How to build a UHF antenna...

[mclapp]

Originally Posted by spokybob

I did my testing tonight in the bedroom with my RCA CECB and the CM4220, 4 bay and a modified 2 bay. All channels lock in.

I noticed your chart showed that the cm4220 was the strongest (if higher numbers mean better)

I also noticed that the cm4220 was pointed in a different direction than the other 2.

Were the directions that you had listed where you got the best signal?

[spokybob]

Yes. On this particular session, I was able to rotate for best reception, without moving the antenna to various locations. 24.1 is usually the weakest, but I have to be check to see if 4.1 is still OK.

[spokybob]

Originally Posted by spokybob

Yes. On this particular session, I was able to rotate for best reception, without moving the antenna to various locations. 24.1 is usually the weakest, but I have to be check to see if 4.1 is still OK.

All of the signal meters I have used, 100 would be perfect and 0 would be no signal. Actually, my 622 box with a 4221 on the roof, does have 100, with 62 the drop out point. My RCA box has never exceeded 76, and the dropout is around 16 or so.
In all my previous tests, I have mounted the Youtube on a wall,either facing east/west or north/south because the elderly folks I am helping either want to hide the antenna, or do not have the room for a floor mount so it must be hung on the wall. I have done 4 installations so far with good results with my 4 bay. There has been 2 days with strong winds that have caused pixelation, oddly enough our 6.1 was the worst one of those days.
Now I am beginning to understand that db & measured signal strength are not related as least as far as measured numbers. I don't understand theory, that's why I like all your comments. It helps me learn about this strange world of antennas.
Bob 61231

[fbov]

Originally Posted by spokybob

...
I don't understand theory, that's why I like all your comments. It helps me learn about this strange world of antennas.
Bob 61231

Bob,
Consider that the IRRL will not accept antenna advertising based on empirical data, only NEC modeling. Why? Because you can arrange test conditions to prove almost anything, and can never (especially in a consumer location) control all the variables.

You, on the other hand, are living in an empirical world; whatever works is best, regardless what theory says. That's what makes this so difficult, and so interesting!

Frank

[Falcon_77]

Attached is my attempt to convert my 1-bay into a 2-bay. I had problems getting the feed lines to stay in place. They kept sliding out when I would tighten the screws. I ended up placing them flat against the wood, but that caused them to sink it when I tightened them. I don't imagine that is the best thing for reception. Are there any tricks to keep the feed lines in place?

With this 2-bay, I recorded strong signals for some channels, but the SNR didn't improve at all and the higher channels (mostly the 50's) saw a drop vs. the 1-bay design...

Perhaps there is a mis-match. I can complete a 4-bay design if I can get the top and bottom feedlines to work somehow. What is the reason that they shouldn't be straight all the way up and down? The top and bottom need a twist for impedance matching reasons?

[johnied]

Originally Posted by Falcon_77

Attached is my attempt to convert my 1-bay into a 2-bay. I had problems getting the feed lines to stay in place. They kept sliding out when I would tighten the screws. I ended up placing them flat against the wood, but that caused them to sink it when I tightened them. I don't imagine that is the best thing for reception. Are there any tricks to keep the feed lines in place?

With this 2-bay, I recorded strong signals for some channels, but the SNR didn't improve at all and the higher channels (mostly the 50's) saw a drop vs. the 1-bay design...

Perhaps there is a mis-match. I can complete a 4-bay design if I can get the top and bottom feedlines to work somehow. What is the reason that they shouldn't be straight all the way up and down? The top and bottom need a twist for impedance matching reasons?

Sort of, but not for imepdance but for "phasing" so all antennas(which each bowtie is really)
"in phase" Imagine the Current and Voltages peaking at different points on
each bowtie or.antenna without phasing . I believe the phasing puts a half wavlenghth at opposite polarity(the twist) to the wave. ,making the wave identical on all the bowties.. I think im right on this :P..

[fbov]

Originally Posted by Falcon_77

Attached is my attempt to convert my 1-bay into a 2-bay. I had problems getting the feed lines to stay in place. They kept sliding out when I would tighten the screws. I ended up placing them flat against the wood, but that caused them to sink it when I tightened them. I don't imagine that is the best thing for reception. Are there any tricks to keep the feed lines in place?

