Trex500 Arducopter w/ 3axis GoPro mount and 3DRobotics FPV

[G. Bird]

I modified a Trex500 with Arducopter from 3DRobotics, along with their FPV gear. I also made a home-made 3axis camera mount for a GoPro, since the Arduino can control and stabilize 3axis camera mounts.

I made a quick video today, the heli was under manual control, I just wanted to see if it would get airborne with all the new gear. Also I quickly enabled stabilization in the roll and pitch axis, with no tuning at all.

Have mixed emotions on the video results. There is vibration in the new mount that needs to be mechanically fixed. The stabilization was overcompensating in banks, though having not tuned anything at all, wasn't bad "straight out of the box". Still lots of things to do, I haven't even hooked up the autopilot yet.

[R_Lefebvre]

Interesting stuff. The result certainly doesn't seem very good. You should try to get some datalogs and I'll have a look at them. I'm not sure if you've got a ton of shaking going on, or the APM is just over-reacting on the roll compensation of that gimbal. Or the shake is actually caused by the gimbal mechanics.

Maybe try doing a test flight with the gimbal powered off, and see if it still shakes that much when the servos aren't trying to compensate for attitude? Or even just turn off the stabilization in the roll axis, and see what that looks like.

I like the size of that 500 machine. I'm sort of interested in building one because I'm interested in building a heli for basic mapping with a down-ward pointing little P&S camera. A 450 is too small, and a 600 is just a little too big and intimidating. If it were possible to get 30 minutes of flight out of a 500 chassis, that would make a great little mapping machine.

I'd been looking at this, but I don't have any suitable batteries. I try to reuse my existing battery supply. I have a ton of 4S 5000's. But the 500 takes 6S right? I'd love to run 8S 5000, but haven't found an 8S rated 500 motor.

[G. Bird]

Yes there is a lot of room for improvement, I'm pretty sure I'll get it going smooth. I know from some nose hard-mounted videos I made it can be much more stable. There are a combination of things that might fix it. If you look at the head-on view of the mount L-shape arm can vibrate in the side to side axis, I hadn't formed an I-beam and will have to reinforce it. I think the torque tube I used as the "downshaft" is flexing, I will have to switch back to a main rotor shaft to stiffen it up despite the heavier weight.

Then the gimbal itself is not tuned, the arducopter is not programmed with anything yet. Also the APM sits on an acrylic mount, I will be swapping to an antivibration mount once I get one 3D printed. I tried to collect vibration logs but when I went to download it after setting "default +IMU" it said no logs. Maybe I did something wrong, or wasn't in stabilize mode. Will read more into it then perhaps ask/post more details on DIYD forum..

Yes I can turn off stabilization and see the video results. Also I haven't enabled the FPV camera yet, it would be interesting to see them both pointing forward and the difference. Another thing is the gimbal itself, I used servos/ belt drives for the yaw and pitch axis, if they are not tight there is play. I plan to replace them with gimbal motors I have on order. The gimbal is not balanced yet forming moments during translational movement which may be causing unwanted movement as well. Basically I was in a hurry to make a 1st video and get it in the air after weeks of construction, rebuilding it after it had sunk in a lake, rebuilding that after the ESC caught on fire. (basically I have experienced destruction via earth, air, water, and fire now).

I also decided on the 500 size heli to put AP on. No way I was going to risk a 700 flying about and having something go wrong. The flight time is about 6 minutes. I made 2 3minute videos like the one I posted on a single charge, and when I recharged put back 1800mAH. Now that I see it can get in the air, plan to make outriggers and carry 2 battery packs before tuning the Arducopter.

My goals for now are just to make long smooth flight videos that don't cause airsickness that some people have noted on some of my videos. Not really out to make professional quality aerial videos like some of the ones I see here. In fact I'm seriously considering looking into an AP balloon/blimp (with parachute and airbags etc..) to avoid some of the controversy surrounding UAVs and safety these days.. The Blade450 I have was probably too small and too flimsy. I want to still carry more payloads to do something useful, like terrain scanning like you, pointing lasers, dropping tennis balls, shooting rockets... There's this new microwave radar technology for finding people buried in rubble I want to explore mounting to a UAV.. Ultimately I want to infiltrate Area 51 and get the TRUTH...

