Dave Richards AA7EE

May 28, 2012

The Etherkit OpenBeacon – A Very Versatile Programmable MEPT

etherkit’s OpenBeacon is the first kit to be offered by proprietor Jason NT7S and I think it’s a good one for him to open up with.  Sure, it’s not the first QRSS transmitter kit on the market but unless I’m missing something (and I don’t believe that I am) it is the most versatile one offered to date. Here is a list of all the modes that it can transmit:

  • Dual Frequency CW with 3, 6, 10 or 120 second dits (3 second DFCW is the most commonly used QRSS mode on the HF amateur bands and is the default mode programmed into OpenBeacon)
  • QRSS CW with 3, 6, 10 or 120 second dits
  • Regular CW (you choose the speed)
  • Sequential Multi-Tone Hellschreiber
  • WSPR – billed as an experimental mode, but I’ve used it with success
  • Glyphcode – a mode that uses Hellschreiber to generate dots and dashes
  • A special calibration mode that is useful for setting the center-frequency and bandwidth of your transmitted signal

There are no jumpers for mode-switching. It’s all done with the software client that is available for Windows, Linux and Mac OS X. This software control opens the door for anyone who wishes to come up with their own code to control OpenBeacon (and hopefully share it with other users in the etherkit OpenBeacon forum.)

First things first. Here’s what you get:

In this shot, I took the rather good-looking PCB out of it’s wrapping so you can see it a bit better:

Construction is straightforward. If you’re new to kit assembly and the art of soldering, it may take a little longer to figure out. The usual guidelines apply – identify all the parts, and don’t solder anything into the board until you’re sure that it’s the right part, and you’re soldering it in with the correct orientation. There are useful checks included at several points during the build and after the first stage has been completed, you get to connect the board to a USB cable and see your computer recognize the OpenBeacon. If you’re running Windows, you’ll need to install a driver before this can be accomplished, but it’s reassuring to have the project recognized by your computer at such an early stage of construction.

Here’s the finished board:

And another view of the completed board:

Apologies – I forgot to include a rule in the pictures to give you an idea of the size of the board, so for the record, it measures 70mm x 90mm.

One thing I really like about OpenBeacon is the on-board connectors. Once you’ve finished assembling it, you can plug in a USB cable and an antenna, program it with your call-sign and you’re on the air. If you want to run higher RF output power than the approximately 40mW max you get from USB power, then you can plug in DC power to the on-board DC power connector and get about 300mW from a 13.8V supply. These kinds of QRSS transmitters are often not installed in conventional enclosures anyway, so it’s really good to be able to begin experimenting with OpenBeacon straight off the bat.

This is a bit gratuitous, but here’s a really close-up view of part of the board:

If you don’t already have a QRSS viewer, Argo is the easiest to set up and get going with.  OpenBeacon is already set to transmit in QRSS3 as the default mode, but you will need to program it with your callsign by downloading the client for your particular operating system and following the directions on the etherkit site. If you have problems with any of this, Jason NT7S, or one of the etherkit forum members should be able to help you out.

Of course, the first thing you want to do with a transmitter like this is to see what your QRSS signal looks like on your own viewer:

The very next thing was to see what my callsign looked like in Multi-Tone Sequential Hellschreiber:

You’ll notice that the vertical scale on my Argo screen captures is calibrated with the frequency.  To achieve this, you’ll need to accurately calibrate your receiver so that you can set it to 10139 KHz. In my case, I know that after it has been running for a while, my K2 receives at about 10Hz lower than the dial frequency, so to receive on 10139 KHz I set my K2 to 10139.01. The next step is to enter an offset of 10139000 in the calibration menu and you’re all set. The QRSS band normally runs from 10140 to 10140.1, so a carrier in the center of the QRSS band will produce a tone of 1050 Hz in the speaker of your receiver.

One thing that’s very important to remember if you run into any difficulties during assembly –

Incidentally, before anyone suggests that I was illegally radiating a sginal without proper identification, the above screen capture was taken as a result of outputting 5mW from the OpenBeacon into a 50 ohm dummy load – not too much radiating was taking place.

I did run OpenBeacon for one night in QRSS3 mode but wasn’t able to find my signal on any grabbers.  It was only one night and I was, after all, running just 5mW RF output (sucker for punishment here). I’ll be trying it again soon, but at the time I was keen to forge ahead and try the WSPR experimental mode.  I did briefly have a go at figuring out Glyphcode but either the instructions in the Client Software Useage Guide on the etherkit site weren’t clear, or I was being a bit thick.  It could well have been the latter. I’ve no doubt that Glyphcode isn’t hard to program but at the time, was more immediately interested in playing around with WSPR on OpenBeacon, so I hoofed it on forward to that section of the online guide.

