Last year a group of high school students, in collaboration with a seminar course on Space Systems sponsored by the University of Florida’s Student Science Training Program (SSTP), gained some real-world experience planning, building, and launching a research payload to the edge of space – all in a couple weeks! Last year’s high altitude balloon launch was covered on my website, and the radio transmitter I built for it was featured on this Hack-A-Day post. Unlike last year’s payload, whose only homebrew device was the radio transmitter, this year’s payload had equipment we assembled ourselves, and instead of launching from NASA we launched from the UF football stadium! There were a couple problems along the way, and the payload hasn’t been recovered (yet), but it was a fun project and we all learned a lot along the way!
Below is a panoramic photo right before the launch - see our balloon on the right? So cool!
Our goal was to take photos from the edge of space, and log temperature, pressure, humidity, and GPS coordinates along the way. On-board were a radio transmitter, an Arduino with a GPS shield, and an Android phone to take pictures every few seconds.
Android details: Most of the Android development was handled by UF student Richard along with high school students Benji, Tyler, Michael, and Kevin. Their GitHub project is here:https://github.com/rich90usa/AndroidSensorLogger. Also note that the automatic photo capture utilized Photo Log Lite. We also used GPSLogger to handle logging GPS to SD. “Both of these programs were chosen for their ability to run in the background - and do so reliably by using the ‘correct’ Android supported methods of doing so.” – Richard
Our code used the phone’s text-to-speech engine to speak out an encoded version of every 90th new GPS coordinate. The data was encoded by connecting every number (0-9) to a word the NATO phonetic alphabet. The code also used text-to-speech to have the phone speak out the phone’s altitude data. –Benji
The device consisted of 4 main components: a payload (the styrofoam box in which all of the electrical equipment was housed), a radar reflector (hanging off the bottom of the payload, to help make this object visible to aircraft), a parachute (at the top, made of bio-degradable plastic), and the balloon itself which measured about 6 feet wide when inflated at ground level (supposedly it reaches approximately 30 feet wide at high altitudes before it bursts). Once the balloon bursts, the parachute fills with air and the device floats back to earth.
Kunal demonstrates the effectiveness of our parachute with a scientific “run test”!
The radio communication system we used this year were a little more commercial than last year. Due to my limited time availability (I had an oral surgery rotation all week the week before launch), I chose to get something pre-packaged. My intent was to use FRS (those little 500mW family radio service radios) to send GPS data back to earth, but I later (after launch) did a little more research and realized that it probably wasn’t the most legal way to do it. However, it was extremely cost effective (amateur radio transmitters and RF transmitter modules are quite pricey). For about the cost of a pizza, we were able to interface a FRS radio to the android phone, and the phone ran a program which polled its GPS, turned coordinates into NATO letter abbreviations, and spoke them through the speaker line. The FRS radio with VOX (voice operated transmit) sensed audio and transmitted accordingly. Although it worked very well, I later learned that this may not have been legal in the US because, although FRS doesn’t require a user license and is legal to use anywhere as long as you use its stock antenna, I violated the rule that it cannot be operated above a certain height (20m I think?). Note that this should not be replicated, and probably shouldn’t have been done in the first place. I know I’ll take a lot of heat over this, but it’s in the past now and will be done differently in the future.
Here are some photos right before the launch. It was a sunny day at the UF football stadium! The Android phone is taped the the outside of the box and takes pictures every few seconds, storing them on a micro SD card. Inside the box is an Arduino with GPS shield, and the FRS radio transmitter.
After launch the balloon ascended at a rate of about 500ft/min. It spat out GPS data often, and altitude (not encoded with NATO abbreviations) was the easiest to hear as I walked from the UF football stadium to the UF Gator Amateur Radio Club to use their equipment (namely an AZEL-rotor-controlled 70cm yagi antenna attached to an I-Com 706) to listen in as the balloon ascended… but not before a group photo!
Here we are in the station… let’s get to work!