[G. Bird]

I mounted a keychain cam at various points of the heli to see if the vibration was in the entire airframe or just the camera mount. In the process I discovered a loose main shaft collar which allowed vertical play in the main shaft. After retightening that down it got rid of a lot of vibration. There is still more to address though.

I had built the mount out of spare helicopter parts and scrap acrylic in the hopes of making one for really cheap. I have learned by doing now what I've read, just can't use servos. There's too much play in them for the mass they have to handle. I may have to give in and buy an aftermarket brushless gimbal motor platform. I could modify my mount with the motors, but I don't think the ardupilot can control/drive them (I think). Haven't read into that, perhaps it will be a project I would like to take on and help contribute to enable it if not..

The mount itself is quite rigid despite what it may look like As much so as the aftermarket one I might end up buying.. The real issue I still need to solve perhaps is the source of vibrations. Perhaps with the added weight and disc loading I will have to carefully balance out blades/shafts/rotors/CG. This is a work in progress..

[R_Lefebvre]

Yeah, your entire camera and pitch assembly is slung off the end of a servo shaft. You can't do that. Use a Servo City Servo Block at the very least.

[heliaddx]

Very interesting! Thanks for sharing with us your process and journey.

[G. Bird]

Thanks for the previous comments guys. Since my last post, I've fixed some vibration issues, and replaced my servo-driven gimbal with an $80 2-axis gimbal off ebay. It seems to work pretty well, so in response to the "any good?" thread- "yes".

Despite the stable video in Flight#1 of the video, there is still some slight vibration in the helicopter that seems to come and go depending on which battery I use now.. I think it might have something to do with headspeed and will be experimenting with different fixed speeds, which I think the Arducopter requires anyway. Right now I'm using "linear" govornor mode.

Next will be to enable the FPV camera and program the miniOSD. Then installing a 3D printed anti-vibration mount for the APM to properly tune it and enable autopilot. I also plan to add another brushless motor on the gimbal's yaw axis to be controlled by the Arducopter APM so I can keep it pointed at a GPS location while flying. The gimbal allows pitch input which I will also connect to the APM to point the camera straight down.

Happy Thanksgiving all!!

GB














FPV/autopilot coming soon...

[G. Bird]

A quick update on the camera gimbal, I think I've solved all vibration and stuttering issues now. I lengthened the gimbal arm using 0.25" standoff's so the GoPro tray was free to roll a full 360 (then I had to modify the mount above to get back ground clearance). I also found I had to add a counterweight on the GoPro side of the gimbal as it still stuttered during hard left banks. The gimbal motors have very low torque and the mount must be balanced to work properly.

The vibration comes and goes, and it is correlated to the battery used. Currently I have a Pheonix EDGE ESC set in "fixed endpoint" mode and not using its built-in governor. Depending on the battery used, the helicopter ends up with different headspeeds, as evidenced by comparing a stable video with battery "A" versus a shaky video with battery "B". This causes the entire airframe to shake. I will be changing the ESC setting to "governor mode w/ fixed RPM". I will download logs from it to find the headspeed during stable flight, then set it accordingly. The governor mode will be necessary when enabling the Arducopter autopilot anyway.


stable video with battery "A" and headspeed "X":



shaky video with battery "B" and headspeed "Y":



new standoffs and counterweight added to gimbal:

[G. Bird]

I've come to realize how important it is to have a balanced gimbal, at least in the roll axis for this particular one. Yesterday the camera stuttered while doing a hard right bank, the first time it has happened. I shifted the camera over by 1/8" and it resolved the issue, plus the video appeared more stable. I think if there is any moment force the motor has to fight it will cause it to stutter or reset. In the video below it almosts seems the airframe video has less jello than the stabilized video at times, though it is easier to see in the raw video than on youtube. Filmed at different locale this time:



Also slight correction to yesterday's post in case anyone really pays attention , both videos were from the "good" battery, I have shakier video from a "bad" battery. In any case I hope to finally resolve any remnants of vibration by tuning the headspeed and balancing the gimbal better.

