Over a year ago a friend lost a Easyglider with a HD camera on board. We spent a week searching about a square Km of grass and scrub but didn't find it. We also flew many patterns and recorded well over 20 gigs of HD video, finding a few interesting things but no plane. Then a few months later an FPV plane went missing at a much worse location with almost no chance of getting it back; all scrub, forrest and steep hills.

We'd talked about lost model trackers for some months before loosing a plane so this was the push needed to do somethig about it. I designed a 433MHz beacon transmitter and receiver specifically for lost RC model tracking.

So far it's worked well in testing and maybe due to Murphys Law we haven't missplaced another model. I did leave a model in the top of a large tree overnight a short while back, but knew where it was. It was a good tracker test as it started transmitting as expected and I could receive the signal from home, almost 2Km away.

Since then I've supplied a few of these. I have just recently been testing a yagi directional antenna that packs up to a small size and weight for easy storage and shipping. All lots of fun.

Well, it's been quite a while since I posted to the blog here.

The weather hasn't been so good lately and the good days haven't often matched up with free time.

But I have been working on some RC electronics projects. Finalizing my lost model tracker and LED lighting controller designs.

I've also built a one-off prototype of a radio variometer. It transmits tone beeps to the same receiver the lost model tracker uses. The beeps change up and down in pitch as the model rises or sinks. So far I've only done a few flights with it, but it seems to be working quite well. It's can pass the throttle control so that I could use it for ALES or to limit maximum altitude.

There is still a bit of work to do as time permits. I want to clean up the beeps and turn them off or reduce the rate when there's no change. Also it needs some sort of relative altitude auto-calibration to be useful for ALES.

Earlier in the year I did some work with the DIY-OSD so that I could see the number of satellites and send some data back to the ground hidden in a couple of unused video lines. This worked but is still a work-in-progress

I'll post some more details later on if there's any interest.

Using a V-Mate to record video transmitted from a model plane I have long had a problem with tearing down the left side of the recorded video only. It’s not seen on the monitor connected to the same video receiver. Finally I investigated and fond that it only occurred at the horizon and when it was near horizontal across the picture. This is where the top part of the image was bright (sky) and the bottom part was darker (ground).

Technical:

In the ground station is a 4-way video splitter that fed the recorder and a couple of LCD screens. The splitter and other components use AC coupling of the video signal; they stick it through big capacitors to pass it to the next stage. With the large brightness shift at the horizon there was a sudden DC average shift in the video signal that the V-Mate couldn’t cope with and lost line sync.

The solution seemed to be to DC correct the video signal. Some web browsing suggested that this was quite involved but I found a amplifier chip at a reasonable price that included DC correction. With a couple of these I built a replacement amp-splitter to replace the old one. Some bench testing indicated that the amp with DC correction generally improved the look of the video so I went ahead and designed a PC Board.

Specs:

75 ohm composite video input
one buffered DC-corrected comp. output
three buffered DC corrected comp. outputs that can be sync corrected **
operates from 5 to 15V DC @ < 100mA
59 x 42 x 7 mm...Continue Reading

I like the idea of flying a very small PC and a USB camera. There are now and will soon be many more cheap and small PCs with the required power. Linux is a great idea for this.

Even though wifi data rates can be easily high enough for this, it would have to cope with fall back to 1 or 2 Mb/s. Even 1 low-res frame per second is enough in emergencies if you have a stability controller assisting with the flying.

I would probably keep the stability controller and auto-return-to-launch system separate. It could would to the PC, getting flight direction and supplying telemetry data etc. but if all else fails it will get the plane back to the launch site.

I've had a question about long wire antennas for low frequency RC receivers for long range. So I'll put a reply here. See earlier posts for a few comments that lead to this.

I've had mixed results from the Corona receivers; some very good and a couple not usable.

The inductor is usually a small surface mount block that connects directly to the antenna wire at one end and into the receiver at the other end. From memory the ones I have bypassed have been a blue color. All you need to do is connect the antenna to the other end of the inductor; the receiver end. It can be left in place on the board.