The results were a bit disappointing, as we believe the Android phone froze/crashed about 10,000 feet in the air! Since that was the device which generated the audio fed into the transmitter, when that phone died, the transmitter stopped transmitting, and we didn’t hear anything else from the transmitter ever again! We included contact information in the payload and it’s possible it will be found one day and we will be contacted about it. If this is the case, we’ll view the SD cards and see the full GPS log and pictures from the edge of space! Until then, we can only cross our fingers and hope for the best. Either way we had a blast, and learned a lot along the way. Next time we can be better prepared for a solid recovery!
Here’s audio of the device’s last words when it was about 10,000 feet in the air: lastwords.mp3
Overall we had an awesome time! I’d like to thank everyone who helped with this project, especially UF students Richard, Kunal, Dante, and all of the SSTP high school students!
Articles typically receive this designation when the technology they describe is no longer relevant, code
provided is later deemed to be of poor quality, or the topics discussed are better presented in future
articles. Articles like this are retained for the sake of preservation, but their content should be
critically assessed.
I was brainstorming some RF circuits today and I had the desire to create a rapid transmitter/receiver pair that anyone would have around their house. I decided that AM or FM radio would be good since everyone can receive that, and pondered how best to generate the necessary radio signal and modulate it appropriately. After a few LC oscillator designs, I thought about the RC oscillators built into most micro-controllers. I grabbed an ATMEL AVR I had on hand (an ATTiny44A) and checked the datasheet. It had an 8MHz RC oscillator, which could be divided-down to 1MHz, and output on a CKOUT pin - all configurable with a few hardware fuses! Note that commercial AM radio stations are between 0.52 and 1.61 MHz, so a 1MHz signal would be smack-dab in the middle of our radio dial! I had to build a prototype to see how well it would work. Once concern was that the RC oscillator wouldn’t be stable enough to produce reliable audio - boy was I wrong!
__The circuitry is textbook simple.__ Appropriately configured, the AVR generates 5V square waves from its CKOUT pin. Although a pretty shape, they're not powerful enough on their own to be heard across a room, so I needed an amplifier stage. A class C amplifier provided by a 2n7000 is commonly done in the low power amateur radio (QRP) community, so I went with it. A 2n7000 N-channel MOSFET with a 220-ohm resistor on the drain and the CKOUT directly into the gate did a fine job (I've used this design for 10MHz QRSS transmitters before), and I was able to modulate its amplitude by feeding the voltage from a MCU pin (turned on/off rapidly) through a decoupling capacitor into the drain of the MOSFET. I couldn't have asked for a simpler result!
This code sends a message in Morse code. It seems too easy! Applications are endless, as this is one heck of an easy way to send audio from a micro-controller to a radio, and possibly to a computer. Morse code is easy, and since we have the ability to dynamically generate different audio frequencies and tones, data exchange is easy too! Nothing’s stopping you from adding the code to turn this into a RTTY (or Hellschreiber?) transmitter.
Again, this transmitter can be heard on a standard AM radio tuned to about 1000 kHz. This is the setup I used with great success:
Here’s the code on the chip! Nothing complicated:
// designed for and tested with ATTiny44A
#include<avr/io.h>#define F_CPU 1000000UL
#include<avr/delay.h>#include<avr/interrupt.h>voidbeep(){
for(char i=50;i;i--){
DDRA|=_BV(PA7);_delay_ms(1);
DDRA&=~_BV(PA7);_delay_ms(1);
}
}
voidrest(){_delay_ms(100);}
voiddot(){beep();rest();}
voiddash(){beep();beep();beep();rest();}
voidspace(){rest();rest();}
voidspace2(){space();space();}
intmain(){
DDRA|=_BV(PA7);
for(;;){
dot();dot();dot();space(); // S
dash();dot();dash();dot();space(); // C
dash();dash();dash();space(); // O
dash();space(); // T
dash();space(); // T
space2();
dot();dash();dot();space(); // R
dash();dash();dash();space(); // O
dash();dot();dash();dot();space(); // C
dash();dot();dash();space(); // K
dot();dot();dot();space(); // S
_delay_ms(1000); // silence
}
return0;
}
THIS IS ILLEGAL to do if you exceed a certain amount of power because you’re stepping on legitimate commercial broadcasters and will have to deal with the FCC. Additionally, you are transmitting on more frequencies than the primary frequency because the signal is heavy in odd harmonics. This means a 1 MHz transmitter, producing square waves, will generate tones on 1, 3, 5, 7 MHz, etc. Don’t do this with much power! Heck, you probably shouldn’t do it at all ;-)
Articles typically receive this designation when the technology they describe is no longer relevant, code
provided is later deemed to be of poor quality, or the topics discussed are better presented in future
articles. Articles like this are retained for the sake of preservation, but their content should be
critically assessed.