GB

[R_Lefebvre]

How would you describe the stiffness of those damper balls? I wonder if they're too stiff. This is a tricky area for helicopters. You need softer dampers than a multirotor would typically use. And I'm guessing that gimbal was designed for multirotors.

I'm really surprised how well that gimbal is holding however, with a lot of yanking and banking of the heli!

Do you know which brushless control system it's using? Is it Martinez or AlexMos?

[Khosravi]

Originally Posted by R_Lefebvre

How would you describe the stiffness of those damper balls? I wonder if they're too stiff. This is a tricky area for helicopters. You need softer dampers than a multirotor would typically use. And I'm guessing that gimbal was designed for multirotors.

I'm really surprised how well that gimbal is holding however, with a lot of yanking and banking of the heli!

Do you know which brushless control system it's using? Is it Martinez or AlexMos?

Hmm, I wonder if this is the reason why I have such bad vibrations in my video footage. I'm trying my first brushless gimbal but having a terrible time tuning it. My gimbal uses 12 of these rubber damper balls, I'm going to now try going with less to see what it does.

[R_Lefebvre]

This is pretty nightmarish for helis though. Because you need fairly soft damping, but not too soft, or the whole camera tray shakes due to the blade frequency. It's just a nightmare. Good luck!

[Khosravi]

Originally Posted by R_Lefebvre

This is pretty nightmarish for helis though. Because you need fairly soft damping, but not too soft, or the whole camera tray shakes due to the blade frequency. It's just a nightmare. Good luck!

Yea I'm just about to give up on it and going back to my servo driven gimbals that seem to be more tolerant to vibrations.

[G. Bird]

The rubber bushings are soft and allow the mount to easily wobble. It would probably take a billiard ball to compress straight down vertically, but to tilt or in the shear directions they are like bending the same diameter silicone tubing.. In my case, however, I don't think it's those bushings at all. A keychain camera helps to investigate that, I have some videos where it seems shaky video is consistent with an entire shaking airframe, but stable when airframe is stable. After looking at past ESC (with built in governor) motor logs and understanding its settings better, I found changing modes and setting to a fixed headspeed of 2400rpm has greatly imroved the stability and consistency of video across multiple flights and batteries.

I use a Castle EDGE ESC with built in governor. In the beginning, I was impatient and didn't want to enter in the motor information and gearing ratios. So I put it in "fixed endpoint" mode, which just scales the throttle signal linearly based on transmitter throttle extremes. It was also in "simple governor mode", which senses where the throttle stick is for 1.5seconds during spool-up, and then tries to maintain that RPM, whatever it is. Since this was unbeknownst to me, and I usually set the stick to approximately 0degrees pitch during startup, I would get different headspeeds each flight, leading to vibration coming and going from flight to flight. In addition, when I fly I flip the "idle up" switch soon after spool-up which sets the throttle at 90%. Depending on how soon I flip this switch, it sometimes led to a consistent "90% throttle headspeed". I suppose I wouldn't have gone down this path if it weren't for noticing that initially vibration came and went depending on battery used, but I constantly was thinking about a post where some guy mentioned "it seems vibration comes and goes from flight to flight".

Below is the ESC log1 for past random flights with different batteries. You can see how the RPM ends up different a lot of times on different flights. In log2, there are 6 flights across multiple batteries (2 flights per battery). Video from flights 1,2 and 3 are noted on them. I'm really happy to report that the video is now acceptably and consistently stable across multiple flights and batteries. I didn't have time to experiment with more headspeeds, though 2400rpm seems to work pretty well, that will come in a future post. I've also started to fly a bit using the FPV camera, more on that soon..

Also, starting at 2:25 in the video, I taped a keychain cam to the stabilized GoPro looking straight down for anyone interested in seeing an electronically stabilized downward view (I'll need to fly higher up next time)..

