Wednesday, September 30, 2009

Fisheye Photo

Here's an image!! The University of Nevada, Reno did a near space launch this week. Onboard their near spacecraft was a camera with a fisheye lens pointed down. Check their Flickr site for this impressive image.

Tuesday, September 29, 2009

Smart Proximity Detector

The PCBs I've designed for the Smart Proximity Detector are int eh mail. I expect them this week. The detector uses a PICAXE-08M to operate IREDs and a 28 kHz detector.

I'm starting to pack kits. I began by creating a spreadsheet with a list of the parts and costs. After that, I'll copy the list of parts into a word document and print that out. I follow what's int he document to pack the bag with parts. I also write the number of items that goes into each bag. That way, after I collect the parts on the list, I do a final count before packing them into their bag. However, before packing each bag, I now affix a label on the outside of the bag with teh name of the kit.

The bag for the prox detector is a sandwich bag, since the PCB is six inches wide.

The Smart Proximity Detector is for any robot that can receive serial data. Look for it under Catalog at the website shortly. I have a video demonstration of a robot using the detector that I am completing. That video along with directions and a copy of the code will be available online for free.

Saturday, September 26, 2009

NearSpace UltraLight

I had a request for the NearSpace UltraLight flight computer. I've started soldering parts together for the board, but haven't completed it yet (or tested).

The flight computer is like a BalloonSat Easy with a built in radio and GPS Port. After building one, you just need to attach the antenna, batteries, and GPS receiver and you're ready to fly. The final kit will include memory (which the version 2.0 BalloonSat Easy doesn't) and a coax cable and wire for making the antenna. It will be my most expensive kit since it includes the $60 radio transmitter. But overall, it should be one of the least expensive ways to ow a flight computer suitable for near space missions.

One day, I hope to enter into the model rocetry field with a line of model satellite payloads.

Friday, September 25, 2009

Tuning a IR Beacon

I've developed a 38 kHz IR beacon for the CheapBot-08. It uses a 556 timer IC to create and combine two square waves. Half the 556 generates a 38 kHz square wave and must be adjusted for the proper frequency (this accounts for the tolerances in the resistors and cap). The second half of the 556 generates a roughly 10 Hz square wave. Combined, the output is a 10 hZ square wave of the 38 kHz beacon. Shutting off the 38 kHz beacon every 0.1 seconds prevents the robot's 38 kHz detector from losing senstitivity to 38 kHz.

I thought I'd just put the 556 on the o'scope in class and tune it up (adjust thwe PCB's pot to get the proper 38 kHz pulse). But I had forgotten that the o'scope would trigger on any signal, whether it was from the 10 hZ of the 38 kHz square wave.

I finally tapped into pin 5 of the 556 to see just the 38 kHz signal. It looks like on the next design, I'll add a test point to the PCB to make tuning the 38 kHz side easier.

Thursday, September 24, 2009

BalloonSat Kits

I think I'd like to start a line of BalloonSat kits. This would be for the airframe and not just the flight computer. A kit would consist of seven sheets of Cellfoam 88, dowels, 1/4-20 nylon bolt, washers, black marker, tubes, plastic sheet, space blanket, and some tape. There would be enough material to make an airframe large enough to hold a video camera, flight computer, and and experiments.

This means I need to get the book together (see my Citizen Scientist articles) for directions and recommendations. The book I plan to publish through Create Space. However, I'll need an ediotor to read over the book first.

Now if there was a way to make a camera kit.

Wednesday, September 23, 2009

CheapBot-08 Radio Slave

Since the PICAXE-08M has such little capability for a robot, I'm incorporating a second into a robot to handle the radio communications. Using a serial port basically converts the two I/O ports required for the radios into one. Now to figure out the two way communication protocol between the nmaster and the slave.

Monday, September 21, 2009

Find Me Spot

Find Me Spot is a GPS tracking system that communicates position reports through a satellite. If you attended GPSL 2008, you'll be familiar with this tracker, as Bill Brown gave a presentation on it. Bill is doing the coolest stuff, you should get to know him.

