Desktop 12 VDC (13.8) Distribution

I go through small 13.8 VDC linear power supplies like they're candy. I keep several around since they're vital to my station. They run my antenna switching system, stack switches, antenna switches, Beverage antennas, BPF (band pass filter), rotator controllers, and much more. They are prone to failure due to lightning and other mishaps.

My problem was ease of connection and disconnection, and cable management. Having a multitude of peripherals makes this difficult. I had to manually disconnect multiple wires on the +ve and -ve power supply terminals while preventing shorts and loose wires. 

It is made more difficult since the power supplies are located on the sub-shelf so that I have to squat down behind or underneath the operating desk which inevitably causes unintended damage, such as to my head. 

Wires inevitably go astray when connecting several to one screw terminal. Some of the DC cables end up snaking and tangling amongst other cables, making them difficult to route and connect. I can never remember which peripheral is connected to which power supply, and it can change frequently. Tracing cables is time consuming. 

I have to do this often since my station is constantly evolving. There are partially built projects on the operating desk while I'm building and testing them. In short, it's a maddening mess. I finally decided to do something about it.

My solution is not novel, or even particularly interesting. But it's necessary, and that makes it worth a brief description. If you're looking for excitement, this article isn't it.

Standards

There are many connectors, both standard and proprietary. We are not talking about high current or demanding applications so there is no need to do anything fancy. My solution is that of all my DC powered peripheral equipment: 2.5 x 5.5 mm DC power connectors. There are similar connectors in different sizes but these are what I've chosen.

They are commonly found in smaller devices on our desktops. They're cheap and effective. However, do take care that the ones you order online are not so cheap that they have unwelcome problems. Some require unreasonably large chassis holes, have unexpected threads, might need the male connector to be jammed in full depth and so forth. Paying double the price for good connectors is still cheap.

Interconnection cables are also cheap to buy if you don't want to build your own. For this project you need male connectors on both ends of the cable. The only bare wire connection is to the power supply terminals.

Power supply choice

I recommend against using the DC supply powering your rigs. Although they may have the headroom to deal with a few extra amps there are other considerations.

In my station, many of my peripherals support antenna switching systems deployed in the field, either on the towers or on the ground. These systems are at risk of lightning damage. In a pinch I once connected my peripherals to my 35A Astron linear power supply and lightning fried several components. I eventually fixed it but I have never made that mistake again. They're expensive to replace and can be difficult to diagnose and repair. 

Small linear supplies of 3 to 4 amps are flea market bargains. Buy a couple of them so that you always have a spare. If you are unfortunate enough to experience lightning damage they are cheap to replace and perhaps not worth the trouble of repairing them. Lightning protection is worth considering, such as GDT (gas discharge tubes) which are inexpensive and effective.

Distribution

I don't want to see the DC power supply and cables. They're necessary and nothing more. However they must be accessible for station changes and periodic reconfiguration for multi-op contests. That requires proximity of the connections to each station and to equipment that is never moved. Long cables should be avoided since they can create a mess and are prone to RFI.

I built two connector strips that are wired together. They are constructed from ¾" aluminum angle with holes for the connectors and strain relief. Holes on the other face are for #6 wood screws to affix the strips to the underside of the desk -- ensure that connectors and wires don't block hole access. Most stations only need one strip, but I wanted two so that the cable runs are kept short when I set up for multi-op contests. 

There are numerous other variations that might be preferred. For example, one power supply per strip, or strip interconnection with a DC power cable (occupies one connector on each strip). It is also possible to add RFI and lightning protection on the distribution strips if it isn't present elsewhere.

Installation 

The only important question is where to place the connector strips. I placed mine under the backside of the desktop. However the power bars and a few other items had to be temporarily removed to make space for a screwdriver. One final electrical test and I was done.

I now need to be more serious about attacking the rest of the spaghetti cabling. I always start with good intentions so that every time I first set up a desk it's pretty clean. It gradually deteriorates with every station change or shuffling of equipment.

That was a boring project, wasn't it? Keep in mind that a worthwhile project does not have to be exciting. More like a chore than a project, but still important to do.

I'll return to more interesting topics after I deal with non-radio matters that are currently occupying my time.

2-Man M/2 in ARRL DX CW

The M/2 category is many contests allows two simultaneous signals on two bands, without restriction on who they can each contact. Thus, you can have both stations running, and not restricted to hunting multipliers. There can be additional stations operating S & P on the same or different bands provided they operated within the constraints of maximum band changes per hour (typically between 6 and 10) and never having two simultaneous signals on the same band. 