GB



log of past flights with different batteries, note different headspeeds due to ESC mode used (Also since motor/battery/gear info wasn't entered in, it assumes 2pole motor thus the inaccurately high "RPMs"):


log of 6 flights from today, 2 each per battery. Flight video from this log using each battery:


Align RCM BL500MX 1600kV 6-pole outrunner motor
Castle EDGE 75A ESC w/ built in governor
162teeth main gear
13teeth pinion gear
Thor 35C 22.2v 6S 3300mAH and 2700mAH batteries


ps- some advice based on my experience so far for video stability-
1) make sure the airframe is mechanically sound. For helis, loosen and retighten the frame/bearing blocks on a glass flat surface, this seemed to help the most out of everything.
2) made sure there are no shot bearings, looseness or play in _anything_ in the rotor shaft and tail torque tube. Use metal servo arms!
3) use "loose" main blade grips, barely enough to hold them up when holding heli on its side.
4) make sure tail linkages are tight. I use loose tail blade grips as well. Advice I got was make sure tail runs smooth, it is spinning ~10 faster and can cause more vibration than the main rotor sometimes..
5) I haven't meticulously balanced my main or tail blades, I'm sure it would help..
6) make sure the gimbal is balanced on all its axis. Don't allow an out of balance gimbal's motors to have to fight any moment forces during "high-G" flight.
7) use a governor for your heli. make sure you are actually setting a fixed headspeed and find one that works.

[G. Bird]

Meant to do some headspeed testing today at 2000RPM, but messed up and flew at the same 2400RPM. Video looks good on all batteries though. Also flew (mostly) via the FPV camera, though I already realize I want to install a diversity receiver before fully relying on it. I don't feel safe fully relying on it yet as it sometimes fades out for longer than "momentary" when flying in certain directions. I also want to enable the autopilot for added safety since it has autoleveling to use when video fades out, or maybe return to home if things go horribly wrong.. I wasn't able to record any of the FPV video which I'm using 3DRobotics miniOSD with because of a driver issue with my usb video capture device.. will have to figure that out for next time. Only flew out about 1000feet today, looking forward to going out 1000meters, flying around the entire preserve, and one day 10 mile long range FPV..

The camera stuttering occurred again because I had replaced the steel nut counterweight with a coin. The recently added nut extended out far enough to hit the mounting plate, and moving it elsewhere would have unbalanced the pitch axis. The flat nickel is 2grams lighter. So I just need to shift the camera position again. Or maybe add a dime, but I'm trying to keep costs down..

I felt a few first moments of blissful fpv flight today, I'm on my way to getting addicted. It makes me think of those drug documentaries on tv, where they interviewed a crack addict and asked him how long it took for him to get hooked. His reply was "The first time I smoked it!!". Well one day when they make pilot rehab videos and ask me when I got addicted to FPV, I'll be saying "The first time I flew it!!"

[Libtekguy]

I'm enjoying your building and progress! I want to build a Helicopter FPV rig, but I'm considering doing a stretched 450. I have a few questions if you wouldn't mind.

1) How well do you think the APM would fit on the 450? Building custom parts or modifying existing ones is within my abilities. A camera gimbal isn't in my plans, and I like the lower prices of the 450 compared to the 500.

2) It looks the a FB rotor head is working great for you, ever tried APM with a flybarless head?

3) How was your experience with setting the APM up (difficult, easy....)?

[G. Bird]

Hi Libtekguy,

Thanks for watching. The APM should fit easily on a 450, in fact I think there has been more testing and data for it than other sizes. There are templates to make a mount for it that would install between the landing gear and bottom of the frame at the 3DRobotics website.

I have a FBL head on standby for if/when I crash and destroy the FB head, so far I haven't crashed (knock on wood). I started with FB because to me, flying a FBL heli is like trying to move a frisbee through the air, and they go faster. Flying a FB heli is like.. moving a whiffle ball through the air. There is a mechanical tendency for it to slow itself down, and for learning 3D/aerobatic flying I preferred that. All my other helis are FBL though actually.

