Constructing a 2m colinear Part 1

Hi all,

My recent 6 summit SOTA trip contained a few QSOs on 2m. All bar one of mine were using a Wouxun 2/70 HT with a 1/4 wave antenna on 2m. This is convenient, but is really only one step up from the stock antenna supplied with the unit. The antenna uses the radio and the operator’s arm as a (in)effective counterpoise, imposing significant loss to the system, well over 10dB. It’s time for a bit of gain.

Why a colinear?

I had a look at N1HFX’s colinear page among others. There are various types of colinear out there, some are based on a jpole, while this is a series of alternating half wave sections. The design from N1HFX is about a permanent installation, but I wanted something that could be used on a SOTA activation. There is a debate about whether to use a beam or a high gain vertical. The beam can be changed from horizontal to vertical and back by changing the mounting of the beam, but here in Victoria, most chasers would have vertical orientation. Horizontal does come out on field day contests, but aside from that vertical does seem the go.

A second consideration is that chasers might come from a variety of different directions. A beam is going to have a main lobe in one direction, with a series of nulls in others. This is good for some use, especially where either the direction of the wanted station is known, or there is some local source that is wanted to be nulled out. On most SOTA summits, there is not large local sources – obviously there are exceptions – but also radios such as the FT-817 do have good rejection of these. The chaser direction is not always known, and can vary quite widely.

For these reasons, I decided to go for a colinear. I also plan to use it as my vertical antenna alongside a Quadruple Quad for horizontal on the John Moyle MFD, which is the only VHF/UHF contest that I spent much time thinking about. Even for that contest, I generally think about running a station, taking advantage of my equipment optimisation for SOTA by going to places that require foot or non-car access to the summit.

Squid Pole and Coax choice

In the last few months, I have been often using a 7m (actually 6.8m) squid pole to provide a centre mount for a 20/40m end fed or the structure for a tunable vertical antenna for the HF bands. I began to think that this squid pole could provide a structural platform for a 2m colinear. The N1HFX design calls for 8 half-wave segments. I have been using LMR195 for a lot of my coax cables. This is a RG-58 replacement with better loss performance. One effect of this is that it has a higher cable velocity, measured with my MiniVNA Pro at around .825 One issue with these higher velocity ratios is the cable needs to be physically longer, compared to RG58. So, I needed to actually use RG58, rather than a drop in replacement for this colinear. I bought some from

The MiniVNA Pro measured the RG58 cable from rfsupplier at 0.66, so this makes for a shorter setup than LMR195. It will actually fit on the Squid Pole with 8 half wave segments, rather than having to go to a 4 segment version. This should make for a nice antenna.

The first job is to cut up 8 lengths of RG58. I settled on a centre frequency of 145MHz. This is to provide for 144.1 SSB and for 145.5 and 146.5FM simplex not being too far away. We’ll see how that goes. Each length needed to be 300/145 /2 *.66 = 682mm. Add an extra 8mm for the overlap needed between two segments as per N1HFX’s design. Here’s a pic of the 8 segments cut and both ends stripped:

8 RG58/U segments cut and stripped

Then, I needed to join up the 8 segments as per the N1HFX design. Here is a look at one of the joins. A reminder that any image on this blog can be clicked for a better view:

Coax join to swap centre conduct and braid

It was a little tricky ensuring that there were no shorts between the two centre connectors. A stray bit of braid could some down (and sometimes did) to effect a short. I needed to continually check for this while soldering each connection. This is a gotcha with this design. I did pump on a good amount of solder to provide a good join and taped up with electrical tape to ensure the braid stayed away, and also to prevent water getting in.

The top end of the colinear needs a 1/4 wave segment, but only connected to the centre conductor. I grabbed some old RG6 cable that I had already stripped away the braid from for this purpose. The bottom end needs a 1/4 wave sleeve connected to the input coax braid. I used some aluminium tubing for this. Finally, there are dreaded common mode currents with a colinear. N1HFX has a choke at 1/2 wave down the tube for this. Obviously they cannot be put at the feed point because of the 1/4 wave sleeve, so 1/2 wave down the feed-line has to do. I would much rather a few turns. I could have gone for more turns on some PVC tubing, creating what is essentially an Ugly Balun. I went for three small toroids with 3 turns. This should provide a very high impedance to a common current, and at 1/2 wave from the feedpoint, this high impedance will be presented to the antenna.

