850 mHz Coax Collinear

[rhombus_000]

So I've been trying for some time to get decent signal strength on a 850 mHz trunked P25 system in my area (Rhode Island RISCON). I'm about 30 miles out, and I just can't quite get clear receive. I made a 5 element folded dipole yagi that will get clear receive if I get on my roof with it directly attached to the scanner, but still not enough signal to make it down any length of coax into the shack.

I figured I would try making a coax collinear as described at this link:

Build A 9 dB, 70cm, Collinear Antenna From Coax

I really didn't expect success considering how precise the measurements seemed to be and the amount of connections to be made, but I figured I'd give it a go anyways. I adjusted the measurements from the above link for the 800 mHz band, and actually it wasn't that bad to build.

Once constructed, I headed up to the roof (BTW scanner is a 436HP) and to my surprise, I've improved from one bar on the s-meter to three!!!

But here's the problem... despite much stronger signal strength, when a transmission comes in the decode is much worse than with 1 s meter bar on the yagi, barely readable if at all.

My question is... could poor construction of the collinear (and I admit it was a bit shabby) result in improper phasing of the received signal, causing a garbled RF signal despite the apparent significant increase in signal strength? Or could the omnidirectional gain be causing multipath reception problems?

The antenna outperforms a 1/4 GP when mounted at the same height and through the same coax on the local MA SP trunked system, so I know at least I am getting some gain.

 

[popnokick]

Since you have improved your reception considerably, the description you gave of what is happening makes it seem likely that the RI system you are receiving is using simulcast transmission. Multiple towers on the same frequency broadcasting exactly the same signal. With your new antenna you are now able to pick up more than one of those simulcast transmitters... and your radio is going to need adjustment to deal with it. The 436 and 536 have P25 threshold and decode adjustment settings that you can change. You may also have to add attenuation to actually reduce the signal from the unwanted simulcast transmitter(s). Some in the RI Forum here on RR should be able to confirm if simulcast is a problem in that area. And here's more discussion of P25 simulcast:
http://forums.radioreference.com/ge...-questions-about-scanning-any-p25-system.html

 

[rhombus_000]

Since you have improved your reception considerably, the description you gave of what is happening makes it seem likely that the RI system you are receiving is using simulcast transmission. Multiple towers on the same frequency broadcasting exactly the same signal. With your new antenna you are now able to pick up more than one of those simulcast transmitters... and your radio is going to need adjustment to deal with it. The 436 and 536 have P25 threshold and decode adjustment settings that you can change. You may also have to add attenuation to actually reduce the signal from the unwanted simulcast transmitter(s). Some in the RI Forum here on RR should be able to confirm if simulcast is a problem in that area. And here's more discussion of P25 simulcast:
http://forums.radioreference.com/ge...-questions-about-scanning-any-p25-system.html

Thanks, I'll give the P25 adjustments a shot.

Although I'm having some issues with this antenna for my intended purpose, after messing with it a bit more I would highly recommend it to anyone looking for a quick way to boost their reception on trunked systems. It took about 2 hours to build and cost me about $3.00 for the PVC pipe, already had the coax scraps lying around.

I was curious how reception would be in other bands, and actually it's no too bad on fairly local (15 miles or so) 450mhz systems, not even mounted above my roofline I can still receive pretty clear.

If all you are trying to do is receive the town you are in at 450mhz and a state police trunked system this would be a great antenna.

 

[prcguy]

The 8 element antenna shown in the link would give you about 5dBd gain, not 9dB. It takes 16 1/2 wave elements in a collinear array to get your 9dB gain and the author obviously has no clue about the antenna and copied the instructions from someone else.

The type of coax you use will greatly affect the element lengths. You really need an antenna analyzer to test the first few elements and adjust the lengths before building the entire antenna. Otherwise it can have an upward or downward tilt or not resonate where you want.
prcguy

 

[ko6jw_2]

This antenna appeared in the ARRL antenna handbook (2m version) about 40+ years ago. I briefly considered building one. It hasn't shown up for decades now and, I think, was considered too difficult to build and make work even though theoretically it was an interesting design. The element lengths are much more critical at 850Mhz and, therefore, the phase relationships between the collinear elements. Do it yourself antenna projects are great, if you have the test equipment to make objective measurements. Unfortunately, most of the evidence of antenna performance in these forums is anecdotal. "I got three bars and I had only one before." The question is, should you have gotten 5 bars with a properly built antenna? If I was interested in signal from only one direction, I would not use an omnidirectional antenna. A well designed commercial yagi would be the answer.

 

[N3JI]

The 8 element antenna shown in the link would give you about 5dBd gain, not 9dB. It takes 16 1/2 wave elements in a collinear array to get your 9dB gain and the author obviously has no clue about the antenna and copied the instructions from someone else.

The type of coax you use will greatly affect the element lengths. You really need an antenna analyzer to test the first few elements and adjust the lengths before building the entire antenna. Otherwise it can have an upward or downward tilt or not resonate where you want.
prcguy

Pulling up a slightly older thread...

I'm actually building one of these myself for 855 MHz, and I *do* have a network analyzer to measure the antenna and more importantly, the coax velocity factor. I'm using some higher-end double-shielded RG58 that I had laying around which is listed as the typical .66 VF, but I'm actually measuring it at roughly .648 (duplicated on two separate pieces of test equipment). Using those numbers, I built the antenna -- brass 1/4 wave bottom sleeve, 8 1/2 wave coaxial elements, and a 1/4 wave top radiator. Problem is that I'm off by about 25 MHz. It has a beautiful 30+ dB dip at ~880 MHz. BTW, it also has a very nice ~30 dB dip at roughly 450 MHz, which would verify that it should work okay there validating the OP's observations.

