SMALL - SMALL - Telemetry SMALL - Radio
landolo's blog
Posted by landolo | Feb 01, 2013 @ 04:28 AM | 1,599 Views
(Part 1 is here)

With my lighting controllers built there was still a lot of work to do on writing the firmware for them; I spent almost the entire weekend of Dec 15 & 16 writing and testing code. Some pics below. By Monday night it was in a workable state. I was going to fly up to Brisbane for Christmas on Thursday, and didn't have any time off work until then, so things were getting a little tight.

The airframe wasn't finished since I still had to seal up the floats and finish the fuse. Pushrods still had to be completed. Hadn't got a power system hooked up. And I needed to modify my shipping crate to hold it. There didn't look like there'd be enough time to finish it all unless I didn't sleep for the next three days.

I didn't really need to get it done, but I really wanted to.
Posted by landolo | Jan 29, 2013 @ 05:57 AM | 2,198 Views
(Part 1 is here)

The master lighting controller for the Mako is based around another PIC microcontroller, a 16F628A. This is similar to the 12F683 used for the slave controllers but with 16 I/O pins instead of 6. The master board drives six LED banks through another MC1413 driver, the same device used in the slaves. It also drives another four LED banks by sending the signals to the slave board darlingtons (see part 5).

The master board connects to five receiver channels. One channel is used to select a lighting mode; the other four are simply routed to the slave boards. It also gets the 5V power supply needed for the microcontrollers (and servos) though the receiver. The 12V needed for the LED strips comes directly from the 3S battery. The boards were constructed assuming the BEC ground (0V) is the same as the battery negative side, which is the case for all the of the ESCs I've used, and so the LEDs share the same ground as the controllers, which simplifies the wiring a little.

The master board takes the control servo signal and uses it to select one of eight lighting programs; it also uses a simple serial protocol to send the current lighting mode to the slave controllers, along with some synchronization packets.

The pictures below show a 12MHz crystal which I installed in case the 4MHz internal clock of the 628 was too slow or inaccurate, but in the end I didn't turn it on.

I've included a simplified version of the firmware source code used on the master controller (without the serial comms and modified to suit a 683) in the zip file attached. It's not particularly well documented but feel free to contact me with questions.
Posted by landolo | Jan 28, 2013 @ 01:02 AM | 1,886 Views
(Part 1 is here)

Ok, after a few weeks break it's back to the Mako build log. This time a few details of the slave LED controllers. Each of the two slave controllers are mounted inside the floats on the underside of the wing, and they control five LED strip banks each - the four banks that make up the upper wing ring and one of the two banks on the tail.

The controller boards also are used to consolidate and route power and other signals from the master board in the fuselage, including the servo signals. There are a total of eight wires from the master board to the slave boards and this loom is run along the same path as the servo wires on the Mako plans. I tend to use very small gauge wire for the control signals, and standard servo wire (24 or 26 awg) for the power.

The board uses a 12F683 PIC microcontroller connected to a MC1413DG darlington array to drive the LED strips. The MC1413DG is a seven channel device that can drive up to 500mA per output, enough for a bit more than a metre of the LED strips I was using. Most of my LED banks are much shorter than that, so I'm in no danger of overloading it.

The 12F683 has six I/O pins; I use one for the serial control signal from the master board, and the other five to control lights. Two additional lighting signals come directly from the master board's microcontoller and are switched by the extra two channels on the darlington array.

As you can imagine, these things took ages to make and wire up. It didn't help that I messed up my first attempt by soldering the chips on upside down . My self-imposed deadline of Christmas was looming, and the Mako looked more like a Christmas tree than an aircraft at this point. A Christmas tree with no working lights at that.
Posted by landolo | Jan 03, 2013 @ 04:26 AM | 3,149 Views
(Part 1 is here)

Bit of a boring update this time; this was the tedious part in the middle of the assembly.

With the two foam tail booms/float sides glued on to the wing, it was time to put LEDs on the tail and the wing topside. The tail lighting was relatively straightforward; I wanted a smallish flashing strobe on the top and a blue strip on the bottom of the horizontal stabilizer (the underside of the aircraft will be predominantly blue LEDs, and the top mainly red). The strip on the bottom will throw light onto the inside of the tail booms as well.