This post describes a project I designed which transmits strings of data from a microcontroller to a PC’s screen using audio beeping in a special mode called Hellschreiber. Although these days it’s almost exclusively used by amateur radio operators, I thought it would make a cool microcontroller project! The result can be accomplished with a microcontroller and a speaker as a transmitter and a PC with a microphone as a receiver and decoder, or with actual radio equipment (even toy walkie talkies) by transmitting the tones over modulated radio frequencies for long distance communication! Ideas anyone?
SPECIAL THANKS: I’d like to think Mike Seese for his brainstorming help in making this project a reality. Mike and I are working on a high altitude balloon project together, and a creative inexpensive radio link is one of our goals. Thanks Mike!
As a professional dental student by day and amateur electrical/RF engineer by night, I’m having a very strange summer. I’m developing rapidly in my experience and skills in both arenas. I finally feel like I have a working knowledge of most fundamental electrical and radio frequency concepts, and I’m starting to see patients and do procedures on humans (no more mannequins) in the student dental clinic. For legal and ethical reasons I do not write specifics about what I do with my patients, but I certainly make up for it by documenting the electronic projects I work on! My goals of doing this are to (a) inspire potential electronics tinkerers to come up with new ideas and attack new projects, and (b) receive feedback and insight from those more experienced than me to help me grow in my knowledge. My eye caught a comment a few posts ago that made me smile: You have been blessed with talent and the drive to attempt things not been tried before, keep it up, great job. –David S While I can’t claim that everything I do is truly novel or never tried before, I appreciate the encouraging words. Thank you David S!
Today’s project is a fun one involving vintage wartime radio equipment, amateur radio computer software, and a healthy dose of microcontrollers! My goal is to design a single chip Hellschreiber (technically Feldhellschreiber) transmitter. “Hellschreiber” translates into English as “Light Writer” and is a pun on the name of its inventor, Rudolf Hell, who built the first device in 1920. It was intended to allow messages to be transferred over poor radio links too noisy for intelligible voice or radioteletype (RTTY) communication. Its cool factor is upped by the fact that it was sometimes used by the German military in conjunction with the Enigma encryption system during World War 2! [As an aside, RTTY is still pretty sweet and dates back to the mid 1800s! Check out hardware receivers in video 1 and video 2]
Seeing a battlefield-ready Hellschreiber receiver gives you a good idea of how it works. (The video isn’t mine, I found it on youtube.) The concept is relatively simple (shown above), and the receiver has only 2 moving parts. A spinning corkscrew presses a ticker tape into ink when it receives a radio signal. As the radio signal beeps on and off, the corkscrew contacts at different positions at different times, and letters are written on the ticker tape!
The designers of these things were extraordinarily creative! The picture on the right shows a Hellschreiber transmitter - basically a typewriter with mechanical wizardry that turns key presses into a series of radio tones corresponding to the pixelated shape of a character.
Almost a century later, people are still sending messages around the world using Hellschreiber! With an amateur radio license and an amateur radio transceiver you can tune around special Hellschreiber calling frequencies and engage in conversations with other people who enjoy using this unique mode. Computers have modernized the process, allowing you to send Hellschreiber text by typing on your keyboard and receive it by just looking at your screen. My favorite program (free) to do this is Digital Master 780, part of Ham Radio Deluxe.