I haven't actually set up the APM yet. Everything has been just direct control through the Spektrum receiver. The APM is mounted to provide GPS/OSD data for now. I've been wanting to get the heli to its final configuration and weight before tuning it, so I don't have to retune again. This is why it has taken me so long to enable it, I've been wanting to get the video stable first. Also I've been waiting for parts in the mail to install a 3D printed anti-vibration board, thinking about building outriggers to mount 2 batteries, building a ground station tripod.. etc..

Happy Holidays all!!

[G. Bird]

I modified a tripod to hold a laptop, fpv monitor, and transmitter using scrap parts I picked up at various hardware and electronics stores. I happened to find a delrin block for $5 that did the job nicely for building a tabletop type support. Screws and aluminum bars were a couple dollars. I used two $1 scrap acrylic sheets from a plastics shop to make the table. I asked them how much it would have cost to cut a rectangle in the same size, it would have been $12...

This is version 1, I'm planning on changing to a 12" FPV monitor. I originally got a 7" monitor with the intent to mount on the transmitter (which is why you see carbon rods on it to make a mount), however I'm finding it's much better to have it held up to eye level and not have excess weight on the transmitter. I was thinking of the 10" monitors, but they are 16:9 aspect ratio, the fpv camera is 4:3, so do I get grey bars effectively bringing it back to a 7-8" screen? Hmm, without looking into it too much I decided to get the 4:3 12" screen, so it would be similar in size to the laptop screen below it.

Also I've ordered a Boscam 4 input diversity receiver, with the intent to use skew-planar wheel antenna and 3 patch antennas. I want to get the most reliable video possible,.. I was considering an antenna tracker but then saw this 4 input diversity receiver and decided to start this way to avoid spending more $$. (Incidentally, why not put more 4 input diversity receivers to each of the first diversity receiver's inputs? Then one can make like a celphone tower with 16 directional patch antennas for 360degree coverage..

I'm going to add standoffs to the table to surround and hold the laptop firmly in place. I put a temporary hook to hang the transmitter on, will probably make it more robust. The bag you see hanging is just from my camping equipment- it's a folding chair. I'm thinking of installing a bottle opener on one of the legs too..

[G. Bird]

Now that I seem to be able to get consistently stable video, I went back to enabling the yaw axis for it again. The thread is titled "3-axis" gopro mount, after all. I got a 360 degree robotic servo online for $12. I cut a mount for it out of some scrap carbon fiber. The rubber dampers for a 3D printed anti-vibration mount used for the APM unit finally arrived, so now I could replace the acrylic sheet you see the GPS/compass and APM unit currently mounted on (and actually start using it).




I bought a new FPV camera, because the old one broke. This one is the 3DRobotics DV-330GH360-N and has higher resolution and sensitivity- and better for night flying.. The reason why it broke, and my current experience with the two cameras, and the rest of the electronics of the entire heli is a whole other story that I will post about in the future. While I had hoped to make some videos to test the new rotating stabilized camera last week, instead things seem to be breaking and I've been debugging various electronic issues including trying to root cause why a camera went dead. This would not be good if it happened during long range FPV flight..




The gimbal originally was hung off a Trex700 tail torque tube. During the course of video vibration debug I had replaced it with a stiffer, heavy steel Trex500 main rotor shaft. The bearing blocks that hold it are from a Trex500. Now I went back to a tail torque tube, which would allow me to run wires out from the gimbal as it rotated to the airframe. The wires I'm running are power, signal for gimbal tilt control, and a wires for video out from the GoPro. The plan is to eventually add an additional video transmitter for the GoPro, so a 2nd "copilot" can pan/tilt it in any direction and watch in real time as the heli is flown.




Here is the newly completed front end. The gimbal rotates freely and wires can twist in the harness for many revolutions before tightening up. A problem right now in the current configuration is that I have no way of really knowing which way the camera is pointing. I can turn on the wifi on the GoPro and view some delayed video for limited range on my celphone I suppose. Maybe I'll stick a little stick to indicate to the FPV camera where it's at, otherwise for now it's will be visual checking