Each of the two tail light sections will be controlled from one of the float slave controllers, so the grounds are taken back along the booms, one on each side, and routed inside the floats. The +12V supply is common, so the top and bottom sections are connected at the tail, and a single wire routed along the right-hand side boom.

The 'rings' on the top of the wing are made of four short three-LED strips arranged in a square, again with common positive and independent grounds, so five wires for each need to be brought through the wing to inside the float. The four grounds will be switched by the slave controller.

Now it was the end of November; next up, building some controllers!
Posted by landolo | Dec 31, 2012 @ 07:10 AM | 1,726 Views
(Part 1 is here)

Time to actually glue some foam together. I glued up the float sides with the bulkheads and the support strips, then attached them to the wing undersides. By this time it was late November and Christmas didn't seem so far away any more.

I had ordered most of the electronics parts I needed for the lighting control from ebay, and they were hopefully speeding towards me from Hong Kong, the US and the UK. There was still the minor matter of designing, building, writing firmware and testing... and my design for the lighting had become more complicated, since, to save on wiring runs, I had decided to install two slave controllers in the floats, controlling five LED banks each along with a master controller in the fuse controlling the other ten, rather than have everything controlled by a central board.

Exciting pictures of glue drying included below!
Posted by landolo | Dec 29, 2012 @ 06:07 AM | 1,749 Views
It took a bit longer than I had hoped, but by early November I had a trim and lighting arrangement for my new Mako sketched out. My last lit build had five independent banks of LEDs; the plan for the Mako was a little more ambitious with no fewer than twenty banks. The final scheme turned out to be:

o Fuselage interior (white)
o Port and starboard float interiors (blue) (2 total)
o Port and starboard navigation (red/green) at front and rear inside the floats (front and rear ganged together) (2 total)
o Front/Rear, port/starboard radial strips on wing underside (blue) (4 total)
o Upper wing 'rings', each with four strips (red) (8 total)
o Fuselage strobe (red)
o Tail strobe (red)
o Tail underside strip (blue)

Most of the LED strips would be attached to the wings or placed within the floats, so the next step in the construction was to attach these before gluing anything together. And also to do just a little bit of soldering...
Posted by landolo | Dec 29, 2012 @ 01:01 AM | 1,792 Views
It's been a while, and the Turbo Flea is still only partly completed, sitting on a shelf. Undaunted by a complete lack of progress, I started on another project a couple of months ago.

West Michigan Park Flyers' Mako seemed like it would be a great addition to my slowly growing fleet of seaplanes. I was also a bit envious of the night flyer Polaris I built for my brother in October, so I also decided that the Mako would be kitted out with a full set of lights, using a (as yet nonexistent) upgraded version of my homemade lighting controller. More on that later.

A trip to the LHS later for more Depron, and with a full-size set of plans printed, I started the build. This was in early October and I had a vague idea in my head that it would be good to finish it before my Christmas trip up to see my brother again. Months away - it should be a piece of cake, right?
Posted by landolo | Oct 28, 2012 @ 10:38 PM | 2,292 Views
I've found myself with a little time on my hands and a spare EDF motor. Time to build a Turbo Flea!

From my spares box: HK2627 4300kv EDF motor, 5x5 prop, Turnigy Trust 45A ESC (chunky!), 2x 6g servos, D4FR 4-ch FrSky receiver, a couple of 800mAh 2s batteries, a fair amount of 2mm Depron and some 3mm CF strips. I'm thinking that it's going to be a bit on the heavy side, but what the heck.

Experimented with laminating 2mm Depron for the main wing for no particularly good reason. Also going a bit light on the spars. The motor & prop look like they will pull around 30A max, which is way too much for my little 800mAh 20/30C batteries, hence the parallel connection.

This project is almost certain to end in disaster. Stay tuned.
Posted by landolo | Oct 11, 2012 @ 04:53 PM | 2,386 Views
Last weekend was a long weekend here, so I took a couple of extra days off work and flew up to Queensland to visit family. Normally I pack a transmitter and a couple of receivers as well; my brother has a full hanger and is trusting enough to let me fly a few of his planes.

This time I crated up my Polaris and I had a surprise for him in the crate as well - I'd spent the last few weeks building a nightflyer Polaris for him, and packed it in the crate on top of mine.

