The View Up Here

Random scribblings about kites, photography, machining, and anything else

  • Flickr Gallery

Archive for March, 2013

In for a Penny

Posted by Tom Benedict on 27/03/2013

In for a pound… Or so the saying goes.

I’m completely, utterly, absolutely hooked on RC gliders now.

I remember back when I was in kindergarten, just before Christmas we had to do an art project in which we drew a boot rather than a stocking – this was Texas, after all – and drew a picture of the thing we most wanted as a present inside the boot. Some of the kids in my class drew pointy-toed boots so they’d have more room to draw stuff. But I knew exactly what I wanted. A stubby-toed boot worked just fine.

When I showed it to my teacher she had to ask, “What is that?” “It’s a remoconchrolplane!” I announced. I think she had to ask three or four times before she finally figured out what I meant. Then she helped me spell “remote control plane”. Never before that day had it occurred to me that it was actually three separate words. Seeing them that way, all sorts of things started to make sense.

I showed the drawing to my parents – this being the entire point of the exercise from the school’s standpoint – and wondered if it might actually come true. Keep in mind when this was happening: 1973. An RC plane of that era had a gas engine, a new-fangled transistorized radio from a company like Kraft, took months of work to build from balsa sheet, and cost more than my parents made in several months. There was just no way.

But my parents tried. There was a plane for me under the tree that year. It was a control-line trainer with a tiny Cox engine that terrified me when it was running. We took it out only once, and not ever having flown before I managed to crash it in the first ten seconds of flight. The plane was made entirely of injection molded plastic, so when the tailplane broke (the only real control surface on the entire airplane) it broke in a way that was practically impossible to repair. My father tried valiantly, but he wasn’t a modeler or a machinist, and the glues he had available to him at that time simply weren’t up to the task. Cox didn’t sell replacement parts, either. In a way it was doomed to fail from the start. But for those ten seconds, I’d flown a remoconchrolplane. My wish really did come true.

Over the years I visited flying fields where people were flying their models. I never had one of my own, but I made a heckuva good spectator. Of all the fields we visited, though, the one that really caught my interest was the glider club that flew in Valley Forge, Pennsylvania. The flying field had enough space to set up a winch, and lots of room for thermals. When the weather was good, people would launch and fly for what seemed like forever. No scary gas engines – though I was over that phobia by then – just silent giants soaring through the air. I continued to dream. But I was a kid in high school, saving up for college. Dream was all I did.

It wasn’t until this past Christmas, twenty six years after Valley Forge, that I finally got a chance to fly again. I asked Santa for a remoconchrolplane, and he obliged. Since then I’ve been flying my Bixler 2 every chance I can get, and in more conditions than I thought possible. Two kinds of flying really caught my interest: thermal soaring and aerobatic slope soaring.

The requirements for the two couldn’t be more different. Thermal planes tend toward the traditional image of a glider: long, high aspect wings, spindly bodies, low weight, and incredible glide ratios. Aerobatic slope soarers tend to be stockier with less traditional shapes, and are usually made out of some sort of foam to survive the inevitable impacts with the ground, or with each other. You can slope soar a thermal glider just as you can at least try to thermal with a slope soarer. But it helps to have a plane that’s really designed for the task.

The Bixler 2 is a seriously fun plane. It has enough aspect ratio to its wings to make it possible to catch a thermal, and it’s stocky enough to survive slope soaring. But it’s not ideal for either. Without massive thermals it’s a little frustrating to fly that way, and it’s delicate enough that risking a crash just to try a new maneuver on the slopes just isn’t worth it. So I decided to keep using the Bixler 2 for what it’s good at – park flying – and look at other options.

I’ve learned that in the world of RC airplanes, there’s no “next” plane. “Next” carries the connotations that you’re supplanting the original – that something is taking its place. Instead, there’s “another” plane. “Another” implies that you’re adding to your hangar, so to speak, and increasing your options for what you get to fly on any given day.

This is akin to how kite aerial photographers considers their kites. I’ve only had one “next” kite. When I started KAP, I was using a kite that was utterly unsuited to the purpose. So I really did replace it with my “next” kite – an Air Affairs Flow Form 16. From then on, I got “another” kites. The 6′ rokkaku my friend and I made was my first “another”, followed by my Fled. After that came the G-Kites Dopero, the Didakites rokkaku, and the PFK Nighthawk. My original KAP kite is long gone, sold to someone who would no doubt enjoy it more than I did. But the rest all live happily in my kite bag, each waiting for conditions to be right for them to be pulled out and flown.

So I bought some “another” planes to compliment my Bixler 2:

The first was a Zagi 5C wing. There’s a group of people who do slope soaring on the north side of Pololu Valley. The favorite among this group is the Zagi. It’s a 48″ flying wing that was designed for slope soaring, and for what’s called “combat soaring” in which the idea is to try to hit each other and knock each other to the ground. When a plane is “killed” in this fashion, its pilot runs out, picks it up, tosses it back off the cliff, and keeps flying. Hey! What better way to learn slope soaring than with a plane that can take that kind of abuse!

