Looking to build a c/ku-band horn antenna
For the lnb side, someone who made some horns for ku found they worked better if you made it to the scalar ring size, not the lnb wave guide size. But from what I have read, the ring is to block noise from past the edge on dish setups. I have also read that putting a rolled edge on the aperture end.
I was thinking of using the DMX741U LNB which covers both ku and c band. The web sites say that it includes a dielectric plate to work with both linear and circular. But when you look at the pdf manual it says it does not work with linear when the plate is used making it one or the other, not both. Others have said you can still get linear but at reduced signal level. It seems to me there would be a way to selectively convert circular to linear. It's ether that or build a diseqc command controlled device that moves the plate in and out the side of the guide.
I use conical horns for KU-Band and 2.4Ghz.. they work great in KU, a 20cm diameter version is enough to receive TV from a 54dBW satellite. I′ve made a 45cm diameter/ 2 meter long conical horn with leftovers of metallic cardboard which work very well.
But a Conical horn designed for 4Ghz will work very badly in KU-Band.,because bigger the frequency, longer will be the Conical horn.
Use HDL_ANT 3b4 to make Horn/conical Horn calculations and create templates up to 25dBi in any frequency. It′s a great freeware..
search in google for: My collection of ku homemade antennas to see some of my versions and information.
I′m no expert in theory:)
Hi, I saw that thread and the one linked to it and I found this: http://www.satcrazy.com/horn_antenna.php. I tried messing with HDL_ANT 3b4, but it only designs to 25db and any time I tried giving it numbers for something bigger, it just repeated the options to enter the values.
From what little I have been able to gather, you design for the lowest frequency and while it may not be optimized for higher, the increased gain at higher frequencies offsets that. Also, when trying to use that program, it gave the same angle no mater the frequency.
3.4Ghz min.52mm 60mm 25db dia 690.5mm len 1620mm flare 1/2 angle 11 degrees
10.7Ghz min. 16mm 50mm 25db dia 200.2mm len 385mm flare 1/2 angle 11 degrees.
Besides, if it didn't work at ku, it's big enough to just mount a smaller horn on it at right angles (no need for the right angle reflector to feed it.) C/Ku lnb's are just to lnbs grafted together.
I am also now looking for info on gaussian corrugated designs. It increases the efficiency alot. But I don't know if that is for transmit only. and using dual corrugation, you can make it shorter. Google 22a-UPNA_AG_2_A4.pdf
HDL_ant has a limit of 25dBi to generate horn/conical horn calculations or templates..but i′ve the formulas that hdl_ant use to generate those calculations (you can find it also in the sourcecode of HDL_ant), if you want i can put here the formulas (it′s simple calculations).
It gave the same angle, but a KU-band conical horn with 700mm of diameter, will be more than +-5 meters long.
There are many variations of Conical Horn, the advantage of Conical Horns is the 3dB extra due to the low noise, but many of those variations have a boost in gain, but it will loose that 3dB extra..
Variations of conical horns that i′ve found:
-Dielectric conical horn (hard to built)
-Multiflare conical horn (easy to built)
-Multiflare dieletric conical horn (hard to built)
-Metallic disc inside Conical horn (very easy to built, +50% of gain(?))
-http://www.afcsat.com/SatelliteConicalHorn.html )cant recall the name of these type of antennas
That corrugated horn is a bit difficult to built.. i recommend first to build a KU-band conical horn (with the template + alluminium cardboard you build one in 30mnts for less than 2€/1usd) just to have some idea how are conical horns..
Imagine you have that 700mm diameter / 1620mm long of C-Band to a satellite at 35o, the "mouth" of the conical horn will be 35o elevation in the air..you will need a heavy support to get a stabilized antenna, unless you burried the cone in the ground (just kidding).
This is the biggest problem with conical horns, for Satellite communications, we need high gain=large diameter=very long conical horns.
I never built one for C-Band, for 2.4Ghz only built 14/15dBi antennas but for KU-Band with <50dBW signal, it′s impractable to built.. i didint tried some variations of conical horns, because for KU we need a antenna that work ok in a large bandwith (10.7-12.7Ghz) and some variations are critical to frequencies..but in C-Band there isn′t that problem :)
Yes, please post it. www.w1ghz.org went down a day or two after I started looking into this. I got the file, but not the source. older version are around, but I think the part I'm after wasn't in them.
A guy with a phd in antenna engineering replied to a post one of my post on another board: http://antenna-theory.com/phpbb2/viewtopic.php?t=86
Not sure what you mean by Dielectric conical horn unless you are talking about corrugated. Corrugating would be a waste of time according to his reply.
By Multiflare conical horn I think are talking about gaussian profiled horns.
Metallic disc inside Conical horn. There are horns with 1-4 ridges that run the length of the horn and they are to increase the bandwidth of the horn, but those where all pyramidal horns. Haven't seen any with metallic disc inside. Have a link to some info?
