At Callsign Software we are always looking for our next radio candidate. We make this decision based on certain criteria. First, we need a large enough customer base (models sold) so we can get a return on our engineering investment. Second, we want the radio to be a solid performer, and, finally, we look for hardware that could significantly benefit from the ability to be software controlled.

Our current test radio is the latest version of the Elecraft K2 . Last month we purchased a fully loaded "kit" to be used for software development. Most amateurs are aware Elecraft has taken over where Heathkit left off years ago in the kit building market. These kits come as a "box of parts" requir ing the user to wind toroids, solder hundreds of parts and assemble a mass of components. As intimidating as this may sound, it really isn't. Elecraft has put a huge amount of engineering into the "user assembly" side of their kits. The manuals are straight forward, parts well laid out and technical support is available and responsive. Once complete, this radio is a no compromise state of the art HF rig .
We started our project on November 18th, and by the end of the month , we completed the base kit, 160 module, Noise Blanker option, SSB option, DSP option and have completed about 1/2 the 100 watt module with RS-232 support. With the exception of one bad capacitor and one diode we had inadvertently reversed, we ran into no issues with the assembly. The alignment using built - in tools (frequency counter, RF probe, volt and current meter, etc.) also went without a hitch.
At this point we have started preliminary software development and are eager to see if this will be our next software release. We have mixed emotions about our success with the K2 for a couple very obvious reasons. The K2 is somewhat limited in the serial department. It really was not designed to be the ultimate software radio . As an example, the volume control is hard wired with no software support and the serial device operates at only 4800 baud. The baud rate will probably be the limiting factor as our software is "real time" and needs a healthy serial connection to make our advanced features work.
On the bright side , if we can take a radio with the performance of the K2 and port our advanced software to it, we could potentially have a combination with features and performance only found in radios in the $3K plus range.
As mentioned, our test rig is operational and software development underway. Check back often for updates on our progress and potential software release. If you have any questions, please do not hesitate to contact our sales department . Regardless of our success with making this work, we highly recommend this to anyone considering a new radio or looking for a kit project. It doesn't get any better than this.
12/3/2003 by Rich KD0ZV
Today we received the missing part (film capacitor) for our DSP module and installed the DSP module in the rig. The DSP initialized and took right off with no problems. We found the best operating position for the K2 to be directly on our lap. This leaves the left hand free to hold the manual and the right hand free to scratch the back of your head. First the good stuff... After verifying the DSP worked we immediately dropped down to 75 meters to listen to some of the locals. Band conditions were pretty noisy in the late hours and this was confirmed on the other HF rig in the shack. While flipping pages in the manual, scratching head and pushing buttons we accidentally kicked in the DSP. It absolutely works. Of course we had no idea how we did it, but it works.
After a few more minutes of manual reading we learned that a quick "TAP" on the BAND+ button while in the AFIL display mode got things rolling. Holding the button longer than a TAP which is known to the experienced users as a PRESS took us to the next menu in the DSP display. Yes, we were lost :-) The PRESS, TAP and two buttons at once commands are necessary to operate the advanced functions considering the radios front panel is small and lacking in controls.
So from a software standpoint we are excited. The rig is without a doubt a great performer but from a new user, or casual user standpoint the advanced features are hard to find let alone operate. I am sure amateurs that operate this rig on a daily basic can zing through the menus like it is second nature but from our less than familiar standpoint......See head scratching comment above.
Callsigns objective is to give the operator access to all the powerful functions of the K2 without multiple button TAPS and PRESSES and do it with a user friendly "amateur like" panel. We have no intention of just transferring the functions of the actual rig to the computer screen. We want to make a fully optioned K2 easy to operate even for the first time operator.
12/4/2003 by Rich KD0ZV
We brought the K2 up under computer control for the first time today using about every program we could find available online including the Elecraft basic program. I have to admit I was excited to be able to control the rig via the PC and not use the cumbersome (authors opinion) buttons and controls on the faceplate of the K2. I think out of all the programs I tried, I liked Ham Radio Deluxe the best. It was not as "radio like" as I would wish and still a little short on K2 specific features but overall it was very nice. My hats off to the author for a fine job.
Mike talked to Wayne on the phone today and hung up enthused that the K2 is going to be something we can work with. In the back of my mind I keep thinking we have the ultimate contest rig. With user friendly filter controls, CW keying, logging/callsign lookup, etc., we have a contesting "system" that will be hard to beat. Only problem I can see currently is that Engineering wants this K2 for initial testing and I will have to build another one. OK that is not a big problem but going to set me back a couple weeks with K2-play !
