rbandrews

A couple days ago I bought Phoenix RC, an RC aircraft simulator. It's an amazing amount of fun, I've been playing with it like nonstop. Thursday night I stayed up Far Too Late messing with transmitter settings (on the transmitter itself; I can use my tweaks on the real quad) and last night I was playing with acrobatics (which I will likely never do with the real, $400 quad, but is a lot of fun on the imaginary one). It has a USB widget where I can plug in my actual controller, and the physics are great. It emulates different kinds of wind, you can tweak the weight / horsepower settings of the quad, simulate random equipment failures, etc.

Anyway, after doing that for two days, and today being the first nice day in a week, I took it out and immediately lost an arm. I was practicing zooming as fast as I could toward the construction site and then back, and I flew too fast too close to the ground and crashed.

That construction site makes me sad, too: we used to have this enormous field behind our office, but now past about 50 yards it's a muddy construction site. For what I'm sure is going to be Joe's Pine Tree, Doberman, and Swimming Pool Emporium or something equally unfriendly to fly over.

It was a pretty impressive crash, to be fair. Even from about 30 yards I could see little bits fly into the air, the thing flipped and rolled a couple times, it was very photogenic, unlike when I lost a prop flying it into a ping pong table. I managed to find all the bits that fell off, and I snapped an arm in half.

Funny thing: the props, including the ones that dug into the ground, were perfectly fine. Carbon fiber is some tough shit. The arm was toast but the prop on it still looks (and works) great.

I have eight (now seven) spare arms, so it's already flying again. But, first major crash! Woo!

I made the lights on the quadcopter operate by remote control. Here's how that went:

There were two general problems to solve here: one is, I needed to control 12 volts to the lights, with a 5 volt signal from the receiver. So there's the voltage mismatch. The second is, the signal from the receiver isn't a simple on-off, it's a pulsed signal, so I needed some way to decode that to decide whether the receiver was saying "lights on" or "lights off."

First, the voltage. The usual way to control one signal with another one is a transistor, so that's the first thing I tried, but the normal kind of transistor (that I have a drawer full of) really wants the signals to be the same voltage. If there's too much of a difference between the base voltage (that controls whether the transistor is on or off) and the source voltage (that the transistor allows through, or doesn't) then the source voltage just goes through the base. I burned up a resistor last weekend playing around with this.

The solution to this problem is called a MOSFET: metal oxide semiconductor field-effect transistor. The difference between a normal transistor and a MOSFET is that once a little voltage starts to go through, that sets up a reinforcement effect that lets the rest through. So a small voltage to the base starts this avalanche going, and so a 5 volt signal will turn on or off a 12 volt signal.

Next problem, the pulses. RC aircraft are designed to have a bunch of analog signals to control servos: some level for the throttle, an angle for ailerons or whatever. An on-off signal isn't very useful on an RC plane, so RC equipment isn't really set up to produce it. What happens instead is you get a pulse: if the throttle (say) is at 0, then you get a pulse 1000 microseconds long followed by 1500 microseconds of nothing; if it's at 100% you get a 2000 microsecond pulse followed by 500 microseconds of nothing. But it's never fully on or off: the minimum pulse is 1000 us, the maximum is 2000 us. So I need something to decode that.

There were a few solutions to this: build an analog circuit which will charge a capacitor with the long pulse but the short pulse won't charge it enough, for example. Or buy another component called a speed controller, which is meant as a go-between to translate this kind of signal for an RC aircraft engine (which the copter already has four of). In the end, I cheated, and took a more expandable approach: I used an Arduino.

The Arduino reads the pulses, looks at the width, and either writes (or doesn't) to an output pin, which is connected to the MOSFET. This is kind of cheating since I used a fairly powerful computer where a few cents of passive components would do, but on the other hand now that I have this, I can use the same board do to other things later, like landing gear or control a camera or whatever.

But for now, what I have is, a board sitting on top of my copter that allows me to flip a switch and turn some lights on or off in flight. Which is kinda useless (I know when I take off whether it's dark enough for lights or not) but also kinda cool.

Today I:

- Went to EPO and spent about seventy bucks. I needed some parts for something I'm doing to the quadcopter, putting lights on it. Got 12V voltage regulators (Radio Shack supposedly sells them but they were out), LEDs (I actually already had these but I found better ones), perfboard, wires (better wires for this purpose than what I already had), stuff to make connectors, and some random odds and ends. I forgot to get a switch, which is a little annoying.

