Fool Resistant Automated Band Switching: A Simple Design Project Using “Old School” Techniques

This is one of a series of “Notes” I published on Facebook. Since Facebook has discontinued the Notes feature, I am publishing that series here on my blog.

The first week of Technician license classes I tell students amateur radio can be a lifelong journey of learning. For some it is a means to an end, be it providing public service or for personal safety while enjoying other recreational activities. For others it is a hobby but not necessarily something that leads to continuous learning or exploration. Then there are those like me. Maybe I have the heart of an explorer. I am quick to dive into exploring a new band or mode, especially if it is challenging. I appreciate and enjoy learning how to do things.

I am by no means an electronics expert. I am in awe of those who understand microprocessor and logic circuits and those who handle RF and microwave design with ease. I wish I understood and felt confident with those things but I have come a long way from the kid whose first project, a key click filter with two inductors and two capacitors, went up in smoke. Today I am able to solve some design problems on my own and occasionally turn junk into something truly useful. As long as my brain will continue to absorb new ideas I will continue to learn about radio and electronics. Sometimes the things I design use older technology because I understand it. Sometimes it is because that is the least costly way to do it. Sometimes both, as is the case with my current project.

I don’t learn well from books or any form of text. I learn better in a classroom, from videos, or best of all from hands on experience. There has been a lot of magic smoke released from electronics here over the years and I have learned much from doing things the wrong way! Many people have contributed to my ongoing education through the years, but Steve K0XP (formerly KO0U) stands out as the greatest mentor I ever had. Through daily email contact in the late 1990s and into the early 2000s, he tutored me on a wide range of subjects involving solid state analog and basic RF design. Steve explained one thing at a time, patiently repeating and rewording until it was absorbed. I have been thinking about that a lot on my recent design project, as it involves a basic concept I finally “got” through Steve’s efforts: using transistors as switches. The drawers full of transistors from which I pulled stock for this were part of an incredible gift Steve found for me.

I call my latest project FRABS (Fool Resistant Automated Band Switching). Everything here needs to be fool resistant. Otherwise the magic smoke gets let out of stuff. This project would be too simplistic for some, too advanced for others. It is within my comfort zone. I admit the design concept using transistor switches is yesterday’s technology. Some would use opto-isolators, some a microprocessor based design. I use transistors because I understand them, I have drawers full of them and they get the job done.

FRABS is almost harder to explain than to build. Last year I started getting back into VHF and UHF using transverters. I have only one station transceiver, which now has to serve for HF and VHF/UHF. I have more than one amplifier sharing a common high voltage supply, and my antenna switching system is somewhat complex, involving both transmit and receive antennas for HF. If I were to forget to put one amplifier on standby before trying to use another, bad things would happen. I once forgot to disable the HF amplifier and dumped 1500 watts of RF into a hybrid transverter drive attenuator rated 250 watts. Zap! So much for that device. Clearly simple wasn’t going to do it here. Simple can be good, as in keep it simple, stupid. But simple isn’t always fool resistant.

What I needed was some means of automating the various tasks involved with switching bands: route RF to the proper places, enable one amplifier while ensuring that all others are disabled, etc. I needed the added protection of having it only available through the automated system, with no manual method that the fool might use instead to blow stuff up. After some discussion with Dave, AA6YQ, developer of the extraordinary DXLab suite of software I have running in the shack 24/7, transverter support in Commander (DXLab’s transceiver control component) was extended to include all bands I planned on adding. Since Commander understands transverters, can control the radio through its CAT command set, and can control external devices through a parallel port I had the method of control. All I needed was to add a parallel port to my PC and to come up with an interface to take signals from that port and control the station switching.

The project started with a concept drawing. I have no time or patience for computer drafting, so I grabbed a piece of paper and started scribbling. This, with its somewhat cryptic notations, is what resulted. It gave me a clear overview of what I was trying to accomplish.

Fortunately I had a vacant PCIe slot in the PC. After some research, reading reviews, etc. I settled on a Rosewill RC-302E parallel port adapter. USB parallel adapters do not work in this application! They are printer drivers and do not provide low level access for “bit twiddling” that is necessary for this application. I was concerned about how much current could safely be drawn from the Rosewill since not all these adapters are created equal and none are designed to source current. I was going to need a couple of milliamps to control my interface. Testing revealed that this one can provide more than three milliamps with virtually no voltage drop. Very good! It took about 10 minutes to set up custom band switch buttons in Commander.

