Andy attended our Club Meeting on Tuesday, May 4th to talk with us about using Linux and various HAM Radio applications which run on it. The page mentioned above contains links to Andy’s Linux HAM Radio files. You can see a copy of Andy’s presentation here.
Get the new CORMAC CQ100 transceiver! Now you can make solid QSOs on Phone or CW all around the world. I already worked several countries during the free 30-day trial period. I understand that is you do not extend your membership, the transceiver is always available on Sundays!
Ira (KC1EMJ) and I helped set up the IC-7300 80, 15, and 10m CW station. We were short a CW operator. Having my license for only 8 months with over 1000 SSB QSOs and zero CW contacts in the log I wasn’t a CW op but the station was available and the field day clock was running. My ability to copy is improving but very limited. On the run up to Field Day I had some experience with N1MM, Fred (AB1OC) added a WinKeyer and loaded his macros, and I had CW Skimmer already installed on the laptop. What more could I need?
It only took a few minutes to get the ICOM IC-7300 and CW Skimmer setup for reliable decoding. It was exciting when the decoder finally started displaying “CQ CQ FD DE CALL-SIGN” for each station I tuned in. I have included a simulated display showing a decoded message. I was in business! Or so I thought.
With Search and Pounce selected, the WinKeyer was setup so the laptop keyboard’s F1 key was QRL?, F2 the exchange (class and section), F3 TU for “thank you” and so on through F9. I assumed one would start with F1, and progress to F2, and then F3, what is QRL anyway? After decoding a “CQ FD call-sign” and entering the call into N1MM I hit F1, and then decoded a “don’t say you do not hear me”.
I moved up the band and tried it again with the same result. It was great I was making contacts, but not so great they was throwing bricks. A quick check with Mike (K1WVO) I found QRL is “Are you busy?”. As a Phone op I had never used QRL, we just ask “is the frequency is in use”. Soon I was responding with N1FD instead of QRL? The Caller returned his class and section.
With his call entered into N1MM it was easy to send his call sign with a tap of the F5 key, F2 for 7A and NH, and after his TU I would send a TU and QSY to the next station and repeat the process. It had transitioned from real exciting (meaning a bit stressful) to real fun fairly quickly. The lesson I learned is to take the time to understand the message stored behind each “F” key even when time is short and the contest or Field Day has started.
The experienced CW operators were using the same process that I was using with WinKeyer, N1MM, and the keyboard. The one difference is they were decoding CW with their ears, and not a decoder.
It wasn’t long before I ran out of new stations to work. I switched from searching for stations in “Search and Pounce” mode to “Run” meaning I stayed on one frequency and called CQ. The F1 key became CQ. It wasn’t long before I had a short run of 5-6 QSOs one right after another, but it quickly came to an end. There are few targets on 80m in the early evening and I worked them all. I was headed home just as the band was heating up at midnight… Next year I’ll take the midnight to daylight shift and plan to copy code with my ears…and not a decoder.
With a little practice, this form of operating is effective for contests and Field Day where the exchange is simple. The high rate you can add new stations to the log definitely makes it fun.
This article discusses some work on designing a matching network to make antennas match well (low VSWR) across the entire ham band. This will be a described in more detail at the September Tech Night.
Antennas have an impedance (or match) that varies with frequency. Transmitters want to see a matched antenna with an impedance of 50 ohms. The antenna has the best match at one frequency and the match gets worse as the operating frequency changes.
Some bands and antennas are more challenging to match than others. Shortened or loaded antennas have a narrow range of match frequencies. The 75/80 meter band has a wide bandwidth in term of percentage.
Here’s a plot of the SWR for my 40 Meter Dipole. It’s a good match at 7.000 MHz and degrades to about 2:1 at 7.100 MHz. Obviously, this is not optimized.
Modern radios have built in automatically adaptive matching networks make the radios work over a wider bandwidth, but networks are lossy and reduce transmitted power.
A manual antenna tuner has a lot lower loss than the built in tuner, but it requires manual adjustment. In fact, the extra tuned circuits generally act to make the antenna have even less bandwidth.
The QUCS RF circuit simulation program has the ability to model SWR, bandwidth, matching networks based on data about antenna performance. The antenna data can come from either an antenna modeling program such as 4NEC2 or EZNEC. Or the data can come from a measurement made by a good antenna analyzer.
QUCS also has a built in optimizer. It has the ability to try hundreds of circuit values and home in on an optimal design.
The optimizer setup needs a definition of “optimal”. For the case of a broadband antenna, “The worst case SWR anywhere in the ham band shall be as low as possible”. In the terms that QUCS understands, “minimize the maximum SWR over the frequency range 7.00 to 7.35 MHz.
Here is the result from running the optimizer on the data for my mistuned 40M dipole. QUCS has designed a broadband matching network that can achieve less than 1.5:1 SWR over the whole band.
QUCS achieved this by varying the components of a filter network. I drew a general filter network and let QUCS tune the component values. This network is designed with coaxial stubs.
The model of the antenna is stored as a file in the X1 file component. Line7 is a 30-meter coax feedline. The actual matching network consists of Line 1, 2, 3, 4. Each line is 50 Ohm coax. Line 1 and 3 are configured as open stubs. The line lengths predicted by the model are…
Line 1: 7.75 meters
LIne 2: 4.47 meters
Line 3: 8.49 meters
Line 5: 8.03 meters
Here’s another example. 160-meter antennas are often implemented as shortened loaded verticals. The loading makes the match very narrow-band. The red curve in the plot below shows a top loaded 160 meter vertical. It only covers a fraction of the band.
The blue curve shows the result of an optimization run that selected the values for a 7 component matching network. It achieves about 1.7:1 across the whole band. This network uses capacitors and inductors because coax stubs would be very long on 160 meters.
The component values for this network…
C1: 3450 pF
L1: 3.954 uH
C2: 3978 pF
L2: 6.951 uH
C3: 6156 pF
L3: 2.831 uH
C4: 4778 pF
I have not built any of these networks to see how they work in practice. The 160-meter network has some extreme values and it is probably very touchy to get right. Building that network to handle Tx power will require vacuum variable capacitors in parallel with high quality stable fixed value capacitors. But, the 160 network doesn’t really need 7 components. Put in one less stage of L-C and the ripple across the passband goes up a bit.
QUCS is a great RF circuit simulator. This shows that it can work with data from an antenna model or analyzer and can optimize matching networks to create a broadband antenna.