Dave Richards AA7EE

November 18, 2015

The Sproutie “SPT” Beacon – A Legal, Unlicensed HiFER Beacon

Note – this blog-post discusses the use of the 13553 – 13567KHz band under FCC Part 15 regulations in the US. Although it is a worldwide allocation, rules vary according to where you are. Off the top of my head, I do know that there are HiFER beacons operating in some countries on the European continent, but that is the extent of my knowledge of this type of operation outside the US.

Before saying anything else, I must note that even though the earlier projects which were named after my cats were not my designs, I did at least contribute enough of my own input that I could perhaps get away with naming them. I’m not sure that is the case with this venture, as I simply re-purposed it for a slightly different band and usage. However, the urge to name things around here after my cats is strong, so what I am calling The Sproutie Beacon, is really an original Hans Summers QRSS TX, modified slightly for the 15553 – 13567KHz HiFER band.

I have long been fascinated with clandestine and pirate radio stations. The UK has a long and hallowed history of pirate operation, since Radio Caroline and the other pirate ships began to grace the airwaves in the 1960’s. When I was in my teens in the 70’s, Caroline was still on the air, as was a newcomer to the pirate ship scene, a station called Laser 558. Laser 558 was, like the other pirate ships before it, stationed just outside British territorial waters, in international waters. It differed from the other pirates in one very noticeable detail though – it had American DJ’s, and was programmed like a US Top 40 station. To a British listener who was used to DJ’s talking quite a lot, the sound of American accents and near-continuous music, as dull as it might sound to a Stateside listener, was quite thrilling to these teenage British ears in the 1980’s. London is well-known for it’s many land-based pirate broadcasting stations on the FM band, but there weren’t too many outside the big cities. As a teen in the 80’s growing up in the Midlands, we did have a fairly high-powered pirate station on the AM MW broadcast band, with a wide coverage area, called Sunshine Radio, which I enjoyed listening to greatly.

When living in Los Angeles in the 1990’s, I was asked to DJ on a local FM pirate by one of the resident presenters, but politely declined, as I was already working in my chosen career, doing DJ, voice-over and production work. I was getting my DJ jollies for about 50 hours a week – and getting paid for it at that point, so said no to an opportunity that a few years earlier, I would most likely have jumped at. Los Angeles was not known for pirate activity at all – the area was almost entirely devoid of it, but this one station was a notable exception. It was known as KBLT. The founder, Sue Carpenter, even wrote a book about it, called “40 Watts From Nowhere”. Written from her perspective, and relating the trials and tribulations of running a pirate radio station that was on the air nearly 24/7 out of her apartment in the Silverlake district of Los Angeles, it’s a good read for anyone interested in the subject of pirate broadcast stations.

Then, in 2008, after moving to San Francisco, I was tuning around the shortwave bands in CW mode from my apartment in Ocean Beach one day, and heard a series of dits on approximately 4096KHz, Further investigation revealed that it was one of a cluster of unlicensed (and not legal) beacons operating from various locations in the California deserts on various frequencies centered around ~4096/4077KHz and 6626KHz with powers of the order of a few 100mW’s. All of them operated from solar power. Some also had batteries and could transmit 24/7, while others had only solar panels and were daytime only beacons. Even Jason NT7S could hear one of them from his QTH in Portland, Oregon – propagation was good on a regular basis back then. There are a number of these beacons, most of them in the deserts of the south west. Some send dits at various speeds, some send letters in Morse code. There is also one that sends the ambient temperature in Morse. They are discussed, with reception reports, over on the HF Beacons forum at HF Underground.

If, from all of this, you conclude that I would still be interested in running some kind of pirate operation, you’d be partially correct. I say partially because, in truth, although I enjoy listening to the clandestine activities of others, I wouldn’t want to do anything that might, even in theory if not in practice, jeopardize my ham license. I’d love to take a QRP solar-powered HF beacon out into the desert and leave it there, sending it’s valiant little signal, day after day, year after year, and checking the online reception reports from time to time. It would be interesting to see how long it would last. I read a report from someone who did install such a beacon, and his description was quite lyrical. He described how, whenever he was out hiking, fishing, or otherwise enjoying the great outdoors, he would take his portable shortwave radio and listen out for his beacon, thinking of the little transmitter out in the remote desert, courageously sending it’s diminutive signal across the great expanses of wilderness. Very evocative stuff.

It turns out that there is a way to operate an unmanned beacon on the HF bands below 28MHz, and to do so legally. The details, in the US, are contained within the FCC Part 15 regulations. These are the regulations which set out the requirements for unlicensed transmitters, among them baby monitors, cordless phones, toy walkie talkies, garage door openers, WiFi and Bluetooth devices, to name a few. In much of the spectrum in which operation is allowed, the power limits are very low, though there are a few bands where the allowance is more generous. The band with the most easily-attainable DX potential is the 14KHz-wide ISM band centered around 13.56MHz. Power limits are specified not in terms of the device output power, but as a maximum field strength at 30 meters. Medical diathermy machines operate in this band hence, I presume, the reason for a field strength stipulation rather than actual power into an antenna. This band is also inhabited by RFID devices. If you listen, you may well hear a variety of odd beeps and carriers, particularly near the center frequency of 13.56MHz. The maximum field strength allowed under FCC Part 15 regulations is 15,848 microvolts/meter at 30 meters. Few among us have access to accurate field strength meters, but John W1TAG has written this very informative paper, in which he runs the calculations, and comes to the conclusion that 2.3mW into a ground plane, or 4.6mW into a dipole would produce the maximum allowed field strength. Now 4.6mW isn’t a whole lot of power, but the WSPR and QRSS folk will tell you that DX results can be had, even within those limitations. In fact, the beacon activity on this band is divided between folk who run beacons sending CW at speeds that can be read by ear, and QRSS transmissions. A few people do run grabbers on this band, and report the results. Beacon activity in this ISM band is a very niche pursuit, but there is a good discussion forum over at the Longwave Message Board. As the title suggests, this site was set up for LF operators, but there is HiFER discussion there too,

My first “proof of concept” at putting together a beacon for this band was to connect an N0XAS PicoKeyer in beacon mode to my Pixie 2 transmitter. With the PiicoKeyer, if you insert the prosign BN at the end of stored message #1, it will automatically repeat. Unfortunately, my older version of the PicoKeyer will not power up again in beacon mode if power is lost. This was taken care of in later versions, but it meant that I wouldn’t be using this particular version of the PicoKeyer in the final version of the beacon. My Pixie 2 put out about 170mW on it’s original frequency of 7030KHz when powered by 12V, but this dropped to an encouraging 5-10mW at 3.6V (3 x 1.2V NiMH cells in series).

For the final version, I wanted all the electronics to be on one single board. At that point, I was thinking about purchasing a PicoKeyer chip from N0XAS and building the keyer, Manhattan-style onto the same board as a Manhattan-built transmitter. Then I remembered the Hans Summers QRSS Transmitter that I had built a few years ago. After a brief flirtation and a lot of fun with QRSS on 30M, the board sat languishing in a box. A little experimentation showed that it would still work down to voltages below 5V, and with the drive trimpot, I figured I’d be able to adjust the drive to give an appropriately low output power. Even better was the fact that on the QRP Labs website, there were details of a mod by Aldo IW2DZX for altering the output from FSK to straight on-off modulation, which I happened to want with this beacon.

Getting the QRP Labs QRSS TX on the HiFER band was straightforward. A pack of 5 x 13.56MHz crystals was purchased from an eBay seller. I chose HC49/U crystals over the more popularly available HC49/S, as I have read that the former tend to pull over a wider frequency range, due to the crystal cut. Receiving this pack of crystals in the mail was exciting. Think of the possibilities!