This is the project I just completed. It takes strings of text stored (or dynamically generated) in an array on a microcontroller (I’m using an ATMega48, but the code is almost identical for any ATMEL AVR microcontroller, and easy adapted for other architectures) and turns it into an audio tone using PWM. This audio tone could be fed into a speaker and a microphone across the room could receive it and use the software to show the received data, or the audio could be fed into a radio transmitter and a PC hooked to the receiver could decode the audio. Either way, the text in the microcontroller is converted to Hellschreiber audio tones ready to be used however you see fit! Although I designed it as a resilient way to transmit GPS/altitude data from a high altitude balloon using a small, cheap, low-power radio transmitter, this project is just the foundation of a plethora of potential projects!
Here’s the circuit I’m using. It’s actually less complicated than shown - all those yellow wires are going to my AVR programmer! The chip just receives +5V and GND, and the audio is generated automatically and output on the OC0A pin, which happens to be pin 12 on my ATMega48. The output (audio level square waves) is fed to a crystal oscillator like this one, which generates square waves with an amplitude equal that to the input. Thus, by audio-frequency AC from the microchip, decoupled through a series capacitor, added to the power supply of the oscillator (provided by the 5V rail through a 1.8k resistor), we effectively produce an amplitude modulated (AM) radio signal!
This is the receiver I’m using. I’m lucky enough to have an all-mode, general-coverage, 100W amateur radio transceiver! It’s a Yaesu 857-D and I’m completely in love with it. It’s quite pricey though! You can find wide coverage receive-only radios called radio scanners (or police scanners), often for $20 or so on eBay which would do just as good a job of receiving all sorts of radio signals! Whatever you use, after tuning into the audio with the ham radio delux software, you’ll be able to decode Hellschreiber like this:
A few notes about the code: Each letter is sent twice vertically and I don’t think I should have done that. It’s easy enough to correct by eliminating the second FOR loop in the sendChar() function, and doubling the height of the pixels transmitted by changing on(1) and off(1) to on(2) and off(2). Then again, I could be mistaken - I don’t use this mode much. Also, horizontal width of characters (increase this and horizontally compress the received image to reduce the effects of noise) is controlled by a single variable, dynamically adjustable in software. Characters are created from a 3x5 grid (15 bits) and stored as an integer (16 bits, 2 bytes in AVR-GCC). Custom characters are certainly possible! This program takes 16.1% of program space (658 bytes) and 25.4% of data space (130 bytes) and certainly leaves room for optimization.
// designed for and tested with ATMega48
#include<avr/io.h>#define F_CPU 8000000UL
#include<avr/delay.h>#include<avr/interrupt.h>/*
character format (3x5):
KFA
LGB
MHC
NID
OJE
variable format:
2-byte, 16-bit int 0b0ABCDEFGHIJKLMNO
(note that the most significant bit is not used)
*/#define A 0b0111111010011111
#define B 0b0010101010111111
#define C 0b0100011000101110
#define D 0b0011101000111111
#define E 0b0100011010111111
#define F 0b0100001010011111
#define G 0b0100111000101110
#define H 0b0111110010011111
#define I 0b0100011111110001
#define J 0b0111110000100011
#define K 0b0110110010011111
#define L 0b0000010000111111
#define M 0b0111110110011111
#define N 0b0011111000001111
#define O 0b0011101000101110
#define P 0b0010001010011111
#define Q 0b0111011001011110
#define R 0b0010111010011111
#define S 0b0100101010101001
#define T 0b0100001111110000
#define U 0b0111110000111111
#define V 0b0111100000111110
#define W 0b0111110001111111
#define X 0b0110110010011011
#define Y 0b0110000011111000
#define Z 0b0110011010110011
#define n0 0b0111111000111111
#define n1 0b0000011111101001
#define n2 0b0111011010110111
#define n3 0b0111111010110001
#define n4 0b0111110010011100
#define n5 0b0101111010111101
#define n6 0b0101111010111111
#define n7 0b0110001011110000
#define n8 0b0111111010111111
#define n9 0b0111111010111101
#define SP 0b0000000000000000
#define BK 0b0111111111111111
#define SQ 0b0001000111000100
#define PR 0b0000110001100011
#define AR 0b0001000111011111
volatilechar width=1; // width of characters, widen to slow speed
#define spd 8300 // synchronization, incr to make it slant upward
voidrest(char times){while (times){times--;_delay_us(spd);}}
voidon(char restfor){OCR0A=110;rest(restfor);}
voidoff(char restfor){OCR0A=0;rest(restfor);}
voidsendChar(int tosend){
char w;
char bit;
for(w=0;w<width*2;w++){ // left column
off(1);
for (bit=0;bit<5;bit++){
if ((tosend>>bit)&1) {on(1);}
else {off(1);}
}
off(1);
}
for(w=0;w<width*2;w++){ // middle column
off(1);
for (bit=5;bit<10;bit++){
if ((tosend>>bit)&1) {on(1);}
else {off(1);}
}
off(1);
}
for(w=0;w<width*2;w++){ // right column
off(1);
for (bit=10;bit<15;bit++){
if ((tosend>>bit)&1) {on(1);}
else {off(1);}
}
off(1);
}
off(14); // letter space (1 column)
}
// CUSTOMIZE THE MESSAGE, OR GENERATE IT DYNAMICALLY!
