Experiments with a lens camera - Tarot 680 Pro

I would like to fly a high quality stills camera without paying the earth, and ideally with an aircraft that is relatively lightweight.

I'm after making artistic still landscape photographs from the sky, so I don't need video. So the conundrum is that flying a full size DSLR requires a very heavy, expensive and potentially dangerous machine.

Currently my chosen compact camera platform is 4/3 so I became interested in a new class of cameras which dispense with the body and controls of a camera and simply provide the sensor and the ability to mount a lens in a unit which is controlled from a smartphone, iOS or Android.

I've experimented with both the Sony QX-1 and Olympus Air 01 and concluded that the Olympus was the more suitable of the two. Firstly it has better image quality in its 16Mp in my opinion than even the 20Mp Sony. Secondly the software programming API for the Olympus allows more control over the camera than the Sony option.

There are quite a few challenges in using this camera on a multi rotor, the main one being that I'll need to do a fair bit of systems integration to get it to work.

My initial plan was to mount the Olympus air on a Tarot 680 pro. I have purchased a DYS 3 Axis gimbal for mounting the camera. Its a little heavy, possibly about 50g heavier than the camera but it will probably do for my purposes.

I have been undecided whether I need a set of E600 or E800 class motors to lift the weight. Now I think I'm going to go for the smaller motors as this means I can build the Tarot in its stock form. Maybe I'll need a different gimbal for the camera if this ends up too heavy for the motors.

I'll post some images later, but for the moment my Avatar is a hello from the camera. Here it is wearing an old Leica CL lens, but if it flies it will have a much lighter weight lens than that one with autofocus. One possible lens I'm looking at is a Panasonic zoom with OIS. I think I'd need to lock down the zoom before flight to keep the lens balanced, but it will give me the option of shooting different fields of view and finding out what works best for my shoots. And the OIS will add another element of stabilisation.

So here's my initial component list

Already purchased

Tarot 680 Pro kit
DYS 3-axis NEX gimbal (may be too heavy)
Taranis Tx/Rx
MK808B Android TV (android stick, headless) to control the camera.
Olympus Air 01 micro 4/3 camera
Arduino Nano (to interface the FrSky Rx to the MK808B)

To decide on / buy

Motors and Escs and Power supply
Flight controller, GPS, OSD
Batteries

I've also been flying a Syma mini toy drone to learn how to fly multi rotors. I've managed to crash it in a number of enjoyable and creative ways. I've retrieved it from a cow pasture and from high in a tree with a 4M tree pruning saw and I have terrified our cats.

As far as I know there is currently no way to control the Olympus Air at a remote distance so this project has the advantage that if the thing doesn't fly I may at least have a way of remotely shooting/filming wildlife using inexpensive off the shelf components.

Ultimately though I want to take some specific pictures, so if this configuration doesn't work I will chop and change as required to meet my goal.
 

I built a Tarot 680. Mine was a ARF kit so the motors and ESC's were included. It came with sunnysky 2216 800kv motors and 30 amp ESC's. I have a Naza M V-2 flight controller with GPS and a mini osd and a fat shark FPV system. I think my all up weight is just over 3Kg. That's probably considered a little light but I only have a gopro with a gimbal and usually use a 5000ma battery. I have 13" props on it and it flys just fine, my motors and ESC's do not get hot. That being said I think the Tarot is capable of a heavier load but from what I have read it's really only a gopro UAV. The Tarot Pro model has the integrated PDB in the frame plate if I'm not mistaken. Mine was an Ironman and I had to buy my own PDB. The kit came with 10" props but I put 12" props on it and then a set of 13"s. I'm very happy with it.
 

I built a Tarot 680. Mine was a ARF kit so the motors and ESC's were included. It came with sunnysky 2216 800kv motors and 30 amp ESC's.

Thanks for your insight John. I have bought nearly all my bits now but am still deciding on the motors, escs and props. I was holding off while I decided if I wanted to build the tarot with extended arms to use the E800 motors from DJI but I discovered that the longer arm version might have more flex and therefore more camera movement so I'm going to build it stock and see what it can do. I may choose your Sunnysky option, they seem well regarded. Which 30A Escs do you recommend?

I've settled on a well reviewed clone of the Pixhawk flight controller with the Neo 8 GPS compass.
Got a Black Pearl Rx/monitor and some 5.8 Tx and inexpensive Rx to experiment with.

Unusually I am starting with the camera system first and then building a drone to fit it, so I started with the Olympus.