With this 2-bay, I recorded strong signals for some channels, but the SNR didn't improve at all and the higher channels (mostly the 50's) saw a drop vs. the 1-bay design...

Perhaps there is a mis-match. I can complete a 4-bay design if I can get the top and bottom feedlines to work somehow. What is the reason that they shouldn't be straight all the way up and down? The top and bottom need a twist for impedance matching reasons?

As long as the elements have good electrical contact to the feedlines, sinking into the wood won't hurt; the screw already does that far better. I made my elements a little differently, attached, so the feedline doesn't actually go under the screw head, but lots of the element does. This element base geometry was obvious to me, but no one else seems to do it.

You've ganged them together in the normal manner for 2-bays. The twist in the 4-bay is to phase-match the bowtie elements on each leg of the feedline. Your elements have independent feed lines, so you just have to make sure the feedlines are the same length (terminals in the middle) to phase match.

As to response, impedence won't be 300 ohms, and will likely vary with frequency. You're using oversize elements as I recall, so upper channels may suffer more, and you're comparison is time-shifted - you used the same parts - so perhaps there were signal field changes?

Regardless, see if my element base geometry improves your feedline capture issues. If it does, add upper and lower elements at the same element spacing you've already established, with a twist to phase match the other element on that feedline. This series-parallel arrangement maintains the dipole's native 300 ohm impedence as well as signal phase. It just strikes me as one of those designs-that-work!

Frank
----
johnied, I think you're right, too! 1/2 wave due to the twist and 1/2 wave due to the element spacing at the design frequency.

[wblynch]

Originally Posted by Falcon_77

Attached is my attempt to convert my 1-bay into a 2-bay. I had problems getting the feed lines to stay in place. They kept sliding out when I would tighten the screws. I ended up placing them flat against the wood, but that caused them to sink it when I tightened them. I don't imagine that is the best thing for reception. Are there any tricks to keep the feed lines in place?

I would try some flat washers. That should keep them from embedding into the wood. You can try nylon if metal is a concern.

[spokybob]

Originally Posted by Falcon_77

Attached is my attempt to convert my 1-bay into a 2-bay. I had problems getting the feed lines to stay in place. They kept sliding out when I would tighten the screws. I ended up placing them flat against the wood, but that caused them to sink it when I tightened them. I don't imagine that is the best thing for reception. Are there any tricks to keep the feed lines in place?

This is ugly, but this is a test board. I have always used washers under the screw heads. I found a few of these fender washers. Mechanically they work well. I'm not sure if they affect the signal.

[Beerstalker]

You guys seem to really know about this antenna stuff so I wanted to ask for some of your input. I tested a bunch of different antennas last night with my new DirecTV AM21. The two that worked the best were the RCA ANT806, and the Phillips MANT940, if I had them outside on my balcony. They did not work well indoors. I realize they are not the best antenna's but they were the ones I could buy locally.

However I also decided to give the old school a try as well, and I found that the bowtie antenna that I used as a kid to get Fox before we got to cable actually worked great for me with it sitting inside on top of my entertainment center. It didn't get quite as good of a signal as the RCA or Phillips but it was only 5-10% lower, and by using it I didn't have to worry about drilling another hole through the exterior wall of my apartment to run a cable in. Right now I am getting the following channels using it (I have also included signal strength)

6 (45-55%) out of the Quad Cities
8 (70-75%) out of the Quad Cities
18 (60-65%) out of the Quad Cities
19 (95-100%) Peoria Local
25 (100%) Peoria Local
31 (85-95%) Peoria Local
43 (85-90%) Bloomington Local
47 (95-100%) Peoria Local
59 (55-65%) Peoria Local

This is using this bowtie antenna and this converter to hook to a 300-75Ohm converter, to a male to male RF connector to a 12' Monster RF cable (I know Monster is a ripoff but this is an old cable I had laying around).



Would I be better off ditching the 300-75 ohm converter and soldering one of the adapters like the Youtube antenna uses?

How about attaching multiple bowtie antennas together kind of like the Youtube antenna? Or should I just leave well enough alone?

If you want to look up my antenna web stuff my address is
7150 N Terra Vista Dr
Peoria, IL 61614

I'm on the top floor of a 3 level apartment complex.

Thanks guys, I'm looking forward to your input.