The goal here is to have red lights on the back arms of the copter and white lights on the front arms. This way I can see how it's facing in the dark, so I can fly it later in the evening, so I can fly it more often. One of the sort of annoying things about flying something like this is that it's easy to lose track of how it's oriented, since it's square and spends most of its time hovering in place.

- Cleaned my room. Cleaned off my desk, cleaned the work table, put things in boxes and in the closet. I hadn't done this since, oh, October sometime? And in that time I've built several small projects, one big project, added a tool (the cutter), spent two weeks sick, et cetera. My desk was actually buried. Like, the level of crap covered up the bottom of the monitor.

- Put the Commando 23 set on a keyboard. This was a group buy for some keycaps that I got in on a while back, and last week they were delivered. They were the final straw for junk piled on my desk. They look pretty all right on a keyboard though.

My transmitter module came on Friday, so today I went to the field behind my office to fly. Mike, my friend who ran the class for building them, came too.

Hovering inside is more or less easy. Holding a constant altitude is tough, the throttle is really sensitive, making it pretty much like every other RC helicopter I've ever flown.

Flying outside is a bit harder. I don't have to care as much how high I am, so it's easier in that respect, but then there's wind. Holding it steady in the wind (even a fairly light breeze) isn't easy. I did figure out how to recover from an imminent crash: cut the throttle, so I start falling, then gun it right before I hit the ground to soften the landing. I saved the thing from being carried away by the wind a few times by doing that.

We found out that there's an RC hobby place about 15 minutes from work, and although they don't have replacement rotors, they do have tools, servos, radio stuff, batteries, et cetera. I bought some tiny screws and a hot air gun (for something else I'm building).

I also found that some (expensive, carbon fiber) propellers can be had from Amazon for about $16. This is more expensive than the $4 / set prop sets from the Chinese site Hobbyking, but they also don't take three weeks to be shipped.

Why do I care about props all of a sudden? Because, naturally, I managed to flip it over (I blame the wind) and break a prop in half. So, it'll be Wednesday before I can fly it again.

...is love, sweet love.

Well, no.

If you take an average sea creature and measure the elements it's made of, you find that it (like everything else) is mostly carbon. But there are also some larger elements in trace amounts, that are in a higher concentration in the creature than in the surrounding environment.

The element that is most over-concentrated like this happens to be phosphorus. Phosphorus is the limiting factor of life: given a source of extra phosphorus, life could find everything else it needs to exist in plenty.

So I would say, what the world needs now is phosphorus. That's the only thing there's just too little of.

On the 13th or so, I started coughing, a lot. I coughed up white stuff, then yellow stuff, then brown stuff. Then I stopped coughing up stuff, but I didn't feel any better: I still had a fever, I was dizzy and light-headed, and I couldn't eat anything. And I kept feeling like this. I took off the entire week before I was supposed to go on vacation, then stayed in bed the entire time I was on vacation. My fever stopped a few days ago, but I've still been light-headed and crappy. I alternate between lying in bed reading and sitting up watching Parks and Recreation on Netflix.

So I have had a kind of shitty christmas. I spent all day either in bed or on the couch, watching the Doctor Who special and Indiana Jones and the Last Crusade. I ate a couple of ginger snaps and Cassie's leftovers from going to the Cleburne Cafeteria. Finally got Cassie to go on christmas day, and I couldn't ever go with her!

Hopefully I'll start to feel better now. I can do stuff I had planned to do over the past two weeks, like mailing out christmas presents.

Silk Soymilk "Seasonal Nog" is not only not terrible, it's actually good.

I had low expectations, I'll admit...

I've been playing on the "career mode" where you start with a few simple parts and have to do experiments in space to get science points to research more. So, last night I tried to make a rocket out of very basic parts to get to Minmus.

I got there, and got into orbit, but when it was time to go home I didn't have enough delta-v. I was able to escape Minmus but my Kerbin perigee was still about 1200 km.

I know that stage separators push the stages apart with some force, so I aimed myself right and separated my last stage. That got me down to 900 km.

Then it hit me: I still had one source of propulsion left. I had Jeb go on EVA, wedged him up against the capsule and fired his jetpack. One full load of jetpack fuel.

That did it, the last 30 m/s or so of delta V I needed.

So, at one point in the history of my glorious space program, yes, I had an astronaut get out and push.

Look at this. It may be my new wallpaper.

Jeb, with his shiny new-model lander, on the plateaus of Minmus, with a crescent Kerbin behind.

Minmus is Kerbin's outer moon, much smaller and with lower gravity. This same lander can actually land on the Mun unassisted, but it needs a refuel to get home. I should be able to return from Minmus easily though, since it takes so much less fuel to land there.