I've determined that the MK808 has not got sufficient guts to handle the live view from the Olympus camera, the LV froze on any size of image (shot images were fine though) so I've ordered the latest model MK903 Android TV which has the latest RK3288 SoC inside. That is quad core instead of dual core and 2Gb Ram instead of 1Gb, as well as uprated GPU FPU etc. If it doesn't work then it will be back to the drawing board and a rethink about the camera.

Ultimately I will have to build a bigger machine to carry a Sony A7 series camera, but that's for the second one.

If I can't get the stills results I want with this, then I may be tempted to make it a high quality video platform instead. I just found this and my mouth is watering.

https://www.blackmagicdesign.com/products/blackmagicmicrocinemacamera

I have assembled the DYS gimbal and the Olympus camera will fit well once I have developed a mount able to support ease forward/rear position setting to cope with different lens sizes. I am planning to 3D print a suitable part for the Olympus camera mount. Yesterday I did some superb tutorials with Shapeways and did a rough design for part of the bit of the mount that supports the camera. I will add a slide underneath with a slot to screw onto the DYS gimbal rail.

widget.jpg


The Olympus camera is available as a 3D model but I haven't yet managed to fuse my design with it, but I can use it to measure the radius of the curve required.

A01_body-bare_preview_featured.jpg


http://www.thingiverse.com/thing:1006758/#files

However its quite an attractive idea to consider making a 3D print of the above body part for the camera with a modification to mount onto a gimbal build in. You could potentially use the side mounting holes (one viewed on the left) to mount to each side of a gimbal directly.

As far as I can tell the really clever bit of the gimbal tech is the AlexMos controller which is available easily if one wanted to 3D print a gimbal designed for a specific camera.

Once I am sure I have a viable set of computers for this lens I may have the cojones to dismantle it and reassemble it with some altered parts.
 
Last edited by a moderator:

I think my ESC's are "hobby wing" but I'm not sure. (I'm at work right now) I bought my ARF kit from foxtechfpv.com I don't know much about your camera. There's a few cameras that are popular with UAV's, you can tell which ones because they make gimbals specifically for them. Going with a different camera platform is cool because being experimental with this technology is a great thing. My next build will be a larger Tarot carrying a Sony NEX 7.
 


Well I admire your intent to do something new with the arduino and the android. Being experimental with this technology is a good thing. What freq is your wifi? have you considered it may interfere with something else?
 

Yes I have considered it and thats why I'm assembling the camera part of the system and electronics first in prototype form. If it doesn't work I won't mount it up on the frame.

The camera also has Bluetooth. Both work on 2.4 Ghz, both use spread spectrum of various types as does the main flight control Rx/Tx. Theoretically there is enough bandwidth for everything but I have doubts. The FPV is on 5.8Ghz.

Unfortunately there is no wired method for connecting to the camera so ultimately it may be back to the drawing board if there is poor connectivity with either the camera or the Tx/Rx.

Initially all the arduino will do is 'press' the shutter button. I already have one and its tiny. If all else fails I may attempt to interface the arduino directly to the cameras shutter circuit (it has a physical button) and use a separate FPV camera, small and light type.
 


Bartman

Welcome to MultiRotorForums.com!!
i'm subscribed to this! watching people do new and exciting things makes all of the work of keeping the site going worth it. good luck!
 

My Tarot has 13" props. There is about 1/2" clearance between them. I got an more than one minute of additional flight time with the larger props on the same battery. I would start with 12's and then move up. But if you buy the DJI propulsion system your kinda stuck with what they give you. Those are nice because the homework is already done and all you do is bolt it on. If you want to be sure you could easily calculate the distance between props. You know the diameter (680mm) so the rest would be easy to figure out.
 

There are 3 build logs for the Tarot 680 that I know of on You Tube. "yellowbonz69" not sure if I spelled that right, will give good insight as to how cramped it will be if you mount your ESC's on the center plate as opposed to the arms. "everything cpo" built one with a pixhawk and is a real good build log. And mine "John Coffield" my first attempt at a build log. I put out 6 videos on my build. You always learn a little something that will save you some time or address an issue that you haven't thought of.
 

violetwolf

Member
There were some serious issues with WiFi on Pixhawk machines last year. I can't recall exact details but recall it was interfering with the compass or GPS... They were losing control of the craft in auto modes and traced it back to the WiFi on a gopro... Pretty sure it wasn't specific to the Pixhawk but rather guidance sensors.
 

