.jpg "TEAM: Probot")
Levitating Potato / Texts
Order by: date ↑ - rating ↑
Flagged! I did it!
Don't be scared to beg ;)
If you need money to pay for the server, say so. If you think it would be better to cut some features and spend less, the do so, but don't assume there's no money available. Just put up a donation drive request, saying "we need $X to keep the site running, we have $Y, please donate!" or some such.
Also, it's at least worth considering asking for some amount of salary for your work on the site. If a little bit of cash made the difference between you having time to work on the site or not, I suspect you might find people are willing to donate.
You've built a community here. It needs you. It's ok for you to need it, too.
For my summer vacation, I decided to turn the Doktor into a cyborg. Well, not completely, but to at least start down that road. And what better way to begin than by giving the Dok a new sense? Any decent depiction of cyborgs will show they clearly have more than the usual human array of senses. So, since Dok's sense of direction is sorely lacking, we decided to give him a sense of North.
Our methodology was a modified version of this belt:
http://infosthetics.com/archives/2005/10/feelspace_tactile_compass_visualization.html
Ours was simpler and cheaper. Rather than the digital approach of having the single northmost motor running, we took an analog approach with four motors. One motor was placed in the front, back, left, and right sides of the belt. Each motor vibrated at full strength when pointed north, not at all when pointed south, and half strength when east or west (and varied smoothly in between, of course). The result is a sensation not unlike a single vibrating object that is always positioned at the northmost side of the belt. The four-motor approach was chosen for cost and simplicity -- no microprocessor is involved, and 4 motors are a lot cheaper than 16.
I'm going to have to agree on that one. Though a couple of my potential candidates seem to have retired. Hmm, I suppose I just have a thing for older players.
Either that or they really want to play spin the bottle.
BARTPA needs more tasks like this.
The while loop's a time-honored verse;
And goto's the programmer's curse.
Today for loops are fun,
if a bit overdone:
For this task I shall tail recurse.
Spartacus is one of my sock puppets!
Mostly because you didn't have a two yet.
You have to collect them all before you can win, you know.
Building a circuit in the middle of the North Woods was an interesting prospect, but I was well prepared with soldering iron, breadboards, and a reasonable selection of components. You can pack a surprisingly complete electronics shop into a fairly small box and stuff it into a duffel.
Early testing was as a hat, but we ended up with the belt design. Nonplayer wearing the hat:
It turns out you really can't adjust a sensitive electronic compass with a metal screwdriver -- even nonmagnetized ones have enough magnetization to cause trouble. So I grabbed a piece of scrap wood from the kindling box and whittled an adjustment screwdriver that I knew was nonmagnetic. Similarly, you can't have the batteries too close to the circuit board -- they're mildly magnetic. So the batteries go in your pocket, with a cord to the belt.
Adjusting the circuit, wood screwdriver in hand:
Along the way, the circuit got a bath:
Solder flux is ever so slightly conductive, and using a generic board rather than one designed for the circuit means there's a lot of it around. Rather than chase strange problems caused by conducting solder flux, I preemptively bathed the circuit.
The circuit design for the belt was an interesting project in its own right. It wasn't terribly hard, but it also wasn't of the sort I was used to. This was a non-precision design, intended to run at somewhat low power from a low voltage supply.
The final project ended up being more expensive than I'd hoped. It was about $30 worth of electronics, including various screw terminals and mounting boards that I could have made do without but vastly simplified things. The motors were another $14, plus there was shippnig on both the electronics and motors. The rechargeable batteries and charger for them added another $20. Since a set of AA batteries lasts about a day, rechargeables are a must for long-term use. Total cost including shipping and mounting hardware was under $100, but not by a lot.
(For those interested, mildly technical details begin here.)
For ease of development, I wanted to avoid having a microprocessor involved -- and was willing to pay for it by having the resulting circuit be somewhat finnicky. The form of the output (four motors, each vibrating with a strength proportional to the cosine of their angle from North) was not chosen by accident. The sensor I used is actually a pair of sensors at 90 degrees to each other, each providing the strength of magnetic field along a well-defined axis. So two of the motors vibrate proportional to sensor output, and two vibrate opposite that.
The circuit starts to get finnicky with the sensors -- they're rather bad, in many ways. The strength of the output is poorly defined (they're not even matched to each other), and they don't read zero when there's no applied field. So each channel needs two adjustment knobs -- one for zero adjust, one for gain.
The degaussing the Doktor mentioned is a result of the sensor construction and the overly simplistic circuit design. For simplicity, I didn't make use of all the features available in the sensor. The chip includes the ability to reverse the polarity of the sensors, and thus correct for annoying things like zero point errors. Exposure to external magnetic fields -- from large motors, for instance -- will also cause semi-permanent changes in the sensitivity of the sensor. The changes last until a strong applied field resets the sensor internals. The degauss switch provides that ability, and it's very closely analogous to degaussing an older CRT monitor.
The motors also need a midpoint adjust -- when the sensor is pointed North, its output is positive, and when pointed South, negative. But we want the motor to output full strength at North, nothing at South, and half strength at no applied field (East / West). So another adjustment knob is needed to set the location of that midpoint away from zero, bringing the total number of adjustments on the belt to 5.
The circuit itself is fairly simple, after that. One op amp is used to generate a reference voltage that doesn't change as the batteries wear down (we wouldn't want the motors to get weaker over the course of the day). A second derives the motor midpoint voltage from that. The reference voltage powers the sensor elements. Each channel then involves three more op amps. The first is a differential configuration that brings the sensor output from the mV level to something reasonable (the sensor zero adjust is incorporated here). The second is an inverting amplifier, which provides a convenient location for the adjustable gain. (It inverts the amplified sensor output about the circuit "zero" voltage, which is the same as the motor midpoint voltage.) This second op amp's output is also the drive voltage for the first motor. The third op amp inverts this output to provide the drive for the other motor. Since the motors are high-current devices, a transistor is added to boost the output capability of the op amps. The second channel is simply a copy of the first.
I don't currently have the schematic or parts list in electronic form, but it wouldn't be difficult if anyone's curious. This is a project that is definitely feasible for anyone who's comfortable with a soldering iron, and it's easily debugged and tuned with only a cheap multimeter.