M/2 is a popular category since it keeps more operators in a multi-op entry active and involved without the burden of competing in the M/M (multi-multi) category. I hosted M/2 teams for the CQ WW CW and SSB contests in 2023 and I operated M/2 at VE3JM at least once. The plan was to do so again for last week's ARRL DX CW contest. There is friendly rivalry with VA2WA about 200 km east of us. Last year we surpassed them, but it was close.

Our plan fell apart a few days before the contest. One team member fell ill. Then we were walloped by two successive snow storms, one a few days before the contest and another forecast for the weekend. About 70 cm fell in the vicinity of Vlad's contest station. One of the operators wouldn't abandon his wife to deal with the second storm. especially since he had no idea when he could return. Our last minute attempt to recruit others was unsuccessful.

It came down to just Vlad and me. I nearly abandoned because I was busy dealing with a family matter. Since that required my presence in a town only 70 km from Vlad's station, before and after the contest weekend, I decided there was no reason not to participate in the contest.

I had a surreal drive down Canada's busiest highway (401) that was still partially snow covered from the first storm. The ditches and median were littered with countless car and truck wrecks. There were even more during the drive home after the contest from the second and worse storm. 

When I got there late afternoon on Friday we discussed the pros and cons of entering M/S or M/2. We read through the rules, considered our expected stamina and the effects of the geomagnetic storm. We chose M/2 because it would be more fun and both of us were accustomed to operating 48 hour contests. We intended to do our best without expecting to be very competitive among other M/2 entries. 

Vlad set us up to participate in the online scoreboard. Then we shared a pizza and got down to business when the clock struck 0000Z.

Note: The rest of this article is primarily of interest to contesters. Others might find it less so. There are no pictures since I was too busy operating. Vlad has updated his QRZ.com page which is full of pictures.

Propagation

A K-index between 3 and 5 for most of the contest did not bode well for the score. Everyone suffered, though not equally. As is typically the case for geomagnetic storms, those of us in the auroral zone feel it more. Indeed, anyone whose signals must pass through the zone feel it. It's just that we see attenuated signal on most paths, including the all important one to Europe.

Those nearby in W1/2/3 and VE1/VO1 tend to fare better since they can "slide" under (south) of the zone. That said, almost everyone was affected. When we're all in the same boat, all scores are lower. Too often we would simply run out of propagation. Openings started later and closed sooner. Signals were weaker and thus many small stations didn't get into our log. Either we couldn't copy them or they simply gave up and turned off the radio. Stations further north (TF, OY, OH, etc.) or in the southern auroral zone (Antarctica) were exceptionally weak.

160 meters is less affected since signals tend to stay below most of the disturbed layers of the ionosphere. Even so it was surprising to see country totals over 50 for many of the big guns. We managed 49, which is pretty good since with just the two of us we couldn't check conditions frequently or make timely band changes. DX openings didn't last long but there were enough of them to make a difference.

When propagation resulted in especially low rates, I might take a brief break to eat, make coffee or check my email. Just 10 minutes is enough to feel refreshed and return to the grind. Towards the end of the contest the conditions improved and we were able to finish strong. 

Station set up

The two run stations were Flex with a PGXL amp, and a K3 with an AL1200. Since there were just two of us we each adopted a station that we used throughout the contest. I am not very familiar with Flex equipment so I took the latter. The third station was for in-band, which could not be used.

We used DXLog which, frankly, is not my favourite. It works quite well but with an unfamiliar interface, the default colours and fonts are atrocious and the function keys never seemed to do what I expected. Other times DXLog would get into a peculiar state (Vlad told me to hit ESC a few times) or a wrong key press would wipe a half completed QSO. Then I'd have to appear an idiot by requesting a repeat of the call sign, unless I was lucky enough to remember it. 

There are other things I could say about DXLog but I'll stop there. It isn't really fair to complain when many problems may be the operator (me) unacquainted with the software. For example, with N1MM I am used to controlling RIT with the up/down arrow keys. Those are used to navigate the log in DXlog. I had to use the radio; if there's a keyboard method I don't know it. The K3 display is quite small so that the RIT value (+0.02 etc.) only display when the knob is turned. I had to manually reset it after every use since I could not reliably remember its value.