Here is a pic of the completed coaxial colinear:

Coaxial colinear for 2m

Now, all I need to do is to carry this and the squid pole on site, mount the squid pole and the end of the colinear on it, and I have a BNC connector to go in the FT-817 or any similar style radio. I’ll test on a SOTA activation in a few weeks.

Wayne Merry VK3WAM

SOTA activations of Mt Matlock VK3/VC-001 and Mt Toorongo Range VK3/VT-026

Hi all,

It’s been three and 1/2 weeks since my last SOTA activation, so it’s time to get out there. Can’t let all the others have the fun.

In the lead up to these activations, I had been keeping a close eye on the weather, and could not but help notice that there had been snow to lower levels. It would seem like my first true on snow activations were on their way. I had also done some work on the antenna setup as described in other posts on this blog.

Mt Matlock VK3/VC-001

First up was Mt Matlock. There is 4WD access to this summit, but I only have a 2WD, so I parked it on the Warburton Woods Point Rd. On the way up from Reefton, I wondered when I’d start to see the snow. There was significant snow from about 1000m. I had parked the car at 1200m altitude, about 1.5km from the summit.

Parking the Prius on the Warburton Woods Point Rd with 10cm snow coverage

It took about 30 or so minutes to climb up to the summit. The summit has a fire tower with a few comms on it, but no mobile phone services. The trig point is nearby, as can be seen in this pic:

Mt Matlock summit with 10cm snow cover

I setup the station to the left of the picture shown above. I was about 100m from the comms tower. I was interested to see if there were any issues with noise from the tower and as it turned out, there were none detected.

The 8 segment vertical was setup as shown below:

My portable vertical antenna at Mt Matlock VK3/VC-001

This was my first opportunity to try the new Ugly Balun and connecting the ATU directly to the feedpoint. In times past, I have used two baluns and about 4m of twin line between the ATU and the feedpoint. I had the twin lead and baluns on hand, just in case. After ensuring that the connections were nice and tight, I used the antenna on 40 and 20. It seemed to work well, including getting some contacts from VK3AFW and VK3PF who were inside 200km. Contacts further out were easy QSOs on 40m. Unfortunatly, there was no DX for me into the US. I wondered about this, but found out after I got home that none of my SMS spots made it on to sotawatch. I could hear plenty of CW at the bottom of 20m, but there was nothing around 14.061 and 2. It seems that most are content to wait for a spot on sotawatch and will not go hunting.

This is a look at the ATU, which is a Z11Pro 2, and the Ugly Balun at the feedpoint:

Antenna feedpoint with Z11ProII and Ugly Balun

Mt Toorongo Range VK3/VT-026

It was time to head off to my second activation. I was able to get the car to about 2km from the summit, which is much better than the 6km that I was expecting. After a walk of about 40 minutes, including about 150m off track, I arrived at the summit. Clearly someone else had been here as there was a rock cairn at the top. It is interesting that this place does not have a proper summit name, even though it is higher by about 10 metres than Mt Toorongo, about 800m to the south west.

There about 5cm of snow coverage here, a little less than Mt Matlock. The weather had also improved, and as can be seen in my two antenna pics, continued to improve as the day went on. It was still only about 1 degree C, but without any wind, it was actually quite pleasant. Before I started, I noticed marks in the snow, which I suspect a wallaby had made:

Suspected wallaby footprints

The vegetation made running all eight radials a little tricky, so I started out with 4, and that is how it stayed. The RD contest had started and this meant that 40m was very crowded. This is where it gets challenging to operate QRP. People either cannot hear you, or do not care that there is a weak station around. They come on frequency and start calling CQ CONTEST. I did not spend a huge amount of time on 40 today. I also found out later that again, I had no spots on sotawatch.

2m was more fruitful. Most contacts here were contest contacts, but it is always interesting to be operating on a mountain in the middle of nowhere and get contacts upwards of 200km away with either a HT 1/4 wave, or my twin lead jim slim for 2m.

As the day wore on, I began to think that it was time for some DX. I put another spot for 20m CW, which did not work, and started calling. I got a contact with KG7E, and he spotted me. Welcome to the pile up! I almost forgot what a SOTA CW pileup can be like. It was good to again put DJ5AV and G4SSH in the log, among others.

The day was drawing late now, and I did not desire to walk back the whole way in the dark, so I packed up around 5pm. A most enjoyable day. If you are reading this and do not know CW, it is not easy to learn, but it is worth it. These SOTA CW pileups are a great deal of fun.