After a long delay (too much work) I'm back on this project in my holiday. Still mainly figuring out which parts to get. Finally decided to go with the DJI tuned propulsion 800 set as the coax power leads are claimed to reduce interference and I like the pretty lights on the ESCs. Thanks to the log by AlienEclipse there is a lot to help me perform this setup.

I've decided to extend the arms on the Tarot 680, as the parts from Tarot do do this are very inexpensive, though very hard to find.

It seems I'll need 4 x TL68B09-02 (323mm) and 2 x TL68B09-01 (333mm), I'll get the retracts as well.

So anyhow, onwards...

Since last writing its become clearer that my original Android TV device was too low powered, so I've upgraded it to the latest model. With this there are still a few video dropouts but its working much better. Ultimately I may use a separate tiny FPV camera instead of the Oly and simply shoot the camera via the Arduino interface. This could be a better solution in total as it would mean I could get good field of view for flying the thing while the shooting lens could be a longer focal length for detailed aerial shots.

I've mastered setting up the FrSKY 8XR receiver so that I can use the additional 8 channels to control the gimbal and the arduino camera interface. That wasn't so easy either, but another step sorted out.

I've also bought a DYS 3-D gimbal which comes with a Simple BGC controller and motors. It seems a bit oversized for the Oly Air 01. If I can get it to work I can mount a larger camera such as a Sony compact or possibly the sexy BMPCC video camera for shooting video. That looks awesome and like the Oly it takes micro 4/3 lenses.

The gimbal seems a bit tall for the standard legs, I'll get the retracts but wonder if I'll need slightly longer tubes. Some pipe lagging might work. Here I've assembled the fixed legs just to get a feel for things. The Oly camera here is fitted with a Panasonic 20mm f1.7 lens. Very light, compact and good quality even at f1.7 but AF is a bit slow and its not possible to lock the focus to hyperlocal distance which may be a better approach.

View attachment 27135
 

Attachments

  • drone.jpg
    drone.jpg
    156.9 KB · Views: 535

There were some serious issues with WiFi on Pixhawk machines last year. I can't recall exact details but recall it was interfering with the compass or GPS... They were losing control of the craft in auto modes and traced it back to the WiFi on a gopro... Pretty sure it wasn't specific to the Pixhawk but rather guidance sensors.

Thanks so much for this I'll check it out. I may have to junk the concept and switch to another camera if this is indeed a problem.

The camera also has a bluetooth interface which might be a better option but so far I can't seem to get it to work, the Wifi connection seems to come on even if I don't want it to.

Research so far suggests the problems were caused by poor placement of GoPro camera too close to the GPS so I may be OK.
 
Last edited by a moderator:

Dawntreader

Member
Arduino and android sounds interesting. I know you said to control the shutter, but forgive my ignorance, what are you wanting it to do?
And what kind of range would that be? would this also be full automation from the Mission planner app?
Also with the Pixhawk flight controller I've heard of some esc's types having problems in the past. I personally don't know, as I'm just about to diving into it as well. but I wanted to warn you. maybe someone else can chime in.
 

Arduino and android sounds interesting. I know you said to control the shutter, but forgive my ignorance, what are you wanting it to do?
And what kind of range would that be? would this also be full automation from the Mission planner app?
Also with the Pixhawk flight controller I've heard of some esc's types having problems in the past. I personally don't know, as I'm just about to diving into it as well. but I wanted to warn you. maybe someone else can chime in.

My plan is to connect some outputs of my receiver to the analog input pins of a nano arduino and then read switch values. Then push this out on the serial output of the arduino via USB to the Android device as a series of commands to do the following.

1. Fire the camera :)
2. Switch mode A/M (I don't use S or I)
3. Set aperture, ISO and shutter speed.
4. Set a custom preset C1, C2, C3

The FrSky receiver has the sPORT serial digital output which would be great to use, but I don't have any understanding of how I could output the switch activity from the Taranis via the sPORT which mainly seems to be used for incoming telemetry.

I'll check on the compatibility of the e800 tuned propulsion with the Pixhawk, good point I need to check that.
 

So now I have the Taranis controlling the Arduino board, reading the PWM input from a servo and switching an LED.

https://www.instagram.com/p/BAFLXtyBRSn/

I've decided to just use two channels for the camera to start with. Once thats working I can upgrade the program on the Arduino to handle what is required on the camera and gimbal. My thinking is that I can program gimbal sequences using the Arduino e.g.: for stitching panoramas and so on.

e800 arrives tomorrow and according to DHL tracking my Tarot extended arms were delivered today and I'll get them when I get home. :)
 

Top