Another quirk of the K3 that I don't like is the receive audio quality. Some have raved about its crispness. That is not my experience. When I get several near zero beat callers, it's sounds like someone scratching a chalkboard with their fingernails. Also pretty bad is that the AGC doesn't compress the audio as much as I'd like. If I turn up the gain to copy a weak signal, when the next caller is strong the audio is painfully loud. The K3 was wonderful when it was new but today we have better rigs.

I am comfortable with a manually tuned amplifier, so it was easy to tune it quickly after band changes. We didn't do that often except to chase multipliers on other bands if we had hourly band changes remaining.

Antennas were selected within DXLog and antennas could be turned manually or via digital entry of the bearing in the call sign field of DXLog. It worked well. One day I'd like to do the same for my customer antenna selection software. That won't happen quickly unless I get impatient and use a commercial product. 

The two tri-band yagis have VA6AM triplexers so we could both use them. However, we didn't always agree on the direction! There were enough antennas that we negotiated solution that satisfied us.

As for the antennas, Vlad's station is different but similar in capability to mine. There are stacks on the high bands, either 2× or 3× 4/5-element stacks to Europe and rotatable top yagis, plus a long boom 7-element on 10. On 40 meters there is a rotatable W6NL Moxon at 140' and an XM240 at ~100' fixed on Europe. For 80 meters there is a 4-square and a low inverted vee that is not useful in DX contests. Like me he has a shunt fed big tower on 160. He added wire parasitic elements that provide a small gain to the northeast and southwest. There are two reversible Beverages for receive.

That is a good complement of antennas. It is not equal to the biggest of big gun stations but well suited to the challenges of DX contests. Vlad turns in exemplary scores in single op contests. He's a better operator than I'll ever be.

Operating

We ran as much as we could. When the pile ups thinned we would click on stations in the band map or available mults & Q's window. Call them if they're there and then a one-key shortcut to return to the run frequency to resume CQing. I lost my run frequency only a handful of times, but you have to be quick.

Vlad and I frequently discussed openings, band changes, breaks, station passing, antenna use and direction. We could do this pretty effectively even while running pile ups. It may seem odd to watch us logging QSOs and talking at the same time.

Regular readers may recall that I often complain about errors by others copying my call. All those dits cause confusion. The most common error is VE3UN. I have since had that error removed from SCP (super check partial) but it doesn't seem to help all that much -- casual operators don't use SCP or forget to update the database, while others don't pay close attention. I had the thought that with so many dashes that VE3JM would be less prone to copy errors. Wrong!

Despite keeping my sending speed to 36 wpm and lower most of the time, errors were common. Typical examples were VE3MM (a fellow contester), VE3OM (dropped dit), VE3OO (?), VE2JM (VE2/VE3 confusion is common), and a few rarer mistakes. That's just how it goes -- there is no good solution. I suppose we should be happy that so many hams are active on CW, even if only on contest weekends. I  wonder whether they pay attention to the LCR (UBN) reports they receive.

Tangling in the pile ups highlighted a difference in technique among the big guns. Most stations click on the spot and toss in their call, big guns and little pistols alike. You might think that the big gun would win every time, yet they do not. Brute force only gets you so far. The sum total of the other callers may be as strong as or stronger than the big gun, or the QRM makes it impossible for the DX operator to separate one caller from the bedlam. 

I would evaluate the circumstances when my initial "brute force" call failed. I might recognize the call signs of other big (and bigger) guns or that the successful callers were in locations with more favourable paths. There are a couple of techniques to deal with the pile up. One is to call well off frequency, around 100 Hz, to make your signal uniquely distinguishable. The other is to call twice, hoping that the second call has less competition than the first. Don't do that if the DX operator is responding quickly because the added QRM slows everyone down! 

If neither works quickly there is the third option: go away and try again later. There are other stations to work and you don't want your rate to suffer. A few minutes later, after the big guns have moved on, go back and one call puts the mult in the log. These are lessons many little pistols have learned and that some big guns have forgetten.

Breaks

Yes, even with a great station you can run out of stations to work. As already mentioned, this is not the most popular of contests. The reason is that for many stations, North America is far away and difficult to work. Contests where everybody can work everybody will always draw more participants. That said, over 5600 contacts during a weekend with challenging propagation is not so bad.

