AdMob

For fun, I thought I would use the AdMob api to include an advert banner in my freeware “Metaloid” application that’s available on Android Market. This is a little banner that sits, rather unobtrusively, above the main screen (see the Figure below).

The Metaloid Field Detector application on Android Market

The Metaloid application has about 20,000 downloads of which 31% are still installed, so about 6,200 active users. The AdMob idea is that you get revenue every time somebody clicks on an advert. Since adding this “feature” (I use the word cautiously), I have earned a grand total of $3.50 for a total of around 80 clicks on 117,000 adverts shown. I shall buy myself a Double Tall Mocha.

Here is the breakdown by day:

WordPress for Android

… seems like a good idea. Let’s see if it works.

First Pizza!

After the adobe oven had dried off for a few days, we removed the sandcastle former by scraping it away from the inside. We lit a fire in the centre of the oven, using scrap wood, newspaper and some barbecue fuel, and let it burn for a few hours.

A couple of days later, we fired it up again, this time using scrap wood and barbecue charcoal bricquets. After a couple of hours or so it had reached a temperature of 550F. The outside walls of the oven steamed as the water evaporated away, and some hairline cracks appeared around the door – this is normal (they closed up when the oven cooled).

After removing the larger pieces of wood, and making enough space to cook a pizza, we cleaned the oven floor’s surface using a damp towel nailed to a long stick. That created a lot of steam! We were careful to remove all the sand and ash from where we wanted to bake – nobody likes a sandy pizza. Then we used some of Trader Joe’s fresh dough to make a simple pizza, and placed it inside – it cooked in a few minutes, and we had to turn it a few times to stop it burning.

(Normally we make our own pizza dough, but for this first experiment we liked the convenience of pre-made.)

The taste was amazing – especially the crust, which had a characteristic aroma and taste particular to pizza from a good Italian restaurant. There must be some chemical change associated with the very high temperature, that you don’t get in a domestic oven.

We cooked a couple more pizzas, and then removed all the charcoal and wood by scraping them out into a metal container partly filled with water. Then we put a small loaf of bread dough in to cook, closed off the door of the oven. It was ready after about 40 minutes. This also turned out to be delicious … with a slight hint of smokiness in the crust.

Charcoal and wood fuel – temperature up to 550F

First pizza cooking – it’s only small

The finished pizza – yum!

First loaf cooked in the Adobe Oven

Making an Adobe Oven

This weekend we set about making an oven for bread and pizzas, using mud. We have a brick barbecue edifice on the patio, and so we used that as a platform.

First we bought and laid 21 red bricks on a 1/2″ layer of sand, making sure to get them even and with small gaps. We could have just used the existing brick oven base, but we wanted something very flat, which it wasn’t. There’s nothing worse than catching your peel on an uneven surface when trying to extract a pizza.

21 bricks laid flat on a bed of sand

Next we marked out on the bricks where the wall of the adobe oven would go: both the inside and the outside edges, using a sharpie. We had bought five 50lbs bags of medium sand, most for the oven wall mix, but some for the oven mould – basically a mound of wet sand like a sandcastle.

The mound needed to be 16″ high, so we first cut a stick to that length and positioned it in a small mound of sand at the centre of the oven bricks.

Then we started to build a sandcastle around the stick, trying to keep the sides of the sandcastle as vertical as possible, and making sure that the sides followed our outline of the inside adobe walls we had marked on the bricks.

Once we had our sandcastle made (in the shape of a dome, reaching the top of the stick), it was time to go digging in the garden for subsoil – the stuff under the topsoil that contains the most clay. We dug about a 3′ deep hole and then extracted a wheelbarrow full of soil.

The clay soil needed to be mixed in one part soil to two parts medium sand. To mix it, we made a ring of four buckets of sand, and put two buckets of the soil in the centre. Then we set about sprinkling it with water and treading it all in to mix it. In fact we found that we had too much sand, and ended up using six buckets of soil to the four buckets of sand.

We checked that the mixing was good when a golf ball sized piece of the mixture, when squeezed tight, and then dropped from shoulder height, didn’t crumble when hitting the ground, but rather just squashed to an oblate spheroid.

At this point we were ready to apply the mud to the sandcastle former. First we placed sheets of damp kitchen paper over the former – this is so that, when we come to remove the sand, after the mud had been applied and set, we could tell when we were at the edge.