Lots of amazing mornings while I was there, so we were out flying every day. We took a lot of video as well, so I've posted some stills below....Continue Reading
Posted by landolo | Oct 10, 2012 @ 08:16 AM | 2,277 Views
A bit of a belated report on my repaired Polaris - I was insanely busy on another project, no time to post. The repairs were completed as reported last time, with the batteryless weight increasing marginally from the original 350g to 360g. The first flight tests were about two weeks ago and went well. Quite a lot of re-trimming was needed but after that was sorted out, she flew almost like new, although maybe it's just me, but repaired models never quite seem to fly as well as virgin ones. In this case, she had developed a tendency to roll out of wide loops. I probably need to balance the wings a little or check the aileron/rudder trim balance.

No pics or video sorry. But in the couple of weeks since the last post, the Polaris has been on a bit of an adventure and travelled more than 2000km.
Posted by landolo | Sep 27, 2012 @ 05:59 AM | 2,449 Views
The final stages of the Polaris repair have been completed. The internal battery tray at the front had been jolted away by the crash, but my battery always sits well back from the forward battery tray, so I decided not to replace it as is. Instead, I added some internal ducting to channel airflow from the forward cooling intake to the right hand side and towards the ESC. Without the ducting, air from the forward intake would have been split across the upper and lower fuselage bays (the battery, ESC and receiver all sit in the upper bay).

I also added some magnets to the hatch so I didn't have to tape it down every flight! That was a real nuisance, sorted now. I also glued the plastic spoon intakes back on.

Finally, more blue tape was added to the upper fuselage to hide some of the gluing. Not the fanciest finish but I actually quite like the slick look of the blue tape. I got the iron out to smooth some of the wrinkles. And I also re-applied tape to the bottom forward of the step.

Finally I routed the receiver aerials out holes in the side so now they're both outside taped to the wings, at 90 degrees to each other. Previously only one was placed externally... I'm hoping that having them both out will prevent a re-occurrence of the control/radio failure that was the prima facie cause of the crash.

Now I'm ready for a test flight and re-trimming...
Posted by landolo | Sep 26, 2012 @ 09:44 AM | 2,575 Views
Poor little micropolaris had another lawn dart episode a couple of days back, flying in exactly the same place as the first smackdown seen a few blog entries back. It had several great flights in a couple of other parks before that on the same day, so I'm wondering if the ley lines converge there or something. It appeared to be another control failure.

To add insult to injury the two magpies that had been chasing it took advantage of their disabled foe and started ripping shreds of foam out of it as it lay on the ground.

So it was time for another rebuild. Luckily most of the damage was cosmetic. Well, ok, maybe a fuselage broken in three places with large beak holes in the sides wasn't entirely cosmetic damage, but nothing a little glue and tape couldn't fix.

While the body was hinged open at the middle, I took the plunge and installed a few red LED strips on the upper wing and inside along the bottom of the fuse, connected to a little homebrew PIC 12F683 LED controller so they could be operated remotely.

The lights (and a new tape livery which hides most of the battle scars) added 10g to the weight, but I guess you can't measure stylishness in such base material terms. Now I hope it still gets off the ground.
Posted by landolo | Sep 25, 2012 @ 02:45 AM | 2,732 Views
Continuing with the Polaris repairs. After reattaching the tail, two side fuselage sections were replaced at the rear, bridging across the old wing and the new replacement piece.

A new aft underside was made, and I decided to add a little reinforcing by gluing a couple of thin ply strips into slots. Having the strips flush allowed them to be placed near the edges, and the slots were surprisingly easy to make with a couple of straight cuts and then excavating with a small flat screwdriver.

The idea is to spread any point impact forces along the length of the bottom fuse a bit more - not that it would save it from a major impact like the last crash, but I've also had the lower fuse cracking from a couple of hard landings.

Sealed up the underside again and it's starting to look almost respectable again.
Posted by landolo | Sep 20, 2012 @ 11:43 PM | 2,467 Views
A few more pictures of my micropolaris. I reinforced the tail join with a couple of small strips of ply after I damaged it while testing. The upper ply tail base is also bigger than stock, overlapping the main spar to try to spread impact forces, but that was less effective than the lower strips. The spars are 3mmx1mm carbon fibre strips, and the horizontal stabilizer spar is a single spar rather than a split spar. The wing is 3mm depron.