To transmit a WSPR signal with OpenBeacon, you first have to load the buffer with the appropriate code. I downloaded wsprcode.exe and placed it in the root directory of my PC.  This is a description of what I did on my PC running Windows. Your mileage will vary if you are running any other OS. In DOS at the command prompt I typed wsprcode “AA7EE CM87 7”  You’ll need to do the same with your callsign, 4-digit locator and power level (in dBM). wsprcode will output the data symbols, sync symbols and channel symbols for your particular data – all you need is the last set of characters – the channel symbols.  Cut and paste this data into a simple text editor (I used notepad on my PC) and edit it to erase all spaces between characters as well as any carriage returns so that all you are left with are 162 characters in a continuous string with no spaces (or anything else) between them.

The next step is to program the buffer of your OpenBeacon with the WSPR code. Using the OpenBeacon client software, the command for this is openbeacon wsprbuffer “flangesprocket” where flangesprocket represents the string of 162 channel symbol characters.

All that remains is to place OpenBeacon in wspr mode, adjust the transmission bandwidth to 10Hz, and the frequency to put you in the wspr band, and manually trigger OpenBeacon (either with S1 or through the client software) at the beginning of an even-numbered minute. OpenBeacon will transmit your wspr signal for the required 1 min 50 seconds.

Here are the unique spots representing 2 evenings of operation:

Date Call Frequency SNR Drift Grid dBm W
by loc km mi
 2012-05-28 02:36  AA7EE  10.140138  -17  0  CM87ut  +7  0.005  K7LG  DM04se  530  329
 2012-05-28 02:20  AA7EE  10.140102  -23  0  CM87ut  +7  0.005  VE6PDQ/1  DO33fl  1867  1160
 2012-05-27 05:24  AA7EE  10.140108  -22  -1  CM87ut  +7  0.005  K7MSC  CN76wv  1021  634
 2012-05-26 07:20  AA7EE  10.140125  -19  0  CM87ut  +17  0.050  W7WKR  CN98pi  1179  733
 2012-05-26 07:10  AA7EE  10.140144  -16  0  CM87ut  +17  0.050  W5OLF  DM78hb  1482  921
 2012-05-26 06:20  AA7EE  10.140138  -22  0  CM87ut  +17  0.050  K7UEB  DN06tb  973  605
 2012-05-26 05:34  AA7EE  10.140129  -26  0  CM87ut  +17  0.050  VE6PDQ  DO34ir  2000  1243
 2012-05-26 04:36  AA7EE  10.140108  -8  0  CM87ut  +17  0.050  KC6KGE  DM05gd  390  242
 2012-05-26 00:08  AA7EE  10.140114  -16  0  CM87ut  +17  0.050  N6RY  DM13id  688  428

Not a lot of unique spots, but I was only running 5mW.  The first spots were reported as being 50mW, but this was my mistake – I was actually running just 5mW and corrected the wspr code as soon as I discovered my error. I decided to stick with an output power of +7dBM as that is the power I feed my diode ring mixers, and it just slightly blows me away that my signals were decoded by VE6PDQ over 1200 miles from me with the same power that my local oscillators put out. Quite amazing. I have no doubt that I’ll achieve even better DX if I keep at it.  As I have developed an affinity for level 7 diode ring mixers, running a power output of +7dBM from the OpenBeacon seems very fitting.

Now, as etherkit is an open source amateur radio company, I’m hoping that someone will write a routine to automatically trigger OpenBeacon when in WSPR mode. It definitely seems to work quite well in that mode.

This is where I stand with my OpenBeacon so far, and I think it’s the most versatile MEPT available in kit form. It boasts many different modes and speeds (as well as WSPR) and to change between them you don’t have to fuss around with wire jumpers. It’s all done in software, like many other things in OpenBeacon. This ability to control it in software, along with the wide variety of modes, and the open-source nature of etherkit make it, in my opinion, the ideal MEPT for a lot of users.


May 11, 2012

A 200mW TX and Regen Receiver Make For Some Very Satisfying QSO’s

It’s been almost 10 months since I published the post on my build of N1BYT’s WBR Receiver – a regen with an unusual design that minimizes re-radiation of the local oscillator signal without the need for an RF pre-amp. It’s fairly sensitive, stable, and only draws about 12-13mA. I’m sitting here right now with this great little regen listening to Bill Crane W9ZN calling CQ in his inimitable style.  If Bill’s the guy I think he is, he was a radio DJ that went by the name of Bill “Butterball” Crane in the 1970’s, and his sending is unmistakable – especially his rhythmic and creative CQ calls.

In the time since I published the post on the WBR, it has been far and away the most popular post in this blog. Every now and again, my build of this receiver is featured on another blog or in a forum, and the page views peak for a few days. Currently the Yahoo QRP-tech group are looking for a regen receiver for one of their group builds, and the WBR is one of the designs under consideration.  I hope that at least some of the members choose to build it. Not all regen designs are created equal; I’ve read of some builders having problems with hard-to-control regeneration, or hum issues caused by the local oscillator signal being radiated by the antenna. The WBR doesn’t suffer from these issues.