Here’s the final schematic for Hans’ little beacon transmitter, modified for straight on-off keying, and with values appropriate for the 13553 – 13567KHz HiFER band –



The jumpers on pins 5, 6 and 7 of the ATtiny13 are used for programming the sending speed. Refer to the original kit instructions for programming the speeds. Hans’ firmware allows 6wpm, 12wpm, and 6 QRSS speeds ranging between QRSS1 (1 second dits) and QRSS20 (20 second dits). The original schematic didn’t include the 3 x 10K pull-down resistors on these pins. I included them, as my TX wasn’t transmitting the selected modes. If these pins are left without pull-down resistors, then an unconnected pin might be incorrectly interpreted as “high” by the chip. The original circuit used a reverse-biased red LED to provide frequency-shift keying. This was removed, along with a 470K resistor and a “gimmick” capacitor, and a 2N3904 transistor, 22K resistor, and 0.1uF capacitor added to key the PA transistor. Also changed were the values of inductance and capacitance in the output low-pass filter. To cap off the mods, a 3.3V regulator was added. Because of the strict power limitations on this band, I wanted to ensure that the TX was running close to the maximum allowed power at all times, with minimal variation due to power supply fluctuations.

Although I modified my existing QRP Labs original QRSS transmitter, if you don’t have one to modify, you could build it from the schematic above, Ugly-style or Manhattan-style. A little transmitter built using MeSQUARES and MePADS would look quite nifty, methinks. The values in the schematic are the final values for the HiFER band. If you decide (with the help of the info on Hans’ site) to build it for QRSS operation as an MEPT on a ham band, you can run it from 5 – 6V for increased TX power output. I believe it can put out up to 150mW. On the HiFER band, of course, we don’t want anywhere near that much power, so a 3.3V regulator does the trick nicely.

Here is a top view of the modified board, with both new inductor and capacitor values for the HiFER band, and the IW2DZX mod for straight on-off keying completed. The speed selection jumper holes to the right of the ATtiny chip have been drilled out to accept a header block. With the original TX, you had to solder a wire between 2 holes to select a given speed. Now, the speed selection is accomplished by plugging in a jumper block (or a combination of jumper blocks). You can also see the 3.3V regulator at the far left edge of the board, in the middle. 3 old parts to the left of the trim cap have been removed, and 2 new ones added. I’ll leave you to figure out what they are :-)  –


The underside of the board, showing the extra transistor (a 2N3904) for the straight on-off keying mod. You can also see the 3 x 10K pull-down resistors, as well as a 1pF NPO capacitor added across the trimcap to tweak the frequency coverage.  The new values of capacitors required in the output low-pass filter were larger than I had on hand, so I made them up by placing smaller values in parallel. You can see 2 of those parts in the photo, placed on the underside of the board, in parallel with capacitors on the topside. I cleaned the board with flux cleaner, but ended up with a white residue. Not sure what it is. It bugs me, but I decided to let it go –


A few more views of the board –






If at this point, you don’t have a functioning keyer chip, you can verify that the transmitter is working by connecting pin 3 of the DIP socket to pin 8 (the +3.3V supply), which will activate the keying transistor and turn the PA on. You can listen to the little transmitter on a nearby receiver, or look at the output on an oscilloscope (or both).

Now to decide on a callsign, or other beacon ID that you want to send. There are no ID requirements when operating under Part 15 regulations. Indeed, this band isn’t even intended for these types of communications, though this usage does fall within the rules. This leaves you, the fledgling HiFER control operator, free to transmit any ID you want. I decided that I wanted the letters “SPT” in honor of my youngest kitty, Sprout. But how to go about changing the firmware? This was no mean feat for a person like myself, who has studiously avoided all types of non-analog electronics my entire life.

This next part of the narrative will be blindingly simple for many, but is directed at people like myself, who are fairly new to the task of compiling and flashing code to a micro-controller. You have to search around a bit to find this fairly basic information in a form that we newbies can understand, so I thought I’d attempt to provide a tutorial of sorts here. For those who know what they’re doing in this area, please feel free to add comments and correct me.

The code, written in C, is posted on Hans’ personal site here. The general page on the keyer chip on his site is here. By the way, if you haven’t seen Hans G0UPL’s personal site, you’re in for a treat. It’s a treasure trove of personal projects and just screams “home-brewer/experimenter”. There are many happy evenings of reading on it!

Somewhere online, (in a Yahoo group, I think), I read a message from Yan XV4Y to the effect that he hadn’t been able to compile Hans’ code, and had made a slight modification to it to correct that, as well as making an addition, to allow spaces to be included in the sent message. Later, Hans told me that he had been using Atmel Studio 4 on an older version of Windows, and that it was possible that some folk might have had trouble compiling it with newer programs. He also said that he wasn’t sure whether the code on his site had been updated or not.  I sent an e-mail to Yan, asking if there was any chance of him sharing his code with me. He responded very quickly in the affirmative, and also said it was fine for me to share it here. Please note that Hans’ code, as posted on his site, may very well compile and work fine. It’s just that I used Yan’s version. Yan was quick to point out that this is really Hans’ code with just a few minor mods from him. Many thanks to Hans for allowing me to post it here, and to Yan for allowing me to post his slightly modified version. For the slightly clueless people like me, the instructions at the beginning tell you what to do in order to insert your own custom callsign. For the record, Yan said that this code compiles perfectly on Mac OS X with Xcode using Crosspack-AVR. I’m a Windows person, so I’ll relate it the way I did it. In the following piece of beacon code, I set the callsign to be GRC. If you’re a fan of grilled cheese sandwiches and bacon, the idea of a “grilled cheese beacon” might be appealing, hence the callsign GRC. I know it’s a bit corny, but it will suffice for this example –

// This is an amendment to the beacon program written by Hans Summers
// G0UPL. It adds the facility for a space character to be embedded 
// within the transmitted callsign character string

// To change the callsign string you have to :-

// alter #define MSGMAX to the length of the callsign string +1

// insert the callsign you want, ensuring each character is seperated by a space
// into the text between the curly brackets after int8_t msg[MSGMAX] ending the string with a _SPC

//  e.g. for AB4CDE 

// #Define MSGMAX 7

// int8_t msg[MSGMAX] = { A B _4 C D E _SPC };

// e.g. for PA5D M M 

// #Define MSGMAX 9

// int8_t msg[MSGMAX] = { P A _5 D _SPACE M _SPACE M _SPC };

#include <avr/io.h>
#include <avr/interrupt.h>

volatile uint8_t msgIndex;
volatile uint8_t timerCounter;
volatile uint8_t counter2;
volatile uint8_t audio;
volatile uint8_t key;
volatile uint8_t bit;
volatile uint8_t pause;
volatile uint8_t character;
volatile uint8_t speed;
volatile uint16_t callsign;
volatile uint8_t keyDelay;

#define PERIOD 6

#define A		0b11111001,
#define B		0b11101000,
#define C		0b11101010,
#define D		0b11110100,
#define E		0b11111100,
#define F		0b11100010,
#define G		0b11110110,
#define H		0b11100000,
#define I		0b11111000,
#define J		0b11100111,
#define K		0b11110101,
#define L		0b11100100,
#define M		0b11111011,
#define N		0b11111010,
#define O		0b11110111,
#define P		0b11100110,
#define Q		0b11101101,
#define R		0b11110010,
#define S		0b11110000,
#define T		0b11111101,
#define	U		0b11110001,
#define	V		0b11100001,
#define	W		0b11110011,
#define	X		0b11101001,
#define	Y		0b11101011,
#define	Z		0b11101100,
#define _SPACE  0b11101111
#define _SPC	0b11101111
#define _0		0b11011111,
#define _1		0b11001111,
#define _2		0b11000111,
#define _3		0b11000011,
#define _4		0b11000001,
#define _5		0b11000000,
#define _6		0b11010000,
#define _7		0b11011000,
#define _8		0b11011100,
#define _9		0b11011110,
#define _BRK	0b11010010,
#define _KEYUP	0b10000000
#define _KEYDN	0b10100000