[spokybob]

I made a conventional 2 bay to test against the 2 bay that was a 4 bay with the center 2 removed.
I have been using this 2 bay for 2 days with a few dropouts on the 2 weakest channels. The newer 2 bay looks a little more directional because 12.1 is off to the side. I think this 2 bay design may have some potential in urban area, setting on top of the convertor box.

May 7_9 test
Date 7-May 7-May 7-May 9-may
Time 9:00 PM 9 PM 9 PM 9 PM
Rec RcaCECB RcaCECB RcaCECB RcaCECB
Loc bedroom bedroom bedroom bedroom
antenna Cm4220 4 bay old 2 bay new2bay
pointed 70 deg 45 deg 45 deg 45 deg
amp no no no no

58 4.1 47 42 32 37
56 6.1 76 50 44 52
38 8.1 67 56 44 55
45 12.1 50 40 32 24
49 18.1 62 55 37 49
23 24.2 35 26 18 18
Old 2 bay

newer 2 bay

[Falcon_77]

Originally Posted by spokybob

This is ugly, but this is a test board. I have always used washers under the screw heads. I found a few of these fender washers. Mechanically they work well. I'm not sure if they affect the signal.

Thanks. The washers I have are probably just too small. Some nylon ones might be a good idea as well. I can't think that having the feedlines getting buried into the wood is a good thing.

[Falcon_77]

Originally Posted by Beerstalker

How about attaching multiple bowtie antennas together kind of like the Youtube antenna? Or should I just leave well enough alone?

If you are getting what you want now, you can probably leave it alone. Looking at the stations on your list, they are staying on UHF. However, if there are any VHF stations you will want, you may need to add a VHF antenna, but that would be a topic for another thread.

[A J]

Originally Posted by Beerstalker

I found that the bowtie antenna that I used as a kid to get Fox before we got to cable actually worked great for me with it sitting inside on top of my entertainment center.

I have a 4 bowtie antenna in the attic but found that I get much better reception with a simple $5 Radio Shack bowtie placed in a window. I have topper windows in my 10-foot ceiling living room so the antenna is about 8 feet above ground level. I use the same type 300-to-75 matching transformer as shown in your first picture, connected to RG-6 run behind the topper curtains and down the side of the lower windows behind the blinds. I get all the local OTA channels (one of which is VHF), ranging from 14 miles to 37 miles distance, 302 to 4 degrees direction (antenna facing midway between), -45.0 to -81.0 dB. Under certain weather conditions I have seen a few (very few) dropouts on the -81 dB station, but never on any of the others.

I am amazed by that little antenna's performance. A couple of pictures are attached.

A J

[spokybob]

AJ: How clever. Thanks for the great pics.

[cpldc]

Originally Posted by Beerstalker

Would I be better off ditching the 300-75 ohm converter and soldering one of the adapters like the Youtube antenna uses?

How about attaching multiple bowtie antennas together kind of like the Youtube antenna? Or should I just leave well enough alone?
partment complex.

I trimmed off the wires from a transformer like this and soldered them where the twin-lead went before (see my post on page #4). I found that it improved reception when compared side-by-side with an unmodified bowtie connected to one similar to your first picture. My guess is by reducing the amount of twin-lead it reduces the amount of signal picked up by the wire after the antenna and thus reduces multipath.

Since they are so cheap, it's worth trying.

[Falcon_77]

Well, my attempt to upgrade the 2-bay to a 4-bay didn't go so well. Attached are pictures of my build, but I think the problem lies in the feed-lines. They don't touch at the cross-over but perhaps they are too close. Also, I have separate pieces for the top and the bottom and not one continuous line.

The results are about 50% worse than the 2-bay build, so I guess it's back to the drawing board.

I think a Hoverman is easier to build than a 4-bay and I'm having very good results with that (picture also attached).

[spokybob]

Originally Posted by Falcon_77

Well, my attempt to upgrade the 2-bay to a 4-bay didn't go so well. Attached are pictures of my build, but I think the problem lies in the feed-lines. They don't touch at the cross-over but perhaps they are too close. Also, I have separate pieces for the top and the bottom and not one continuous line.

The results are about 50% worse than the 2-bay build, so I guess it's back to the drawing board.

I think a Hoverman is easier to build than a 4-bay and I'm having very good results with that (picture also attached).