With all the talk about the MIT flight that used a cell phone for tracking, I'm wondering about writing an article on using Find Me Spot to track. This would allow someone who does not have an amateur radio license to launch near spacecraft. Of course, the tracking unit would have to go up as a back pack with a Tickle Me Elmo to create a Find Me Elmo.

Sunday, September 20, 2009


The smallest, simpliest robot controller I have created is based on the PICAXE-08M. With only five I/O pins, it's a tough one to build a robot controller around. Of those those five I/o pins, only three are truely I/O. One is an input only and the other is only an output.

Problem number one was the H-Bridges. With a code space of only 256 bytes, it was important to incorporate two H-Bridges into the design. That way it only takes two short lines of code to drive or steer the robot and the PICAXE could do other tasks simultaneously. But it takes four I/O pins to operate two H-Bridges. What could a robot controller do with one I/O pin?

My solution was to double-up the function of one of the I/O pins. It controls the drive of two motors and therefore leaves the robot controller with two I/O free pins. So now the robot can walk and chew gum at the same time. However, it can't pivot about its center - it has to turn centered on one wheel.

The turning subroutine has the robot taking turns pivoting on the left and right wheels. I call it psuedo-zero turn radius. Since it takes so little memory to turn the robot, the extra code space required doesn't fully offset the gains in the extra I/O.

Saturday, September 19, 2009

Electronic Supplies

What we need around here is an electronics supplier. I ran out of parts filling an order and had to order them online last night. Wouldn't be nice if enough people did hobby electronics to justify stores (like the old Radio Shack) carrying parts again.

It's up to small businesses like NearSys to specialize in types of electronics that hobbyists might like to purchase.

Do we have a problem in the US that not enough people are doing hobbies?

Friday, September 18, 2009

Smart Proximity Detector Movie

This weekend I plan to begin editing a movie on my Smart Proximity Detector. The detector is programmed to send range information from its left and right side sensors. The robot in the movie (a CheapBot-14 controller on a CheapBot body) is programmed to drive away from the nearest obstacle. It's fun to watch the robot steer away from the plastic wall I swing in front of it.

Thursday, September 17, 2009

GPS Simulator, Part 2

The GPS Simulator only creates GPGGA sentences. That's because the memory in the PICAXE-08M is limited to 256 bytes. And it was tough to cram that much into that limited amount of space. Eventually I'll get around to creating a simulator that produces all the GPS sentences in a way that looks like a near spacecraft on a mission. That's the one I'd like to produce as a kit.

Wednesday, September 16, 2009

GPS Simulator

I've been building new GPS Simulators for a customer in Texas. The simulator uses a PICAXE-08M to send sentences that look like the GPGGA sentences of a balloon in flight. Eventually I'll move this to a larger micro and have it simulator all the sentences of a GPS receiver on a balloon flight. Using the simulator lets you test a flight computer's program on the ground.

A ground test like this lets you find errors without leaving the ground (where it's a lot cheaper to correct).

Tuesday, September 15, 2009

Line Follower and the Creation movie

I've updated my online catalog with a movie about the CheapBot line follower. It's a four inch wide PCB with two pairs of IR LEDs (IREDs) and phototransistors. The IREDs constantly emit IR. If the neighboring phototransistor detects a reflection, then the surface below the pair is white. No reflection detection means there's a black strip below that side of the robot and corrective action is necessary.

The phototransistor and its collector resistor behaves like a voltage divider. In the non-conducting state, conventional current flows from the positive supply, throught the resistor, and to the robot's I/O port. This means a positive five volts is detected by the robot controller (AKA a logic high). When the phototransistor is conducting, current from the collector resistor flows through the phototransistor to ground. This leaves no current (or more accurately, very little current) to flow to the robot controller. As a result, the controller see less than 1.4 volts, or a logic low.

I read earlier today that the movie, Creation, which is about Charles Darwin, can't get an American company to distribute it. Amazing, we live in the 21st century, but some throw-backs are going to make it too contraversal to show a movie about Darwin's most important contribution to science.