#define MSGMAX 4
#define SHORTSTART 0
int8_t msg[MSGMAX] = { G R C _SPC };

uint8_t speeds[8] = {1, 2, 10, 30, 60, 100, 150, 200};
uint8_t dit[8] = {150, 150, 150, 150, 150, 150, 150, 150};
//uint8_t speeds[8] = {1, 1, 1, 10, 30, 60, 100, 200};
//uint8_t dit[8] = {150, 36, 30, 150, 150, 150, 150, 150};
// 150 		1		12wpm
// 150		2		6wpm
// 150		10		QRSS1
// 150		30		QRSS3
// 150		60		QRSS6
// 150		100		QRSS10
// 150		150		QRSS15
// 150		200		QRSS20
// 36		1		50wpm
// 30		1		60wpm

int main(void)
	DDRB = 24;
	TCCR0B |= (1<<CS01) | (1<<CS00);	// Prescale by 8
	TIMSK0 |= (1<<TOIE0);
	msgIndex = 0xff;
	return 0;

	if (audio == 1)
		if (key) PORTB |= 0x08;
		PORTB &= ~(0x08);
		audio = 0;
	// 1500Hz here
	if (timerCounter == dit[speed])
		// 10Hz here
		timerCounter = 0;
		if (keyDelay)
			if (counter2 >= speeds[speed])
				counter2 = 0;
				if ((character == _KEYDN) || (character == _KEYUP)) 
					key = 0xff;
					bit = 0;
					if (!pause)
						if ((!key) && (!bit)) pause = 2;
				if (key == 0xff)
					if (!bit)
						if (msgIndex == MSGMAX) 
							msgIndex = SHORTSTART;
							if (callsign > 6000)
								msgIndex = 0;
								callsign = 0;
								speed = 0;
								msgIndex = SHORTSTART;
								speed = (PINB & 0x07);
						bit = 7;
						// Get character from message
						character = msg[msgIndex];
						// Look for 0 signifying start of coding bits
						while (character & (1<<bit))
					if (character == _SPC)
						key = 0;
					else if (character == _KEYDN) 
						key = 1;
					else if (character == _KEYUP)
						key = 0;
						key = character & (1<<bit);
						if (key) 
							key = 3;
							key = 1;
					if ((character == _KEYDN) || (character == _KEYUP)) keyDelay = 100;
				if (key)
					PORTB |= (0x10);
					PORTB &= ~(0x10);
	TCNT0 = 156;

Copy and paste this code directly from here into a simple text editor, such as Notepad, if you’re using Windows. You can include the instructions at the beginning if you want – the compiler will know to ignore them. In the text editor, you can alter the code to include the callsign/message of your choice (no more than 8 characters, including spaces), then save it. You can name the file whatever you want, but make sure that the file extension is .c so that the compiler knows what it is.

Before compiling and flashing this code onto the ATtiny13 micro-controller, the other thing you will need to know is how to set the fuses on it. This beacon circuit uses an ATtiny13V, but I believe the ATtiny45 or ATtiny85 could also be used, as the only significant way in which they differ is that the later versions have more memory. The fuses determine basic operating parameters of the chip, and only need to be set once, though they can be reset, if you wish. After setting them you can re-flash the firmware as often as you like, and the fuse settings will remain the same, unless you purposely change them.

To find the fuse settings, you can use a fuse calculator such as this one. I used the default settings, with the exception that I disabled the internal divide-by-8 divider for the internal clock, and set the BOD (brown-out detection level) to 1.8V. The piece of code we are using assumes use of the internal 9.6MHz clock.  If you don’t disable the internal divide-by-8-divider, your keyer will send the code 8 times too slow. You can read elsewhere as to why the BOD level is set at 1.8V – try this page, under the heading “Brown-Out Detect (BOD). The resulting command line argument to set the fuses, as given by this calculator, is -U lfuse:w:0x7a:m -U hfuse:w:0xfd:m

The code for the beacon, as written in c, cannot be flashed to the ATtiny – the chip wouldn’t have a clue what to do with it. Before it can be flashed, the program has to be converted into a format that the micro-controller can recognize, through a process called compiling. It might be overkill to use such a big suite simply to compile a program, but Atmel Studio was the first free one I came across, and it worked, so I used it. Download the latest version of Atmel Studio (at the time of writing, it is version 7). It’s a big download – several hundred MB, if I remember correctly, so depending on the speed of your connection, it may take a while.

After opening Atmel Studio 7, select File>New>Project

A dialog box appears. On the left-hand side, under “Installed”, select “C/C+++” and then on the right-hand side, select “GCC C Executable project”. At the bottom of the window, you can name the project “grc-beacon” (or whatever you want to call it), and select where you want the generated files to be stored, unless you want to stick with the default location. Then click “OK”. Then a device selection box appears. You’ll want to pick ATtiny13, unless you’re using an ATtiny45 or ATtiny85. I haven’t tried the latter 2 devices, but believe they will work for this application. Then click “OK”.

You can insert your code where indicated, but at this point, I chose to completely delete everything that appears on this screen, and paste the code into the window. If you have already edited the code in a text editor to include your desired callsign, then no further changes will be necessary. If you are still using the code exactly as displayed on this page, you can at this point edit callsign “GRC” out and replace it with your callsign. Remember to also alter #define MSGMAX to match the number of characters in the callsign +1 (if a change is necessary). For the callsign GRC, that will be 4. If, for instance, you were using “DOGGIE” you would set it to 7. That’s it. Simple!

In the next step, we will generate the hex code that can be flashed onto the valiant little micro-controller chip in our beacon. Go to Build>Build Solution. As soon as you click “Build solution”, you should see all sorts of activity in the window at the bottom of your screen, as the compiler goes about the business of compiling the code. Hopefully, after the bottom window has finished scrolling, you should see –

Build succeeded.
========== Build: 1 succeeded or up-to-date, 0 failed, 0 skipped ==========

Then, towards the top right-hand side of your screen, in the solution explorer, after clicking on the little arrow next to the “Output Files” section, you should see the coveted hex file. Note that I called this project “grc-beacon-3” (I think the original version was called “grc-beacon” but this was my 4th attempt at getting it right) –

If you double-click on the hex file in the Output Files section, a new window will open up, and you’ll see the code in hexadecimal format. Mine looked like this. This is the code for the “grilled cheese beacon” :-)

Now you have the code in hex format, and the command line argument for setting the fuses. All that remains is to flash this onto the ATtiny micro-controller. SparkFun make a Tiny AVR Programmer that includes the target board for plugging in the ATtiny chip. I already had a USBTinyISP AVR Programmer from AdaFruit, so decided to make a target board, which cost me nothing extra, as I already had the parts on hand –


The ribbon cable that connects the AVR programmer to this target board can be inserted the wrong way, as the header connectors are not polarized. I opened up my AVR programmer and traced the pins from the ATtiny45 in the programmer to ensure that they would be connected to the correct pins on the ATTiny chip plugged into the DIP socket on the target board. Like goes to like, i.e. reset pin is connected to reset pin, MISO is connected to MISO, MOSI to MOSI, SCK to SCK, +vcc to +vcc, and gnd to gnd.

Here’s what my version looked like when finished (made with Rex’s MePADS). The thin strip of solder at the top left-hand side of the board in the next shot was put there as a visual reminder of which way to plug in the ribbon cable from the USBTinyISP –



Here’s the target board plugged into the AdaFruit USBTinyISP –


P.S. – when programming the ATtiny chips, I don’t fully insert them. With high quality machined sockets, a gentle push makes good enough contact, and makes it easy to remove the chip without deforming the pins. In fact, I did the same when plugging the chip into the beacon board and it has been running fine now for a few weeks. I do this in case I decide to reprogram the chip a few times before deciding on  the final callsign. Another way of treating the pins gently would be to use a zero insertion force (ZIF) socket when programming the chip. Tayda have them for a low price, or you could use this target board from John KC9ON, and his company, 3rd Planet Solar.