Mike: Thanks for the pictures. On the 4 bays that I experimented with, I bent the wires up and over, and also laid them nearly flat, with tape on the crossovers. I found the latter to work better.
On your 4 bay, it looks like you went with 8" whiskers & 8" spacings. This design was my poorest effort, even though Frank's tests showed that it should work good on all but the highest RF #s.
Before you tear down your good looking 4 bay, try routing the connecting wires like I did in post #94. I leave the spacing far apart, then bend sharply towards the crossover.
Keep us posted on the Hoverman. It looks like fun.

[Falcon_77]

Bob, thanks for the tips.

I'm using 10" elements and 10" spacing on my UHF bow-tie build, to optimize reception for "band IV" as it's rarely called in the US.

I will try to re-work the feedlines and see what happens. Right now they are perhaps too close or bending one instead of the other causes the lengths to be different. Regardless of the problem, it seems clear that these bays are out of phase at the moment.

[Falcon_77]

Well, I spent a lot of time rebuilding the feedlines tonight only to find out that the balun is faulty. I discovered that signal strengths fluctuated significantly depending the angle that I held the cable, which caused me to find out that the leads into the balun were loose. Replacing the balun helped significantly, but I had already dismantled the other two bays by then.

However, not all is lost as I've discovered that mounting the feedline to the back may actually be viable (see pic). After I make sure the 2-bay is ok again, I will re-build the other two bays... but not tonight!

[johnied]

Originally Posted by Falcon_77

Well, I spent a lot of time rebuilding the feedlines tonight only to find out that the balun is faulty. I discovered that signal strengths fluctuated significantly depending the angle that I held the cable, which caused me to find out that the leads into the balun were loose. Replacing the balun helped significantly, but I had already dismantled the other two bays by then.

However, not all is lost as I've discovered that mounting the feedline to the back may actually be viable (see pic). After I make sure the 2-bay is ok again, I will re-build the other two bays... but not tonight!

May I ask what material you are using for your bowtie elements.
These look like they might be something i want to give a try.
Thanks .


John.

[Falcon_77]

Originally Posted by johnied

May I ask what material you are using for your bowtie elements.

It is aluminum suspended ceiling wire, 12 gauge. I found them in Lowe's (thanks to a tip from Frank). They are sold in bundles of 50 6' straight pieces and have been easy to work with. They were about $8 for the bundle. This is cheaper per foot than the 100' coil that I have yet to use (which was about $7).

The following is the part number:

181510 6' USG Hanging Wire 12GAU

[fbov]

Looks like I should have stopped at Lowe's instead of Home Depot ... 12 ga. is closer to the commercial gauge than the 18 ga. I used, and you got it straight!

Optimum antenna size depends on the channels you can receive. 10" should have peak gain around Ch 38, and would be great if all stations are Ch 50 and lower. I saw a drop in signal strength for my UHF 58 (my weakest at -77.5dBm) using a 10" design to mid-70's compared with mid-90's with the 7" spacing. YMMV, and we know Bob's does!

For my "experimental" set-up (read nice looking) I run the balun straight out the back. The feedlines are on cross-pieces mounted on the spine, so I just put a nut on the back side and attach there. I see no reason this approach won't work.

There's also a method behind CM's feedline routing. As I recall, optimum propagation requires a specific spacing between parallel conductors that's a function of feedline diameter. That's why the insulator in twin-lead hold a constant spacing. This is clearly a second order effect, given the results of those who ignore it, but one CM just as clearly built into the CM4221/4228. CM's approach also gives you equal length feedlines, another plus.

It may be time to stop at Lowe's ...

Frank

PS Mike, good catch on the balun; a practical reminder to check for continuity in coax, baluns, splitter/combiners, etc.

[Falcon_77]

Originally Posted by fbov

Regardless, see if my element base geometry improves your feedline capture issues. If it does, add upper and lower elements at the same element spacing you've already established, with a twist to phase match the other element on that feedline. This series-parallel arrangement maintains the dipole's native 300 ohm impedence as well as signal phase. It just strikes me as one of those designs-that-work!

So, when I re-build the feed lines for the top and bottom bays, I should keep the lines parallel and only have a sudden twist at the end, right? Previously, I had them crossing over in the middle (between the bays).