Conditions on 20m HF were quite good today, but I still feel that I am about 2 to 3db ahead on 20m using the ATU and the Ugly Balun than the two 4:1 current baluns and the twin lead. Both setups do not have any evidence of common mode currents and RF at the radio.

I’ll leave you with the two antenna pics showing the improving weather from VK3/VT-026. Regards, Wayne VK3WAM

Antenna at VK3/VT-026

Late sun at VK3/VT-026

Fixing the loading coil and building an Ugly Balun

In my earlier post at: I still had two pending jobs:

  • Fix the crocodile clip on the tapped loading coil
  • Build an ugly balun

Crocodile clip

Clearly they don’t make ’em like they used to. I had not used this clip for very long, but already there were signs of stress, shown below:

Stressed crocodile clip

The top part of the clip is bending and cannot take any force without further yielding. The material is “stooffed”.

A simple job, and also I can fix a problem with the tapping coil being a little too short. I recently purchased some LiPo chargers that came with crocodile clips for their DC power, but I use alternative cables with anderson poles – no clips for my DC power thank you!

The job nearly finished is shown below, with just some electrical tape over the solder joint. Much easier to connect to the bottom of the coil for bypassing, which I need to do on 20 and above. Bonus!

Ugly Balun

The job of an ugly balun is to be a choke to common mode currents. RF is supplied on the centre conductor in coax and the return is on the inner of the coax shield (through the skin effect). There is effectively a third conductor which is the outer of the shield and this is where common mode currents can arise. I have no doubt that they do arise with a coax fed vertical – if I do nothing about them, there is RF at the radio. It interferes with my CW key and it significantly raises the noise floor on receive. This is especially if there is other stuff around nearby, such as a commercial comms tower on the top of a hill or mountain when activating nearby for SOTA. Common mode currents also mean that the coax acts as an antenna rather than a transmission line. This antenna is lying on the ground – not so good.

It is clear that we do not want common mode currents, so we need to get rid of them, and one approach is to use baluns which is what I have been doing, but now I want to put the ATU right on the feedpoint. Solution is an Ugly Balun. This is not really a balun at all, it is merely a current choke.

Designing the ugly balun

In an ugly balun, coax is used to wind around something. This does not affect the inner conductor and the return on the shield. It does expose common mode currents to reactance, and if this is high enough, it will effectively block it. I have read in a number of places that 500ohms of reactance is what is needed to be a good block. Another way to look at this is the reactance should be 10 times more than the resistance of the antenna – presuming the antenna is resonant.

There is a good inductance calculator at: I have used this many times and it gives good matches to what is measured using a VNA.

I settled on 21 turns. This gives the following reactance values:

  • 154ohm @ 1.85MHz, which is not very high, but will get rid of the bulk of it
  • 300ohms @ 3.59MHz, which should get rid of around 90% + of it
  • 600ohms @ 7.1MHz, looking good
  • 3000ohms @ 21MHz
  • -11000ohms @ 28MHz. This point is just above coil resonance and it has effectively just turn capacitive – will still block the common mode currents though, so it’s all good.


I used 50mm internal diameter PVC pipe and cut it about 125mm long. I drilled 2 holes at the appropriate spots with a M5 drill bit. After penetrating with the bit, I put it on an angle so I would not excessively bend the LMR195 coax to go through the hole.

Here is what the ugly balun looks like after coiling and threading the coax:

Ugly Balun cable coiled and threaded, but no termination

Next up was to put some connectors on the coax. I use UHF connectors for HF in the field (I also use it for 6m as well). I know that BNC is the QRP standard, but sometimes I’m QRO. I used a male UHF connector on one end to plug into the ATU, and a female connector at the other, for a run of coax to wherever I will operate the radio – keep in mind, this whole setup is for portable activations – typically Summits on the air (SOTA). These connectors I obtained at

UHF connectors for ugly balun, one male and one female

UHF connectors need soldering of the inner conductor, but the shield is crimped. After doing this, the ugly balun is ready for use:

Completed ugly balun


Now time for testing. The next activation is soon enough, and that will be the real test, but for now I wanted to see what was the insertion loss of this thing. I hooked up to a VNApro in transmission mode:

Ugly balun under testing with VNApro

A screenshot of a test between 1 and 30MHz is shown below:

Ugly Balun test results

Essentially, the insertion loss is around 0.25db from 15m and up, which is close to coil resonance. Lower down, the insertion loss is lower, around 0.19db at 20m, 0.14db at 40m and lower loss on the lower bands. I am happy with this, it is not much higher than the pure loss of the coax itself. It is interesting that there are some loss factors aside from the losses to be expected from LMR195, hence the flat line around 21MHz to 30MHz, which would not happen on a simple coax run. It will be interesting to see how this goes in the field.