Even with 5600 contacts spread over ~40 hours operating time is not as busy as it sounds. Consider that there were two run stations. That's like 80 hours for a single op or M/S entry. Divide 5600 by 80 and the average rate for each operator is only about one QSO per minute. The average is itself misleading since during the morning high band openings to Europe the rate can easily surpass 3 QSO per minutes per operator. The rate at other times can be painfully slow.

So, what to do? We hunted for stations and multipliers as much as we could, trying to keep an eye (or two) on the AMQ window (available multipliers and QSOs). Many of those required careful tactics and operator communication to pick suitable antennas, turn them if necessary and decide which station is best to chase a station based on the band change quota. For whatever reason, Vlad made many more band changes than I did, mostly because he was considerate about not interrupting my runs.

Yeet there were still dead times. We both took sleep breaks soon after European sunrise (after ~2 AM local time) when the low band rates plummeted and the high bands turned up little. During the afternoons, a quick break might be taken to get refreshed by doing something different, be it to check email, progress of the snow storm or to prepare and deliver food and drink to the other operator. One way or another we kept busy. But we took no sleep breaks other than a few hours overnight. We got up before sunrise to chase mults on the low bands.

Perhaps what was most frustrating were the poor rates when the propagation was quite good. You know it's good when the DX stations are loud but there are so few of them. Casual and only moderately serious contesters keep to a more human schedule for contests like ARRL DX. Tthat means regular sleep, meals and attending to their families and other responsibilities. They're in bed at 3 AM local time.

Taking a break at those times is not sacrilege. What you find is that the stations will still be there after the breaks. They want to work you and they will. Usually there would be a flurry of activity after breaks when we were newly spotted by the CW skimmers and by stations that had just "rotated" onto the band. After 5 or 10 minutes it was back to the grind.

Problems

Other than the problem of not enough operators, we had a few others. Some I've already touched on in this article, which I'll elaborate upon. They can be surprising.

Snow is precipitation, and therefore can cause precipitation static. Especially a major snowstorm accompanied by high winds. The upper yagis are most affected. The highest antennas on all bands from 40 to 10 meters were periodically unusable once the storm intensified on Saturday and continued through Sunday. We had to choose lower yagis. That wasn't too dreadful since there were fixed stacks to Europe on the 20, 15 and 10. It was unfortunate that on 40 meters we had to use the less effective XM240 rather than the Moxon. The Moxon is much smaller than my 3-element yagi but I was impressed by how well it worked.

As the saying goes: you can't work them if you can't hear them. It believe it is possible to select different receive and transmit antenna via the DXLog interface but for some reason we didn't do that. Perhaps that was an oversight on our part. I don't know the setup well enough to know for sure and I never thought to ask Vlad.

There was occasional severe interference between the stations. I don't know if it the Flex station was responsible or if it was the K3 receiver. Both have a pretty good reputation. The towers at Vlad's station are closer together than in mine, which might have had an impact. I know the VA6AM high power BPF work well, as do the triplexers. When one of us accidentally chose a run frequency half that of the other station it was easy enough to move. It's a common problem on CW; for example, 7.002 MHz and 14.004 MHz. It happens far less often on SSB since the phone band segments are mostly not harmonically related.

We didn't self spot, which is permitted in ARRL contests, even for unassisted entries. Many casual operators solely rely on human spots so this might have helped raise our score.

At the height of the snowstorm the lights flickered quite often. We crossed our fingers and hoped that the power would not go it. When it happened we'd exchange worried looks. Luckily the electrical supply held up. That was one big problem we escaped! You just never know with distribution networks in low density farm country, even though the grid is well maintained.

Results

Vlad had the online scoreboard on his screen but I didn't bother. The only page I had open was the space weather dashboard. When the going was slow and I was curious I'd ask him how we were doing. Mostly we were tracking our usual close competitors at VA2WA. With more operators they kept ahead of us though the margin was small. 

In the end we had about the same number of multipliers as VA2WA and they had 4% more QSOs. That's not bad for two operator M/2. Had activity been higher or the propagation better, their advantage would have been greater. Maybe I shouldn't complain so much about the poor propagation since it may have helped us. It'll be interesting to see whether errors change the comparison after log checking.