The second was a Le Fish from Leading Edge Gliders. This is a 66″ wide fully aerobatic slope soarer designed by Steve Lange. It’s almost completely neutral in all attitudes of flight (meaning it flies inverted as easily as it flies upright, and can even knife-edge), and like the Zagi, it’s made from EPP foam. And like the Zagi it, too can handle the rough impacts that are part and parcel of learning to fly aerobatic maneuvers close to the ground.

The third was a Raptor 2000 Advance from R2 Hobbies. This is a 2m motor glider that’s designed for thermals and for high speed. R2 Hobbies also sells a non-motorized R2k fuselage for pure thermal flying, but since I don’t have room for a winch or high start, I opted for the motor. The plane is offered in several versions. I went for the one with the Dbox wing construction and flaps. The control surfaces of the Raptor and the Bixler are similar enough that I hope to be able to try out full house glider mixes on the Bixler before applying them to the Raptor.

Which brings me to the last bit of kit I bought: a new radio. The radio I’ve been using – a Turborix 6ch 2.4GHz radio I picked up for KAP work – is a good solid piece of equipment, but it doesn’t offer enough options for setting up a full-house glider. The new radio I chose, a Turnigy 9XR, offers a lot more options for gliders, and should let me set up all four planes just the way I want. And if not, it’s one of the most hackable radios on the planet. The guts of the thing is an AVR processor from Atmel running open source code developed for an earlier Turnigy radio. Want a new mix that the radio doesn’t offer? Write it. And since I did AVR development for robotics several years ago, the language and development environment is familiar territory.

So I’m in for a penny and in for a pound. I’ve got my next radio, three another planes, and probably several months of building and covering to go. True, it didn’t happen in kindergarten when I first caught the spark. But it happened.

– Tom

Posted in RC Airplanes | Leave a Comment »

WWKW 2013 – Take Two

Posted by Tom Benedict on 26/03/2013

Ok, I lied. It turns out I will have new gear to try during WWKW 2013. I’m replacing the monitor on my ground radio.

Two things led to this. First, the monitor I have is one of those auto-switching “I go to blue screen when confused” monitors that show up so much on Ebay. It was a great investment at the time because it only cost me $40. But it also means that any noise in the video signal, and it fritzes out. Radio links always have noise in the video signal. This is a fact of life. So… It was constantly fritzing out.

But that wasn’t enough to make it useless. That flight I did over the school garden in town was proof of that. Without the video feedback, I never could’ve done that session. Just enough to be useful, but just unreliable enough for me not to trust it. OOOOOH that’s frustrating!

The second thing that led to this change is that my eyesight is going. As my optometrist put it, I’m already far-sighted. Add in that my eyes are hardening prematurely, and the answer is: I’m screwed.

I’ve been able to ignore this to some degree, but it was really brought home to me yesterday when I had the rare opportunity to play some pool. I haven’t played even semi-seriously since college. But I was appalled by how bad my game was! Every time I scratched, another ball came back out on the table. At one point I realized I’d run out of balls to put back on. GAAAAH!

That’s when I started to look at the problem analytically. I realized I could focus on the target ball if I took my glasses off, but then I couldn’t see the cue ball or my cue. If I put my glasses back on, I could see the cue ball, but not the target. Like I said, I’m screwed.

With KAP my eyesight is rarely an issue. By definition, things are far away: kites are high, cameras are distant, subjects are more than an arm’s reach away. And my glitchy little blue-prone monitor? It’s right at that distance where it’s impossible to see without glasses!

So I splurged and picked up a 7″ monitor. It doesn’t do the blue screen thing, and it has a very solid mount for attaching it to my transmitter. Even better, it has a VESA hole pattern on the back so I can package my video receiver and battery pack into a nice neat enclosure that can bolt to the back of the monitor. It even comes with its own sun shade.

I doubt I’ll have all of this changed over by WWKW 2013. Making custom enclosures is a royal pain, especially if you have to be able to get a battery pack in and out for charging. (No, I don’t charge LiPo batteries without removing them!) But I’ll have something cobbled together enough to test with. That’s good enough for me.

Oh! And the repairs to the Fled? One word: AWESOME! I took it out flying yesterday and got it up in a thermal. Like the song goes, she was climbing the stairway to heaven. I swear I couldn’t stop laughing. I never realized how much I’ve missed that kite.

– Tom

Posted in Kite Aerial Photography | 2 Comments »

Getting ready for WWKW 2013!

Posted by Tom Benedict on 24/03/2013

World Wide KAP Week 2013 is right around the corner. The dates this year are from Friday, April 26th to Sunday, May 5th. One month to go!

I really like to have new gear to fly, or new things to try, or new places to go for WWKW. But with life the way it’s been the past few years, I haven’t managed to pull this off. The damped KAP pendulum is still in the design phase, my video gear is more or less dialed in, and what with turning into a homebody since Rydra’s surgery, I don’t have new spots planned out.