I saw that link you posted. It's a larger version of what I have in mind. check this one out. It's the one that gave me the idea to go right angle so the horn is rotated instead of pivoted: ../imgqa/eboard/Antenna/rf-yj1qv55lbv3.jpeg
Nice thing about making it from styrofoam and foil. Keeping it from blowing away in the wind will be the main problem and a con has less wind loading then a dish sail.
These are the formulas to generate a conical horn (according to gain) all measures are in mm:
A= wavelength × ( 10^ ( (Gain+2.82)/ 20)
B= A/PI -this gives the diameter of the antenna
C= B×B / (3 × wavelenght)
D= SIN-1 × ( (0.5 × B ) / C )
E= 0.5 x (( B - Waveguide diameter ) / (Tang D) -> this gives the lenght of the cone
--------------------------------------
Example to confirm it: 26dBi , 11750Mhz , 25.5mm , Waveguide 19mm
A= 703.947
B= 224.073
C= 656.324
D= 0.1708
E= 594.4
After you done the calculations, use HDL_ANT to confirm.
-------------------------------------
Great straight-forward answer that guy gave :D
About the metallic disc inside the conical horn, you can find some infomation here
http://www.satellites.co.uk/forums/f...sh-home-6.html (see post 141). On post 202 (page 9), you can find some information about conical horn reflector.
Dieletric coated conical horn, is a conical horn made of Dielectric material (er=2/2.2, etc) coated with metal. there is alot of information on the web about these antennas.
The problem is almost 90% of the articles are in IEEExplore, which is a paid site, unless you have the oportunity to go to a university, you can download it for free.
Thanks for the info. If it does come out to 14m+, the project is sort of dead :(. Need to check about focal point also. Different for different lnb's. Also need to dig up more about Gaussian profiled horns. Found some more info that indicates that has far more to do with improving a horns efficiency then corrugating, pushing them up to 80-90% as well as improving bandwidth.
Found another long thread about home made dishes. nothing about horns, but nice stuff on dishes:
http://www.vetrun.net/forums/showthread.php?t=1077
http://www.vetrun.com/forums/showthread.php?t=1700
Yea, Alot of the sites that looked like they would have usefull info are pay. Defeats trying to keep the cost down.
That was reason that i "abandoned" conical horn antennas, the antenna becomes too long for satellite TV. I still use a 26cm diameter conical horn to a 54dBW, works very well.
You can also look for a Lens antenna (only work one polarity at time), HDL_ANT gives all calculations and it′s a multibeam antenna. I have tested those to KU, and worked like a charm for multibeam links, but since the antenna only work with one polarization, it was not usefull.
Now i′m working on this antenna http://www.alshawibt.com/files/skycone26.pdf , 32dBi, 26cm diameter= +150% efficiency? i dont know how that′s possible..
I don't see where you get 150%, but I don't think that is possible. It would have to have active gain (amp) and that boost noise as well as signal. Did you try the root address to that pdf? they have a skyhorn2 also :) Looks like some kind of offset horn.
The metal lens antenna not only would only be good for one polarity, it would only work for linear, not circular. If you really want to look into lens:
http://www.farragsat.com/
http://en.wikipedia.org/wiki/Luneburg_lens
http://www.integratedsoft.com/News/R...g-lens-Antenna
http://www.sumitomoelectricusa.com/s...Q/luneberg.cfm
http://www.radar-reflector.com/wp-co...09/jina_gb.pdf
I don't know what Farrag uses for materials, but most of the stuff is pricy and hard to work into proper shape without compressing it any.
26cm /27.5dBi /11.75Ghz= 54% efficiency
26cm/32dBi / 11.75Ghz = 152% efficiency 8O
He afirms that conical horn with only 26cm diameter have a performance of a 70cm dish...if it′s true, i want to make one!
I already made a luneberg antenna, it′s my proud...because it was made of sugar lol
search for "Lutenna" in google. The problem with my single dielectric material version (sugar ER=2.0), only work on smaller versions (25/30cm), >30cm doesnt work better than 25cm. I study a lot Farragsat antennas and never came close to the 25cm with 140% of efficiency like he affirms.. a 10cm diameter lutenna have a efficency of 120%, but bigger the diameter, smaller the effiency
The best explanation that i found was this:
http://img338.imageshack.us/img338/295/77096461.jpg
But you′re correct, it′s hard to make a Luneberg original antenna, it needs at least 5/7 different dielectric materials, guys like us, who make homemade antennas is impossible to acomplish that. But check the farragsat antennas, he uses a globe, so he must be using only 1 or two dielectric materials too achieve that performance.
Got a chance to mess with this a bit more. Started running through the formula and using that, the gain is solely based on the diameter and it puts it in the same size as dishes. First I went to some sites and noted the rated gain which was listed at 12.5Ghz. For dishes close to 1m the gain was around 40db
36" - 90cm 39.6 dBi ----------------------------------------------------------------------------------------- 39.36996" - 1m Gain at mid-band 11.2 12.1 12.6 39.83 dBi 40.50 dBi 40.85 dBi
I'm puzzled about formula for D. SIN-1 should produce a negative number. And the number in the sample for D didn't calc that part.