So, this thing actually still works okay on the 855 MHz design frequency, as it's only about 3% off. My question is where the (bleep) is the 3% error coming from?? Is it simply that I'm stuffing it in PVC? I haven't measured it yet without the PVC radome over it, so that could be it. My gut says that a 3% shift is possible in the PVC, but I haven't any experience to confirm. Can anyone validate that? I'm about to lengthen the top 1/4 wave element to see if that makes a difference, but I have my doubts. When I pushed the antenna up so the last element or two plus the 1/4 wave top radiator were well above the PVC, it made no perceptible difference.

 

[N3JI]

UPDATE: I had some time during lunch today, so I measured it again outside the PVC radome. No significant change (in fact it actually went up very slightly indicating the PVC added somewhat to VF, which is what I'd expect). I also lengthened the top element by an inch or so and it still didn't change much...

SO, just for grins, I'm going to take the numbers I have, increase the lengths by 3%, re-build it and see where that puts me. I REALLY want to figure out where the mysterious delta is coming from though.

BTW, I used 492/Freq for the 1/2 wave elements (x VF), and 246/Freq for the 1/4 wave parts.

I may also build one in the VHF range (with fewer elements) as another experiment to see what the numbers yield. Once I nail this down, I'll build one for the 900 MHz ham band as well.

 

[prcguy]

The diagram I have for this type antenna is different than the link in this thread and as I said before, the guy in the link doesn't know much about antennas. My diagram shows a 1/4 wavelength in coax above the choke balun point, then eight or 16 half wave elements and a 1/4 wave whip on top where the base of the whip shorts to the center and braid of the coax under it. The link says to connect the top whip only to the top coax center conductor and not to short that to the associated braid, which is how these antennas have been made for the last 50yrs or more.

Usually you would make a smaller version from calculations like three 1/2 waves plus the choke balun, 1/4 wave base section and 1/4 wave top whip, then tweak the dimensions depending on how that plays. I believe the final dimensions will be a compromise between a free space 1/2 wavelength to resonate each 1/2 wave element and the velocity factor adjustment which helps keep each radiating element in phase so all their patterns converge for maximum gain.
prcguy

Pulling up a slightly older thread...

I'm actually building one of these myself for 855 MHz, and I *do* have a network analyzer to measure the antenna and more importantly, the coax velocity factor. I'm using some higher-end double-shielded RG58 that I had laying around which is listed as the typical .66 VF, but I'm actually measuring it at roughly .648 (duplicated on two separate pieces of test equipment). Using those numbers, I built the antenna -- brass 1/4 wave bottom sleeve, 8 1/2 wave coaxial elements, and a 1/4 wave top radiator. Problem is that I'm off by about 25 MHz. It has a beautiful 30+ dB dip at ~880 MHz. BTW, it also has a very nice ~30 dB dip at roughly 450 MHz, which would verify that it should work okay there validating the OP's observations.

So, this thing actually still works okay on the 855 MHz design frequency, as it's only about 3% off. My question is where the (bleep) is the 3% error coming from?? Is it simply that I'm stuffing it in PVC? I haven't measured it yet without the PVC radome over it, so that could be it. My gut says that a 3% shift is possible in the PVC, but I haven't any experience to confirm. Can anyone validate that? I'm about to lengthen the top 1/4 wave element to see if that makes a difference, but I have my doubts. When I pushed the antenna up so the last element or two plus the 1/4 wave top radiator were well above the PVC, it made no perceptible difference.

 

[N3JI]

There are apparently two ways to do this type of antenna. One with 1/4 wave coaxial elements at the top & bottom, and one with a decoupling stub at the bottom and an open 1/4 radiator at the top. I find that the latter is a bit easier to build, and I'm very close to having it done except for that dreaded 3% error... Here is a link and an image from the page of another built like the one linked above, and like the one I built:

Development of VHF Collinear Ant

 

[nanZor]

Remember too that the gain increase also comes with a narrower lobe. Depending on your environment, you may have a lot of gain much higher or lower than your desired target.

I've never built one of these, but always wonder about the velocity factor being applied to all the elements? Doesn't sound right, since you have alternating transmission line elements (vf should be applied), and common-mode radiating elements where NO vf should be applied. Or at least a very small compensating factor of .95 or so to account for the jacket on the common-mode elements.

 

[prcguy]

That's also my thoughts but I've taken apart Phelps-Dodge/Celwave 5dB Stationmasters and all the coax elements are the same length. The only difference there is their solid copper jacket coax vs braided coax on home made versions.
prcguy

Remember too that the gain increase also comes with a narrower lobe. Depending on your environment, you may have a lot of gain much higher or lower than your desired target.

I've never built one of these, but always wonder about the velocity factor being applied to all the elements? Doesn't sound right, since you have alternating transmission line elements (vf should be applied), and common-mode radiating elements where NO vf should be applied. Or at least a very small compensating factor of .95 or so to account for the jacket on the common-mode elements.

 

[mancow]

I wonder how large a 42 MHz lowband model would be?

 

[prcguy]

Looks like an 8 element version would be about 70ft long and would give about 5dBd gain. A 16 element would be about 119ft long.
prcguy

I wonder how large a 42 MHz lowband model would be?

 

[mancow]

Looks like an 8 element version would be about 70ft long and would give about 5dBd gain. A 16 element would be about 119ft long.
prcguy

I've often thought about doing this and hanging it from a high tree.

 

[N3JI]

Just build a J-Pole. I did that for that freq a few years ago supporting a military op that was quite successful.