Instead I’m using WWKW 2013 as motivation to go through my gear, fix the things that are broken, mend the things that need mending, and generally take care of as many minor annoyances as I can.

Today I started with my trusty Fled. Just over a year ago (a whole year!) I broke the spreader on my Fled. Rather than replace it with the 6mm tube it came with, I replaced it with Skyshark P200. So far so good. I only got into trouble when I tried to replace the Velcro pocket the spreader fits into. The new pocket I made was too big, I didn’t get it aligned quite right, but worst of all I was using a sewing machine that had no business getting near a kite. I had no way to set the thread tension the way I needed it to sew that @$%! pocket on. In the end I was swearing, shaking my fists, and threatening to throw the kite and sewing machine out the window. I shoved the Fled in its bag, crammed it in the closet, and stewed.

Rydra, in her infinite wisdom, used that as an opportunity to order a new sewing machine. We wound up with a lovely Janome that not only handled ripstop nylon with grace, it also helped fix the jib on our sailboat, made several light baffles for work, and has generally been a workhorse of a sewing machine.

But I’ll be damned if I pulled that kite out again!

And for another year I didn’t. Then only a few days ago my younger daughter asked about it. “Yeah yeah… I know where it is. Grrr…” Then she hit me with her Kid Magic: the eyes… the pleading look… the words! “I’d love to see it fly again.”


So today I pulled it out. The snarl of stitching I’d originally done came clean after a few minutes with a seam ripper. A quick trim and melt job with a lighter, and the pocket was a more appropriate length. And sewing the new pocket back on? It was an absolute DREAM with the new machine. It came together so well, and looked so good, I had to keep going!

I re-capped the P200 spars with a pair of “Bic Pen” style end caps. These are caps I make myself on the lathe using Delrin. They look sort of like the plugs in the end of the Bic Pen: outside diameter matches the OD of the spar, a hemispherical rounded end, and a press-fit into the spar leaving no seam. Hands-down this is my favorite end cap for slipping spars into a kite. Nothing to snag, nothing to cut the fabric.

With the Fled back together, I started to look at my other kites, too. I’m planning to re-cap my Dopero’s spars with Bic pen caps, and replace the dying Velcro on the bottoms of the longerons with new Velcro, backed by nylon webbing. I might do the same to my Didakites rokkaku, too (though so far I’m pleased with the caps they used.) Then it’s time to rummage through my kite bag and find other stuff that can be fixed.

I’ve got about a month before WWKW 2013. No new gear this year. No new techniques or places to fly. But if I can get everything fixed before I head out, maybe it’ll be just as good as if I did.

– Tom

Posted in Kite Aerial Photography | Leave a Comment »

Offroad Camera Project Revisited

Posted by Tom Benedict on 22/03/2013

I did some more work on the offroad camera platform I wrote about back in 2011. As delivered back then, the platform had a couple of serious weaknesses. First, having a highly compliant suspension meant that it never truly sat flat. So no matter where you drove, no matter what you took a picture of, your horizon was never, ever, going to be level. Second, the pan/tilt head from Servo City was inherently unbalanced, so the servos were constantly fighting the weight of the camera. In addition to burning through the batteries and making a pretty loud racket as the servos fought for position, this also meant that when power was removed, the camera slammed down either forward or backward onto its hard stops. Toss in a number of other minor annoyances like that bundle of cables on the side of the camera, and you can see it was far from ideal.

Crawler 1

So the 6WD camera platform came back for a mod job. Here’s the list:

Remove the Servo City tilt axis, replace with a tilt/roll unit:

The new unit is a variant of the HoVer KAP rig I made for my A650 years ago. In the case of my KAP rig, it was set up so the two extremes of motion on the servo gave me horizontal or vertical orientation for the camera. By flipping a switch on the transmitter, I could toggle between those two modes.

HoVer Axis

In the case of the offroad camera platform, the roll axis was centered on the horizontal position, with +/-45 degrees of motion to either side to level the horizon. The only real mechanical differences are that the tilt frame had to be a little wider to clear the video and shutter cable on the side of the camera, and I used a beefier arrangement for the roll axis using Servo Blocks from Servo City.

Servo Block Built

Unfortunately, when I put everything together I found that the Brooxes KAP gear was too flexible to work in this application. But let’s face it: In the world of KAP, having a slightly compliant rig is a bonus, not a detractor. And lightweight is the name of the game. So it actually works really well for what it’s designed for. But this thing is intended to roll over rough terrain and potentially get thrown around a lot. Rather than start over from scratch, I beefed up the main frame with corner bracing and a brace across the back to stiffen it. The tilt frame didn’t get any bracing, but if it looks like it’s necessary down the road, it’s easy enough to add.

Roll Axis

Replace UBEC voltage regulator:

In testing the changes to the camera mount, I ran across another problem: Occasionally the servos would jerk and shake, and sometimes while driving at full speed the motors would slam to a stop, then resume at full throttle a moment later. Scary! I wound up checking the servos, the speed controllers, and finally checked the voltage regulator I’d installed when the unit was first built. GAAAH! It was noisy! Time for new hardware.