16″ measuring stick

Building the sandcastle around the measuring stick

Preparing to mix the clay soil and sand, on a black plastic sheet

The well trodden sand and clay soil mixture

We made an oven door from a piece of old wood. Shaped like an arch, it had to be 12″ at the base and 10″ high. We folded a sheet of paper in half, drew half an arch against the folded edge, cut along the line with scissors, then unfolded the sheet to use as a template on the wood, which we marked around the paper’s edge. Then we cut the wood with a jigsaw.

The door was placed against the front edge of the sandcastle, before we started placing the mud. Then we started packing mud around the sandcastle, being careful not to press inwards on the castle, and trying to maintain a mud thickness of 4″ all the way around. After we had surround the sandcastle with mud, we smoothed out all the bumps, and patted the surface down with a flat paddle made of wood, for an even finish.

Now we need to wait for the mud to dry – after which we will remove the door, and scoop out the sandcastle.

To be continued ….

Packing the mud around the sandcastle. Note the oven door.

Finishing the mud packing

Porsche 914s

It turns out that there are a quite a few Porsche 914 owners at Caltech. A few of us met up today: here are the assembled vehicles -

Quadcopter

I have just finished ordering the parts to build a Quadcopter, similar to the one shown in this clip:

The design I’ll use is based around an Arduino Duemilanove processor, and is called an AeroQuad. Here’s what I ordered, for a total of around $250:

Motors, Speed controller and Propellors

Four TowerPro BM2410-08T / 18A BEC / 9×3.8 Prop Combo

According to the HobbyKing table this motor should run at up to around 8900rpm at full voltage (10V). Plugging the size of the propellor (9 inches with a pitch of 3.8 inches) into this static thrust calculator, the thrust from each motor should be a maximum of 0.6kilos, so a total thrust of 4×0.6 = 2.4 kilos.

Battery

One Turnigy 2650mAh 3S 20C Lipo Pack

At around 11 Amps for each motor, maximum, the total will be 44 Amps. This battery should give a minimum operating time of (60*2.65)/44 = 3 minutes.

Radio Control Transmitter/Receiver

One Hobby King 2.4Ghz 6Ch Tx & Rx (Mode 1)

The “Mode 1″ refers to how the controls are laid out. Apparently in the USA, most pilots use Mode 2. Mode 1 looks like a more logical layout to me, and since I am not a pilot, is what I chose. The Transmitter/Receiver works at 2.4 GHz. One idea I have is to put an X10 RF Spycam I have on board the quadcopter: this also operates at 2.4GHz, but in analog, so there should be no interference. I also have an ICOM IC-R3 handheld I will be able to receive the X10 video stream on:

One Hobby King 2.4Ghz 6Ch Tx USB Cable for Win2000/XP

This interface cable allows the Transmitter to be controlled from a laptop/desktop.

Processor

Arduino Duemilanove with ATmega328

Sensors

One 5DOF add on board (X,Y,Z, Roll and Pitch)

One Dual Axis Gyro

The concept is that the sensors detect the position and orientation of the quadcopter, and the Arduino board runs software that uses this information, together with motion control signals from the Transmitter, to adjust the power going to each of the four motors. This in turn affects the thrust from each of the propellors, which produces the desired (or not!) motion of the quadcopter.

Nash Equilibrium

I was in a discussion the other day where the topic changed to Nash Equilibria, about which I knew nothing. Later investigation led me into Wikipedia vortex in which I discovered not only what a Nash Equilibrium is, but also that if you add a new road to a transport system then it can increase the congestion, rather than alleviate it. Who knew? Apparently this is a real phenomenon, and has been observed in various large cities around the world, and is called Braess Paradox.

Will I ever use my new found understanding of Nash Equilibrium? Quite possibly not, but it’s generally true that time spent investigating stuff like this is well spent. I recall reading about genetic algorithms years ago, and spending quite some time writing my own for fun (looking at the Travelling Salesman problem), never thinking that it would turn out to be useful in the future. Nowadays I use GAs a lot, for all sorts of problems.

Earlier on today, a chap contacted me and asked if his 1913 British Penny was worth anything. He included a picture of a very worn specimen, probably essentially worthless, and I told him so. Just because something is old, doesn’t mean it is valuable

On more pressing matters, Microsoft Security Essentials is asking me to reboot my machine, and I must obey. I love this MSE, so much better and less intrusive than that awful Symantec/Norton thing that I had before, and which seems to snarfle a significant fraction of my system resources.