12/9/2003 by Rich KD0ZV
Continuing to get familiar with the "available" features of the K2 trying to figure out what operating enhancements we can bring to the party using our software. At this point, it looks pretty promising.
I have to retract my comment about how well the DSP works removing noise from last week. Turns out the DSP was creating the noise !! Had to send the DSP board back to Elecraft to get checked and repaired. So, will have to hold my final thoughts until I get it back and can try it out.
12/20/2003 by Rich KD0ZV
DSP unit came back from Elecraft. For the most part it seems to be OK and it does work but the rig is still quieter without it. Have yet to talk to them again to verify it is OK.
Still trying to find out how to use the rig. The rig basics are somewhat easier to operate now that I am getting familiar with it but the the DSP unit reminds me of my rubiks cube 30 years ago. Decent software would be a great help. Starting to compile my notes for Mike so we can get a preliminary package together. Some of our early conversations have been around two products. The first being a full featured TOTL rig similar to our other Callsign products and the second being a contact rig. The latter would be quick access to features you would use in a contest environment, (scratch memories, filters, CW keyboard, logging, etc.).
1/31/2004 by Mike WA1EYP
While Rich is trying out the electronics of the K2 I have been preparing the port of our
control software to the K2. There are approximately 40 tables that characterize a radio along with a programming interface to the radio that consists of about 30 APIs. These APIs along with
the tables permit our radio control system to function on top of a "Radio Platform" so to speak.
In addition to the tables and APIs we have about 1800 user configurable settings that must be mapped through functionality supported on the K2 in the form of a set of basic commands that
we use. This is truly the dawn of the "software defined" radio. The basic premise is that the radio like the K2 can support a basic set of commands that we use to control the radio. Commands that set the radio frequency or read a signal report are considered very basic. We are in the process of mapping our "virtual" radio's command set to a set of commands available on the K2. This will get us running with the K2. The next step is to take advantage of enhanced set of capabilities in the K2 and provide support for the most useful to all.
I guess the real question is why did we go to such an extent when we could have just
slapped together a bunch of buttons on a window pane to support the radio's commands. Well
first of all we do NOT want to go backwards in terms of features that we present. The full featured memory architecture based radios from "offshore" still use a basic command set that
is packaged in a box with "features" and sold as a unit. Our packaging is the software we write combined with the radio we use. These are the features you see. Some radios radios are well suited and some not for this type of interface. We want to offer our users an "Integrated Station" with "ALL" the features you would expect and none missing. Some "Offshore" radios make you pay for those "MISSING" features even though today it is primarily the software packaged with the box.
A friend one said, "I do not give a darn about the radio, its what I can use in the software that counts as long as the radio's performance is reasonible!"
Because of the rising costs of front panels, knobs, displays,and the cost to get Type and Safety Certification, and
the falling costs of the PC, the "software defined/controlled" radio is a way in which we can get performance at a
reasonible price! The K2 is one of the early pioneers in the area of the "Software Defined" radio along with the
offerings that TenTec had but it has the advantage of going much further in my opinion! It offers performance at a
reasonible price.
3/18/2004 by Mike WA1EYP
Needless to say we have been quite busy breaking out all our low level APIs (programming interfaces) to the lower level
radio commands that we use to support our overall radio functionality. We have about 45 different base functions
that we must fill in to give you complete functionality. The K2 is quite rich in capability in some areas and in
other areas just a little short. For instance the audio gain control is not software controllable. A comprimise
here is to feed the audio into your PC's line jack on its sound card and control the audio through the PC. Our early
glances at the DSP unit given us stars in our eyes over the potential set of DSP commands that are possible. Oh well
back to the more mundane API work we must complete. More news as we go!!
4/08/2004 by Mike WA1EYP
K2 Users upgrade your firmware!! We are working on our initial panel and are targeting it
against the firmware version 2.04 at the minimal release that we shall support. Our panel's operation will
be in native K2 mode, not a Kenwood look alike! It will operate in the Extended command
set mode. Simultaneous use of both the front panel and the software control panel will NOT be support. So you will have to decide one way or the other but not both.
Initial releases will not support all the commands. However free updates will be released periodically
to our licensed users.