Wayne VK3WAM

Improving the backpackable vertical antenna

My SOTA activations on HF use a vertical antenna which is all home brew. I decided on using a vertical because:

  • I wanted an antenna that could be used anywhere, whether there were lots of trees or not.
  • It needed to be able to be disassembled into segments no longer than 75cm in length so they could be put into a backpack. Some SOTA summits need significant off-track walking to access, and carrying a large 1m or more item outside the backpack – including being strapped on – was not a great solution for those situations.
  • I wanted something with a low angle of radiation to work SOTA DX. Perhaps this would be at the cost of working local stations, but if the antenna was efficient enough I should still be able to work local, but get the DX as well.

Bunnings sell cyclical al in 10mm, 12mm and 16mm segments – and larger. By using the workshop at my radio club, I have cut these to size.

State of play in June

In June, I had a 6 summit activation long weekend in the Grampians. The final activation was at VK3/VW-007 Near Baroka Lookout. Allen, VK3HRA joined me for the second half of my activation. I was working 20m DX on CW while he worked 40m SSB. Some stations he worked had also worked me earlier on 40m, so I was very interested in his signal reports, which were generally about 2 points higher than mine. It’s not a contest about who has the best antenna, but when conditions are poor, a better antenna could be the difference between a successful activation or not, so it was time to do some work.

Allen was using a Buddistick with one radial. Now some get into the counterpoise vs radial wording thing, but at the end of the day it is about providing some kind of ground to minimise ground losses with a vertical. Now both our antennas were short for 40m, but mine was less short. I also had eight radials! So why the poor performance? His antenna was loaded, mine was unloaded. Answer: Rather than relying on an ATU to match the antenna because because of the short capacitive reactance, I needed a loading coil. Time to make one.

The Loading Coil

I like working multiple bands, and putting a loading coil on the antenna could restrict me to one band, and this would not do. The loading coil needed to be tapped, so I could bypass it, also select tapping points suitable for various bands. With 8 segments, the antenna is 6m long – longer than a 1/4 wave on 20, but short on 30 or lower. I used an online inductance calculator at Hamwaves, designing it on 50mm PVC pipe with a flat piece of aluminium to allow it to be screwed onto the antenna at one of the coupling points. I had already made up a piece of wood about 100mm long with two M8 screws at either end, so this was perfect to mount the coil. The coil would be long enough to have enough turns to resonant the antenna on 80m, as well as making tapping points lower down to resonant the antenna on 30 and 40.

My plan was to put the coil between segments 2 and 3 on the vertical, about 1.5m up from the feedpoint. On the net, and in NEC2 modelling I had done, having the loading coil up as much as possible is better, but it needed to be low enough that I could physically access it to change bands without needing to lower the antenna. I would still use an ATU, but this would be more about matching the resistance, rather than having to deal with all the reactance as well. A short antenna has a feedpoint resistance well below 50 ohms – even 1/4 wave is about 36 ohms, and shorter is lower. As the ground plane is improved, this resistance is lower than otherwise, leading to greater mismatch – but this is good because the missing resistance is missing ground losses. The job of the ATU is to match for this mismatch, but the losses across the ATU would be small.

First use in anger VK3/VC-003 Mt Richie

Now with my new loading coil, it was time to get it into action. Mt Richie was the first activation, with Mt Donna Buang VK3/VC-002 later in the day. Reports on 40m were a good two to three s points up on the usual, with many stations reporting that I had a good signal, rather than the usual “I’m digging you out of the noise”. So it looks like I’m better in the game. I could also tell during the setup that when I tapped the coil on 40m at a suitable tapping point – the radio was hearing a lot more signals. Loading a short antenna certainly seems the go.

One impact of the loading coil was that I could not use the antenna with 8 segments. The loading coil added enough weight that the antenna, despite 2 sets of guy ropes, could not stay up. I had to use the antenna with only 7 segments – 5.25m high from the feedpoint.

