nite613

My friend and I have recently started stumbling. Not much luck so far, but I think we're having cabilng problems. This is our setup:

-Some old Pentium-class IBM thinkpad running Linux
-Dlink 650 with capacitor modification for external antenna
(I know, all you winstumbler boys are orinco, not prism2)
-A helical PVC yagi (http://www.guerrilla.net/reference/a.../2ghz_helical/)
-VW with black-tinted back windows
-Camera tripod fitted to antenna

Our plan so far has been to drive around scanning likely buildings/stores/microwave dishes on roofs/etc. Will this work? Can I point the Yagi through the car window, through a wall or two in a building, and directly at the base station? The theoretical gain on the Yagi is 15dB, are we nuts?

When we get just out of card range of a known AP (through walls/windows) and point the Yagi at it we start to see packets again, but they seem to be corrupted as the sniffer doesn't know what to make of them anymore. We're using 10-base2 network coax and connectors, is that the problem, or is our entire MO flawed?

-=nite (any help greatly appreciated)

Thorn

-<snip>- We're using 10-base2 network coax and connectors, is that the problem, or is our entire MO flawed?

nite,
The rest of your MO is fine, but the radio gods will be smacking you upside the head any minute now.

You need 50-ohm cables and RF (Radio Frequency) components, not the 75 ohm coax used for 10-Base2. This is completely different stuff. You need N connectors and LMR (or equiv.) cables. Substitutions using stuff that "kinda looks the same" won't do. The RF side of this is tricky, especially if you're only used to dealing with computers. You have two different disciplines working together here. Read through these forums, and you might also want to look over some ham radio sites.

Your antenna should be outside the metal of the car body, otherwise the signal is reduced or blocked. Also, I don't know about the VW, but some auto window tints are metallic films. If yours is, you may have a double-whammy against you for the RF signal.

Get a small omni directional antenna for most of your driving, and use the yagi only to determine direction.

One last thing: Creating an good antenna, such as the yagi you describe, is something that has been known to reduce seasoned hams to tears. It isn't as easy as it looks, and from the sounds of it, you're tackling it with no radio experience. Buy a good known antenna and cabling. Work with that, and then start to experiment.

BTW, what's with the camera & tripod?

Cheers,

__________________
Thorn
Sex and Violence. You can't enjoy one, if you don't survive the other. (And that works both ways...)

nite613

Thanks for the reply thorn! I've read all the forums and a couple months of history and am beginning to get a feel for this radio stuff. I've done a bit of radio before and quite a bit of electronics, so I have a bit of confidence when building stuff like this. I had a feeling the cable/connector was wrong, but my friend asked the guy at the store and this is what he acme home with. :P

I have a few more specific questions if you don't mind. First of all, the pigtail from the card. We've built it using a short length of BNC like the card you see on this page: http://kevlar.burdell.org/%7Ewill/antenna/ . I was wondering if using a converter as seen here: http://www.tux.org/~bball/antenna/connectors.jpg would work, and why, if the connectors/cable are so sensitive, the BNC connector/pigtail dosen't screw up the signal.

We intend to built an omni such as this one: http://homepages.wwc.edu/student/kyleti/discone.html for the roof of the car for sniffing out APs and then use the yagi to put some distance between our car and our AP... we're hoping for AT LEAST 500m, does that seem unrealistic?

Finally, the tripod is for helping to aim/hold the directional. In the back seat we can get easy 120degrees rotation out of one side of the car. It looks pretty elite.

One other thing I was wondering. I've seen many dishes on roofs that look like the ones on fab-corp, some of them I HIGHLY suspect to be 802.11. Do I have to get my yagi directly into the beam from the dish (likely disrupting their normal link) or can we just point up at the front side of the dish from a 30degree angle or so from the ground?

-=nite (thanks for your help!)

fordem

Those radio Gods will be smacking him same time the LAN Gods are smacking you.....last time I looked 10base2 was using 50 ohm coax

OneGuy

I've made plenty of terminators to prove it.

That coax will bleed like crazy at 2.4g as it was designed for 10Mhz, big diff. I won't even go into the bnc connectors.

Using that coax for 802.11b is like shooting yourself in the feet, yea both of them.

OneGuy

p.s. Hey Thorn, your not thinking of ArcNet coax at 75 Ohm are you. If you are your an old bastard like me.

fordem

Guess I'm showing my age here - but anyway ARCnet used 93 ohm RG/62. Only LAN I ever saw with 75 ohm cable was a proprietary system - Orchid PCNet.

On a different note - I was chatting with a guy here who's setting up a WISP, using 802.11b, and the cost of LMR came up - he said you don't need it - use RG-6 - the satellite stuff ..?..!

Believe me - that's when he lost my business

Thorn

Yup fordem, I guess I am showing my age. The last time that fiddled with any coax on a network was at least 10 years ago, and I thought it was 75 ohm... On the other hand, it could be senility is just starting in.

OneGuy, I AM an old bastard. Hell, I'm a grandfather, so it's official. It says so right in the fine print (Grandfather = Old Bastard) I'll have to read that contract again, I think I'm supposed to be grouchy, too.

nite, as you can see from fordem's and OneGuy's replies, we all agree in principle; the network coax is really not designed for the 2.4GHz range. You will get some losses from the correct cable and connectors; you'll get a lot more loss from the wrong cables and connectors. The 802.11b RF is very low powered to begin with, why kill off more of the signal than you have to?

I'm betting that that pigtail you describe does screw up the signal to a certain extent. Just because someone has published an account of how the hacked a pigtail onto their card, doesn't mean they are not getting some loss. The connector converter you mention MIGHT work, but it will definitely introduce another loss into the equation.

Too many little losses will eventually add up to no signal.

If you are just out wardriving, 500m is probably unrealistic under most circumstances. I have done it but it is rare. 50 to 100m is probably a lot closer to average. However, I've made point-to-point connections at well over 2k, using directional antennae.

Dishes are usually a pretty tight beamwidth. About 15 degrees or so. At 30 dregrees, you may or may not get any signal. It is hard to say without actually trying.

Cheers,

__________________
Thorn
Sex and Violence. You can't enjoy one, if you don't survive the other. (And that works both ways...)

nite613

It's the senility. I'm a network guy and I also am pretty damn sure that it's 50ohm. That's actually why we picked it! Forgot about the frequency issues though... oh well.

That's exactly what I expected. We will replace all the cable and connectors from the pigtail right through to the antenna with the proper gear and give it another go. Hopefully our yagi will then take us out of card range without passing only corrupt packets. :P

Damn. Oh well... sitting 50m from a building with a big funny antenna doesn't seem as fun as what we had in mind. We'll see what we can do though. I guess having a yagi with good gain doesn't help all that much when you're trying to talk to a low-power, low-gain omni. Make sense, I just don't like it!

Maybe I'll look into building a cheap repeater.

Fair enough. I expected the tight beamwidth (hey that's what they're for after all!), but I'll definately try out the more likely looking ones just in case.

Thanks a lot for the replies, I'm more confident than a few days ago, but still a bit disappointed that we can't sit with the antenna in the back seat 3km away from our mark Oh well, we'll figure something out.

-=nite