4-bay Using Metal Tubes Instead Of Bowties

[BOUVAL]

On this rainy lousy day (my birthday), please let me introduce to you my new build called CEB 4 bay antenna. CEB for Chubby Easy Build

The 4 bay antenna is made of 8 lengths of 7 inch aluminum tubes (1 inch O.D.), 4 lenghts of 7 inch PVC rigid conduit (1/2"I.D.) and 16 rivets.

Made 4 grooves on the PVC spine and glued the four 7 inch PVC conduits. The phase lines are made of 14AWG electrical wire. As you can see on the pictures, the 16 rivets are holding the tubes together and 8 of those rivets are holding the phase lines.

As for OTA reception, it's as good as the bowties and I still get real channels 14, 22, 32, 38 and 43 at 86 miles. Forget about Balm's channel 13 Did not have time to check the quality of the signals.

Compared with the bowties, this antenna used with the same reflector (20" wide) has a narrower beam width. So an 18" wide would be needed for a wider beam or 24 inches to make it more directional.
http://www.tvfool.com/?option=com_wr...267eee22a6d100

[ota_canuck]

Hey! Nice simple structure! Good job... I don't think I"ve ever seen anything with straight horizontal bay elements modeled.

CEB = Chubbtenna

It would be interesting to see the effect of having 9" elements at the top and bottom bays with the 7" elements in the middle bays. Maybe 13RF would be possible.

________

Edit:
Second thought!,... 29" narods placed 1/2"above the top elements and 1/2"below the bottom elements would likely produce some VHF enhancements.

[BOUVAL]

Thanks Ota_Canuck,

That was one of my thoughts when I built it, because 2 times 7" is near the quarter wave length (λ/4) for RF13.

As for 9" mixed with 7", just sorry to deceive Balm again.

AntennaCraft U1000 and Winegard HD 4400 have horizontal elements, don't know if they were modeled. With chubby elements it's another ball game.

29" narods placed 1/2"above the top elements and 1/2"below the bottom elements would likely produce some VHF enhancements

That I can try this week.
.

[nezdepain]

The bowtie is supposed to provide more bandwidth than a dipole. See the "bowtie" topic :http://www.hdtvprimer.com/ANTENNAS/glossaryA.html. I don't know if that applies to db4s

[300ohm]

The bowtie is supposed to provide more bandwidth than a dipole.

Yes, but FAT (relative to the wavelength) dipoles behave like bowties.

The 4 bay antenna is made of 8 lengths of 7 inch aluminum tubes (1 inch O.D.), 4 lenghts of 7 inch PVC rigid conduit (1/2"I.D.) and 16 rivets.

Post the other dimensions, so it can be modeled.

The reflector to element distance looks extremely short, but that could be an optical illusion.

[ota_canuck]

300ohm, are you planning on doing nec modeling this?

What do you think about adding 29" narods?

Those straight horizontal elements vs wiskers? The only downside with straight horizontal bays is that it becomes an eight-seater for birds Can it be modeled with eight birds?

And,... should those elements be lengthened closer to 8" to avoid extending into the 700Mhz+ ranges?

[rabbit73]

BOUVAL:

What I mean for my 8 bay stack is, instead of using a co-phase harness arrangement like you did, I'd cross the phase lines BETWEEN the 2nd and 3rd bay from the top and 2nd and 3rd bay from the bottom.

The description of your 8-bay is not clear to me. Is it a vertical stack of 8 bays with just one feed point in the center? Can you please post a diagram of the antenna that shows how you intend to cross the phasing lines?

[BOUVAL]

I know that I'm french and spanish speaking, but my english was clear.

bay 1

bay 2

Crossing

bay 3

bay 4

Center feed point

bay 5

bay 6

Crossing

bay 7

bay 8

.

[mclapp]

Yes, I guess front view would be a better description of that co-phase line diagram.

I should have said bay spacing not phase line seperation as the critical factor in VHF-hi tuning, 8 1/2" or so would be best for ch13.

I had a 8 bay 9 1/2 x 9 vertical stack that would pull in distant ch's 11 and 13 when a same sized 4 bay would only intermittently lock at best, the 9 x 8 1/2" 4 bay would lock them most of the time but still not as good as the 8 bay. I'm thinking of putting together a 9 x 8 1/2" 8 bay to try at a similar location to see how much better that would be on those channels.

I tried a 6 bay once that used one phase line to tie all bays together and found it performed marginally better on a few channels than a 4 bay at best.

Your english was clear enough I just wasn't getting it, the last description cleared it up for me perfectly.