The inductor makes the antenna "appear" longer to the receiver than it actually is; sort of like a very inefficient long wire antenna. By removing the inductor and connecting a true long wire you get much better efficiency but pay for it by having to drag around about 4 meters (at 35MHz) of antenna wire.

Calculate about a half wavelength and connect that to the receiver. It should hang clear from the receiver; no parallel wiring, carbon rods or metal push-rods. I have tried different things tied to the end of the wire to make it follow straight out behind or hang down more. About 45 degrees is best or flapping around between 30 and 60 degrees. Straight out is much worse.

You can tune the wire length but it takes a while and you need the RSSI signal; FPV OSD is very handy. This is how I did it:

In a clear open space with at least 20 meters to any objects...Continue Reading

OSD: On Screen Display

Overlays data in text and simple graphics form on a video down-link from the plane. Most often used with FPV flying.

I've been watching the OSD scene for 18 months or so. I've been flying a Vizion OSD with Zlog and have been considering building an OSD to interface to my autopilot. I'm looking at this as a flyer, builder and developer.

I think the requirements for what I'll call "FPV fun flying" and "serious FPV" are different. For fun flying altitude AGL, voltage, current and time are good to know. A GPS is an expensive add-on and may be cost-reduced with a pressure sensor altimeter. Accurate altitude is not required; +/- 10 meters is OK. For serious flying (which does not have to mean long range) GPS and RTL are required as the plane is probably expensive, heavier, capable of longer duration flights and carrying an expensive camera. This OSD will have to display more, voltages, currents, power used/remaining, position of aircraft/home, speeds, altitude, temperatures etc..... Being able to cycle through pages of info from simple to detailed and pop up alerts would be great.

I don't see configuration via an on board switch or laptop as an issue. The Vision joystick works well. I don't like installed software configuration tools and would prefer a simple text menu system via a terminal on the laptop. It must work with Windows, Linux or any other OS.

IMHO to design an OSD now you can keep it very simple and very low...Continue Reading

On the back of my transmitter I have a slide switch that sets the RF output to normal or low power. I installed this ages ago for testing antennas, receivers and installations so that I didn't have to walk 100m for each test. On the low power setting with the TX antenna half collapsed the range is 5m good to 8m nothing.

Before flying I check the switch is in the up (high) position. But a few days ago my check failed and I launched the plane with the switch set to low with the TX antenna fully extended. The plane flew off as planned and at about 100m out and 30m up I started a gentle left turn. It didn't feel right; the plane was very sluggish to respond but otherwise seemed OK. I thought maybe a servo problem and continued the turn to head back. After landing and checking the control surfaces again, I found the switch was set to low power.

I wouldn't have done this test deliberately as it's too high-risk and I wouldn't expect more than 50m or so. I was pleasantly surprised it worked out past 100m and that loss of control was minimal and gradual. A lucky save.

I've seen a few comments on-line about the desire to build a "standard" UAV/FPV airframe. This is an interesting question; is it possible, would enough people use it, with so many other choices available, why bother.

It's probably possible to build the ideal aerial photography platform, or the ideal course-following, competition-winning UAV platform.

The airframe has to suit the mission or intended use. At this stage I'm not wanting to fly GPS controlled tracks within a virtual box nor carry a 2Kg payload long distances. I think one of the most useful and realistic uses for a small FPV/UAV aircraft is in aerial searching and photography and I think these can be combined into one airframe.

Basic requirements are small size (for ease of use, transportability, safety and low cost) and long flight duration (for searching and station holding). This does not rule out carrying a payload of 1Kg or more over shorter distances.