The servos are hobbyking S0361 3.6g servos and the motor pictured is a hobbyking 1404N 2290Kv 9.2g motor, which draws about 3A on 3S with the 4x4 prop. I've also tried a larger 2800Kv 30x26 motor which draws closer to 8A. I'm using a 10A ESC.

The batteryless weight is 69g, a little heavier than I wanted but still giving a low wing loading. No water flights yet, it floats but it gets blown over or pushed over by smallish ripples really easily as I found out while float testing.

After a couple of dusk flights last night, I think it would look cool with a few lights on it. That might be a stupid idea but I'll probably give it a go anyway.
Posted by landolo | Sep 20, 2012 @ 11:13 PM | 2,369 Views
Continuing with the Polaris repair. A section of the upper fuselage was put back in place, and the heavily damage area of wing around the tail base was replaced. As mentioned before, the tail was completely unharmed.

Not too happy with the wing/tail base repair, I think I've altered the stabilizer angle a bit because the tail didn't sit flat on the new section and I didn't clamp it enough. But I don't want to take it apart and do it again, it took at least an hour to do it the first time...
Posted by landolo | Sep 19, 2012 @ 10:22 PM | 2,441 Views
Some video of a fun flight of the Polaris before last weekend's crash. The resident hyper-aggressive magpie appears in the last minute and a battle royale ensues. It turns out that foam is faster than a magpie, but a magpie is smarter than a chunk of foam.

All good clean fun. no magpies, foam or egos were harmed.

I like my Polaris (5 min 16 sec)

Posted by landolo | Sep 18, 2012 @ 08:28 PM | 2,549 Views
I'm having a go at repairing the polaris crash damage and recording it in way too much detail here. You probably want to skip to something more interesting.

First steps are to cut away the most extensive damage and glue up the biggest cracks in the wing. Also remove the tape covering the forward underside and the nose and remove the gravel!
Posted by landolo | Sep 16, 2012 @ 09:49 AM | 2,381 Views
After crashing my big polaris, I needed something to cheer me up, so I quickly re-liveried the micro (really hated that yellow nose, what was I thinking) and had a quick fly before dark. On Sunday morning, before the winds picked up, I took the micro out again.

I think something is wrong with my radio.

Micropolaris crash! (0 min 23 sec)

The little white bit flying off at the end before the bounce is the battery hatch. The micro is a tough little beast, there was almost no damage apart from a detached firewall and a related prop scar.

I hate it when the Fates stalk me like this. My computer crashed last night as well.
Posted by landolo | Sep 16, 2012 @ 08:46 AM | 2,427 Views
While the micropolaris had a good day, my 90% Polaris didn't fare so well. Flying my fourth battery at the lake in the evening, I had what appeared to be a radio glitch. The aircraft was heading towards the shore at low-ish level when I heard the motor cut briefly, then it banked and control became erratic. End result - a powered impact with a boat ramp. It was a bit depressing because it's my favorite plane at the moment:

An initial post-crash analysis was unenlightening. My Tx aerial was not in the best shape after dropping it a while back, but it was still functional. The plane was only about 20m away when loss of control occurred.

I wasn't sure whether to attempt a repair or just build another airframe; the damage was extensive but repairable. I think it crashed level but nose down, so the tail escaped almost unscathed, but the wing was cracked in several places. All the electronics were still working.

Time to sit down with a glass of wine and take a good look at the damage.
Posted by landolo | Sep 16, 2012 @ 07:37 AM | 2,415 Views
I spent the last couple of weeks building a 50% scale Polaris inspired by Otto Dieffenbach's videos of his tiny version. Mine was built from 3mm and 2mm depron, with a wide nacelle containing elevator and rudder servos. Ailerons are pushrod activated on the top.

I was hoping to use a 2s power system but didn't do a good job on the specs and ended up with a choice of two underpowered motors. A first attempt at getting it going last weekend on a 2s battery was a dismal failure, it wouldn't move on grass and gentle low level hand launches resulted in the classic polaris tail separation.

This weekend I put together a 250mAh 3s pack from some old cells and maidened it. First flights on Saturday were a success! It does suck the juice though, with the battery only lasting a few minutes.

Micropolaris Maiden (0 min 49 sec)