BTW you would have to do a cross between every bay except the center feed point, maybe that's why it wasn't making sense to me. The wave changes polarity every 1/2 wave and the bay spacings are approx. a 1/2 wave apart so you have to swap polarity between the bays to keep them all working in phase.

If the RF wave and bay spacing are not perfectly matched things could get of of phase quite fast when trying to use one phase line to feed the whole 8 bay vertical so that doesn't seem to work very well, the outer bays will get out of phase worse and worse as you go away from the frequency the bay spacing is perfectly tuned for.

[rabbit73]

mclapp is correct, a cross between every bay except the center. It didn't make sense to me either, which is why I asked. I wouldn't want you to make an antenna that doesn't work well, but it might be interesting to make one each way and compare them.

He is also correct about the outer bays. The antenna will work OK at the design frequency, but as you move away from that frequency the phase errors at each bay add together at the end bays.

The purpose of the phasing line is to make the direction of the current flow the same in each halfwave element, so that all elements are working in unison, that is, in phase.

Note that the current reverses direction after it travels a halfwavelength.

Reference:
ARRL Antenna Book (not Handbook), 18th Edition, Chapter on Multielement Arrays, and Chapter on VHF and UHF Antenna Systems, section on COLLINEAR ANTENNAS.



The phase line crossing must continue in the same manner if more bays are added at the ends, as in your 8-bay, so that the current flow in all the elements is in the same direction.

[holl_ands]

It's pretty simple if you follow the signal flow:

Obviously, the two inner Bays must be fed IN-PHASE, where
the signal at their feedpoints is electrically phase shifted the
SAME amount as it flows from the center to each feedpoint.

From the Inner to the Outer Bays, there is a phase reversal
due to the length of the feedline, which is cancelled out by
the physical crossover.

If another pair of Bays are added to the top and bottom,
each will need a physical crossover to cancel out the electrical
reversal due to the feedline length. If yet another pair of
Bays are added even further out, physical crossovers are
needed yet again.

However, mclapp's stacked 4-Bays (forming a Vertical 8-Bay)
use a SEPARATE feedline structure that connects the center
feedpoints of each 4-Bay antenna. Obviously, each 4-Bay
would need to be fed an IN-PHASE signal, so the feedline
lengths must be the same length with NO crossover.

It is ALSO possible to feed the array from the BOTTOM Bay,
which would require crossovers between each and every Bay,
presuming that they have equal separations between them.

You would need to model each to see which was better....

BTW: Since all of the bays are in the same vertical plane,
they all need to emit RF signals with SAME phase in order
for the resultant signal to be directed forward as desired.
So crossover reversals are ONLY needed to fix phase
reversals in the Feedlines and NOT to overcome the
physical Bay separations.

[BOUVAL]

Thank you gentlemen for sharing your great knowledge, I really appreciate it.

I've decided to go with the crossing between every bay except the center feed point. As I had plenty of aluminum tubes for free and had good results with my 4 bay, I decided to keep on using them. I started my build a few weeks ago and I'll explain what I've done so far.

Step 1: I hacksawed 16 aluminum tubes. Tubes lenght 9 3/4", 7/8" OD, 3/32" thick wall
http://i1140.photobucket.com/albums/...g?t=1358223694

Step 2: I bought two 10 foot lengths of 1" ID rigid PVC conduit. I'll use them as spines, one in front for the elements and the other in the back for the reflector. Then I filed 8 notches (flat indentations) on one to receive the elements.

Step 3: As I could not find any rigid PVC conduit to fit into my aluminum tubes to hold them horizontaly on the spine like my first two builds, I bought a 1/4 inch acrylic sheet and had someone cut 8 pieces 13" by 1 1/2". I laid those acrylic strips on the basement floor and used the plywood sheet to align them perfectly and glued them to the spine.
http://i1140.photobucket.com/albums/...g?t=1358224203
http://i1140.photobucket.com/albums/...g?t=1358223096
http://i1140.photobucket.com/albums/...g?t=1358223091

Step 4: I hung the antenna at the basement ceiling to put the phase lines
http://i1140.photobucket.com/albums/...ps234a3e43.jpg
http://i1140.photobucket.com/albums/...ps0f516bc4.jpg

Mr 300ohm, yes I'll put broomsticks in the PVC conduit.

To follow . . .

[mrvanwinkles]

Bouval, and all ... I watch and read with great interest your 8 bay project.

I have many tubes from lawn chairs cut to 8 inches long - but am concentrating on other projects and am busy with other things right now.

