70cm Collinear Antenna from RG8X

Hello All!


Yet another project from N3EAQ..... This time I'm making a home-brew 70cm Collinear Antenna from RG8X Coax!

If you do a web search you will find about a thousand copies of N1HFX's instructions for building a 70cm collinear antenna from RG58-A/U coax sections. I dedicate this construction article 10,000% to N1HFX and those instructions! However, I've added a twist, and a correction or two to that oft repeated tome!

Here We GO!

I have a local repeater on 444.350/449.350, I've been feeding it into the 440 leg of a Diamond tri-band antenna. Well, just cuz that's what I had in the air! The performance in both Xmit and Receive sucked! I want looking for a better way, a stand alone 70cm repeater antenna.... Well guess what? They cost a LOT of $$$$$ !!! I was really surprised to discover what a mono-band 70cm antenna cost these days! OK, fine, off to the internet, and a suitable project. That's where I found the many incarnations of the N1HFX 70cm coax collinear. (http://www.rason.org/Projects/collant/collant.htm)

I decided to build this 9dB wonder except I decided to make a couple of changes:

1. I decided to use RG8X coax instead of RG58 A/U

2. I realized there is a measurement mistake in the original instructions, which I hope to correct here.



I'm NOT going to Re-Write the original N1HFX instructions, they are quite good, and very complete! There are a few places that need clarification and correction though. At this point I encourage you to go to the above link and print out the original instructions so you can follow along......

First Change! Using RG58 The Formula given for the 8 sections is:

wavelength of coax = 300 / F / 2 * V

Where F = Frequency in Megahertz

V = Velocity factory of Coax

300 / 444 / 2 * .66 = .2229 meters or 223 millimeters

In my case I decided to use RG8X, only a bit fatter than RG58, BUT the Velocity factor is DIFFERENT!  The velocity factor for RG8X is about 82% (depends on what you have or buy, check the actual specs!) thus the formula changes to:

wavelength of coax = 300 / F / 2 * V

Where F = Frequency in Megahertz

V = Velocity factory of Coax

300 / 444 / 2 * .82 = .2770 meters or 277 millimeters

From the original text:

"To allow for cutting the ends of our coax, we will need to add 8 millimeters to each wave length for a total of 231 millimeters."

Yes, add the 8mm, but be aware that the instructions are not exactly correct here. The author forgot there are TWO ends to each coax section!


You need to do the above surgery to BOTH ends!! So, as you see this is 8mm on both ends! The fact that you do this on BOTH ends does NOT hurt anything in the end of the project, but it is confusing when you start to do the build! Suddenly you think you've cut things too short!

From the original text:

To get started, we will need 8 half wave lengths (231 millimeters) of RG58/U coaxial cable to be cut and connected in the manner shown in Figure 1. First cut back 4 millimeters of the outer jacket, braid and dielectric exposing the center conductor as in Figure 2. Now cut back the outer jacket another 4 millimeters to expose the braid and push the braid back about a millimeter to prevent it from shorting with the center conductor. It is best to lightly tin the braid with solder at this point. Now solder each half wavelength as shown in Figure 1. Attach a few feet of RG58/U to the bottom of the array as in Figure 1 for feeding the antenna.

Hold on here, just wait! Solder together the 8 sections you cut. Go as far as the next paragraph, BUT, DON'T ADD ANYTHING TO THE BOTTOM YET!

Now its time to add the additional elements to the top and bottom of the collinear array. First add a wave element to the top of the antenna as shown in Figure 3. Use #16 solid wire or similar and solder it to the center conductor only. The length of the wave element is calculated as follows:

1/4 wavelength radiator = 300 / F / 4

Where F = Frequency in Megahertz

300 / 444 / 4 = .1689 meters or 169 millimeters

Yes, add the stinger at the top.... NOW, you should have 8 coax sections soldered together, a stinger at the top. I took my DVM in "Buzz mode" and made sure that there was no shorts from any shield to center sections... When I was sure that all was clean I got out the heat shrink tubing and shrunk across all the joints I just soldered/connected. THEN I got a few more feet of RG8X as my "bottom" connection and got it ready for the rest of the process.

I cut an appropriate length of copper 5/16 tubing as instructed. I then took the FEED coax and cut back a few inches (~4") of cover off of it. I shoved that through the tubing and peeled back the braid over the tubing. I soldered the braid to the tubing all around. BE CAREFUL!!! Too much heat and you will make a melted mess of everything! Just use enough heat to tack the braid firmly to the tubing. 

Now, cut the feed center conductor back and connect the feed section to the "8 Antenna Sections" as described in the original instructions. (i.e., the feed section center conductor to the 1st antenna section braid, and the braid/tubing section of the feed section to the "antenna section" center conductor)

I put heat shrink tubing over ALL the connections that have been made above! From the stinger connect at the top to the tubing connects at the bottom.

All 100% heat shrunk!

NOW... I laid the whole assembly out on the floor and as instructed, I put a torroid assembly at about 1/2 wave down from the top of the tubing, and started to take readings with my MFJ analyzer. Also, I added about 18+ inches of coax from the torroids to the end of the feed coax. Because at the end of the day I'm going to mount this like a vertical/whip antenna, and I want room at the end for mounting, that will not interfere with the antenna itself. 

Back to torroids... With the analyzer, I moved the torroids up and down the coax to get the best SWR at my desired freq. (449.350). When I found that spot I put a wire tie in front, and behind the torroid, and then did some shrink wrap on top of it at that spot.

All I have left to do is wire tie the whole assembly to some dowel rods, stuff it up a 16' section of PVC pipe, cap the ends, and attach the feed section to an "N Type bulkhead connector" connected to the bottom cap....

SO, More to come!


My readings with the MFJ were 1.2:1 at my freq of 449.350 and less than 1.5:1 from 440 to 450 MHz !!!! VERY NICE!!!!


That's where I'm at right now.... Will let you all know how it goes, and update from here as I complete the project!

Mike - N3EAQ