The USBTinyISP Programming Adapter from 3rd Planet Solar. Photo reproduced with kind permission of KC9ON.

(Note – all my instructions here are for Windows. I know next to nothing about Macs, but if you’re a Mac person, the AdaFruit instructional linked below can help you out.)

AdaFruit have a useful instructional on how to use their AVR programmer, which applies to any USBTiny ISP. If you haven’t done this before, refer to their instructional, install WinAVR, and become familiar with it’s use. I’ll assume you know this stuff in the following paragraphs.

I burn the fuses first, in a separate operation. That way, I know they are set, and it makes subsequent programming operations simpler (with fewer things to potentially mistype at the command prompt). With the USBTinyISP plugged into your computer via a USB cable, as well as the target board, make sure the ATtiny13 (or ATtiny45 or ATtiny85) is plugged in to the 8-pin DIP socket on the target board, and you are ready to flash.

At the command prompt, navigate to the directory where your hex file is located. If it is on the desktop, for example, at the command prompt, you type

cd desktop

– and just to the left of the blinking cursor, you should see


– indicating that the Windows Desktop is the current directory. You’ll also see some other stuff to the left of the word “Desktop” but exactly what, will vary, depending on your particular set-up, so I won’t confuse you.

Just to check that your programmer is working, with it plugged into a USB port on your computer, type


and you should get a list of all the commands that it recognizes. It should look something like this –


Then, at the command prompt, type

avrdude -c usbtiny     

Then hit return, and because you didn’t specify the target part, the programmer will tell you so, and give you a long list of all valid parts. I’m not showing it here, because the list is too long to fit on the screen without scrolling but near the bottom, you’ll see the ATtiny13, and it’s abbreviation, which is simply “t13”.

Now that avrdude has slapped your wrist for not specifying the part, let’s give it what it wants, by typing

avrdude -c usbtiny -pt13  (or -pt45 if you are using an ATtiny45, or -pt85 for an ATtiny85)

Hit return, and you should get something like this, which indicates that your USBtinyISP is accepting commands, and recognizes the ATtiny device. In other words, it is ready to flash the firmware –


Then to set the fuses, type

avrdude -c usbtiny -U lfuse:w:0x7a:m -U hfuse:w:0xfd:m

Hit return, and if you get something like the following, it means you have successfully written the fuses. Congratulations – you don’t have to do it again!


If you want, you can set the fuses when you are flashing the hex file, but there is the potential to goof up, set the fuses incorrectly, and render the ATtiny incapable of further use. I’d rather do it in a separate operation and then not have to worry about it again.

Now to flash the beacon firmware onto the chip. At the command prompt, type –

avrdude -c usbtiny -pt13 -U flash:w:grc-beacon-3.hex

The above example assumes that your hex file is already in the directory that you have navigated to (in these examples, I have navigated to the Desktop), and that your hex file is called grc-beacon-3.hex  It probably won’t be called that, so make sure to substitute the name of your hex file. After hitting return, if you get something like this, you have hit the jackpot, and it looks like you are in business –


If you have already built/modified the beacon transmitter, you can plug the ATtiny chip into it, and should hear the sweet sounds of your beacon ID being sent repeatedly on a nearby receiver (with a brief pause between ID’s). You can also connect a crystal earphone or other piezo-electric transducer to pin 2 of the chip to hear sidetone, as a check.

Tayda Electronics is now carrying a small range of enclosures, including some diecast ones, and they have great prices. I ordered a couple of sizes to see how they were, and ended up using the smaller one for this beacon. Here’s the board mounted inside it’s enclosure –


That’s a small dummy load plugged into the BNC connector. Once connected, you can measure the peak to peak voltage across it with an oscilloscope, and use that to calculate the output power.



Although you can’t see them, I fixed 4 little vinyl bumpers to the bottom of the case.


Once you have this little powerhouse in an enclosure, you’re ready to set the output power with the 2.2K drive trimpot. Ideally, you’d be able to accurately measure the field strength at 30 meters from the antenna and use that as your yardstick. This is what K6FRC did when setting up his “FRC” HiFER beacon. IIRC, he runs 1.8mW into a groundplane. I saw an online posting from him in which he said that he was running very close to the maximum permitted field strength at that power level (he has access to a field strength meter). As I don’t have an FS meter, I chose to go with the results from W1TAG’s paper and chose 4.6mW into a dipole as my goal.

If you have an oscilloscope with a bandwidth high enough to measure voltages at such frequencies, it is a useful tool for measuring the output power of your beacon. As the power is specified in terms of the field strength it generates, there is no need to locate the transmitter close to the antenna feedpoint in order to minimize losses. If the regulations specified a maximum power out of the transmitter final, then this would be a worthwhile approach. This is the case with some Part 15 allocations (such as the one for the MW AM broadcast band). However, in this band, we are free to calculate the loss of the feedline and adjust the transmitter power accordingly.  This means that the transmitter can be located indoors, and away from the extremes of weather and temperature.

With my MFJ-259B, I measured the loss of my 50 feet of RG8-X at about 0.7dB, and figured that a transmitter output power of 5.4mW should result in about 4.6mW at the antenna. Using this online calculator, 5.4mW translates into a peak-to-peak voltage of ~1.47V into 50 ohms. With the 3.3V regulator in circuit, the maximum power output was only 10mW,so adjusting the drive to produce 1.47V peak-to-peak on the scope was fairly easy.

Incidentally, the backwave is very audible when you are close to the transmitter. The backwave is the carrier that is still radiated from the antenna when the keying is off. This happens because we are keying the final, so that when the key is “up”, some of the signal from the oscillator still leaks through the PA and into the antenna. I measured the backwave on this transmitter as 01.mW, and it remains at the same level regardless of where the output power is set. Granted that at lower output levels, such as 5.4mW, it is a greater fraction of the power when the key is “down”, but although I could hear the backwave in my immediate neighborhood, it gets lost in band noise pretty quickly. 0.1mW is about 34dB lower than 5.4mW, meaning that if someone is hearing the beacon at S9 +30dB, then the backwave will be a little over S8. Realistically though, anyone who is not really close to it will not be hearing the mighty 4.6mW signal at anything more than a few meager S-points at most, relegating the backwave into the noise. If it really bothers you, you could run the transmitter from a 5V regulator, set the output power higher, and then reduce it with an attenuator pad in the output circuit. That would lead to less backwave in the antenna. I didn’t bother about it.

Here’s the antenna – a Buddipole vertical element, mounted on a painter’s pole on the balcony of my house, putting the base of the L-shaped dipole at aobut 25 feet above ground level. The other element of the dipole is a length of wire. It’s a pretty good take-off to the north and east, but it is blocked by the house to the west and south –


This is what the beacon sounds like on the K2 in my shack. I purposely took steps to reduce the signal level into the receiver, so as to get an idea of what it would sound like at a distance –

Here’s Mingus the neighborhood cat, listening to the Sproutie beacon on my K2, from across the street. Apologies for the cat butt! You can hear the backwave in this video –

The “SPT” Sproutie Beacon is now sitting in my shack, pumping it’s plucky little signal into the ether 24/7, and has received 2 “DX” reports so far. The first was from Jeff KF7RPI, who heard it at his QTH in Portland, Oregon, briefly at a 239 – 339, before it faded back into the noise. He is about 530 miles from me as the crow flies, which is pretty good for such a QRPp signal. The second report was from Bill Hensel on the LWCA message board. He was hiking in Pike National Forest when he heard SPT one day at 1845utc (also briefly) on his KA1103 portable receiver. Bill was about 900 miles distant from me, so that is also exciting. These are the only 2 reports SPT has received so far, but it is encouraging. Some folk do run grabbers on this band and look for QRSS signals. I’m thinking that if SPT’s 4.6mW signal can be heard at 900 miles while at 6wpm, it could go a lot further if it were sending much slower. However, I do like being able to decode it with my own ears, so will keep it at 6wpm (or maybe 12 wpm) for the time being.