Thanks . Nice job.

I look forward to hear any reception results - etc.

[holl_ands]

Bouval's UHF Vertical 8-Bay using Fat Dipoles:

Using what I THINK are Bouval's dimensions (let me know if there are any corrections), I uploaded 4nec2 analysis results to fol. link:
http://imageevent.com/holl_ands/mult...aydipolenorefl
PS: Included 4nec2 file started out as a 4-Bay Bowtie, which hasn't yet been cleaned up to eliminate some duplicate variable SYmbol definitions. Asymmetry in the pattern on the highest frequencies could probably also be reduced if I changed the middle crossovers to Left Hand vice Right Hand in all of the others...although it will still won't be PERFECTLY symmetric due to the odd number of crossovers...

SWR is under 2.0 and Max Gain = 14 dBi, however there is significant Gain Loss above about 650 MHz.

This Gain Loss can be reduced or eliminated by Rescaling the entire Antenna so that it is somewhat SMALLER (shifts Max in Frequency Response curve to the RIGHT). However, I'm investigating some simpler mods that would involve 1) shorter Dipole Elements (only) and 2) shorter Dipole Elements plus wider Feedline Separation and "optimized" placement of the Crossover (close to the Outer end of each Feedline). So standby for a few days before grabbing a hacksaw.....

[holl_ands]

UHF Vertical 8-Bay Dipole - NO Reflector:

Finally uploaded 4nec2 analysis of several alternative for fixing the Gain Loss at high frequencies in Bouval's original DIY project. Simplest fix is to simply hacksaw off the ends of the Dipole Tubes so that they are 3-in shorter...this was pretty good, with both improvements and degradations if the Feedline configuration were changed. The best configuration was to change both the Dipole-to-Dipole Separation and Feedline configuration, as illustrated here:
http://imageevent.com/holl_ands/mult...aydipolenorefl

Dimensions per Bouval's DIY posts starting with this one: http://www.digitalhome.ca/forum/show....php?p=1469742
Original (9.75"x9.625"): SWR Max = 3.1. Max Raw Gain = 14 dBi, but significant Gain Loss on channels in the mid-40's and above range.

Mod1: First optimization to fix Gain Loss problem was to contemplate simply cutting 3-in off the ends of the Dipole Tubes. This yielded a fairly flat frequency response curve, except for some loss on lowest freqs and higher SWR on highest freqs.
Mod1 (6.75"x9.625"): SWR Max = 3.3. Raw Gain = 9.8 dBi (470 MHz), 12.0 (674 MHz), 11.0 (698 MHz).
So a classic trade-off of Gain for wider Bandwidth. Mod1 provided higher Gain on 698 MHz than Mod2a or Mod2b below.

Mod2: Second optimization found optimum Dipole Tube length while modifying the Feedline Separation and position of the Crossover relative to each OUTER Dipole. Bouval's Dipole-to-Dipole Separation=9.625-in was NOT changed.

By adjusting optimization parameters it is possible to influence the relative importance of Gain vs SWR:
Mod2a (7.0"x9.625"): SWR Max = 3.2. Raw Gain = 9.8 dBi (470 MHz), 12.0 (674 MHz), 11.0 (698 MHz).
Mod2b (8.0"x9.625"): SWR Max = 2.1. Raw Gain = 8.7 dBi (470 MHz), 13.2 (674 MHz), 8.7 (698 MHz).
So Mod2a provides "flattest" freq response while Mod2B provides highest Max Gain but at expense of high/low freqs.

Mod3: Third optimization found optimum Dipole Tube length while modifying the Feedline Separation and position of the Crossover relative to each OUTER Dipole. Dipole-to-Dipole Separation was also allowed to change in this more generalized optimization search process to compare an "optimum" configuration to the above alternatives. To minimize Azimuthal Pattern asymmetries, Crossovers alternate as to whether Right or Left Feedline Wire hops positive or negative on X-Axis.

Provides excellent SWR with more Gain on highest freqs to overcome extra loss in the Balun, Coax and (if used) RF Splitter(s), although maybe more than needed for a mast-mounted Preamp with only Balun Loss a significant factor.
Mod3 (8.0"x8.5"): SWR Under 1.8. Raw Gain = 9.1 dBi (470 MHz), Max = 13.8 (686 MHz), 13.4 (698 MHz).

BTW: There are only a few of the many possible optimization results...which can vary quite a bit depending on the chosen optimization Gain & SWR Targets and relative importance weightings for these factors.