bmoore314

I made me yet another antenna, and so far this one works the best. For 15 bucks I got me a dish network dish, and converted it, using a single pineapple can for the waveguide. So far it is working great, pretty directional, but once I found a direction, I no longer had to adjust it (unlike my cone type antenna) I just finished it, so its a bit late for pictures, but I will have some pictures up tomorrow.

Brian

bmoore314

As promised, here are the pics of my new satelitte antenna. Seems to work very well, signal I used to not be able to get during the day, it was very sporadic is now a solid yellow in NS, sometimes a light green. My estimate on gain is around 18 dbi, maybe more. It is a strong performer, and with the fairly stiff adjusting, it was much simpler to get it to stay pointed in one direction. It seels to vibrate slightly in the wind, but that doesnt seem to affect the signal much at all. The bas hasnt slid any, it is pretty sturdy. Its not nailed to my roof or anything, just sitting on there with its base ( a 2x4 and some 1x3's ) I used a protractor and measured the angle and placed the wavguide in the same position and angle as the original lnbf's and then slid it up and down along the wood to get the best signal.

http://www.forgottenfords.org/dish1.jpg

http://www.forgottenfords.org/dish2.jpg

http://www.forgottenfords.org/dish3.jpg

http://www.forgottenfords.org/dish4.jpg

lincomatic

nice setup. i like your portable mount...good idea to avoid damaging the roof. too bad my roof is sloped instead.

btw, you might want to try trevor marshall's biquad for the feed instead. it kicks the can's butt in gain and is simple to build.

where did you get the dish? ebay is impractical because the shipping costs too much. i'm thinking of getting one of the cheap grids from www.aerialix.com to try instead.

__________________
~lincomatic

bmoore314

I picked the dish up at a yard sale for 15 bucks, came with the mounting pole too. My roof is actually sloped pretty good, the fourth picture shows it, I held the camera pretty much level, also in the fourth picture it has the peak of roof compared to the fence. Even though its not level, it stays put pretty well, that 2x4 is almost 4 foot long and it is redwood, weighs more than your average cheap pine 2x4. The hardest part was making the little 'A' mount to get the can in the right position, since my miter box only goes as low as 45 degrees, but the angle of the cut needed to be 30 degrees. Also trying to figure out how to hold it together was pretty tough. I ended up mashing it in a vice, and drilling some pilot holes, then putting some drywall screws in (i have a whole bunch of em). Do you have a link for that bi-quad antenna, I tried a biquad design, and it failed horribly, maybe I did something wrong or something. One idea I had that I might try out is building a mini dish as a collector or something similar.

Brian

lincomatic

maybe you did the feed wrong. check out the biquad on my homebrew page:

http://www.geocities.com/lincomatic/homebrewant.html

i messed up the feed at first and had horrible performance. there's a link to trevor marshall's site from there too.

__________________
~lincomatic

bmoore314

That 2nd picture you have helps a bunch. My bi-quad that doesnt work is exactly the same as your first one, seems that tonight I will have myself another project I'll let you know how it turns out. Looks like that bazooka cantenna you made turned out pretty nice, aluminum foil sure looks better than my hacked together metal screen (i didnt have any aluminum foil at the time) Im sure if you made the conical section even larger, it would have an even higher gain. One thing I found out is that having a strong signal in netstumbler doesnt always mean a good connection.

So far on the satelite antenna, its been connected for a few days straight now, with only a few 10 - 20 second dropouts a day (maybe once every couple hours) but other than that, it seem to be a very fast link. I'll try and post some more pictures when I get my biquad completed.

Brian

lincomatic

you were right, about the reduced xmit power due to the conical collector; that's why i kept mine small. since the remote AP is lucent, i was able to run the CM link test, and it shows the received signal at the remote site is 3db lower than the received signal at the cantenna. 5 dB is pretty marginal already; i'm afraid that making the collector larger will make me lose too much xmit signal.

actually i think your screen idea is better for holding up against wind & rain, i only used foil because all the screening i have in the house is fiberglass.

i've been having trouble coming up w/ a decent way to mount the cantenna on the roof, too. since it's rented property, drilling is out of the question. i was thinking of using heavy wood like you did, but i still can't come up w/ a decent rig for tweaking the aim until i get it right.

i wish i could think of a way to gain about 3dB more w/o going to a dish; the best LOS i can get is from the front of the house, and my bro doesn't want an ugly dish there.

__________________
~lincomatic

bmoore314

You might want to try rotating the cantenna, so the probe comes out the 'side' instead of the 'bottom' this will change the polarization of the cantenna, and even though you will have wrong polarization combo, it seems that having the probe parallel to the ground, it reduces reflections and noise, which gives you a cleaner signal. Another idea is (if possible) put a conical cantenna on both ends, the reduced xmit will be picked up by the high gain of the antennas on both ends. True story, one guy made a conical antenna out of a hot water heater, and put it in his attic to recieve hbo, when it was broadcast at 2.3 ghz. The gain was huge, around 40-50 db, but it was recieve only, didnt have to transmit.