Incidentally, if you want to put a HiFER beacon on the air with the minimum of fuss, the Ultimate 3S QRSS/WSPR transmitter kit from QRP Labs will operate on any frequency in the HiFER band, thanks to it’s Si5351 frequency synthesizer. The LPF for 20M should work fine for attenuating harmonics. As this kit is capable of producing far more power than Part 15 regulations allow, it is your responsibility to limit the output power if you operate this transmitter on the HiFER band. The Ultimate 3S will do multiple modes and bands – it’s a do-it-all-in-one MEPT, really, and at a very affordable price.


If you hear the SPT beacon on 13558KHz, please send a report – either to the e-mail address listed on my QRZ account, or as a comment underneath this post. Reception reports will be very eagerly received. One gentleman in Seminole County, FL, reported that the area around the SPT frequency was a cacophony of noise in his area, and he stood no chance of hearing it. Those kinds of reports are useful too.  If you put your own HiFER beacon  on the air, do introduce yourself on the LWCA message board, and John can include you on the list of known active HiFER beacons.


4.6mW of legal, unlicensed pluckiness and grandeur, hiding out in a diecast box.

September 30, 2012

Building and Installing The Internal ATU, SSB Adapter and 160M Receive Options For the K2

F5VJD’s DSB80 is not completely finished.  I need to get rid of the chirp, and the first step will be to clear up some of my shaky grounding practices. I’d also like to build a sidetone oscillator. However, I spent a lot of time and put quite a bit of work into getting it into the case and wiring it up, and I really needed to take a breather before continuing with it.

In the meantime, Eric WA6HH posted a heads-up to the Elecraft reflector that there would be a slight price increase on several of their products in mid-September.  No more information than that was provided.  I knew that at some point I wanted to assemble several options for my K2 and, not knowing whether the options I wanted would be affected by the price increases, decided to go ahead and order them. As it turns out, mid-September came and went, and the options I ordered are still the same price but that’s OK – I wanted them anyway.

After working on the DSB80, assembling a few Elecraft K2 options would be a nice bit of relaxation. All I had to do was follow the instructions and everything would work – kind of like knitting a sweater according to a pattern. You do what they say and as long as you execute well, it works out perfectly.  That has been my experience with Elecraft so far, and these options were no exception.

The options I ordered were the K160RX 160M receive option, the KAT2 20W internal ATU, and the KSB2 SSB adapter. The 160M receive option was a bit of a no-brainer at just $40. I don’t do that well on 80 with my current antenna, so the chances of me doing OK on 160M are pretty slim, but one day I’ll be in the position to put up an antenna for 160M and on top of that, I was keen to see how far beyond the band edges I’d be able to receive – it was a way to increase the general coverage capabilities of the K2 just a little. Plus, the ability to have a different antenna for receive could be useful. The internal ATU would make operation simpler.  I’d been using an MFJ manual tuner, and wanted to be able to hop bands and frequencies more quickly (yes, lazy I know).

The SSB option wasn’t in my head when I first built the K2. I was so into CW, and loved the simplicity of a high-performance rig assembled solely with CW in mind.  However, very occasionally, I have heard stations on SSB that I wanted to contact. One station in particular was in the Philippines who was calling CQ with no-one coming back to him.  I don’t remember the band, but he was so loud that I knew he’d be able to hear even a QRP signal from me. It was one of those moments where I thought it’d be neat to speak into the mic and make an instant, easy contact with a DX station. Plus, even though I don’t use digital modes much at all, it would allow me to use WSPR occasionally if I so desired. I was also harboring a fantasy that I might want to start rag-chewing on phone one day but in retrospect, that was probably an unrealistic hope :-) As it turned out, the novelty of having SSB was starting to wear off after 2 QSO’s, and was mostly gone after the 3rd.  It’s still a good extra capability to have though, and I do like going through the process of assembling, installing and aligning circuits.

The K160RX 160M receive option is  a simple little board that doesn’t take long to put together, so I decided to have a go at it first. The board is upside-down in the first shot.  Sorry about that. Hmmm – looks like a couple of my foster cat’s hairs made their way onto the relay on the right-hand side. Those long-haired cats leave their marks everywhere…….. :-)   (Scroll to the end of this post to see where those hairs came from.)

Installation of this option is straightforward. On receive it tunes down to about 1600KHz, though with reduced sensitivity.  I’m not yet sure what the upper limit is. The board adds a 160M lowpass filter for the separate RX antenna, and switches in extra capacitance in the bandpass filter on the main RF board for 160M.  I’ve only heard one station on Top Band so far, though I’m sure the next Stew Perry contest will change that. Here’s what it looks like installed in the K2:

Looking from the rear (the plastic “boot” over the antenna connector is to remind the user not to use that BNC when the internal ATU is installed – it is supplied with the internal ATU kit –

Next up was the KAT2 Internal 20W ATU. This next picture of  the top-side of the partially completed LC board (the ATU is comprised of 2 boards) is posted here to illustrate a point.  I had a Twitter conversation recently with a ham who wants to build the K2, yet has a problem with his soldering generating a large amount of residual flux.  He couldn’t figure out why after soldering, his boards have so much flux on them.  I referred him to the Elecraft soldering guide and after reading it, he concluded that he was putting too much solder on his joints. This reminded me of Don W3FPR, who repairs and aligns a lot of K2’s, saying on the Elecraft reflector that he sees a lot of K2’s with way too much solder on them.  If you use a thin and mildly active solder along with a small tip on your soldering iron, you can easily regulate how much solder flows onto the joint, and ensure that you use the right amount and no more.  I use Kester RMA 285 in .02″ diameter (the RMA stands for “Rosin Mildly Active”.)  My tips are 1/32″ and 1/16″ diameter chisel tips – that’s 0.8mm and 1.6mm.   With boards that have plated-through holes, thin solder and a small tip will allow you to apply enough solder to fill up the hole and just a little more. From what I’ve read, ideally, you want to avoid a concave shape of solder in the hole – you should have a very small fillet of solder leading up to the lead, with the operative phrase being “very small”. I veer towards making the joint almost flat with the board – though flush cutters won’t allow me to make the joint totally flat. If you are applying the minimum amount of solder needed to make the joint, you wont have problems with solder bridging adjacent pads and causing unwanted shorts. You also won’t need to clean the flux off with a flux cleaner, because there won’t be much, and mildly active rosin isn’t corrosive enough to cause a problem if left on the board.  I think I did end up applying a little flux cleaner to this board eventually, but this picture was taken before applying it. What little flux there is, is honey-colored and hard to see in this picture, but I hope it illustrates the fact that unless you’re very particular about appearance, you don’t need to use flux cleaner if you’re using the right kind and amount of solder:

Another view of the top-side of the partially completed LC board:

Incidentally, if the board you’re soldering on doesn’t have plated-through holes, you will need to apply more on top of the joint in order to make a reliable connection. The capacitors on the above board are bent over per the instructions in order to achieve the necessary clearance. The underside of the LC board contains all the relays. If you look closely, you’ll see that relay K13 is not quite parallel with the other relays. These kinds of oversights on my part drive me potty but are of no consequence to the performance of the circuit:

I don’t have a picture of the whole of the top-side of the completed control board, but here’s a view from the side. It’s not a very good photo technically either but it’s all I have. Sorry about that. Note that the NPO cap behind the toroidal transformer is not the stock part (in case you were wondering why yours looks different.)  The stock part is a monolithic cap. I thought that I might have damaged mine (long story) so replaced it with this part. As it turned out, the stock part was fine, but as this NPO cap does a perfectly good job in it’s place, I left it in –