As was the case with our Pegasus program, Licensed users will be able to participate in our pre-release
program and get updates before they are released. Along with their testing efforts they will be entitled
to participate in the development of the K2 software panel with their valuable suggestions many of which
were incorporated in our panels for the Pegasus, Jupiter, Rx350 and Rx320.
We are not the "supermarket" of software control panels just a small set of high quality ones primarily
designed for and by our user base.
5/02/2004 by Mike WA1EYP
Well here is one of the first snapshots!! Slowly getting the functions laid out.
There may be less filters depending how well I can set them up. Currently our main system clock and control loop if functioning without a hitch. The deadman is running. we are getting S-meter signal reports live. I can tune the radio, switch bands and perform some limited extended K2 functions. We are well on our way with great radio.
5/06/2004 by Mike WA1EYP
I have some new functions going; PRE, ATT,Spectrum Sweep, some MODES, and in the middle of SPLIT (VFOA and VFOB access and configuring on the fly for the RXvfo and the TXvfo). Trying to finish the extended commands as they do not require the menu steps. This weekend we shall try to get the structure working that suports the menu commands that require a query then set, query then set, etc. now that I have the main control loop working with a subloop ququeing system.
7/03/2004 by Rich KD0ZV
We just took delivery of our second K2 for the testing lab. As Mike continues working on the extendedcommand sets for the K2 I will be starting to build the second test unit. So far our preliminary software looks promising and we have high hopes of this being the utlimate black box.
8/30/2004 by Rich KD0ZV
I started building the second K2 about two weeks ago. Just finished it yesterday including the SSB, NB, DSP, 160 and 100PA options. The build went great but ended up with a bad DSP unit which is being replaced. I also tested my "stupidity" and installed an IC on the 100 watt PA backwards and IGNORED Elecrafts instruction to check orientation and heard it crackle :( So waiting for a replacement IC.
So tonight fired up the new K2 using our latest BETA software and thought I would update the site and post a screenshot of the program. Mike is currently working on secondary advanced menus as well as an advanced DSP interface. All is good :)
9/14/2004 by Rich KD0ZV
We have "most" of the primary functions working on the radio and it is working quite well. Well underway is the secondary menus and DSP functionality. We are struggling with timing issues as the K2 just does not respond well to traffic on the serial line. Our intention is to query the radio and write all the configured filter widths to the FL buttons,etc. This is working great but slow. The DSP module although fast in itself is really choked by the K2's auxbus and MCU. It currently is responding slow to our commands and we have seen response times in the 600ms range. Although this is much faster than clicking through the K2 menus it seems "snail slow" when using our zippy mouse and software.
We are currently working on ways to reduce these delays. I must say, the DSP is much nicer to use via software than it is using the hardware interface. If we were not building this for production K2's I probably would mount the DSP remotely and use software to control it via a serial cable.
9/25/2004 by Rich KD0ZV 
Last week we took shipment on a couple automatic antenna tuners and completed building them today. For those of you looking for another project we highly recommend this one. It took about 7 hours to build each tuner. Of course after just finishing a couple K2's and a couple of their small kits we are starting to understand their language. I would say that pretty much anyone without any Elecraft experience can build one of these in a weekend. Since the K2 project follows our TenTec software packages we have had the opportunity to use both TenTec's original LDG tuner and the latest LDG memory tuner. We must say this tuner out tunes both of them hands down. Not only is it faster, but always yields a better tune. This is the first ATU I have had in the shack that would tune my 160 antenna and it did it effortlessly.
OK, back to the main topic :)
The software is really coming around. As we have mentioned previously, we have had some timing issues talking to the radio and have had to slow some commands down. Anything that goes across the AUXBUS is extremely slow. This includes DSP and ATU commands. Currently we are running everything as fast as we can and slowing down just the commands that are slow so we don't degrade the performance of the entire radio. As you can see from the image on the left (click for a larger image) we have added quite a few buttons and features. The K2 is fairly robust in its command set and we are trying to leverage everything available. We have a new learn command that goes through and reads all your filter information including AF and DSP and writes them to the software for every mode. So after configuring your filters the software updates automatically with this new information.
I would say we should have our basic package ready for release in about 30 days. Check back. 73.
11/27/2004 by Mike WA1EYP
Well we have been busily testing the K2 and for the CW operators we have implemented a CW Memory Keyboard. It has
100 memories, repeat, contest confirmation numbering, Rx notes buffer, Tx and Repeat Indicators, special characters
and other features.

|