A while back I converted the video system on my KAP rig to a 3S LiPo battery, but kept the 4xAAA battery pack for driving the RC receiver. On the advice of a fellow KAPer (hi, Bill!) I’ve been wanting to install a UBEC to power the radio instead. What a perfect opportunity! I ordered two 3A continuous / 5A burst UBECs from Hobby King, and when they arrived one went in my KAP rig and the other in the robot. The KAP rig worked perfectly first try (thanks again, Bill!) and the installation on the 6WD robot was a snap. Most of the noise went away. (Yeah… most… more on that in a sec.)

Replace the antenna support:

Back when I started on this, there was limited selection for radio and video gear. The only video link that fit the bill was a 2.4GHz system. This precluded using a 2.4GHz RC radio, so we went with a 75MHz ground radio from Hitec. I’d post a link, but it has since been discontinued. The one gotcha with dropping to a lower frequency is that the antenna gets longer. And for best reception it needs to be mounted well clear of the vehicle. The first rev used a fiberglass rod I mounted in a Delrin block to support the antenna. But it was hard to transport, was fragile, and finally snapped off at the base.

For this rev I looked to the RC car and truck world for inspiration. Not only did I find it, I found nice off the shelf components that fit the bill perfectly! Now the radio antenna is supported by a Dubro antenna tube that mounts in a nice machined aluminum collet system from Integy. It holds the antenna out of harm’s way and up in the air for good reception, and is easy to break down for transport. In the end I bought two: one for installation and one as a spare. But I wound up mounting both because it’s handy to have a second mount on the drive system for those times when the upper deck is removed for servicing.

Antenna Collet

Build a new video cable

As you can see in the photo from the original 2011 build, the cable from the camera to the video transmitter was the stock Canon A/V cable that came with the camera. “Ungainly” is a kind word to describe that bundle of wires. But with the addition of the roll axis to the camera head, “disaster” came a little closer to the mark. The stock Canon plug wouldn’t clear the tilt frame, and the cables constantly tangled on everything. So I took the last of my Ebay cables from building the T2i video downlink for my KAP rig and made a custom cable for this camera. Since there’s already a shutter cable running from the video transmitter to the camera, I made them both the same length and zip-tied them together. Now they’re one neat little bundle that fits fine and doesn’t cause tangling.

Custom Video Cable

Track down that @$%! NOISE!!!

I mentioned the servo noise in the section on replacing the voltage regulator. I got most of it, but not all. The pan axis continued to exhibit noisy behavior that I couldn’t track down. After double-checking the UBEC (fine), I tore the transmitter and receiver apart.

Here’s another one of those areas where hobby bleed-over is a good thing. A while back I bought a cable on Ebay that lets me use my RC transmitter as a controller for RC flight simulators on the computer. It came with a bunch of different plugs for different transmitters. When I got it I plugged in the one for my radio, and I was up and running.

The cables work by picking up the pulse coded signal off of the radio and turning it into positions for each of the axes via the USB port. Aha! What a great way to test the RC transmitter! It turns out I had a Hitec plug for this cable, so I plugged in the transmitter and took a look. The potentiometer driving the pan axis tested out fine on my meter, and when I tried it with this cable, it came out completely clean. Most radios tap the pulse coded signal right before it goes into the transmitter electronics, so at that point I figured the transmitter was fine.

The receiver also turned out to be fine. For what it’s worth, the noise filtering on the Hitec 75MHz receivers is remarkably clean – almost pretty. This one was no exception, and tested out fine. No blown caps, no cold solder joints, no scorch marks. Nothing!

Eventually I broke down and tested everything using an independent radio, battery, and servos out of my parts drawers. The only pattern I found was that the pan servo on the camera head started to have jitter if I plugged in the tilt servo from the camera head. No other combination caused the noise. And it didn’t matter which channel either was plugged into. The tilt servo was the culprit!

Unfortunately the only matching servo I had was in my KAP rig. For the sake of the project I gutted my KAP rig and installed the new servo in the camera head. It tested out fine. But I had to order another servo for my KAP rig!

After re-installing all the servos, tying down all the wires, and bolting on the upper deck for what I hoped was the last time, I gave it a test run. Jitter free at 250′! YAAAAAY! The noise was finally GONE!

Lock it down and send it home!

The last step was to remove every single screw in the thing, apply blue Loctite thread locking compound, and replace it. No sense having this thing fall apart in the field! I also neatened up the wire runs, tied down the servo wires, and applied rubber to the spots where the camera would contact the camera mount. At long last, the offroad camera was ready to go home.

6WD DSLR Revised

This was, and continues to be a fascinating project to work on. I’ve learned a lot in the process, and will doubtless keep learning as I service this in the years to come. But for now I’m ready to fix my KAP rig and get back in the air.