This is why I'm now working on a powered glider of about 1 to 1.5 Kg that can be transported in pieces and very quickly assembled, launched and flown from a small ground station. In standard setup, endurance (air time) must be at least 30 minutes and able to be extended to 120 minutes or more by adding flight batteries as payload. This is based on some operation at full power but mostly operating at half power or less. Airspeed of 80 Kph or so may be possible but a cruise speed of 30-50 Kph is desirable.

Basically an electric foam version of the U2, with detachable wings that will fit in the boot of a small-medium sized car.

It looks to me that most of the "standard" airframe projects are too large for my plans and are not designed with maximum endurance in mind. One of the primary design goals seems to be maximum internal payload space. My idea is to simplify and shrink the on-board systems and have an external payload.

Still thinking about this.

Not an ideal setup but a simple first test from home to a distant hill.

At home, one of my slightly modified transmitters connected via 20m of 50 ohm coax to a 1m scanner antenna about 6m up on the roof. The receiver was taken to a road about 9Km distant and 120m higher with line of sight across a populated area to the transmitter. Good signal, no problems.

This is the longest safe test I can do without leaving the area as we are surrounded by hills. It gives me confidence that baring outside interference my 35MHz RC range is more than enough to test the video range in the air.

I've been asked about the RC receiver mods so here it is.

The idea was to improve or maintain the receiver performance without modifying the receiver itself while running the relatively noisy ESC and video TX. It's not complicated but is time consuming and fiddly.

I made a few small enclosures from circuit board laminate that would provide some shielding. One compartment holds the stripped down receiver and another is a filter. The servo connections are on the end and servo current is kept away from the receiver. Ferrite beads and small capacitors filter the channel PPM lines and the RSSI. Power passes through a LC filter to the receiver. The antenna loading coil is bypassed and connects via a simple smd bead-capacitor low-pass filter. Ground is common right though.

These are 35MHz Corona receivers:
RP8D1 : synthesized, dual conversion, long range.
RS810II : crystal, single conversion, long range.

The RSSI connection consists of a 10K resistor and 100n capacitor to ground. See picture (it's small, smd). It comes out through what was channel-8.

So, does it work:

Without the mods the esc and vtx can halve the effective control range when in close proximity to the RX or with connected power common.

The mods seem to virtually eliminate the range reduction that was caused by esc and vtx proximity when on a separate power supply. There is still a slight reduction (just noticeable) when running on the same battery.

With a standard Dynam TX I can maintain control with some minor glitching at 4Km in a ground test about 10m AGL. Maximum air range is still untested (and will remain that way until the RTL is perfected) but so far no problems.

MAP_1A is just a designation I use. It's a circuit board I've designed to combine the basic electronics and interconnections. At the moment I'm in the process of assembling and testing the first prototype.

This is a double-sided pcb for surface mount components that I've designed to be hand assembled and use soldered wire vias rather than through-plating. This makes it low-cost and reasonably quick to assemble. Once I get it sorted out I'll probably get a few pth/screened boards made.

Features:

• Size: overall 140mm x 50mm x 15mm and approx 50g
  flight pack direct connection for 3 packs
  esc connection power and control
  esc powers servos
  filtered ppm to servos
  master on/off mosfet
  current monitor for osd
  voltage monitor for osd and cpu
  2 aux power out connections
  electronics power from flight pack or dedicated pack
  2 stage filter for electronics power from esc/vtx/servos
  ldo 5v regulator for electronics
  remote RC receiver connection includes channels, raw ppm and rssi
  audio to osd, mic amp, raw ppm, alerts from cpu
  gyros for aileron and elevator
  ATMega324 main cpu for autopilot and/or other functions
  mostly compatible with Ardupilot
  64K bytes of serial eeprom for logging and/or waypoints
  status LED
  icsp port and bootloader via uart
  gps via uart (5V power switched by cpu)
  X,Y,Z IR sensors to ADC
  altitude and airspeed pressure sensors to ADC
  ATTiny25 ppm decoder feeds main cpu
  3 driver outputs for LEDs etc.
  voltage and temperature sensor to ADC

...Continue Reading

This is a new fpv plane that I'm currently working on. As already mentioned, the plan is to simplify the equipment, make it as reliable as possible and as low-cost as possible. I'm not too bothered if it takes a bit longer to build and test because it should be quicker and cheaper to duplicate.

