up the image is purely digital and the image smears and lags badly. They're great for looking at some thing where neither the camera or object being viewed don't move.
After I got the board laid out in my head, it was time to put it together.
The pins for me have got to be the most difficult thing to solder and get straight. I cheated ok? I had one of these styrofoam take out boxes laying about destined for another project and just pushed the pins in to the board then rammed them in to the styrofoam box to hold them in place. Hey, it's works. LOL I did the same thing to get the pins in the tiny little stepup regulator board.
If you don't have what you need, improvise.
This pic is not the best I ever took but it shows you when the pins got soldered to the regulator, the result was a small cone shape blob of solder that didn't allow the board to sit flat. A piece of double sided tape trimmed out with a x-acto knife proved to fit just fine.
The rest of the pins didn't come out perfect but hey, it's a proto board. The out of focus wiring is camera one's video connection soldered to #1 pin set.
I added another set of pins for the power out to the OSD board.
This is how it wound up. This shot has all but the GPS board in it. It's got a really long lead on it. Everything is anotated.
The real magic comes from how the IC chip breakout board was put together. I just took advantage of the premade traces and designed the circuit around it. By using the 20 breakout traces seperatly I was able to reduce the over all wiring. Now keep in mind, this is a prototype. A proof of concept so to speak. It's still a little messy.
And here's how all of it came together to do its thing on the monitor screen. The colored lines in the picture are called park lines and are a function of the camera itself. I forgot to turn them off before taking the picture. Hmmm Maybe use them for a artficial horizon. They can be adjusted on the screen for size, placement, and color. The thing that got my attention is the 6 descimal place GPS co-ordinates. Pretty danged accurate.
Late Edit 09/17/2015: I was mulling over this project and thought it would be beneficial to add the following.
I had decided to try and automate when the cameras switch instead of having to do it with a knob on the transmitter. The position of the knob would throw off the break point between front and back cameras. After I made this video I mapped a 3 position switch whereas top most position is the front camera, the bottom most position is the beck camera, and the center was a neutral position letting the NAZA gimbal controller trigger the camera switcher according to position data from the IMU sensors in the flight computer.
The video is just to show the problem I had run in to and was asking for some advise as to how to solve it. But it does show it in action.
There's a lot more of this project but has little to no bearing on the electronics. I did some radical tweaking on the Hex frame moving the flight control computer to the top deck to get better stability over all.
What I did to resolve the knob resetting the end points on the gimbal outputs is switch over to a Spektrum DX9 transmitter. It has a lot more 3 position switches and the programming is a lot more flexable. I wrote up the settings for the platform and then mapped a 3 position switch to manually take over the camera switcher. In position 0 (top) on the switch, the switcher is locked to the forward looking camera. In position 2 (bottom) the switcher is locked to the rear facing camera. In position 1 (center) there is no manual control. The switcher is controlled by the gimbal outputs of the flight control computer.
I did manage to get it flown a couple of times but then the weather turned to mush and I couldn't get out to do any more engineering flights. Maybe this fall I'll drag it and the hex platform out and tinker with it some more.
Boards from Radio Shack.
RX/TX from Horizon Hobby / Spektrun
Cameras from security camera 2000
Regulator boards from Polulu
Goggles / FPV TX from Fat Shark
Monitor from Westinghouse / NewEgg
Skylark OSD from Helipal
Servo leads came with the RX and DJI NAZA Lite kit
Total cost around $700.00usd. The DX9 was $400.00usd by itself and is factored in to the total cost.