The finished KAT2 Internal 20W ATU:

At the bottom is the control board, and on top, the LC board, containing all the capacitors, inductors, and the relays that switch them all in and out of circuit when finding a match. The pink piece of foam underneath was simply to raise the ATU to the right angle to get a good shot.  I have become a bit lackadaisical recently, as I really should have looked for a way to support the board that wouldn’t be visible to the camera. I’m slipping.  Sometimes I just want to take the pictures and get to the next part of the project:

The completed unit showing the underside of the control board, with the 2 trim-pots that set the readings for forward and reflected power:

The KAT2 20W Internal ATU installed in the top cover of the K2. I’ve placed black electrical tape over the unused holes to prevent dust from getting into the K2.  My top cover was supposed to have been supplied with green tape over these holes, but it had already been removed when I received it.  Not a biggie:

Phew – only one more option to go – the KSB2 SSB adapter. I loved assembling the KSB2 board, as it is a little more densely packed than the K2 or any of the other options I had built so far. It’s fun building small things. Here’s the top of the completed board, showing all those lovely crystals (it’s a 7-pole crystal filter) –

The manual says to install the crystals flush with and tight against the board.  I usually try to build my circuits so they will be as reliable as possible and although there was an insulating solder mask on the top of the traces that connect to the crystal terminals, there was, in my estimation, the slightest chance that if the integrity of that solder mask were to be breached, there was a possibility of the pads being shorted out by the metal underside of the crystal casing.  It’s a long shot but why risk it if you can avoid it? I spoke with Richard at Elecraft who confirmed to me that if I were to space the crystals above the board by the thickness of a piece of paper, it wouldn’t adversely affect the performance of the filter. That was all I needed to hear, so that is what I did.

Although I have definite OCD tendencies, I do keep them in check on a regular basis.  Functionality is an important factor and although it’s nice to have all my toroids looking neat, I recognize that there is no difference in functionality between a toroid with perfectly spaced turns and the ones that look like mine in the picture below. Some of my toroids turn out looking really nice and some of them look just OK, but they all work just fine :-)  –

The underside of the SSB adapter board. RFC1 and RFC2, though marked on the underside of the board, are actually installed on the top-side. They are wound on very small FT23-43 cores and the only reticence I had about top-mounting them was that it was hard to avoid the windings coming very close to the metal cases of the adjacent crystals. I know the windings are covered in insulation, but I’m not overly keen on it. Unavoidable though. You can see how close the mini toroidal chokes are to the metal crystal cases in the direct overhead shot of the board, 2 pictures above. Here’s one of the underside of the board –

There are several reasons I post many pictures of my projects. I know there are times when I’m assembling something and I want to know how other people do it.  Occasionally, if an instruction in a manual is unclear to me, or can be interpreted in different ways, I’ll go trolling the internet to see what other people who built the same thing ended up with. Perhaps I can help a few people by offering up my pictures. Also, it’s fun taking pictures!

Before adjusting the carrier balance, the instructions in the manual say to place the wiper of the carrier balance trim-pot at approximately mid-travel. I did this and found that very little further adjustment was necessary to null out the carrier. I have only performed a rough adjustment using the S-meter of the K2 so far, but I’m pretty close.

All these options required a small number of modifications to the main RF board on the K2. Some builders who are reticent about desoldering components from their beloved K2’s have opted to use the Rework Eliminators.  I did think about it but decided against using them for 2 main reasons. Firstly, I’m cheap frugal and was already stretching my finances a bit even by buying just the basic K2. Secondly, I’m quite good at soldering and desoldering and don’t find it a nuisance at all to have to partially disassemble a rig and desolder a few components in order to fit a new option.  I have actually enjoyed the partial disassembly of the K2, as I got a chance to renew my acquaintance with it’s innards. The K2 PCB’s are high quality boards. You’d have to apply an awful lot of heat for a long time in order to damage them.  I did turn the flux darker in color in one or two places, but a quick application of flux remover with a plumbers flux brush took care of that.  The places on the K2 board where I have removed jumpers and desoldered components still look really good.

As a continuation of the last topic,  I hope you’ll allow me to air a few more opinions about soldering and the K2.  I do understand how someone who hasn’t built a K2 before would want to protect their investment. The Rework Eliminators are very appealing in this regard. In my opinion, they could be useful if you are going to be changing options a lot. Even a good board in the hands of an experienced tech will start to look shabby if soldered and desoldered enough times. Now that I have installed the 160M receive, internal ATU and SSB options, I doubt very much that I will need to remove them. If you’re experienced at soldering, the lack of Rework Eliminators  in my opinion is not a problem. If you’re not experienced at soldering, in my opinion also, you shouldn’t be building a K2.

Please don’t interpret the last sentence as a discouragement from building one of these absolutely brilliant rigs. If you have any inkling to build a K2 – please do.  It might be the last chance you’ll get to build this kind of a kit at the component level.  It’s certainly a classic of our times.  If you’ve ever thought that you’d like to build a K2 – definitely do it, but make sure that you can solder well before you start.  $740 plus tax + shipping (at time of writing) is a lot of money to spend on something if you’re just going to drown it in solder!

If your response to this is is something along the lines of “But my soldering isn’t very good and it’s always been that way” then I challenge you to improve it. It can be done.

As part of installing the SSB option, some extra components have to be installed on the control board – some capacitors, a couple of transistors, and a set of header pins for configuring the mic wiring for whatever mic you will use. For configurations that require adjacent header pins to be connected together, you can use jumper blocks.  The mic I decided to use was an old Heil Traveler headset.  I didn’t like the bulk of the thicker wire and adapter cord (that in my case was configured for an FT-817) and the small plastic box that contained a PTT button and freq up/down buttons, so I chopped it all off and got to decide how I was going to wire it to the 8-pin mic plug.  I decided to go with the straightforward configuration that the Elecraft MH2 mic uses i.e. adjacent pins connected to each other.  If I decide in the future to use any different type of mic, I’ll probably use a home-made adapter cable so that I don’t have to go into the K2 and re-configure for each different mic –

Here’s the KSB2 SSB Adapter installed on the K2 main RF board.  There is a hole in the SSB adapter board that allows the frequency counter probe to be attached to TP2, but I’m a little concerned that although it can also still make electrical contact with TP1, if I want to seat it fully into TP1, it looks like I’m going to have to file some of the edge off the KSB2 board –

Those blank holes on the back of the K2 are starting to fill up.  The blank holes are supposed to be covered with green masking tape as supplied, but that tape had already been removed on the covers supplied to me, so I have covered the few remaining holes with black electrical tape. Looks OK and keeps the dust out –

I’ve had a few SSB QSO’s with the K2 so far and out of 5 QSO’s, 4 of the stations gave me unsolicited reports of excellent sounding audio. One station said that perhaps I could use a few more highs and a little less on the low end, but that I should be careful about messing with the audio too much as it already sounded very good.  I may adjust the carrier placement in the filter pass-band just a little to accentuate the highs, but I think I’m pretty close.

Assembling and installing these options has been an enjoyable exercise.  It has given me the opportunity to re-familiarize myself with the insides of the K2 and has only solidified my liking for this rig, as well as my respect for Elecraft. The ham community is very lucky to have this company making these great products for us.   My budget has taken a bit of a beating this month, but I’m hoping to be able to get the K60XV 60M option pretty soon :-)

Oh – and those hairs on the relay in the very first picture in this post? They came from Chala – a 10 year-old kitty that I’m fostering right now. There is a group that goes out every night and feeds the feral population here in Oakland. Chala was amongst those cats, but one of the volunteers brought her into the shelter because she was not doing at all well on the streets. She’s a shy kitty with a very sweet nature, and was unable to defend herself from the other cats. She was in really bad shape when first brought into the shelter, but is slowly starting to settle down and get comfortable here. She’s a sweetheart and as you saw from the photo of the K160RX option (if you were looking carefully), is already starting to leave her mark on my construction projects :-)

Chala is pondering how the K2 will perform on 60M when I build the K60XV option.