Monday, September 14, 2009

Thermal Test Results

After about a 45 minutes test, the interior of the lunch sack was 50* F warmer than the chamber's air temperature (-10 vs -60). And this was with one sheet of foam rubber missing and no power applied to the tracker. I did notice the antenna elements did get stiff. But they are silicon jacketed 14 gauge stranded wire, so it should be just fine.

I need to look up some information on Rocketman parachutes to finish this article for Popular Mechanics online.

Sunday, September 13, 2009

Thermal Testing

The movie 9 wasn't that great, but it's animation was. The story had promise, however.

I built a thermal test chamber years ago that circulates cold air around a test subject with air fans. Dry ice is loaded into a metal in basket and then two smaller in baskets sit on top of the dry ice. The experiment sits on top the stacked baskets allowing cold air to flow around the experiment.

Right now the Popular Mechanics near space tracker is inside. I plan to let it chill for 30 minutes (at least) and compre the internal temperature of the reusable lunch sack to the outside air temperature. The dataloggers for this test are Hobos.

Saturday, September 12, 2009

Popular Mechanics Tracker

Thanks to help from VHS Products, I have a current version of the MicroTraker 300 to use for my Popular Mechanics article. I was using their version 1.3 initially, but having trouble with it.

The Pop Mech article I'm writing explains how to build a near space tracker inside of a reuseable lunch sack. This is similar to designs by NSTAR and TVNSP. The goal is to keep it cheap, but leave room for expansion and flexibility.

This afternoon I'll watch the movie 9. It takes place in a post-human world were humans are destroyed by their technology. Ummm, should I be building robots?

Friday, September 11, 2009

CheapBot Beacon

I've started designing a smart beacon for robots. The beacon will flash one of four signitures in IR based on the setting of a jumper. It uses a PICAXE-08 that will measure the voltage generated by the jumper setting. The jumper portion of the circuit consists of four resistors in series to make a voltage divider. So a jumper setting creates either 1.25, 2.5, 3.75, or 5 volts for the PICAXE -ADC. When the PICAXE powers up, it will measure the ADC input voltage.

Now the PICAXE can't send enough current to all the IR LEDs (IREDs). So it triggers a transistor when then provides the power for the IREDs. A chromed ball above the IREDs (which are pointed inwards and up) will fan the IR out in all directions.

After completing this design, I'll build a PCB for a IR beacon detector that robots can use to locate beacons.

A beacon placed on a robot locates the robot. But a beacon can also be placed on a destination for the robot to find.

Thursday, September 10, 2009

Smart Proximity Detector

A PICAXE-08M operates the proximity detector. It takes turns flashing two IR LEDs (IREDs) at six different frequencies. The detector between the pair of IREDs is most sensitive to 40 kHz flashes. So when the IREDs are flashed at a different frequency, its ability to detect reflections is reduced (it can't see obstacles as far away). So the PICAXE flashes an IRED and checks with the detector to see if it saw the reflection.

The PICAXE cycles through all six frequencies until there are detections on both the left and right IREDs (the only time it runs through all six frequencies is if the IR detector doesn't ever detect a reflection). The PICAXE then sends a serial message with the range to the obstacles as thety appeared to the left and right IREDs. The range of dfistance is from 1 to 6, with a value of 7 indicating there was no reflection or that the obstacles are infinity far away.

The robot reads the data from the proximity detector as it needs to and then acts on the results.

Here's the code as it is currently written.

symbol RightDetect = B0
symbol LeftDetect = B1
symbol distance = B2
symbol counter = B3
symbol RightDistance = B4
symbol LeftDistance = B5
symbol left = 1
symbol right = 4

RightDistance = 7
LeftDistance = 7
for counter = 1 to 6

if RightDistance < 7 then CheckLeft
low right
high left
gosub Flash
RightDetect = pin3 '0 = detect, 1 = no detect
pwmout 2 off
pause 2
if RightDetect = 1 then CheckLeft
RightDistance = counter

if LeftDistance < 7 then FinishCheck
low left
high right
gosub Flash
LeftDetect = pin3 '0 = detect, 1 = no detect
pwmout 2 off
pause 2
if LeftDetect = 1 then FinishCheck
LeftDistance = counter