I started off with a standard brushless Dynam HawkSky kit for about NZ$250. I wanted just the foam parts but when added up as parts and freight it was only about NZ$60 more for the RTF, with 3S LiPo, motor, esc and RC. I can use the ESC in the new plane and the rest for testing. The foam fuse comes already stuck together so it takes and hour to separate and clean the glue off. The Dynam servos are OK for bench testing but strip too easily.

The next step is to lay out the gear and modify the fuse. It has to be as streamlined as possible because I want maximum flight times. None of the normal equipment is external but it will have mount-points to attach payload equipment under each wing.

I want a 4000mAh 3S in the front so it will take a bit of weight at the tail. Using 35MHz RC I want to keep the RX away from the rest of the electronics and have the antenna pass straight out the tail. The RX is installed in the fuse just in front of the horiz-stab. To keep the weight low it's stripped and all wiring passes through a filter on the way forward. The RX antenna has a connector on the end of the tube allowing different length antennas to be used. A second tube provides a mount...Continue Reading

Transmitters:

I use Dynam radio gear because I got a 4-channel set with the first HawkSky RTF kit. When I tested the transmitter it seemed OK; construction and function were suitable for the basic flying I was doing. There were no programmable features but that wasn't a concern. I bought a Dynam 6-channel set because I planned to modify it and didn't want to cut into an expensive set. I ended up with 3 TXs on 35MHz, two modified and one stripped to make a remote TX head.

I have the original 40MHz set which I don't currently use. Here in NZ 40MHz is shared with other things and IMHO it's risky for FPV or stretching the limits. 35MHz is currently dedicated to RC aircraft; I have noticed a few other users, but not on the RC channels.

I have a 72MHz 6-channel Turborix set that I bought because a field I fly at during summer is adjacent to a cable TV provider HQ and they wipe out the 35MHz band for about 500 meters across the field. It took a bit of driving around with the spectrum analyzer to work out why my plane got a mind of it's own and decided to perch in the top of a large tree.

The other set I use is a 2.4G 6-channel Turborix which works with Turborix and HobbyKing receivers. It's great for park flying (practice) and I've never had any problems with other 2.4G systems. However, it can't be used with 2.4G video. The VTX upsets the RC RX and the RC TX upsets the VRX (if that makes sense).

Receivers:

I started off with the Dynam single conversion receivers....Continue Reading

The History:

I've been interested in RC since I was a kid; 30+ years ago. Planes were a lot of work between crashes and too expensive so I built a boat; Brave Boarderer MTB, about 3ft long, ply and balsa. Since RC gear was so expensive and in NZ model shops were few and far between we (Dad and I) built a multichannel PPM 27MHz AM system from a construction series in I think it was Popular Electronics magazine. I've still got the series stored away here somewhere.

Late 1980s I briefly had a 52 glow trainer with 35MHz radio gear, from memory a brand sold by Dick Smith Electronics for a short time. The servos sounded like electric coffee grinders. The only bit that remains is the OS 52 engine now installed in a Long-EZ that I haven't flown yet.

I'd done FPV RC cars before using 2.4G video and then mid last year (2008) while watching YouTube I came across Jettpilots videos of FPV RC flying. That looked like much more fun than cars and I decided that I could keep the cost down by building some of the gear.

I bought the Long-EZ and a Predator, both crunchies (ply and balsa) and then decided that I needed a foam plane to learn to fly. I bought a Dynam HawkSky because it was cheap and a spare set of foam because I thought that teaching myself to fly would be quite hard on the plane. As it turned out, it was hard on the plane but it could take it. I'm still flying the original foam; although with a lot of hot-glue and a few chunks missing.

After basic flying I added...Continue Reading