– Tom

Posted in Electronics, Photography | 2 Comments »

First RCAP Video

Posted by Tom Benedict on 17/03/2013

I know I said I had no real plans to stick a camera on my RC airplane. But I went and did it anyway!

I actually had a legitimate reason to do this. I found, during testing, that my plane suffers from a good bit of adverse yaw. I won’t go into a real discussion of adverse yaw and the reasons for it, except to say when you do an aileron roll, your tail wags out to one side. (The other side is “proverse yaw”.) This can be corrected by mixing in some rudder with your ailerons, but a better fix is to dial in differential aileron throw. This means that you set up your ailerons to move up more than they move down (or vice-versa). This can be done mechanically if you’re good with tuning linkages. OR you can do this using a computerized radio that lets you change the throw for each servo, in each direction. I chose the latter.

But the question comes up: how do you measure adverse yaw? The classical experiment is to fly the plane directly away from you, throw the ailerons hard over one way, then the other, and watch which way the tail wags. If there’s no adverse or proverse yaw, it shouldn’t wag at all. Mine wagged. So I dialed in some differential.

That’s when the wind picked up. It got harder and harder to see what the tail was doing. That’s when it hit me! Stick a camera on the plane!

In the kite aerial photography world, video mode on modern digital cameras has been a real life saver when trying to find out why your shots are blurry. Put the camera in video mode, put it up, and see what it’s doing! If it’s nodding up and down, the kite is pumping the line. If it’s swinging rotating back and forth, you might spread out your Picavet clips to help damp the motion. And if it’s just going all over the place, maybe it’s not such a good day to fly!

My hope was that I’d be able to do something similar with my plane: stick a camera on it, start video mode, and go through the motions. If I had any remaining adverse yaw, I should see the nose swing out as I entered a turn.

My camera “mount” was a copious application of blue painter’s tape to hold the camera in place. Sorry, no pan/tilt or head tracking. I’m too low tech for that. The camera was aimed up slightly because of the geometry of the plane. This kept most of the nose out of the shot, but it also meant that until I cut the motor, the ground wasn’t very visible. Still, I got the information I wanted. I still have some yaw to deal with, but not enough that I’m going to pursue it any more while I’m still using this radio. And I got a nice view of the town as well.

I’m also thoroughly convinced that KAP will still be my go-to technology of choice for doing aerial photography. Flight videos are fun, but nothing beats a tack-sharp high resolution KAP photo.

– Tom

Posted in Kite Aerial Photography, RC Airplanes | Leave a Comment »

When is KAP your best bet?

Posted by Tom Benedict on 15/03/2013

In my last post I mentioned that a particular subject really needed KAP to get the job done. Truthfully, a helicopter or multirotor probably could’ve done the job more easily, and in less time. I guess what I meant was I couldn’t do it from a pole, a rig suspended from a tree, or any of the other aerial platforms I’m currently using. (Nope, no multirotors in my stable of AP platforms yet. YET…) Which leads to the question: When is KAP the platform of choice? Or is there always some other aerial platform that’s more appropriate?

Before going into my reasons why I think there are cases in which KAP is the most appropriate platform, let me clarify: I’m talking about low altitude aerial photography. No full-sized airplanes, helis, satellites, or alien spacecraft. I’m sure you can see the garden I photographed yesterday from a thousand feet. But getting that kind of resolution from that altitude isn’t easy. For low altitude work, a low altitude platform is almost always the better choice.

One case that comes to mind is radio quiet zones. This may sound odd, but I work in one. Our building is close to several radio telescopes. Using a cell phone, ham radio, or RC transmitter near one of them essentially blinds them. It’s like shining a flashlight into an optical telescope. Bad idea. I know that with a GPS and waypoint system I wouldn’t even need the radio. But if the aircraft got close to the beam of any of the radio telescopes, the RF coming off the motors would be enough to raise the background on the telescope’s signal. Still not great. But a kite, a KAP rig, and an autoKAP controller, and you’re golden.

Gemini North Sunset via GoPro

Another that’s more common is whenever the aerial photography is done directly over people. For me, aerial portraiture is a fairly specialized form of KAP that requires 100% concentration on what you are doing. Let’s face it, getting hit with a KAP rig because the KAPer wasn’t paying attention would be quite painful. But getting hit with a multirotor or a generously sized RC helicopter? That’s closer to deadly. When I’m flying over people, give me a kite every time.

Sand Play for Two

Any time I naturally want the camera to be tethered to the subject, I think of KAP. A good example of this is a boat that you’re photographing from above. I know it’s possible to launch and retrieve multirotors from boats. The grassroots mapping of the British Petroleum Deepwater Horizon is an excellent example of this. But after watching some of the behind the scenes videos of those launches and retrievals, I have to think that using a kite is a lot less stressful and more likely to succeed. Besides, once the rig is airborne, you really can sit back and enjoy the ride.