Another idea I have tried is to make a slot for the probe in the cantenna, instead of a hole, and adjust the probe forward and back slightly in order to find the spot with the highest gain. Your drilling could be off by a slight bit, and at these frequencies, a small ammount makes a large difference. Also make sure to buy high quality connectors, and high quality cable. The N connectors on my cable are expensive, silver, with a gold plated center conductor. Each connector runs about 8 bucks or so. They have low loss compared to most of the cheap N connectors. My cable is some rg8u (i think) which isnt the best, but it works. When I make my cable longer, I am going to use some lmr400 (i think, its the one that is about 3/8" diameter) which is expensive, but works the best. Compared to having my antenna hooked directly to the pigtail, and having the 10' section on it, I get about 6-7 db less SNR. (2 male n connectors, one female-female n connector, and the male n connector on the pigtail)

One interesting method of getting a signal, is an indirect signal, if you have a rather solid object directly between the two points, that is blocking it, aim both antennas at the peak of the solid object, and you might get a signal. The object difracts the sigal a bit, and makes kind of a shadow where the signal is weaker, but present. You can also try and get reflections of objects, for example, you may not have a direct line of sight to each other, but you both might have a direct line of site to someone's garage, or other large, flat area (side of a large building maybe) the signal could reflect off that building and possibly give you a decent connection as well.

With my cantenna's, I could get a signal by pointing in three different directions, one directly at the source, another about 45 degrees left of it, (pointing at the edge of a tree that is in the way) and for some reason if I pointed it about 15 degrees to the right of it, and about 20 degres above the horizon, I could get it as well.

Brian

lincomatic

brian,
thanks for the detailed reply. the probe location tweaking idea is good. i plugged the into my calculator, and since the AP is set to Ch11 and my antenna is tuned for CH6, there is a .1" diff in the probe position. wonder how that affects the SWR & gain. would be nice to be able to plot SWR vs channel for a particular design.

i'm using decent quality N jacks w/ gold plating, too, although the bodies are chromed not silver (won't silver corrode outdoors?). i'm hooking it up to 30ft of LMR400 to get it into my bro's house, which makes about 2dB loss not including connectors & pigtail.

i've tried just about every pointing direction i could and so far the best is about 12dB directly connected to the pigtail. i don't have control over the remote AP (no i am not stealing a connection...my bro is a prof @ the college, and they just haven't gotten around to installing AP's on his street yet), so I'm pretty much stuck w/ whatever the antenna setup is (which, by the way is a Yagi pointed away from me).

thanks again for the help. seeing your dish setup also has put some ideas into my mind about how to hang the antenna on the roof w/o drilling into it.

__________________
~lincomatic

rogerRabbit

Could be the shape of the radiation pattern is a major factor, as well? I've yet to test my cantennae but I suspect that the pattern will be strong around 30 - 40 degrees from centre, very especially if the diameter is sub-optimal for the frequency.

bmoore314

That could be the case, but I dont believe that it is the major factor. The signal that I got from an angle was equal to, sometimes even greater than the signal from pointing directly at it. It still could be the case that the signal spikes are from the radiation pattern of the cantenna, but I dont have anything close to the proper equipment to test that with.

Another good method of attaching an antenna non permanently to a roof is with some pipe and some of thoes cement filled lifting weights. Stack a bunch of weights on top of each other, and set the pole in it, should hold up pretty well if you can find a pipe that is a good fit for the diameter of the hole in the weights.

Brian

bmoore314

I also would like to add, having someone in your house plug in a hair dryer on the same circuit that is running 6 computers (one of them has 3 towers) 5 monitors, multiple clocks, really sucks ass. I was right in the middle of writing the last post when the breaker popped. My main machine (3 towers and a 21" monitor) is on a rather large ups, and was able to be shut down without issue. I still lost my post cause both my main gateway to the net, and my backup (dialup) gateway's died from the lack of power. Just a rant, ignore if you wish

Brian

lincomatic

i'm glad my main machine is a laptop. %$# power is always dying here for an hour at a time. lucky i can run off the batts for a long while. one time i had to retreat to my car, though, and run it via an inverter.

__________________
~lincomatic

rogerRabbit

I think cantennae have a tendancy towards patterns that are relatively uniform up to 30 degrees, in both planes, and fall off quickly over that. That could make the signal sensitive to objects within that cone.

fordem

Dedicated circuits for all your computers - it's the only way to go - and of course UPS to protect them all.