I pulled the following from Vlad's submission to 3830:

Call: VE3JM
Operator(s): VE3JM VE3VN
Station: VE3JM

Class: M/2 HP
QTH:
Operating Time (hrs): 40:50

Summary:
 Band  QSOs  Mults
-------------------
  160:  110    49
   80:  612    81
   40: 1156   105
   20: 1154   114
   15: 1445   117
   10: 1138   112
-------------------
Total: 5615   578  Total Score = 9,660,114

Again, complain as I might, we did pretty well. The multipliers were there and we worked almost all that we could. Mostly we lost out on those in Asia on paths through the auroral zone. DXCC on 4 bands is quite good, as is the large number of mults on 160 and 80. I didn't expect to work nearly so many on 160. Vlad worked most of those, even running Europeans at intervals when signal levels peaked.

Last year we scored over 13 million points, and other stations had similar results.

The storm continued after the contest ended so we both spent the night at the station. In the morning, Vlad dug us out when his neighbour had tractor trouble and couldn't help. Typical for Canada, the roads were quickly plowed and we were both on our way, Vlad heading west to Ottawa while I drove east to resume my familial obligations. 

When I arrived home in the evening I only had the time and energy to clear enough of the driveway to get my car off the road. Then I slept. A lot. The rest of my 100 meter long driveway had to wait for another day.

VA6AM Switchable 6-Band BPF

I have mentioned building and using these low power BPF (band pass filters) over the past several years. What I haven't done was discuss the BPF themselves. I didn't want to do so until VA6AM fully productized the BPF for sale. 

He's a busy guy, it is not high priority and so it may never happen. Since these are good products I decided to not wait any longer. I will do a deep dive into it in this article.

The 6-band BPF units are kits. That is, the individual filters are kits, the relay boards come bare, while the enclosure, connections and controls require fabrication by the kit builder. 

Pavel makes it easy to order the parts since he created a project on one of the major electronics distributors for ordering the parts. I doubled the quantities and ordered a few extras of some parts. I am not linking to the project since it may no longer be current. I built my units in 2021.

The kit is not the same as purchasing his low power BPF kits for the 6 contest bands: 160, 80, 40, 20, 15 and 10 meters. There are differences:

• PCB for the filters are narrower so that they fit into a not too large enclosure.
• Relay boards, including bypass relays, run on both sides of the BPF PCBs.
• The control board design is incomplete and not available. You have to develop your own control system, whether manual or automatic.
• There are interactions between BPF and via to the relay boards that the design addresses but had to be experimentally dealt with to optimize performance.

There are numerous commercial 6-band low power BPF products that I could have bought to avoid the challenge and difficulty of build these "kits". These are my reasons for not choosing from among them:

• Cost: High performance commercial units would have cost at least $2000 (for two). I spent about 35% of that to build my own, and they measure better than most. Although most of the saving was eaten by the purchase of the VNWA3 by DG8SAQ, I've used it for so many other projects so that it isn't a direct cost. An inexpensive NanoVNA can also do the job but not as well. I borrowed a friend's before taking the more expensive route.
• Learning: Everything from winding toroids, precise and accurate measurements with a VNA, filter tuning, designing control circuitry, and so much more. I love to learn new things and there was much to learn and skills to master.
• Accomplishment: Just as for towers and antennas, I like the challenge and accomplishment of completing a difficult task and seeing the results on the air. 

I understand that few contesters would make the same choice. They would rather assemble the station using commercial products and get on with the business of operating. Both are equally valid options. I did what I judged was best for me. I have no regrets, but it was a major undertaking.

What you get

The "kit" is comprised of two each of the BPF kits and sets of relay boards from VA6AM. You supply the parts for the relay boards, enclosures, connectors, hardware, wiring and control systems. 

Should Pavel decide to make this a product there will be revisions based on our experience with my prototypes. We both learned from the experience. Pavel was very generous with his time for the several months I spent on the project. He also shipped extra parts for me to experiment and improve performance.

Pavel had not completed design of the control board so I developed my own solution. My simple solution was probably better for me in any case. An 8-position rotary switch is used to manual selection one of the 6 BPF, BPF bypass and automatic band switching via my custom antenna selection software and hardware. 

Toroids

There are a lot of toroid coils in these BPF! There are 4 on each BPF for 160 through 15 meters, totalling 20, each with 2 interleaved windings, and double that for two units; solenoid coils are used on 10 meters. I've wound small toroids many times, however it is more difficult to do it well with heavy gauge stiff wire. This chore was probably the most difficult part of the project.