int message[]={AR,AR,AR,S,W,H,A,R,D,E,N,PR,C,O,M,SP,R,O,C,K,S,
SP,AR,AR,AR,SP,A,B,C,D,E,F,G,H,I,J,K,L,M,N,O,P,Q,R,S,T,U,
V,W,X,Y,Z,n0,n1,n2,n3,n4,n5,n6,n7,n8,n9,BK,SP};
voidsendMessage(){
char i;
for(i=0;i<sizeof(message)/2;i++){
sendChar(message[i]);
}
}
intmain(){ // ### PROGRAM STARTS HERE ###
// this sets up CPWM in CTC mode,
// it may be slightly different for other chips
DDRD|=255; // OC0A is now an output
TCCR0A=0b01000010; // toggle on match, CTC mode
TCCR0B=0B00000011; // set prescalar
for(;;){
width=1; // fast mode
sendMessage();
width=3; // slow mode
sendMessage();
}
return0;
}
Articles typically receive this designation when the technology they describe is no longer relevant, code
provided is later deemed to be of poor quality, or the topics discussed are better presented in future
articles. Articles like this are retained for the sake of preservation, but their content should be
critically assessed.
This is one part of a multi-post project
Last week I had the crazy idea of sending data from a PC to a microchip through the monitor, using javascript and a web interface as a ridiculously simple data transfer platform that would work on virtually any computer! While I quickly hacked together the hardware, I struggled with the web interface (I’m a little slow with javascript) and I got a lot of help from people around the internet, especially after my project (and need for assistance) was mentioned on Hack-A-Day!
This is part two of a multi-page project. To fully understand what I’m trying to accomplish and why I want to accomplish it, read the first part of the project.
Finally, I have a working javascript! I’d like to thank Tom, Riskable, Ben, and Mike for their input on this script. We got it to a point where we think it’s friendly to the majority of browsers and platforms. The idea is simple - enter two bytes to send the chip, it generates it’s own checksum (an XOR of the two bytes), and it flashes it out. Here’s a photo of the interface, click it for a live demo:
Here’s the code that goes on the microchip:
#include<stdlib.h>#include<avr/io.h>#include<avr/pgmspace.h>#define F_CPU 12000000UL
#include<util/delay.h>#include"lcd.h"#include"lcd.c"volatileint times=1000;
charreadADC(char pin){
ADMUX =0b1100000+pin; // AVCC ref on ADC5
ADCSRA =0b10000111; //ADC Enable, Manual Trigger, Prescaler 128
ADCSRA |= (1<<ADSC); // reset value
while (ADCSRA & ( 1<<ADSC)) {}; // wait for measurement
return ADCH;
}
intmain(void)
{
lcd_init(LCD_DISP_ON);
char lastClock=0;
char thisClock=0;
char thisClock2=0;
char thisData=0;
char buffer[8];
char lastNum=0;
char bitsGotten=0;
int msInactive=0;
/*for(;;){
itoa(readADC(5), buffer, 10);
lcd_gotoxy(0,15);
lcd_puts(buffer);
itoa(readADC(4), buffer, 10);
lcd_gotoxy(8,0);
lcd_puts(buffer);
}*/for(;;){
thisClock =readADC(5);
if (thisClock<250){
_delay_ms(1);
if (readADC(5)>250) {break;}
_delay_ms(1);
if (readADC(4)<250) {thisData=1;}
else {thisData=0;}
lastNum=lastNum*2+thisData; // left shift, add data
itoa(thisData, buffer, 10);
lcd_puts(buffer);
msInactive=0;
bitsGotten++;
if (bitsGotten==8){
lcd_gotoxy(1,1);
lcd_puts("= ");
lcd_gotoxy(2,1);
itoa(lastNum, buffer, 10);
lcd_puts(buffer);
bitsGotten=0;
lastNum=0;
lcd_gotoxy(0,0);
}
while (1) {
if (readADC(5)>250){
_delay_ms(10);
if (readADC(5)>250){break;}
}
}
}
else{
msInactive++;
if (msInactive==400){
bitsGotten=0;
lastNum=0;
lcd_clrscr();
lcd_puts(" TIMEOUT");
_delay_ms(1000);
lcd_clrscr();
lcd_gotoxy(0,0);
lcd_puts("________ =");
lcd_gotoxy(0,0);
}
}
_delay_ms(1);
}
}
Articles typically receive this designation when the technology they describe is no longer relevant, code