Speed Boat 2

One case that comes to mind is one in which the subject isn’t necessarily comfortable with the idea of being photographed. No, I’m not talking about paparazzi, or stalkers, or stealthy surveillance. I’m talking about wildlife photography. I’ve found that many animals that would otherwise register my presence if I approached them on foot don’t react at all to a KAP rig floating overhead. This has let me photograph them from a vantage point that’s not available with a ladder, pole, or anything with a spinning propeller.

Honu 3

So when else is KAP the platform of choice? I don’t know. For me, it’s whenever I want to relax while doing aerial photography. For others that may mean RC airplane, or helicopter, or a hundred foot pneumatic mast. But for me, it will always mean a kite, a string, and a sky to fly it in.

Fine Guidance System

– Tom

Posted in Kite Aerial Photography | 2 Comments »

A Challenging KAP Session

Posted by Tom Benedict on 14/03/2013

I’ve had a long-standing request to photograph one of the gardens in town. I’ve tried with a kite, I’ve tried with a pole. The first time I photographed it from above, I slung a line from a nearby tree and used that to hoist my KAP rig. But since then the tree I used was cut down, so I really had no way to pull it off.

Then I got a new request to photograph a different part of the yard in order to help plan a project at the school. That location clearly needed KAP, so I gave it another go.

These photos are straight off the camera, except for a re-size in Photoshop. The last one is a little dark because the sun went behind a cloud.


It all started innocently enough in the field near the place where I work. The school is just on the other side of a big stand of eucalyptus trees to the right of this photo. The trees are the real obstacle when photographing the school. They’re 80’+ high, and far enough away from the school to make the geometry really unfriendly to KAP.


That’s when I saw The Gap. Ooooh yeah! Not only was there a gap in the trees, it was almost at the perfect angle to fly a rokkaku through it! W00t!

The trip through the trees was hairier than I like, but I managed to get through with all my gear intact. The camera was hanging from the kite line for all these photos, but I didn’t let line out to raise the camera until I was past the trees.


Aaaaah! Subject #1! For the record, I couldn’t actually see anything during most of this session. My rig was in the sun, so I couldn’t see where it was or which way the camera was pointing. I had 10-20′ high bushes between me an the subject, and trees at my back. I was flying blind. Except for the fact that I was using a video downlink! Now that I’ve used it on a flight that would’ve been close to impossible without it, I’m starting to like it more. I used my A650IS for this session, so rebuilding the video downlink to interface to both of my KAP cameras came in handy.


This gives you a better idea of the challenge involved. In the foreground is the school. In the mid-ground is the garden and yard. Behind that is the line of bushes I couldn’t see over. Past that is a dry creek bed, and past that is the path I was standing on for most of this. Just out of view at the top of the frame is the big stand of eucalyptus trees.


This one is a little dark, but it further illustrates the challenge with this session. Yep, that’s me in the upper left corner with the kite line stretching up. Those trees are big!

Once the KAP work was done, I reeled in my rig and got it off the line. The trick after that was to reel the kite down until it was below the tops of the trees. Past that it came down like a big floating elevator. I managed to hand-catch the kite without dipping the line to the ground. A good end to a good session.

Hey, I almost forgot! I flew on the other side of the eucalyptus trees back in 2010:

St. James Circle and Waimea Country School

You can see how far the garden has come since then. It was just fallow beds waiting to be planted.

– Tom

P.S. Yeah, I didn’t mention the OTHER challenge with this session. In the next to last photograph you can clearly see what look like utility lines just above the school. Those really are utility lines, and the upper line really is electricity. It’s insulated (I checked) but I was very VERY concerned about the possibility of touching it with my kite line. If the wind hadn’t been perfect, and the kids hadn’t been out of school, and things hadn’t been working perfectly, I never would’ve risked it.

Posted in Kite Aerial Photography | 4 Comments »

Home Shop Fabrication Tips #1 – Panel Labels

Posted by Tom Benedict on 13/03/2013

In the perfect world, we’d all have silk screen setups for making beautifully inked labels for our panels, enclosures, or anything else where a connector might poke through a wall. Unfortunately I don’t have one of those in my home shop. I don’t have on at work, either. Instead, here’s a trick we use at work that serves just as well in a home shop:

It all starts in CAD. I use a commercial CAD package, but there are inexpensive or free CAD packages out there as well. Since you have to cut the holes in the enclosure in the first place, might as well draw up all the holes in CAD anyway. Making labels is just one more step that’s well worth the effort.

Design Label in CAD

Next, print the labels 1:1 on your printer. Black and white works fine, but I like to use color whenever I can. Most of the stuff we make at work is used at 14,000′ of altitude, so by definition we’re mentally challenged whenever we use it. Having big bold labels with clear names, lots of color, consistent connector orientation with keys all facing the same way, etc. is a really good idea. This set of labels is for some fiber connectors. We put a band of electrical tape of the appropriate color around each fiber so there’s no confusion about which jack it plugs into. On larger electrical cables, we’ll name them, number them, and still color-code them, just to be sure we don’t screw up. (Actually, we use one other trick: For electrical connectors, we use a different shape connector for each cable that needs to be plugged in. That way we can’t screw it up!)