I also got an airspeed indicator from 3DRobotics. It consists of a Pitot tube and differential pressure sensor for connecting to the APM. Knowing airspeed can helpful in long range FPV, for example, when say maybe you are trying to return home and have a headwind, and can't really tell if you're making progress. However, 3DRobotics firmware does not support airspeed indicator for helicopters/multicopters, because they are unreliable in the models' downwash and ability to fly in different directions not parallel to the tube. Well I was just interested in getting a sense of speed, and definitely not hooking it into the autopilot system (becoming another point of failure). They are used in real helicopters, it must work to some extent on a model.. Anyway for now I mounted it, perhaps someone can enable the software just to read it for helis. This is another reason I like the ardupilot platform, not only is it hundreds of dollars less than others, it is open source. People can contribute to its development. It's not some proprietary black box where you wish there was a feature and then just have to hope and wait for it.






When the APM is mounted on the new anti-vibration mount, the wiring harness sits higher, and now the flybar comes very close to striking it. I could cut the case, remove the straight headers and install right angle ones (p.i.t.a). I could cut all the servo plugs and just solder wires directly to the existing headers to make it lay flatter (also p.i.t.a.) Nope, instead I just bought another APM- the version for wires coming out the side. I have plans for another UAV in the future so it will be re-used. Also, it was $50 off due to Christmas sale . For now, since I don't actually have the APM connected to servos, I can remove that wiring harness if I decide to fly it. The new APM will likely arrive before that happens though..




Now that the APM is on its mount and the approriate lengh wiring harnesses arrived in the mail, I moved the GPS/compass further back on the tail to get it further away from the electronics, and also aid in counterweight. With the latest mods it allowed me to remove the steel bolt previously used and replace with a 5gram nickel.




With everything done, the airframe CG is checked on a high-point balancer. Previously I balanced the airframe by holding the flybar, I finally dug up the balancer to ensure a precise CG. I hope this will make flight and other things in general better.. will have to fly and find out. With some heavy parts removed but a servo and heavier FPV camera added, I think the overall weight remains the same or is lighter. My scale was out of batteries though..




So the heli is updated, but right now the gimbal doesn't work (I fixed an issue, then rebroke it, now waiting for an LDO in the mail), a camera got damaged (would like to find out why), and I'm doing other electronic mods to the 3DR MiniM OSD board and maybe the power system (you know that spark you get when you connect/disconnect a battery? That's due to inductance in the wiring causing back EMF and a spark as the circuit makes/breaks. Been using some equipment to see if there are spikes on the 12v rail that may cause intermittent damage to components not protected by a BEC, and if so the proper circuit mods to protect against it.)

Meanwhile the 4 channel diversity receiver arrived for the ground station, and the 12" monitor should get here any day. This frees up the 7" monitor and receiver for a copilot to control the gopro.. hmm.. Hope this post isn't too long, a lot has happened lately, and I've been learning a lot. Hopefully showing issues I've gone through will help others who are interested in doing a similar thing. In future post(s) I plan to draw out the entire wiring diagram for the heli, and an overview on the parts and materials used, where to get them, other specs on the heli... Looking forward to a great 2014 in UAV'ing!!

Happy New Year all!
GB

[G. Bird]

I was ready to make some flight test video of the rotating gimbal last week but things started to fail and I ended up doing electronic debug. For one, the gimbal started shaking on all axis while testing on my desk. Another is the camera also died while testing at home. I was able to determine what was wrong in both cases, and in the process did a careful review and overhaul of the electrical system to make it more reliable and safer.






The gimbal consisted of a "Martinez" type open source gimbal controller board and IMU (sensor), cnc aluminum frame and brushless DC (BLDC) motors. I noticed the red LED on the IMU would sometimes go dull then bright as I wiggled the wire to the sensor. It must have been damaged, causing intermittent power and/or I2C connection from the sensor to the controller. This seemed reasonable as the wire guage is very thin and subject to fraying or breaking even within the insulation when repeatedly bent/twisted. After resoldering leads on, the gimbal worked again- for awhile. Then all of a sudden on the bench the pitch axis started stuttering, then completely died after awhile. I probed the I2C bus to ensure the IMU wires were soldered on correctly, it was fine (i2c data waveform in image above). I probed the motor controller outputs, found the 3-phase DC signal was present on both of them, so the controller board was ok. I then ohmed out the motor and found the pitch motor open, apparently the coils in it had burnt out. I had to order a new gimbal to resolve this issue. In the process though I learned a lot about BLDC motor control, and programming the controller

nice app note on BLDC control:
http://ww1.microchip.com/downloads/e...tes/00857a.pdf