provided is later deemed to be of poor quality, or the topics discussed are better presented in future
articles. Articles like this are retained for the sake of preservation, but their content should be
critically assessed.
Several days ago I had a crazy idea. I was driving to Orlando to pick my wife up from the airport and it was dark and stormy on the highway and I was thinking about the backlash I got from my Sound Card Microcontroller/PC Communication project, where I used an embarrassingly simple hardware to accomplish the simple task of exchanging a few bytes of data between a PC and microcontroller (in the face of many people who adamantly prefer more complicated “traditional standard” methods). The car in front of me drove with his emergency flashers on, and at times all I could see were his lights. At that moment the crazy idea popped in my head - I wonder if I could use a PC monitor and phototransistors to send data to a microchip? I can’t think of any immediate uses for this capability, but perhaps if I make a working prototype I’ll stumble upon some. Either way, it sounds like a fun project!
The circuit is as simple as it gets.
A phototransistor is exactly what it says, a photo (light-triggered) transistor (uses small current to trigger a large current). It’s a photodiode with a small transistor circuit built in. Make sure you give it right polarity when you plug it in! For some reason (likely known to electrical engineers, not dental students) the larger metal piece in the plastic part, which I normally associate as negative for LEDs, should be plugged in the +5V for my photodiode. Again, make sure you hook yours up right. I purchased mine from eBay quite cheaply, but I’ll bet you can find some in RadioShack. Note that the value of the 22k resistor is important, and that your needed value may differ from mine. The resistor relates to sensitivity, the larger the value the more sensitive the device is to light. If it’s too sensitive, it will sense light even when aimed at a black portion of the screen.
Initial tests were done using the pins as digital inputs. This was difficult to achieve because, even as transistorized photo-diodes, it took a large difference in light to go from 5V to 0V (even past the 2.5V threshold). After a few minutes of frustration, I decided to use ADC to measure the light intensity. I use only the most significant 8 bits (ADCH). I found that in ambient light the readings are 255, and that white monitor light is around 200. Therefore my threshold is 250 (4.88V?) and I use this for logic decisions. Here’s my setup showing the ADC value of each phototransistor translated into a 1 and 0 for clock (C) and data (D). Both are aimed toward the lamp, so both show a logical 1:
My first test involved reading the data from the image above. The clock is on the bottom line, data is on the top. Every time the clock transitions from black to white, the value of the data at that point is read (white=1, black=0) and the number is placed on a screen. Here’s what it looks like in action:
Hopefully soon we can get a JavaScript interface going! Rather than swiping I’d like to just point this at the screen and let JS flash some squares for my device to read. This will allow virtually unlimited amounts of data to be transferred, albeit slowly, to the micro-controller. Here’s a preliminary sketch of how to send strings.
Remember now we’re using a time domain, not a 2d barcode. I really stink at writing JavaScript, I’m going to have to pull in some help on this one!