As if all this attention online isn’t enough, my build of the WBR (Wheatstone Bridge Regenerative) Receiver is about to be featured, along with photographs, in a new edition of an ARRL book to be released soon.  As I mentioned in an earlier post, Daniel N1BYT is either very gratified, or thoroughly fed-up, with all the attention his design is still receiving over 10 years after the date of it’s first publication!

After building it, I spent a month or two happily listening to, and fiddling around with it, then got caught up with building a beta version of NT7S’ CC-series QRP transceivers, and then building a K2.  Since finishing it, most of my other projects, including the WBR, have been boxed away, leaving pride of place at the operating position to the K2.

Then a couple of days ago, I pulled the WBR out of storage and hooked it up. I connected it to a 2 transistor TX powered by a 9V battery that I built based on W1FB’s implementation of the Pixie 2, and this evening called CQ with a whopping 200mW coming from the plastic 2N2222A in the final.  The Reverse Beacon Network showed a 20dB spot from N7TR 180 miles away in Reno and I figured it was worth continuing to call.

Eventually, NY6G in Tracy, CA gave me a 429 and we chatted for about half an hour. Stan lives only 46 miles away from my QTH in Oakland so although it was quite gratifying to know that he was hearing the flea power signal from my little 2 transistor transmitter, the thing that was really satisfying about the QSO was more to do with the fact that it connected me to the world of amateur radio that I remembered from my boyhood in the 1970’s; a world where there seemed to be more camaraderie and gentlemanly behavior.  Stan told me that he was born and grew up in Oakland, where I currently live.  He then moved to Alameda, served in the Air Force and left the immediate area after WW2, but remembers his Mum taking him to Lake Merritt to watch the boats. (I was watching the boats on Lake Merritt just 2 days ago.)  He also remembers when the streets were lit by gas lamps, a job which was performed by a man carrying a long pole.

Sitting at my desk with a separate transmitter and receiver that I had built from schematics Manhattan-style and using them to hear about what the city I live in was like in the 1930’s from a former resident was pure magic.  Stan was getting a kick out of my home-brew separates too:

The Altoids tin on top of the receiver is the N0XAS PicoKeyer.

This is a fun little set-up. The next day, I powered the Pixie 2 TX from 12.8V which allowed it to go QRO and output 500mW. Over the next couple of evenings I worked KI6ON Joe in Santa Clara CA, KE7GKM Bob in Boise ID, W6VJJ Rex in Half Moon Bay CA, N0EK Ed in Bergen ND, KF6RMK Bill in San Rafael CA and Jerome VA7VV in Vancouver, BC

Jerome had a couple of traditional receivers in the shack on which he was having trouble copying my flea-power signal – an FT-950 and a ProIII. However, on his SDR-IQ (a state of the art SDR receiver) with the bandwidth dialed down to 75Hz, he was copying with no problem.  He said it was the first time that he had experienced such a substantial difference between copyability between conventional receivers and SDR receivers. The thought of the signals from that little 2 transistor TX being heard in Vancouver almost 800 miles away was pretty neat.

It was great to QSO with KF6RMK too. Bill lives locally and we run into each other on the air from time to time.  We exchanged e-mails after our QSO and Bill sent me a picture of an old regen that he bought as a kit from Radio Shack many moons ago. He put it into his own case and wound some coils for it to turn it into an all-bander. Neat looking regen Bill! He said that our QSO has inspired him to drag it out and see if he can get it going on 40M  (photo credit Bill KF6RMK):

It’s been fun putting the WBR and Pixie 2 TX back into service for a while.

Now what can I build next……?

May 8, 2012

4SQRP Come Out With Another Useful Kit For Homebrewers

4SQRP recently announced another addition to their line of kits for the QRP’er and home-brewer. This one particularly caught my eye, for reasons that I’ll explain in the next paragraph.

Designed by David Cripe NM0S, the Hi-Per-Mite provides a -3dB bandwidth of 200Hz with no ringing. The center frequency as supplied is 700Hz, but this is user selectable at time of building. The filter response is bandpass as opposed to lowpass, meaning that it will help to filter out any remaining traces of 50Hz hum you may have in that homebrew direct conversion receiver of yours (or an older vacuum tube receiver perhaps?)  After the filter is an LM386 audio amplifier with a gain that, once again, can be set by the builder. If you set it to the maximum of 50dB, you have a combination bandpass filter and audio amp that can provide the complete audio chain for a regenerative or direct conversion receiver. If you already have the receiver, you can set the gain to a lower figure and you still have a really useful CW filter.

I’m thinking that for those who like to build simple receivers, this could be a very handy kit to have around.

All this for US$19 including shipping within the US (more for overseas).

You can find the 4SQRP Hi-Per-Mite here (opens in a new browser window).

Blog at WordPress.com.