Thursday, May 20, 2010

On Google Ads and other "ad revenue streams"

Updated Sep 7, 2010
Today I received an email from Makersmarket. They ceased operation. Their business model didn't work. I guess that the high monthly fees they charged made other sellers to pull out as well.

Updated July 9, 2010
The book promised by makersmarket never showed up. (I also closed my "store" on makersmarket because the fees were really eating the little revenue I made there.) Therefore, I took out their logo as well (shown below for "commemorative" purposes).

This has nothing to do with Arduino, but I thought it is still a piece of information (that others may find useful) and not just noise.
Out of curiosity, I signed up with Google Ads (aka AdSense) when I first developed this blog. I could have filled the screen with "AdSense gadgets" as they are called, by I chose the most modest of the advertisement modes, the one that displays a few line of ads between postings (plus one at the bottom of the page).

A year later, I made US$30. Well, I did not really cash it in, since Google sends out checks of a minimum US$100. So I will have to wait patiently (and write diligently) about 2 more years to see any money from Google ads.
Also worth mentioning is the number of hits ("page impressions") for the last year: 37K.

A little rough estimate of potential revenue from Google ads (if they would display Google ads) is somewhere around US$3,000 a year, which would be 100 times more than this blog (I remember them mentioning 1M pages served; that would be roughly 30 times more, which looks hugely underestimated). Again, $3,000 if the ads would be lightly displayed. Not bad, but not great either, considering they have sales in excess of US$1M a year.

On the same line (ads), also out of curiosity, I signed up with Amazon, although I did not start using their service yet. Expect some postings on books I read (or plan to) in the future, when I will find the time.

Thirdly, makersmarket, where I listed assembled wise clocks, offers a book (a value of $25) to the first 50 fellows who display their logo. Which is exactly what I did yesterday. This would be a much better ("efficient") deal than Google Ads, obviously.

And lastly, I got hooked up, also out of curiosity, with yet another "ad revenue stream", a business model developed by a friend who is trying to take over Google (I am kidding, evidently). The result is the "Rent this space" box on the right column.

Now you have all the explanations and excuses for all the ads you see on this blog. As they say, "don't blame the player, blame the game". Seriously, it is just a way to get a little understanding of what's going on.
I cannot resist to close with another saying: "if you are not outraged, you are not paying attention".

Thursday, May 6, 2010

New version of Wiseduino

The new Wiseduino board offers several improvements (I would like to think) over the old one:
  • prototyping area, with pads for an 8-pin SOIC chip; this part can be sawed off along the fine holes, to match the dimensions of a small shield (and of the previous version); the intention was for this piece to be broken off by hand, but the PCB is too thick to make this possible;
  • two mounting holes for screws/standoffs;
  • power (5V) provided through the FTDI connector as well;
  • through-hole power switch.

The prototyping area can be used to add buttons, buzzer, LEDs, or even a voltage regulator.

Eagle files are available: schematic hereboard here.

Related posts:

Saturday, May 1, 2010

Prototype of an analog clock with LEDs

Inspired by Makerbot Watch, I designed my own version of an analog clock with LEDs.

The prototype uses a Wiseduino with DS1337 and a protoshield with 24 SMD LEDs.
Physically, on the board, the LEDs are placed in 2 "rings" of 12, the outside ring designated for minutes (red LEDs), the inside one for hours (green LEDs).

In the circuit, the LEDs are organized as a matrix of 5x5, as shown below.

The challenge is to display the time in an intuitive manner. For example, an analog clock moves its hour hand proportionally with the minutes. This is hard to achieve with relatively few stationary LEDs. Tricks need to be "invented" and even those may not be enough for an "accurate" display. Watch this video and let me know if what's shown is intelligible as a clock reading.

As seen in the video, the time is shown after the button is pressed. There are two reasons for that:
  • since LEDs are the biggest consumers on this board, try to light them up only when necessary;
  • for counting the flashing minutes there should be a start moment, and this is what the button provides.

Since this project will eventually evolve into a wrist-watch, a very important aspect is the power consumption.
Here are a few numbers so far (idle current/current with LEDs on):
  1. normal loop, with no sleep: 4.7mA/12.2mA
  2. "unoptimized" sleep mode most of the time: 0.13mA/12.2mA
  3. "optimized" sleep mode: (to be completed)

To be practical as a watch, it will need a way to set up the time, which may require a second button (plus software, of course, to navigate the hours and the minutes).

Optionally, the watch may also have an alarm. This feature would require a small buzzer (plus, again, software to set up the alarm time and to activate the sound). The board size will be bigger and current consumption will be greater.

Somewhat similar projects on the interwebs (both featuring Microchip microcontrollers):