Cette page était sur le site de W0LMD; mais le lien ayant disparu, j'ai reproduit ici l'article que j'avais conservé. (F5AD)
Making a Circularly Polarized 2401 MHz Patch Feed
My W0LMD patch version has the usual 2 metal plates. The reflector backplate is a 4" in diameter piece of 1/16" aluminum. A 1/8" hole is drilled in the center, and 11/16" out from the center is a 3/8" hole through which the center conductor of a type N connector extends. Most of the time I use a female type N chassis connector but sometimes I use a male type N connector for directly attaching the patch to the input of a downconverter. The type N connector, on the side of the reflector away from the patch, is attached to the reflector plate by four pan head 4-40 bolts with the heads on the side of the reflector facing the patch plate. The connector is slightly skewed clockwise, about 30 degrees, when looking at the type N connector from the patch side of the reflector, which provides better clearance for the tuning screw on the patch.
The new patch plate was smaller, 2 3/8" in diameter, than my non circular 2401 MHz patches and was made from some hardware or hobby store 1/16" thick brass plate. I drilled three 1/8" holes in this patch plate. The first hole was in the center of the patch plate, the second hole was 11/16" out from the center (mates to the type N center conductor on the reflector), and a third hole is 1/2" clockwise from the type N hole, also 11/16" out from the center hole. I then carefully tapped this 3rd hole with a 6-32 tap.
I put a 1/2" long pan head 4-40 bolt through the center hole of the reflector from the side away from the patch, and fastened the center bolt to the reflector with two 1/4" size 4-40 nuts. This provides the patch spacing, which is about 3/16". You can trial fit the patch to the reflector, verifying that the second hole in the patch plate just fits over the center conductor of the type N connector. The third hole, tapped to 6-32, should be clockwise from the 2nd hole for LHCP. If counterclockwise (anti-clockwise for you Brits) from the 2nd hole, the patch becomes RHCP. Since the dish reflection inverts polarity, the desired RHCP dish needs the clockwise LHCP location for the third hole.
Now carefully screw a 1/2" long flat head 6-32 bolt into the threaded third hole on the patch, from the side of the patch that will be facing the reflector. Leave the head about 1/8" above the patch surface for now. Attach a 1/4" 6-32 nut on the other side of this tuning bolt to act as a tuning lock. Finger tighten only as this 6-32 flat head bolt is riding on a single thread. If you strip the threads in the brass patch plate you can always solder the lock nut to the plate. But be careful.
Fasten the patch to the center bolt with another 4-40 nut, and solder the Type N center conductor to the 2nd hole in the patch. I experimentally determined that the best signal from this patch feed resulted when the flat head of the 6-32 bolt was spaced .020", the thickness of three postcards (direct mail "Bingo cards") from the surface of the reflector. Lightly snug up the tuning lock nut when this spacing is achieved. Later, if you are happy with the performance, you can seal the nuts with some fingernail polish.
Two small disks, a type N connector, 6 bolts and 8 nuts later, I had a really good performing 2401 MHz patch feed for my big center fed dishes. It takes me about 1/2 hour, tops, to make this great patch feed. When the squint angle is high (AO-40's transmitting and receiving antennae aren't pointing my way), this feed will produce a strong uniform signal with much less QSB compared to a linear feed.
Show an' Tell
The disk in the upper left is a 4" reflector showing the side that faces the patch. The pan head 4-40 bolt heads are on this side and the center 4-40 bolt is facing upwards with two 1/4" 4-40 nuts providing the 3/16" reflector to patch spacing holding it to the plate. The reflector plate to its right is the outside view of the reflector with a type N female connector attached to this side.
The 2401 MHz patch disk at the bottom left shows the side of a patch disk that faces away from the reflector. Notice the hole in the center, a hole at the upper left that mates with the center conductor of the type N connector in the reflector and the clockwise third hole that was tapped 6-32 has the threaded end of a 1/2" long flat head 6-32 bolt with an attached 1/4" 6-32 nut. The patch to its right is the flip side which will face towards the reflector, with the center hole at the bottom, the type N center conductor hole now at the right when flipped, and the 6-32 threaded hole has the 1/2" long 6-32 flat headed bolt and locking nut with the threads facing upwards. The 2401 MHz patch plate to its right is the flip side that faces the reflector, showing the shiny flat head of the 1/2" long 6-32 flat head bolt.
The 2 disks at the bottom right are finished 2401 MHz circular patches, a normal LHCP one and the one at the bottom right has the tuning screw at counter-clockwise for RHCP. If truly circular, it should have a definitely weaker signal when attached as feed for a dish receiving a RHCP signal. It does show about a 10 db weaker signal.
The 2401 MHz patch feed to left of the helix is my "Safeway Special", which has a 4" peach can cut to 1" high side ring reflector and a Campbell's soup can lid cut down to 2 3/8" with LHCP polarization. This patch feed is slightly weaker than the patch below it due to under illumination, but would be good feed if you are experiencing off axis QRM from 2400 MHz devices.
In the upper right is my 2400 MHz test signal generator. It has a 50 MHz TTL crystal oscillator (available 3 for $1 from Electronic Goldmine (800) 445-0697 -$10 minimum order (That's a lotta xtals!) - they have other junque listed at http://www.goldmine-elec.com - or put a 50 MHz TTL crystal oscillator, p/n 76831, from Jameco (800)-831-4242 for $2 - no min order) inside the die cast box along with a 4.8v NiCad rechargeable battery. 50 MHz is handy as the downconverter picks up the harmonics, but the 2M xcvr doesn't. The 3 turn RHCP helix is tied directly to the output pin of the TTL oscillator, heat sinked to the top of the die cast box. The wider angle RHCP patch at the bottom right can also be attached to the signal source. It puts out a weaker signal over a wider beam pattern .I have another similar test signal generator that uses a 16 MHz TTL crystal oscillator, Jameco p/n 27908. I made 1/4 wave verticals for 145 MHz, 435 MHz and 1296 MHz for testing antennas on these bands. I also use a Gardiner 2400 MHz signal source with a linear dipole & reflector output, available from K5GNA, mentioned elsewhere as an AIDC downconverter source, for $35. This puts out a very strong signal on a 9.6 volt rechargeable NiCad battery that should not change in received signal level when rotated if the dish & feed are really performing in circular mode.
I made the 2401 MHz patch feed, shown at the left, that had a reflector the same size as the C band Chaparral feed that had been on the TVRO dish. The bolt hole spacing (4 1/8" square) was the same, so it was a simple unbolt and replace operation.
A single band patch antenna for 2401 MHz can be made with a 6 3/4" shiny aluminum disk, a 2 3/8" soup can lid (or get fancy and obtain some 1/16" brass sheeting @ ACE hardware), and a type N, male or female, connector. I make this circular polarized by capacity loading it out of phase. I concentrate on fully utilizing the dish, netting a few extra db from a higher efficiency dish. Making it really circular polarized gains 3 db which is equivalent to doubling the surface area of your dish and results in less QSB under marginal conditions.
Retour au menu antennes