On the left is a toroid built by VA6AM. It's nearly perfect. My first attempt is on the right. That is not so good. Luckily the performance impact of ugly toroid windings is small. Perhaps the biggest risk is excess wire length due to loose winding which increases the inductance. I tried various winding techniques (Pavel made good suggestions) so that my toroids improved over time. 

There was no need to go back and redo the earliest ones since they could be adjusted to the required resonance. Indeed, toroid adjustment is, in all cases, done by compressing/expanding windings. It fussy work that is mystifying at first and soon becomes routine.

Construction and tuning

The input and output LC resonant circuits are tuned using a VNA or (less accurately) an antenna analyzer. The objective is to make both equal. A home built VNA probe is shown in use. I found that the 1 kΩ resistor at the end of the VNWA transmit port worked very well. It was suggested by Pavel.

Another common probe is a loosely coupled one turn coil passing through the toroid. It is important not to have the measuring system couple so strongly that circuit resonance is affected.

After each BPF is completed, both ends are connected to the VNA to measure port impedance and insertion loss. For the first BPF I temporarily soldered on BNC connectors (see earlier picture). Using wires is not recommended since they add inductive reactance that will not be present in the final product. The tuning will be incorrect and may be impossible to achieve. These filters are precision devices. The kits come with testing and tuning instructions.

Knowing little about filter design, there were curiosities during the tuning and measurement process. For example, I was convinced that the 40 meter BPF wasn't working properly because the lower band notch was on the high side of the desired pass band on 80 meters. Pavel explained that is expected since for Cauer filters that results in the best rejection across the 3.5 to 3.8 MHz segment typically used in contests. His explanation alleviated my concern, and in use it isn't an issue.

Another was the insertion loss on 15 and 10 meters. It was difficult to get the insertion loss below 0.5 db while also keeping the port SWR close to 1. Pavel tried to explain the reasons but I have to admit I still don't understand. 

It was to reduce the insertion loss that he switched to solenoid coils for the 10 meter BPF. He considered using solenoid coils for the 15 meter BPF but that increases the risk of interaction with the 10 meter BPF that also uses solenoid coils. 

The complete set of VNA plots can be found below. You can see a glitch on 10 meters for the 15 meter BPF. From additional testing per Pavel's direction it appears to be due to capacitive coupling via the relay boards, possibly between relay contacts and contact wiring. We did a test to ensure that is was not due to direct coupling to the solenoid coils on the 10 meter filter.

Visible in the picture of the 10 meter BPF at right are the back panel connectors and bypass circuit. There are UHF connectors for the input and output, 12 VDC power and a DB9 for automatic switching (not yet wired when the picture was taken). 

Pavel noted that the filters are symmetric so the connections can be swapped, however it is not unusual for one permutation to give slightly better performance. In that case it is best to label the connectors as radio and antenna (or amp) and always use them that way. The difference is negligible for my units so I didn't add labels.

Performance

Ordering of the BPF affects performance. 10 meters is closest to the connectors and 160 meters is at the other end of the enclosure. Although the relay board traces mimic a transmission line, they are not perfect and there are the deviations via the relays to consider. The primarily stray inductance is why 10 and 160 meter BPF are placed where they are. 

The BPF for the other bands are interspersed so that none is next to that of its second harmonic. For example, don't place 40 and 80 meter BPF next to each other.

Final tuning of every BPF must be done with the full unit assembled in the enclosure and the cover in place. It is also necessary that wiring of the control system (manual or automatic) is complete and functioning.

Another subtlety that affects performance is the soldering of wires from the BPF to the relay PCB. There is a ground connection via the metal stand offs to the enclosure bottom that should not be solely relied upon. One or (better) two short wires should bridge the ground strips at the PCB edges. 

The bypass circuit is simple enough, with relays and a short length of RG58. It isn't quite that simple since there is, again, stray inductance. The usual method to alleviate the problem is to place a shunt capacitor at one end of the coax. I experimented and got my best result with a 10 pf capacitor on one unit and (if I recall correctly) 8 pf on the other.

The VNA plots show the difference -- you may have to click on the picture to read the numbers. The scales are different but the marker data tell the story. It is easy to tell which plots are before and after the correction. 

The reason this works is that the shunt C and stray series L form an L-network that transforms away the stray inductance. That also changes R but only very slightly for the small amount of compensation required. The same technique is used in wide bandwidth T-bias and other circuits, and I used it to good effect in my stack switches.