The Bunnings 10mm segments were just too weak and I needed something stronger. I had used 16mm segments for something else, but I was not that happy with how ridged it was either. I made up two 16mm steel segments with some 10mm aluminium tubing inside for the M8 threads. The 10mm tubing was held in place by a 6mm nut and bolt about 10cm up from the coupling. This has worked well in my experience, and makes the couplings stronger than other parts of the antenna. This approach would allow me to have a drop in replacement for the bottom 2 75cm segments on the vertical, and should get me back to an 8 segment antenna with the loading coil.

Rubicon forest triple activations

A few weeks after Mt Richie and Mt Donna Buang, I was activating three summits, Mt Bullfight VK3/VN-002, Pyramid Hill VK3/VN-005 and Bill Head VK3/VN-004. Conditions were poor on 40m to say the least, with a large skip zone during the day and attenuation on hops on 40 during the day high. If it was not for the loading coil, I would have not got away with the points from the three summits that day – as described in by blog post. I might not even got a single contact on some of those summits.

By now, I had used the stronger steel segments down the bottom of the vertical, and these had made the antenna structurally sound. There was no buckling anywhere lower on the antenna, which was the reason it used to fall over.

While conditions were poor, and a few days later, Peter VK3PF only got two contacts from VK3/VT-003, underlying the challenge, it made me think that things are still not quite right.

Further improvements

I had always not put the ATU actually on the feedpoint, but somewhere on the ground nearby. This was because the antenna would not be able to bear the weight of the ATU – it does not weigh much, but I was always struggling with the structural integrity of the antenna. My two steel segments seemed to have resolved that problem, so now I have begun to think it is time to put the ATU directly at the feedpoint.

When I first started doing portable activations with this antenna last year, I first had a 10m run of coax from the ATU to the feedpoint, with the radio near the ATU. It is convenient, but performance was terrible. I then had a short, 1m bit of coax from the feedpoint to the ATU, and then a longer run to the radio. This worked better. It does not significantly matter how long the run is from the radio to the ATU. It is all about what is between the ATU and the feedpoint and coax here is bad.

So I built up some homemade twin line from enamelled 1mm wire I had available. The run was 4 metres long. It was this length because another vertical I had been using has a feed point 1.5m off the ground. Even with my 8 segment vertical with a feed point 30cm off the ground, the 4 metres of twin line worked much better than 1 metre of coax. There was nothing wrong with the coax – I could use it quite happily between the radio and the ATU!

Still, the antenna is unbalanced and so is the tuner, and twin line is a balanced transmission line, so running an unbalanced signal down it must not be good. I built two Guntella 1:4 baluns, one to go at the tuner, and the other at the feedpoint. This further improved things – there was evidence before of RF at the radio, and these baluns removed it, also my RX was much better – further evidence of common mode currents before. One time I forgot a balun, and using the setup without them was certainly worse than what I was used to.

Reading W2FMI Jerry Sevick’s Transmission Line Transformers made me think that I still have not got this thing quite right. The Loading Coil gets rid of capacitive reactance, but it can not be said that there is 50 ohms on one side of the balun and 200 ohms on the other. This would mean that the losses of 0.05db that I had measured using a network analyser on these baluns would not be the case because of these mismatches. Over the Gippstech weekend, I did some Keith Roget activations at 100 watts in Morwell National Park and Tarra-Bulga National Park. The baluns got warm. Not hot, but the hotter one had it’s temperature raised about 20 degrees. This tells me that the loss is much greater than 0.05db, more like about 2 to 3db. There are two of these baluns, so there is 4 to 6db of signal improvement waiting to be taken.

So my current plan of attack is to do the following:

  • Put the ATU directly on the feedpoint of the antenna, now that the steel segments have removed the instability problems. Get rid of the twin line and the two baluns. The ATU is unbalanced, feeding an unbalanced antenna.
  • There will be some common mode currents reintroduced because the baluns are gone, and we do not want these. We certainly do not want them anywhere near the radio, but across the ATU is not such a big deal. So a 10 turn on 50mm PVC pipe Ugly Balun with coax – which is really a 1:1 unun will do the trick.

This Ugly Balun will get rid of the common mode currents that run on the outside of the shield, but because it is made of coax, the inner conductor and the inner surface of the shield are unaffected. There will be enough reactance on the Ugly Balun to block common mode currents even at 80m, but not too many turns to not work at 10m. Being on the radio side of the ATU – but right next to the ATU, means that the RF inside is going along a matched transmission line with low SWR, so the loss of this Ugly Balun should be very low.

That’s the plan, we’ll see how it goes at the next activation.

Regards, Wayne VK3WAM