[fbov]

Yeah, but Bob just crosses his, so it can't be a strong driver, more a best practice. I'm a copycat; I cloned the CM-4221, partly because their design details made sense. I find a certain joy in trying to make things that match the requirements of the physical world.
----
with apologies to cpcat ...

[mclapp]

I ran a model of the wires crossing when I built my 4 bays and it really didn't show much difference. I also ended up using the CM cross method mainly because it is a good working design.

The ones I built were modeled for a lower frequency so the phasing lines are a bit longer.

One thing I will say though is running the wires on the wood the entire length of the phasing line may slightly change the way they react electrically. It would be best, if possable to space them away from the wood maybe with some plastic washers or something.

I will attach the results of the models I ran with different wire crosses. The first set will be with the cross in the middle and only 1 wire bent up and over and the other running flat.

All the plots are done in free space and the gain measurments are in dbi

[mclapp]

Here are the plots for the wires crossing in the middle with both sets of wires being bent away from each other where they cross the same amount.

[mclapp]

Here are the plots and diagram for the method I used which is similar to the factory designs.

[Don Bouldrey]

I have a question about the antennas y'all are building. They all seem to have the same length of "bowtie". Don't you want to have an array of different lengths to better capture different frequencies? Such as nearly 30" for channel 9 to around 8" for channel 49?

While I'm here, let me request some opinions for a new antenna system. I'm actually in a good signal area but in a lousy local situation because of large, old trees. My house itself cannot see in the direction of the towers so an attic, rooftop, or sidewall antenna is out of the question unless I go extremely high.

What I do have is the remains of an old dish system out in the yard, away from the house and underneath a huge old cedar tree. It's about 100 coax feet away from my TV. At this point there's a nice aperture that allows for a relatively clear shot at most of the transmitters. What I've been using the last 10 years is an array of twinlead dipoles hung up in this tree. A few of these dipoles covered the range between channels 3 and 55. This worked great for analog. No problems in the worst of wind and weather. I was able to watch the doppler radar on TV all night one night when we had winds that peaked out the wrong side of 135.

Now I'm going digital and there's no more low vhf to worry about but the old way isn't doing all that well. Most of the time I'm getting fine reception but when the weather gets bad, it goes down the tubes with a lot of dropouts, stuttering and pixellation.

Here's what I have to work with at my exact coordinates. I've also included the good weather signal strengths.

ch 17 at 9.4 miles/233° dBm -37.8 signal 87-97
ch 27 at 18.5 miles/269° dBm -44.8 signal 85-90
ch 9 at 16.6 miles/268° dBm -49.6 signal 85-90
ch 47 at 7.8 miles/222° dBm -45.8 signal 75-80
ch 31 at 7.8 miles/222° dBm -41.4 signal 90-95
ch 41 at 22.3 miles/264° dBm -60.3 signal 65-77
ch 45 at 7.7 miles/195° dBm -48.5 signal 82-87

What I've done is thin down to a 8½" dipole aimed at 265° and a 24" dipole aimed at 220°. Hanging dipoles don't work with digital. As an interim solution I have the two dipoles with opposing plain wire reflectors strapped to pvc pipes stuck in the tree. Great reception under ideal conditions but not stable enough for severe weather. After the changeover, we'll have one more on 25 @ 7.8 miles and 222°. Right now, 41 is the channel that gives me the most difficulty. There are other channels but I don't want them and have them blocked.

So anyway, I'm planning a better system and thinking about a single unit consisting of a grate/mesh reflector and bowties like you guys are building. Ideally this thing could be kept at a low altitude, like 10' or under.Do you have any recommendations for reflector and element sizes?

[mclapp]

Here are the models and pics of the 4 bay antenna I made recently. It uses 9 5/8" bay spacing with 9 3/4" wiskers.

The reflector is 40" tall and 36" wide the outside edges are bent forward about 2" on the ends and is spaced 5" back from the elements. The elements are also swept forward 2".

It seems to work pretty good, I get TV stations that both antenna web and TV fool don't even list for my location.

I also built a hoverman which worked pretty well on UHF but I need VHF-hi reception and the hoverman didn't work at all on VHF-hi (nor is it supposed to).

The 4 bay with the oversize reflector will get all the UHF & VHF-hi stations I have in my area with-in 40 mi. with solid signals.