The picture shows the shunt capacitor on the bypass circuit (top left). It also shows another on the 15 meter BPF (lower left). I found that I could slightly improve the insertion loss SWR with it. That was possible on 15 meters but not on 10. Since the explanation is a little esoteric and I don't fully understand it, I will say no more. 

Pavel was surprised that it worked and understood full well why it cannot be a general solution. It was an accidental yet successful experiment based on my measurements of R and X in this unique case.

I think I've given more than enough detail about how I built and tuned these units, indeed more than most readers want to know! I thought it worth taking the time since the details are quite fascinating to me. Designing filters with such excellent performance is well beyond my ability. All I can do is marvel at what is possible.

If you are in the majority of contesters that use commercial units or a set of high power BPF (VA6AM's are excellent as well), you might now understand more about what goes on "under the hood". Those who haven't read this far, well, I guess you'd rather not know.

The following plots for each band are for one of the units. The plots for the other are almost identical. That's a good indicator that small construction differences have little impact on performance. Careful kit builders should be able to achieve similar performance. As before, click on each picture (there are 3) if they are not readable in your browser.

4 years later

Despite my praise, it has to be said that these devices are not pretty. I could have painted the bare aluminum enclosures, added LED indicators, used a modern knob and added professional looking labels. I didn't see the point since I only touch the units twice in each contest: to switch it to automatic mode at the start back to bypass at the end -- they're in line full time. My preference (as it is for other contesters) is to get them of the operating desk. They are just clutter during contests, even attractive commercial products.

It seems a shame to do this after all the work put into this project. That's how it goes. I also don't need to see patch panels, rotators, antenna switches and all the rest of the station while I'm operating. The desktop looks tidier compared to the last time I showed a picture of it.

Low power BPF are not the highest performance solution to inter-station interference (SO2R, M/2, M/S, M/M). High power BPF are superior. However, that's a much more expensive solution and you can't easily carry them with you to other stations as a guest operator. Harmonic stubs are a good complement to low power BPF to deal with harmonics and spurious emissions from amplifiers. Construction, tuning and placement determine their performance.

Of course you don't need high power BPF if you operate low power or QRP. Low power BPF are sufficient in the majority of cases -- VA6AM low power BPF are good for 100 watts if the SWR is not excessive. I've done QRP contests without any BPF. You just need to avoid frequencies close to the other's station's harmonics.

In closing, I am very happy with these 6-band low power BPF kits. I've been using them for almost 4 years and I am not tempted to upgrade to high power BPF. It's a shame that they are not real products, and may never be. I was the first customer and it is entirely possible that I still have the only ones in active use.

VE3CRG (SK)

In January I lost one of my oldest friends, Brian VE3CRG. His passing was not unexpected since his medical challenges were accumulating and he was just shy of his 88th birthday. Nevertheless it is a sad reality. The family held a celebration of life a few weeks ago.

I won't make this article too long. It is my opportunity to share a few reminiscences of a fellow ham of an older generation, a person who was dear to me. Those who are active on the HF bands may have heard him as he rattled away on CW, often at speeds above 40 wpm. The CW sub-bands are a little quieter today.

We first met in 1980 when the multi-op contest team at VE3PCA was looking for CW operators. Brian had already been an avid CW DXer for more than 20 years, first licensed as MP4BCV in 1959 when deployed by the RAF to Bahrein. You can read more about him at his QRZ.com page, from which the above composite photo was retrieved.

I had been a ham for 8 years with a penchant for CW DXing and contests. Our 20 year age difference was no barrier. We paired up well since (as we often laughed) he liked to talk and I liked to listen -- an introvert and an extrovert. We did many road trips (Dayton etc.) where he did almost all the talking. We also both had careers in Ottawa's high tech industry -- mine had just begun when we first met -- and we shared an oddball sense of humour. No matter how many times he retold some stories they were still enjoyable.

He had an extraordinary exuberance for life, even recently when his prospects were dimming. When opportunities arose during his life, he jumped, whether it was marriage to Mollie (for 67 years!), moving to Canada, a business opportunity, or an invitation to visit hams elsewhere in the world. It was easy to convince him to pamper himself with some modern equipment, which he enjoyed immensely in his final days and months.

I had been working on towers for several years before meeting Brian, although my knowledge was modest and my safety practices abhorrent. When his family moved to a small acreage in the early 1980s it was an opportunity to learn and do. Being single, young and eager I stepped up to help make his antenna dreams come true. 