October 8, 2010

The SSTRAN AMT3000 – A Part 15 AM Transmitter

In the last post you saw my first steps towards putting a low power unlicensed (yet legal) AM broadcast station on the air from my house in Oakland, California.  I spent several months putting the programming together using a piece of free radio automation software called Zara Radio. Adding songs is not quite as easy a process as ripping songs into your iTunes.  As well as ripping the songs from CD, I have to trim any excess silence from the beginning of the song and then mark the exact segue point, so that the software knows at which point to start the next song.  It takes a little while to do this, and I have now done it for 1650 individual songs, as well as producing sweepers, promos and other little doo-das that all go to make a radio station sound like, well, a radio station. The software also automatically plays a newscast (grabbed from an internet feed) at the top of each hour after playing a news intro. It all sounds pretty nifty and I’m quite happy with the way it has worked out. So with the programming worked out, the next step was to build and install a transmitter. I had originally decided to use the Hamilton Rangemaster, but that was going to set me back the best part of a grand for the transmitter and cabling alone, not to mention the cost of outboard audio processing.  The AMT-3000 Part 15 AM transmitter is made by SSTRAN, has onboard limiting and compression and comes as a kit for around $100.  On checking reviews and write-ups, it seemed to be a high quality kit with a stable well-modulated signal and decent onbaord audio processing. The relatively high cost of the Rangemaster had dissuaded me from continuing with my low power broadcasting aspirations,  but the thought of getting on the air for $100 meant that the financial barrier to going through with this had just been removed. $200 is a little closer to the mark actually, as I would also have to fabricate a vertical antenna with loading coil and buy the cable to connect the outdoor transmitter to the studio, but this was within the realm of justifiability for me.

So this is what the SSTRAN AMT-3000 looked like when it arrived at my house. Exciting eh?


This kit arrives in a box. Who would have thought it?


On opening the box the first thing you see after the packing slip, is a clear and detailed instruction manual sitting on top of a well-packed kit. First impressions are very good.  At this point I definitely want a second date:


The instruction manual is detailed and clear.


When you open up the box, you see all the parts, including a wall-wart transformer, all knobs and connectors,  a high quality silk screened PC board and a plastic case with printed front and rear panels. You even get antenna and ground wires that are sufficient if you’re just intending to broadcast to the radios in your house (50-200 foot range):


This is what you get inside the box.


If you look closely at the above picture, you’ll see that instead of a front and rear panel, I have 2 rear panels.  An e-mail to SSTRAN solved the problem and a front panel arrived in my mailbox a few days later. It didn’t delay my building the kit as I could still complete the board, place it in the case if needed, and use the transmitter.

Although a pretty straightforward kit, this is not a project for the absolute beginner. The parts density is fairly low, but some of the individual parts have leads that are quite close together. If you have some experience soldering parts onto circuit boards you should be OK with this kit though. There is one SMT device which came pre-soldered to the board in the version that I bought.  There is also a version of the kit for $3 less that doesn’t have the SMT chip pre-soldered, but this wasn’t available at the time of ordering so I took a deep breath and decided to pony up the extra $3 (I’m a cheapskate and happen to like soldering SMT devices).

Another thing about this kit is that there are no coils or toroids to wind. I joked in the forums over at Hobby Broadcaster that I almost felt as if I hadn’t really built a transmitter because I hadn’t wound any inductors! The kit uses pre-wound inductors, which does make the whole assembly process faster and more straight forward. This is what the board looks like when fully assembled:


The finished board. At this point, you can plug it in and transmit!


Looking at the board above, at the back from left to right are the 2 audio input jacks.  It is a mono transmitter, but if you have a stereo feed you can plug both channels in here and they will be summed to a mono signal.  If you have a mono feed, it can be plugged into either connector. Hiding behind the 15V regulator heatsink is the power connector.  The kit comes with a wall-wart that outputs 16V AC, but you can also run this from a DC input; I used 24V DC from 2 gel cells and it worked fine. Finally at the far right is the antenna and ground connector. You can see the 3 RF chokes that can be placed in or out of circuit with jumpers in order to combat hum due to stray RF. 2 of the chokes serve to isolate the power input and one isolates the audio input ground from the board ground.  Because these are RF chokes, they allow audio to pass, so if you have an audio ground loop you will still need to fit an audio isolation transformer. In front of the regulator heatsink is the blue 8 position DIP switch that is used to set the transmit frequency. The transmit frequency is derived from a PLL synthesizer which gets it’s reference frequency from a 4MHz crystal (for the US version with 10KHz channel spacing), making this transmitter stable enough for most Part 15 purposes. The 4 position DIP switch to the right controls the switching in and out of circuit of several inductors for use when tuning the supplied indoor antenna. When using a base loaded outdoor vertical, the on-board inductors are not used.

The whole thing looks pretty nifty when you put it in the case:


Your own mini AM broadcasting station (substitute the phrase "medium wave" for AM if you're in the UK.)


On the front from left to right is the audio gain control, the pot that sets the modulation level (it sets the point beyond which limiting occurs), and the compression level control.

Oh, here’s the bottom of the board too, just to prove that I can solder:


Look Ma, I can solder!


Many folk buy these in order to broadcast programming to their vintage radios. I don’t know what the AM band sounds like in other countries (or medium wave band as it is more correctly referred to in the UK), but in the US it is mainly conservative talk radio produced with ratings (and not quality of content) in mind.  The conservative part doesn’t concern me, but the fact that much of it is highly opinionated banter designed solely to push the emotional buttons of listeners does bother me no end. John N8ZYA refers to it in his blog as drivel and to my mind he’s right on the mark.

Anyway, if you just want to broadcast around your house, you’re pretty much done at this point.  You can connect the supplied pieces of ground and antenna wire, tune them up per the instructions, and you’re ready to blanket your homestead with good sounding AM broadcasts. If, like me, you’re hoping to cover a slightly larger area, then the next step is to build an outdoor vertical antenna with a loading coil.

The instructions for a base loaded vertical made from readily obtainable parts are on the SSTRAN website, and there is a drawing and parts list also. All the info is contained on the site, so I’m not going to repeat it here, but this is the loading coil made from 16 AWG magnet wire wound on a former of 3″ white schedule 40 PVC:



This loading coil is BEEFY.


PVC has a tendency to absorb moisture over time, so the pipe was painted with 2 coats of varnish, the coil wound, and the finished coil coated with varnish.  The ends of the pipe were masked with tape so that the end caps could be cemented into place afterwards.

Here’s the finished antenna and transmitter installation.  The transmitter was housed in a Rubber-maid container.  If I decide to make this installation more permanent, I’ll search around for a white plastic box with weatherproofing seals. You can’t see the top of the antenna, but it’s just a length of copper plumbing pipe with a cap soldered on the end:


Schedule 40 pipe and Rubber-maid containers - the giveaway signs that this is not a high-tech installation on the roof of some government building, but just another radio experiment at my house.


Here’s a closer view of my Home Depot/Orchard Supply Hardware/Safeway special:


At this early stage, the antenna wire is still connected to the coil tap with an alligator clip; tuning has not yet been finalized. Audio and power cables enter the box through small drilled holes and will be sealed with silicone caulk.


Some gratuitous beefcake; the loading coil gets it’s close-up:


If this Part 15 thing doesn't work out, perhaps I could use it on Top Band......