I didn’t need to use any sort of decoder or digital to analog device. The parallel port has 8 data bits which can be controlled by Commander. I needed only seven “states” for this project: HF, 50, 144, 222, 432, 903, 1296. That meant that I could use a single data bit for each state by having Commander write the appropriate value to the port to make a single bit (pin) go high. I used the following seven values: 1, 2, 4, 8, 16, 32, 64. A value of 1 causes pin 2 of the DB25 port to go high, or about 3 volts, while all others remain low, zero volts. A value of 2 causes pin 3 to high, 4 causes pin 4 to go high, and so on. Nothing could be simpler than that. I have the most significant bit, pin 9 (128) left over for some possible future use. I may end up using it in some way for our new 2200 and 630 meter bands.

I had already decided this was going to be a low budget project, making use of parts I already had. I was going to need both NPN and PNP transistors for the switching circuits. I had drawers full of PN2222, PN3904 (NPN), PN2907 and PN3906 (PNP) transistors. A quick look at the relevant data sheets indicated the PN2222 and PN2907 would be the better choice for this project. I got down to business working out the details of the hardware interface.

For HF (160-10 meters) and 6 meters, there wasn’t much to be done. Those bands are native to my station transceiver, the transverter select relays would all be de-energized for these bands, and I didn’t need sequencers. All that would be needed is a simple switch to enable the proper amplifier.

HF and 6 meters were easy, since those bands are native to my transceiver and very little switching would be needed. The higher bands (144, 222, 432, 903, and 1296 MHz) would require a bit more. There I would have to route RF to the proper transverter, one path for transmit, another for receive. I would also need to use a sequencer for these bands since fast (vacuum) relays become problematic or completely impractical and there would eventually be complications such as tower mounted receive preamplifiers. This means things need to be switched in a specific order with time delays when going from receive to transmit and vice versa. T/R sequencers are the usual way of doing that, and since a kit is readily available for $20 I wasn’t going to design my own. After evaluating what would be involved with using a single T/R sequencer on multiple bands, I opted to use one for each band. The switching would be simplified and I could adjust the step time delays independently for each band if I needed to.

The 2 meter (144 MHz) switching circuit. This same circuit is dupliated four more times for 222, 432, 903, and 1296 MHz. It’s a good thing I had those drawers full of transistors!

I was also going to need a drive attenuator for the transverters. They require a few milliwatts of 28 MHz drive. Putting the full 100 watts from the transceiver into them would surely let out some of the magic smoke! I was fretting about not having a needed resistor when some tutoring obtained through one of the VHF discussion forums reminded me of something Steve had taught me years earlier: the capacitive voltage divider. Straight out of the junk box an adjustable drive attenuator was built. The final implementation involves having the band select buttons in Commander set the transceiver power output to 10 watts and using 30 dB attenuation to drop that to 10 milliwatts. Perfect.

Circuit for the drive attenuator. C1 allows this to be adjustable from 27 to more than 50 dB attenuation. I’ve set it for 30 dB in my application.

FRABS isn’t completely finished at this point. It is up and running to the point of proof of concept. I have been using it to toggle between HF and 2 meters for the past few days and it is doing what I designed it to do. There were a couple of glitches on this mission. The smoke came out of an ancient transformer I used in the FRABS power supply, but digging a little deeper my junk box provided a more modern unit that turns out to be better suited anyway. I also had a brain fart and forgot to include base current limiting resistors for the PNP transistors in the first draft design. Poof! Oops! I just wasn’t thinking. I know they are required, and I knew instantly what I had forgotten when the first test failed. What is not shown on the schematics is a liberal sprinkling of bypass capacitors to keep any stray RF out of semiconductor junctions where it could cause all sorts of mayhem.

Inside the transverter drive attenuator. I liberated the type N and BNC jack and cable assemblies from an old Motorola MICOR UHF antenna network.

Thge assembled transverter drive attenuator. The heat sink was also liberated from an old Motorola MICOR UHF antenna network.

Inside the partially complete FRABS control interface during testing. There is one more large board to be added (it wil stack on top of the one at the upper right) and four more of the small green baords (T/R sequencers), not to mention a lot more wires.

Bottom view of the FRABS control interface during early testing. Eventually there will be cables plugged into most of those connectors.

The relays used to toggle the station from HF/6m to the higher bands are shown here mounted on a rack panel. It’s a bit messy in here. Cables going here, there, everywhere.

The transverter drive attenuator (left) and transverter select relays (right) mounted on the inside of a a multi-function rack panel. The wiring across the top of this panel is part of 12V DC power distribution to network components (unrelated to FRABS). Cables will be color coded by band. Here, white heat shrink tubing is used on cables for 144 MHz.

The 2 meter gear (minus that unrelated thing at the lower left) currently sits atop the rack mounted station PC.

This project is neither pretty nor elegant, but it does make station operation much simpler and more enjoyable. It also helps pave the way for adding bands with a minimum of fuss.

Leave a Reply