Tandem FPV

[tungym]

I found my tandem trex is so stable that it would be a great platform for FPV (first person view) flying.

I am learning the video taken and will try keeping the horison in the middle.

The tandem has so much payload that putting heavy gears are no problem.

[Gra55h0pper]

Cool! Show us some of the resulting videos... :D
I've considered going FPV, but the thought to use the Twinn-Rexx hadn't crossed my mind yet. I'm sure you must be (one of) the first!

[tungym]

I just forgot to bring a DV tape and so cannot take the video on the first flight.

Range test of a stripped home 2.4G AV transmitter only reached 200m. So not reliable for long range FPV yet. Also spektrum DX7 can cause much interference on the AV receiver if the Tx is closer than 1 meter. So I would buy a 1.2G 2W instead.

Trex tandem is a rather good starting FPV platform because of the inherent stability. You can actually hands off the transmitter for almost 10 seconds in calm condition.

[billyd]

I have an fpv setup on my 450se. It's very cool, BUT it's important to have a friend with you. The limited field of view makes it difficult to have good situational awareness. You could be backing up to a tree or the like without knowing it. Also an important safety measure in case you are floating toward a person.

But it's alot of fun. I am working on a more sophisticated version in one of my planks that will have pan and tilt (attached to a head tracker) to improve the view problems associated with one camera.

Also, I'd love to get artificial horizon superimposed on the image (like a hud) along with a radar altimeter... ahh to dream. You could make a home-made horizon pretty easy, although I have no idea how to do an altimeter with low power and weight. lol. That would be a great project.... for somebody else :YeaBaby:

[tungym]

I was thinking of either using a wide angle lens (may be difficult to find one that fit on the CCD), or using car convex side mirror (or dental mirror)putting in the front. I am not sure whether pan/tilt is the best , or vice versa. My feeling is pan/ tilt is annoying if you are flying in confined area, but perfect in open space.

Since I plan to fly in open air space, my worry is the AV transceiver range rather then pan/tilt.

One more important issue is that the annoying vibration comes again when I attach some gears on the frame.

[v22chap]

cool touch tungym ,,,keep it going :thumbup:

[billyd]

Well I wasn't going to mention it, but I am working on an idea I've had for some time.

Stereo cameras with two seperate signals sent to an eyeglass monitor capable of two inputs to create 3d vision. Each camera would have a 90 degree fov which is about equal to a human eye. Set the two cameras about 75 mm apart and turn them at about a 5 to 10 degree each divergent line from center. That will give an approximation of what a pair of human eyes see and give actual depth perception allowing much better situational awareness and make judging distance to objects and the ground much easier. However, that still won't tell you what is behind you.

[v22chap]

Keep it comin billyd:hug::thumbup: sounds like a neat idea .

[tungym]

Well I wasn't going to mention it, but I am working on an idea I've had for some time.

Stereo cameras with two seperate signals sent to an eyeglass monitor capable of two inputs to create 3d vision. Each camera would have a 90 degree fov which is about equal to a human eye. Set the two cameras about 75 mm apart and turn them at about a 5 to 10 degree each divergent line from center. That will give an approximation of what a pair of human eyes see and give actual depth perception allowing much better situational awareness and make judging distance to objects and the ground much easier. However, that still won't tell you what is behind you.

That is cool but definitely it needs to be adjusted pecisely as stereo image can cause great brain strain !!(like changing a new pair of glassess) But my gut feeling is that stereo image (depth perception) is of lower priority than a wider angle of view, in terms open area piloting. Definitely it would be nice to have a laser guided missile @!!!!

[billyd]

Well in addition to the great depth perception that a stereo camera system would provide, by using 2 x 90 degree fov lenses set on divergent lines, you would get the same view that a human being gets looking straight ahead, about 160 degrees !!

If done properly, the view would far surpass any single camera system and cause less strain on the eyes. Don't forget we are not simulating 3d here, this would be true 3d, just as your eyes work right now.

[tungym]

Thanks for your information! So it will deliver higher resolution and less distortion than a wide angle lens.

[sokal]

this site has a Hardware worth looking in to for FPV
http://www.gentles.ltd.uk/gentvu/altiosd.htm

[zphere]

There is an OSD board ("OSD" henceforth) that you can build, if you want to. A previous poster linked to the prebuilt version.
http://www.kapelec.com/altiose1.htm

SparkFun has a nice ultrasonic range finder for $28 USD/pc.
http://www.sparkfun.com/commerce/product_info.php?products_id=8502
This reads from 0 to 255 inches, with 1 inch precision, and outputs the reading as 10mV/inch. (0.01 volts per inch)

The OSD has two input pins where it will read the voltage and display it on screen.
It starts displaying voltage at 60mV, and then displays in 20mV increments. (0.06V, 0.08V, 0.10V, 1.02V, 1.04, 1.06V, ...) Using it, you would have two inch precision, starting at six inches. You could point one range finder down to get altitude from 0 to 21 feet (in inches), and then one aft so that you know if the aft end is getting close to anything. (tree, building, personnel, et al.) The OSD displays the voltages as (a.bcV) so you can just imagine the decimal point as not being there and then read off integer inches (from 6 inches, on up jumping by 2 inches each time)

About the version you build, vs buy; I think that the voltage reading starting at 60mV and jumping in 20mV increments is purely a software restriction. I am fairly (75-80% ??) confident that the PIC16F876 would be able to read with at least 1mV precision. I've not read the assembly source, but if it is purely a software limitation, removing it would not be a huge task. Furthermore, and this would be even easier, would be remove the decimal point, and change the "V" (for volts), to double quote inch symbol.

If you build it, you can modify the source to display what you want, or buy the prebuilt version and accept the 6 inch, jump by 2 inch limitation. The source is 193KB of assembly, the first 40K of which is only lightly commented in FRENCH! The one the previous poster linked to is http://www.gentles.ltd.uk/gentvu/altiosd.htm and costs about $208 USD, so that's a fair chunk of money.

Fun Stuff, eh? :YeaBaby:

-zphere

[billyd]

Great ideas zphere. Very simple to implement. :thumbup:

Assembly language... ugh. Haven't looked at that stuff for around 10 years. It would take me about that long to figure out uncommented source code in assembly.

I think if I were to do this I would have to deal with a simple voltage display, unless some other kind soul would like to post a program, lol.

Anyway I've got too many fires going right now before I upgrade my fpv setup. I'm still trying to get my tandem to fly in a straight line, lol. Of course if the wind ever stops blowing here, that would make life a little easier.

[zphere]

Here's just a quick update;
The output of the sonar is not 10mV/inch as stated at SparkFun, but actually Vcc/512 per inch. To get 10mV/inch out, it would need to be powered at exactly 5.12 volts.

-zphere