I was over there many weekends, even during winter, for a few years doing tower and antenna work. Mollie was always ready to seat me at the table to share meals with them and their teenage sons. For a single young man far from home those home cooked meals were really appreciated.

The 100' tower on the right was erected in the mid-1980s. (I took several photos of the prints in his pictures albums.) Since both of us came from very modest beginnings, we enjoyed getting the job done as cheaply as possible. The total spent for the tower, concrete, steel and antennas was about $1000 (not adjusted for inflation). 

A local ham, a retired welder, helped us fabricate parts needed to assemble and strengthen the old 80' windmill tower, and the 20' of tower we fit on top. The Hy-Gain yagis (204 and 155) were purchased used for a good price. The tower also supported wire antennas, including 2-element delta loops for 40 and 80 meters. He enjoyed operating this station for many years. 

The tower was still standing when the property was bought by a developer almost 15 years ago -- it had badly rusted and collapsed into the trees on then now vacant lot about 2 years ago. He took the smaller tower and hex beam to their new house some 20 km to the west, escaping the relentless expansion of urban sprawl. It's hard to believe that the busy residential thoroughfare used to be a quiet county road and that the neighbour's horses come up to wood rail fence to see what we were up to. That's all gone now.

There were 3 of us who did almost all of the work, with the third being Dave, now VE3KG. It was a lot of fun and we learned a lot. I have to say that I learned lessons about how not to build large guyed towers! But nobody was injured and I came away with knowledge and skills that would come in handy later, for helping others and more recently for myself. 

For example, instead of using hand tools I rented a pneumatic breaker to cut through the shale to plant my tower in 1985. Another was making a gin pole to lift 20' long 200 lb tower sections!

I still remember the exhilaration when I first stood atop the tower with nothing more than a steel mast impeding my view of the countryside. The reinforced top section was so narrow that my work boots overlapped the bearing plate. 

That's me 40 years ago posing with the 15 meter yagi after assembly. I notice that I still have that toolbox. Raising the yagis did not go well. I was still a little unclear on how to construct and operate tram lines (to lift yagis over the guys) and mistakes were made. Nothing was seriously damaged but it took us far longer to get them up the tower than it should have. More lessons learned.

One thing about Brian was that he was easily distracted. His acreage was littered with my tools, which he frequently put down or dropped on the rough ground and in the bush, to be lost forever. Oh well. I saw it as an opportunity to buy better tools! 

Often it was safest to put a hardhat on his head, sit him down in the shade with his favourite lawn chair and elect him supervisor. That brought smiles all around and the work would then proceed quite smoothly.

Brian was never a serious contester. For him the joy came from working lots of DX, teamwork and sharing good times with friends. He especially enjoyed keeping track of country multipliers -- well, he was a DXer! -- and cheer leading. Had we had online scoreboards back then I'm sure he'd have tracked the competition and constantly urged us to work harder.

He would enter contests from time to time after our multi-op VE3PCA group disbanded in early 1984, but never seriously. Dave operated from his station in a few contests although I never did. I was happy enough with the smaller station I built at my first house. I believe the SB220 near Brian's elbow is the one I have stored in the next room!

In recent years he came over a few times to help with the antenna work to, as he said, repay the help I'd given him over the years. Unfortunately that's hard to do when in your 80s. He occasionally operated my station and surprised his regular QSO mates with a signal far louder than usual. Most times he and Mollie would simply come to visit.

One of his final objectives was to gather the cards he needed to apply for DXCC Honor Roll. He was one short. Unfortunately cards had gone missing over the years and he couldn't get replacements, despite his efforts to follow the leads to those who held the logbooks for DXpeditions from years ago. 

While he didn't succeed at that quest he did achieve one last goal: the 2500 endorsement for the DXCC Challenge award. In mid-December he emailed me a picture of the LoTW screen showing that magic number. Mollie is waiting for the endorsement to arrive so she can paste it onto the plaque. The plaque and lots of rare QSL cards were prominently displayed at his Celebration of Life.

Brian was operating high speed CW or talking to old friends on SSB until the end. We spoke often on the phone and I periodically visited when I was in the area.

He was in the hospital for a week after the final crisis without regaining consciousness. I wasn't able to have that one last eyeball QSO with him.

His shack is now empty and the equipment will be sold. The tower and beam will be taken down when the winter weather relents. I'm glad that I can help in this final stage of his long ham career.

73 my friend.