Print Label 1:1

The next step is easier if you have your own laminating machine, but even if you don’t it’s well within the capabilities of most copy shops. We keep a stock of adhesive-backed laminating films at work for making labels. Simply laminate your 1:1 printout, and cut out your labels on a paper cutter. If you don’t have your own laminating machine (hey, I don’t!) you can take your 1:1 printout to a copy shop and have them do this step for you.

I always draw a pinstripe outline where I want to cut. This makes it tough to screw up when using the paper cutter. These outlines also tell me where to cut holes for each of the connectors. Since I already have the outlines in my CAD file to tell me where to make the holes in the panel for the connectors, this is really a no-brainer.

The next step is to cut out all the holes. The toughest for me are the screw holes. Luckily Fiskars and several other manufacturers make a line of hole punches with hole sizes varying from 1/16″ up to much larger sizes. If in doubt, grab ’em all! Here I’m cutting out holes for #2-56 screws using a 1/16″ hole punch. This is smaller than the clearance hole size for these screws, but I haven’t had a problem using a slightly undersized hole. The screws just push through the label when you install them.

Hole Punch Clears Screw Holes

The rest of the outlines can be cleared with a craft knife or razor knife. I tend to freehand mine, but using a straight edge makes for a cleaner cut.

Craft Knife Cuts Outlines

Once the labels are all cut out, they can be installed on the panel, enclosure, or whatever else you needed to put a connector on.

Installed and Looking Good!

Voila! Nice, readable labels that are quick to make and use. In case you’re wondering, these are the labels for the fiber connectors on one of our instruments. The people here started using this trick back before laser printers existed. We still have labels that were made on a pen plotter! With a good lamination job, they’ll stand up to all kinds of abuse.

– Tom

Posted in Home Shop | Leave a Comment »

Damped Pendulum KAP Suspension

Posted by Tom Benedict on 06/03/2013

In an earlier post I alluded to a pendulum test bed I wanted to build for experimenting with pendulum suspensions for kite aerial photography rigs. I haven’t cut metal yet, but I’m far enough along with the design to start sharing details.

Before I get into it, first a note about designing long spindly structures: Put simply, it’s a pain. The areas of interest wind up being separated by large swaths of open space that need to be navigated repeatedly in CAD. And visualizing the entire structure in one go becomes difficult at best. Shown below is a rendering of what the test bed is likely to look like when built. As you can see, it’s not fun to work with in CAD.

Damped KAP Pendulum - Uncompressed

Rather than put myself through this, I did the design work using 2″ long spars in place of the full-length spars I intend to install in the final hardware. Keep this in mind when reading the rest of this article. The design is compressed for the sake of making the CAD work easier. The final hardware will use full length spars! Shown below is the “compressed” version of the design using 2″ spars:

Damped KAP Pendulum - Compressed

Much easier to work with.

First and foremost, this is an experimental test bed, not a finished design. The whole idea is to build a kit of parts that can be used to test various ideas quickly and easily without having to make new mechanisms for each test, preferably just by swapping parts in the field. As much as possible, parts are interchangeable and should be easy to remove and install. Global geometries are defined by only two 2D parts to make it easy to test additional geometries without remaking every component of the design. And the dampers – one of the key elements under test – are tunable. Finally, the whole thing should be able to fold into a compact shape that will fit in a typical kite bag.

Moving from top to bottom, there is a stage at which the pendulum attaches to the kite line, a parallelogram stage, and a single pendulum stage. The entire design is modular, so stages can be re-arranged, or omitted altogether. This allows for experiments with a single pendulum, a double pendulum, a parallelogram pendulum, or a double pendulum in which the upper pendulum is a parallelogram pendulum.

The spar that the kite line attaches to is not completely drawn. The spar can be rotated into position against the locating pin, and is then locked into place using a 1/4″-20 screw (not shown). This is so the rig can be collapsed for transport. Also not shown at either end are the hook plates that keep the kite line from coming loose. I plan to use something very similar to the arrangement I had on an earlier gimballed suspension rig. In use the line is hooked under one end plate, and is then wound around the spar and under the other end plate. This arrangement worked remarkably well on the gimballed rig, though I plan to use a slightly different geometry for the end plates on this rig, and plan to add Velcro “keeper” straps.


Each pivot is made using a zero-play double row ball bearing cartridge, housed in a custom machined ball end. The stem on the ball end extends into the spar by at least one inch. This is deep enough that the clamp blocks never clamp on an empty spar – there is always a plug filling the spar to keep it from being crushed. The bearing cartridges are a slip fit inside the ball ends, and are fixtured using a screw that fits into a tapped hole in the side of the bearing cartridge.

Several of the pivots can be damped using stacks of slip plates, made from FR4 and packed with Nye Lubricants damping grease. The damping force of each damper can be tuned by adding or removing slip plate pairs, and by changing the grade of damping grease in the stacks. The stacks are fixed at either end. One end is fixed in the red bracket, and the other on an aluminum clamp block that is attached to the spar being pivoted. Any combination of dampers can be used, ranging from none installed to all installed.