While testing video, suddenly I got a black screen on my FPV monitor, though I was able to see the OSD (and YES, the lense cap was removed). After some research on the web and debugging at the bench, probing for video signals and ohming out boards, I determined the camera was dead. Along the way, I learned some lessons on AV cable compatibilities.. (ok if you look carefully at this picture, it was actually taken later when I had installed a new camera. The green and yellow waveforms are video in and video out of the OSD chip. Notice the green waveform from the camera is not actually flat line and has negative video sync pulses.. I thought I had damaged another camera until I notice them- in this case, the lense cap WAS on )






I saw no video signal coming out the camera, so I decided to open it up and look inside. I found that the thru-hole crystal oscillator component was a metal can sandwiched between two mezzanine boards. In my case, it seemed it was contacting 2 components on the opposing side, shorting them out. This seemed a reasonable explanation- as with movement over time the can could migrate over until it came into contact with another part. I put kapton tape around and and reassembled it hoping it would work again, but it did not. It looks like it was permanently damaged, and I had to buy a replacement.






In the process of getting the new camera to work with the new diversity receiver and new 12" monitor, I discovered a few things as well- and that is not all AV cables wiring harnesses and/or AV phono/RCA cables are the same, leading to compatibility issues accross equipment and brands. For example, though I got 2 cameras from the same company, the video and power pins are different between them, you can't just buy the new camera and drop it in where your old one was. The same thing went for those phono jack to AV cables- at least for the random one I picked up at the store (I must have picked up a camcorder cable), it turned out the plugs did not lead to the same terminals of the phono jack. Both these cabling issues added to the initial confusion of debugging camera issues..






Next I turned my attention to the Minim OSD board that overlays telemetry data onto the video signal. While it turned out it had nothing to do with the video issues I was having, there were valid concerns and posts on its reliability and the fact its MAX7456 video driver chip kept burning out for some people. The main thread I was looking at was:

http://www.rcgroups.com/forums/showt...2#post24876545

While the thread offered a solution that seemed to work for people, it didn't give any satisfactory explanation as to why the problem occurred or what was really happening with the proposed workaround. After doing some research and circuit reviews, I think I came to understand what was causing the issue and modified the board a bit. I ended up removing a switching regulator on it and joining two power planes. This would ensure equal voltages and simultaneous voltage sequencing on the AVDD, DVDD, and PVDD pins of the MAX7456 monochrome display driver chip on it..






Since I had not yet determined WHY the BLDC motor on the gimbal burned out, I had a slight suspicion that maybe the power rails were spiking and causing damage. (The camera sits on the same 12v rail had burned out, maybe due to its faulty mechanical design, but may be not..) I decided to overhaul the power system slightly. Below is a block diagram of the updated wiring, I'm still waiting for parts in the mail to complete it.






Changes include moving the gimbal to be powered off the main motor batteries via a 22v to 12v switching regulator. If any short develops on the gimbal during flight, the regulator has short circuit protection and will prevent it from collapsing the rail, bringing the heli down. This also then puts all motors and servos onto a "noisy" dc rail. In addition I've changed the 2A ubec to a shielded, 5A UBEC designed for FPV use to power the APM and provide more overhead. Lastly I've added a power filter to the 12v going to the video camera and transmitter. I would prefer to use a switching regulator, but that would involve changing the 3s battery I use to 2s and using a boost converter..

Still waiting on some parts in the mail to complete the mods. I might add some Zenar diodes on the gimbal's 12v rail to protect from (so far unobserved) voltage spikes.. Also I'm adding LED navigation lights.. Hopefully soon though I will be able to make some FPV videos, and stabilized rotating gimbal videos..Already there is some static in some of my initial FPV video captures. I now foresee a lot of RF testing and FPV video debug/optimization in the future..