So how does it work? Weeeelll……some people get great coverage and results right off the bat.  I am not one of those people, and I think I’m going to have to put quite a bit of effort into perfecting this system if I want it to work to my satisfaction. The transmitter itself seems to be doing everything it should.  The carrier is stable. The audio processing sounds good.  Considering all the processing is taken care of by one chip I’m quite happy with the way it sounds.

I’m experiencing two problems. One is that there is an AC hum on the carrier that I’m pretty sure is happening because RF from the antenna is being re-radiated by the house AC wiring. This will be a tough and maybe impossible problem to solve, as I am one tenant out of 10 in this house and don’t have access to the other tenant’s rooms.  There are a couple of solutions in my head, but they may not be possible given my current living situation.

The external antenna certainly does increase the coverage area over the supplied wire antenna. With the supplied wire antenna I could receive my transmissions all over the house, but not too far outside. With the external antenna coverage  seems to go as far as 1000 feet in some directions, but only a block or two in other directions. I don’t believe that this is in any way a fault of the transmitter but of my imperfect installation. I’m using a cold water pipe as a ground connection and don’t know how good a ground it is.  Although the antenna is about 15 feet off the ground, it is shielded on one side by the house and on another side by apartment buildings next door.  If I could mount the antenna on my roof I think I’d get better coverage (as long as I were able to ensure that the ground connection is not radiating in order to keep the FCC happy), but roof access isn’t too easy here.

I’d prefer to ground mount the transmitter so that I can ensure that the ground lead is very short and connects directly into the ground to avoid any possible misinterpretation of the ground lead rule if the FCC were ever to inspect, but ground space is limited on my plot and I have to be careful not to overstep any boundaries with my landlord and fellow tenants.

So…….I’m going to sit back and not do too much with this project for the time being. I may decide to use this transmitter with a short wire antenna just to broadcast around the house, or I may get another burst of enthusiasm and decide to try a different installation in the hope of increasing the coverage (and eliminating the hum on the carrier.)  It sounds great with a short wire antenna and I know that with the right installation it will sound great with an external antenna also. I’m hoping to cover an area of radius 3/4 mile around my house and I do think it can be achieved.

I’m also realizing how even my QRP ham radio activities are easier than Part 15 operation. With QRP, I run 5 watts and have no restrictions on antennas at all.  In fact, if I want, I can run up to 1500 watts on most ham bands into any antenna I want. Engineering a Part 15 system is truly a challenge.

Now if the FCC could just relax the Part 15 antenna restrictions and allow me to hook this thing up to an ATU and long wire antenna…………



May 27, 2010

A Different Type Of QRP

My radio interests have taken a different course in the last month or so and I’m not sure whether it’s appropriate to include them in this blog or not, so let me tell what I’ve been up to and you can decide;  I’d welcome your comments.

I had been earning my living as a DJ/announcer/voiceover guy since 1987 until last year when the paid work all came to an end for me. I really thought I’d gotten over the DJ bug and in some ways, I think I have.  As a youngster, I very much wanted to prove myself – to both myself and my peers.  Being perfectly honest, there was a need for a certain amount of ego gratification too. Well, I don’t feel the need to prove myself to anyone anymore and as for the ego gratification part – well, I think that age has cured me of the need for that.

But interests picked up at an early age never go away completely and so it was that a month or two ago I became interested in setting up a Part 15 AM broadcast station.

Most people reading this blog will know about the FCC regulations that cover Part 15 devices, which can be broadly divided into 2 types – non-intentional radiators and intentional radiators.  Non-intentional radiators are devices that emit RF as a by-product of what they do and then happen to leak some of it. This category includes the local oscillators of radio receivers and the crystal controlled clocks in computers, for example.  Intentional radiaters are devices for which the RF radiation is the main point of the device.  Baby monitors,  cordless phones and garage remote controls are examples of intentional radiators.

Part 15 regulations also cover something which is quite attractive to me,  and that is the ability to operate a low power unlicensed yet completely legal broadcast station. On the FM broadcast band, there is a strict field strength requirement at a set distance from the antenna that makes it unlikely you’ll be able to achieve a range much greater than 200 feet.  This is fine if you want to play music from your iPod on your radio, but you’re not going to build much of an audience with a Part 15 FM station. The criteria for transmitters on the AM broadcast band are much more lenient; instead of field strength measurements at a specified distance from the antenna, the main rules are that the input power to the final amplifier cannot exceed 100mW and the total length of the antenna and ground lead cannot exceed 3 meters.

100mW to a 10 foot antenna on the AM broadcast band isn’t much, but if you engineer the whole system for maximum efficiency, it can be possible (from what I’ve read) to achieve a range of up to a mile (or even more) from the transmitter. What makes this even more appealing is that the FCC allows the use of multiple transmitters to increase coverage. Place a few transmitters around the edge of your primary coverage area, and now you’re starting to cover a significant part of a city.  Some small town residents can cover a significant part of their entire community with one well-placed transmitter. The same discipline of maximizing the efficiency of the whole system that applies to QRP ham operating also applies to Part 15 AM operating. After all – it is also QRP. In fact quite a few operators of Part 15 AM broadcasting stations are licensed hams.

Many Part 15 AM broadcasters are either people who want to broadcast old-time programming to their restored antique radios, or Realtors with their “Talking House” transmitters broadcasting details of houses for sale to prospective buyers parked outside.  A smaller, but very enthusiastic subset of  Part 15 AM operators are the folk who run their own radio broadcasting stations.  The FCC don’t recognize these outfits as radio stations; they are simply classified as intentional radiators.

At this point, you’ve either completely tuned out or have at least some level of marginal interest. Here’s a picture to break the monotony:


Ant Radio broadcasting from the Pill Hill district of Oakland, California


Not a great photo I’m afraid.  At some point I’ll take a better lit and processed picture, but at least you can see what it looks like. This is where I do my ham operating too.  In the shelf unit to the left is the FT-817,  KK-1 straight key and Bencher paddle along with the Fort Tuthill 80 and 2N2/40 as well as my soldering station, 13.8V regulated power supply – oh, and my DVD player to boot! There is also my Signalink USB sitting on top of the Tut 80.

On the right is AM broadcast central. At the bottom is a Mackie mixer on a pullout shelf. Above it the Denon dual CD player (2 separate units – the control unit at the bottom and the CD trays in the unit above it.) The microphone is an EV RE27N/D. There is a Shure SM7 (also on a boom arm) out of sight of the camera for guests.  At the top of the rack are the 2 audio processing units.  The lower one is the first in the audio chain after the mixer.  It is an Aphex compellor which provides compression, leveling and peak limiting of the signal. Above it is an Inovonics 222 which provides pre-emphasis, a lowpass filter (to limit the bandwidth of the transmitted signal) and more peak limiting. It supports asymmetrical carrier modulation to modulate the transmitter to as much as 130% – another way to maximize the range of this QRP signal. The Inovonics 222 is quite popular with AM amateur radio operators to help them squeeze maximum efficiency from their signals.

If you have any interest in this, I thoroughly recommend Hobby Broadcaster (link opens in a new window) – the site for Part 15 AM and FM broadcasters.  It’s run by broadcast engineer Bill DeFelice who also actively participates in and moderates a great set of forums. There are other sites that deal with Part 15 broadcasting,  but Bill really sets the tone in his forums with helpful friendly comments and advice as well as equipment reviews.  He also posts his online finds for those who are looking for deals on good affordable gear for their stations.

When I’m not DJ’ing live (which will be most of the time) the computer runs the whole station in automation and so far, it’s not sounding too bad at all. I’m currently spending a lot of time recording and producing all the station ID’s as well as adding to the song library. When Ant Radio hits the airwaves, I’ll probably post an air-check so you can hear what my little broadcast station sounds like.

Maybe I’ll even get the occasional DX report :-)

The Rubric Theme. Blog at WordPress.com.


Get every new post delivered to your Inbox.

Join 1,638 other followers