if LeftDistance < 7 then IsRight7
goto Repeat_Flash

if RightDistance < 7 then Report


serout 0,T1200_4,(255,"L",LeftDistance,"R",RightDistance)
goto Proximity_Detect

if counter > 1 then Check45
gosub kHz46
goto End_Flash

if counter > 2 then Check44
gosub kHz45
goto End_Flash

if counter > 3 then Check43
gosub kHz44
goto End_Flash

if counter > 4 then Check42
gosub kHz43
goto End_Flash

if counter > 5 then Check41
gosub kHz42
goto End_Flash

if counter > 6 then Check40
gosub kHz41
goto End_Flash

gosub kHz40
End_Flash: return

pwmout 2,24,50 ' 14 inches

pwmout 2,23,49 ' 10 inches

pwmout 2,23,48 ' 9 inches

pwmout 2,22,47 ' 6 inches

pwmout 2,22,45 ' 5 inches

pwmout 2,21,44 ' 3 inches

pwmout 2,21,43 ' 2 inches

Wednesday, September 9, 2009

Proximity Detector

I've developed a smart proximity detector for robotics. A PICAXE-08M flashes two IR LEDs at 40 kHz (one at a time). The 40 kHz IR detector between the IREDs determines if there is a reflection from the IREDs. The PICAXE monitors the output from the detector. Since it knows when it flashed each of the IREDs, it knows which of the IREDs caused the reflection. The IREDs are mounted on the extreme left and right sides of thePCB giving the smart proximity detector the ability to determine if a wall is located on the left side, right side, or across the front of the detector.

Eventually I'll flash the IREDs at different frequencies to determine if the wall is located farther away or closer.

but first, I have to set my wait period between left and right flashes. I don't know how long the detector "remembers" that it saw a reflection.

Monday, September 7, 2009

Easy Star Gazing

One of my hobbies is astronomy. Currently I'm giving introductary astronomy lessons quarterly at two public libraries. My latest presentation is available online at There you will find a Power Point and supporting handouts.

Saturday, September 5, 2009

BalloonSat Launch

I have a near space launch scheduled for September 26th. This will take place in Salina, KS with the help of Pete Sias, the guy who got me started in near space. The launch will carry five BalloonSats my CT442 class is assemblying. I also hope to carry the lunch sack tracking module I'm developing for my Pop Mechanics article.

Friday, September 4, 2009

Popular Mechanics Article

The PM online artcile I'm writing uses a MicroTrak 300 and Garmin GPS-18X. The container for the electronics is a insulated lunch sack. The lunch sack is an Artic Zone bag that should provide some insulation and cushioning.

The batteries recommended are photo lithiums. I've used them before in near space.

Thursday, September 3, 2009

LabPro UVA and UVB

Two of the LabPro sensors I'm nmost interested in seeing operate are the UV sensors. They're mounted inside a thin tube, so I suspect they're direction sensitive. I'll recommend my students mount them inside a ping pong ball (acts as a diffuser) and point them up through the top of the BalloonSat.

Wednesday, September 2, 2009

LabPro Dataloggers

The latest datalogger I'm playing with is the LabPro by Vernier ( The logger/programmer is smart, it detects the sensors plugged into it. After collecting data, the programmer then processes the data appropriately for the sensor and displays the results in a chart. You can export your data into a text file for processing in a spreadsheet. That's great for incorporating GPS data. I'll post my notes on using the datalogger on my website alogn with some pictures on making a LabPro BalloonSat.

Vernier is popular in science classrooms.

Tuesday, September 1, 2009

Calculating Distance to the Horizon

You can calculate the distance to the horizon with a simple equation. The real equation involves trig, but this one gives a nearly identical answer.

Dist(in miles) = sqrt[height(in feet) X 1.5]

Multiple the altitude in feet by 1.5. Then take the square root of the answer. The distance to the horizon is in miles.

Since my eyes are 6 feet above the ground, I multiply 6 by 1.5 to get 9. The square root of 9 is 3. So the distance to the horizon for my eyes is three miles.

A near sapcecraft at 90,000 feet sees 367 miles to the horizon. Because of refraction, radio should reach a little bit further.