By changing spar length and damping force at each stage, the oscillation characteristics of each stage can be individually tuned. For a given weight KAP rig, it should be possible to choose a primary frequency of oscillation, and to over-damp, under-damp, or critically damp each pivot.

So far all of these pivots damp motion in the pitch direction. Because of the bearings at each pivot, yaw should be quite rigidly constrained to track the kite line. But I am designing another damper to fit on the pan axis of the KAP rig to take out line wobble in yaw, should there be any. As with the rest of the design, this can be installed or removed for each test.

This leaves the roll axis. There is one additional experiment I’d like to try for damping oscillation in roll, but it can be attached to the hardware already shown using clamp blocks similar to the ones used on the dampers. It amounts to a damped pivot with two spars, each carrying a counterweight. This can be fixed to any of the spars shown, including the line suspension spar. It will use a custom mount block that allows the orientation of the pivot point to be changed anywhere between 90 degrees to the spar all the way to inline with the spar.

What I hope to gain out of this is a better understanding of the mechanism that causes rig shake, and the information I need to build something that’s easy to transport and use, and results in wobble-free KAP.

– Tom

Posted in Engineering, Kite Aerial Photography | 4 Comments »

Sitelle Getter II

Posted by Tom Benedict on 01/03/2013

Two posts ago I alluded to a design change to the cameras at work. Since I posted pretty fab pictures of the first rev of the getter, it’s only fair that this one has its turn to shine.

The original design was novel in a couple of ways. It had two chambers, one on top and one on bottom. Each was filled with its own charge of activated charcoal, and each had its own filter, protective screen, and polished aluminum cap. This made for a fairly high volume / high throughput getter in a neat little hockey puck sized package.

Sitelle Getters

Unfortunately two problems killed this getter. The first was that it simply didn’t fit. There’s a fair bit of material between where the cold head bolts on and where the cold strap attaches. In the original design of the camera, the flex circuit for the electronics feedthrough passed quite close to the bottom surface of the getter. In practice, though, the flex wound up being longer than expected, which caused the flex to touch the getter. The getter is maintained close to -190C, and the flex is maintained with one end at -100C and the other at room temp (typically 0-5C). Having a thermal short to -190C right in the middle was going to cause problems. To get past this, we machined off the bottom chamber of one of the getters, and moved on.

The second problem that killed this design was that the cold strap attached through the force of a magnet embedded in the getter attracting a second magnet, embedded in the top of the cold strap. Unfortunately the cold strap was too stiff to let this arrangement work well. The magnets were too weak to overcome the springiness of the cold strap, no matter how carefully we arranged things during assembly. And moving to stronger magnets would’ve meant the camera was harder to disassemble. So instead we ditched the magnets and moved to a conventional bolt-together design.

All of which completely invalidated this design for the getter. So I started over. I was planning to make another CNC-heavy design with lots of sexy curves, vented screws, stainless mesh, etc. But right as I got ready to fab the thing, I found out our CNC mill was completely tasked to making new filter frames for one of our other instruments. Whoops!

So I re-designed it (again) to be made entirely using conventional hand-crank tools. And I got rid of the screws. And used thinner material – a good thing in a getter since it means the getter is cooling faster, and is pumping before nasty stuff like water vapor can condense on the face of the CCD. The new design is made using 2mm plate and the same 0.010″ thick copper strap we use for our cold straps. They’re not as nice looking as the original design, but they should work.

Sitelle Getter II

And truth be told, they’re probably better than the original design in a number of ways. The overall cavity size is larger, and I managed to pack almost twice the carbon inside as the original getter. There’s a third copper strip hidden inside the getter that provides a thermal path up inside the bulk of the carbon, so it should cool quite quickly. I kept the filter, so still no chance of carbon dust getting into the camera body. And considering I made four of these in three days, it was a quick solution to a nagging problem that’s been hanging over my head for months.

The only real drawback to this design is that there’s no way to service the thing. If the filter tears or finger oils get on the top, there’s no real way to fix it. Toss it and grab another. Which is why I made four of these for two cameras. Spares!

I probably won’t have a chance to test these until next week. Meanwhile I need to make the bolt-together cold strap for the two science cameras. The prototype has been cooled down twice, and is working perfectly.

So now all I have to figure out is what to do with the two surviving first generation getters. They work! They just don’t fit inside these cameras. And it’d be a shame to turn them into paperweights.

– Tom

P.S. Yes, in case you’re wondering, all those holes were drilled by hand. On four parts. Nothing fancy, just a sacrificial plate bolted to the mill with the part clamped on top. The rest was all trig, and a lot of triple-checking my position so I didn’t screw anything up! All four turned out perfectly. But I swear, if I can skip drilling any holes for a while I’ll be ecstatic!

Posted in Astronomy, Engineering, Machining | Leave a Comment »