marine antenna choices?

"Bill Kearney" hath wroth:

Panel/patch antennas are probably a good choice. Most are easily sealed and inherently waterproof. They're also fairly cheap.

big question is how big and how much gain. Too much gain will make the antenna excessively sensitive to vessel rocking and swinging. Not enough gain will result in insufficient signal strength if the signal is too low. My guess(tm) is to size the antenna by the appropriate -3dB beam width and take whatever gain results. From a random assortment of panel antennas from the above URL. Gain -3dB beamwidth dBi degrees 8 60 10 60 12 40 18 18 I would think 8 to 10dBi gain would be about right. If the antenna is easily removable, it might be useful to have two, low and high gain.

Sigh. There's also the blockage that might be presented by the radar radome and other antennas up there. It appears that an adjustable fixed mount isn't going to work. That leaves the motorized camera mounts that I suggests might be more useful. When underway, the panel antenna can be pointed downward or edgewise to minimize wind resistance. White epoxy paint should help with the aesthetics problem. However, this is an engineering project, not a simple installation. Some method of control will need to be arranged. I suspect that the usual computer controlled gimbal mount won't work.

Speaking of engineering projects, I'm thinking of something like a miniature version of the TV studio van with the telescoping dish antenna mounted on the roof. Probably too big and ugly, but interesting.

It also seems you don't have much room on the arch. That's going to be a problem even with the original 8dBi omni antenna. The problem with end fed colinear antennas is that the bulk of the RF comes out from the bottom of the antenna. Of the bottom part of the antenna is blocked by other antennas or structure, it's not going to work too well. That means a mast, which was your original idea, putting us back where we started.

Perhaps a conglomeration of the adjustable gimbal mount, in place of the usual tilt over VHF antenna mount, with a small fiberglass mast, and an 8dBi omni antenna on top. It still has to be adjusted manually, but it will clear the other antennas, tilt over for bridges, and can be aligned roughly vertically.

Do you have the room to swing the antenna? My guess is that the panel will be about 10" across and mounted 2" from the center of rotation. That means you need an 11" diameter clear area on the arch.

It's easier to obtain forgiveness than permission. I think what you're looking for is simplicity. How about a stainless "gooseneck" contraption with a panel antenna on the end? It would look kinda funny, but is very easy to adjust (and move out to the way).

I suggest a mechanical joystick or something simple. Maybe a TV remote control running a few relays that drive the motors.

There are larger versions of such gimbal mounts. John's suggestion of using a camera tripod head is one variation. I have a few similar mounts (forgot the manufacturer) installed as panel antenna wall mounts on mountaintops that get fairly high wind loads. I have had problems, but not from lack of rigidity.

I don't know anyone that sells a commercial biquad. However, a panel/patch antenna is a good equivalent.

One possibility that might work (no guarantees) is to use a client radio with diversity. Something like a WAP54G or WRT54G with DD-WRT firmware. Two omni antennas, on opposite ends of the arch, tilted about 8 degrees outward from vertical. That will give double the roll angle. Again, I'm not sure this will work. If the client radio cant switch antennas every time the vessel rolls, it won't play. I guess I can test it on the workbench, but don't have the time.

Well, we're back to the stainless Shakespear 4187 tilt over mount. Hire a machine shop to make a suitable pipe and adapter for the antenna. I've used while ABS plastic pipe to make masts and adapters, but I think that will fail inspection by the high command.

I'm beginning to think that mounting an antenna permanently is a bad idea. Perhaps a small tripod, suction cup, or rail mount with a panel antenna and radio attached. CAT5 cable to the computah. Leave it stowed when underway. Deploy when anchored.

The N-feedthru can be used as a mounting base. A piece of LMR-400 is stiff enough to support a small (6dBi omni antenna). Wrap a spring around the coax cable for additional stiffness. It will act like a gooseneck and take a set in any position. However, it will look terrible, probably not survive much wind, is not very water proof, and probably too crude.

Not that I know about. There are adapters between the 1" - 14 thread but not for an N connector. However, somthing can be fabricated by a machine shop. The big problem is that the omni antenna has the coax cable along the center line, while the Shakespear mount has a solid center core. The coax will need to exit somewhere. The typical VHF antenna uses 0.25" RG-58c/u which is flexible and easy to work around the mount. However, 2.4GHz wants LMR-400 which is about 0.4" in diameter, is fairly stiff, and is not easy to work around the base. The fairly large bend radius of the LMR-400 isn't going to help much. This is not going to be easy.

Need a starting drawing or sketch? (Say no. I'm overloaded).

"Bill Kearney" hath wroth:

Sure, but that also ruins the ability to use a directional antenna at anchor. The vessel could easily swing well beyond the -3dB beamwidth. Only an omni will work under those circumstances.

Oh that's just great. Yes, it's back to the notion of putting it up on some sort of extension.

One thing you may not be taking into account is vibration. This is a powerboat, not a trawler or a sailboat. Granted it's only a 37' express cruiser, not an offshore racing catamaran. If something's not *seriously* well attached and tight it'll work loose in pretty short order. Thus my on-going resistance to a panel or shape other than a simple cylinder/mast. The combination of windage and higher center of gravity makes using panels or something in boxes a less-than-good idea.

I suspect we're zeroing in on using a standard mount, at least to start with, that has an extension on top of it. To which I'd use the usual pole mount held with U bolts. I'd just want to find ones made of a material that's not going to corrode and make a mess. I'll have to measure just how much higher than the radome it would have to be.

If the mast idea doesn't work it'd be worth considering. I'm really looking to avoid having to ever adjust this thing. But in the few situations that might warrant it the ability to do so would be nice.

Marine-grade? From what supplier?

As I mentioned before, it's not soley about wind. Wind load put on it at

35mph is one thing, couple that with pounding through waves and it's a lot of stress.

As opposed to the offerings from hyperlinktech?

Yep, already planning on dd-wrt on a wrt54g.

Hmmm, this might also help deal with obstruction by the radome.

Again, an interesting idea. I may just experiment a bit using a couple of

9db rubber ducks I've already got on a WRT54G here at home. Besides netstumbler what's a good signal strength charting program?

Glomex and others sell both fiberglas and stainless extensions in a variety of lengths. I'd probably go with fiberglas to avoid having the weight make the center of gravity higher.

Eh, there's enough crap to stow and deploy all the danged time. I'm looking to actually have a laptop in the boat permantently. As a gateway of sorts of web, mail, proxying and the like. That and scanning for wifi sources and cross-referencing them against the boat's GPS. Automagically handling it. Having to deploy a tripod wouldn't make that very convenient. I'm planning on our laptops to connect to just the on-boat SSID and then have the gateway (either the laptop or the WRT54g) handle the interconnecting to the short points.

The wind/waves/speed would probably break the copper center conductor in pretty short order.

Some omni have coax from the center, not all. Some are just a connector (N-style) on the bottom. But even the ones with the center wire most extensions are slotted in some fashion or another to accomodate the wire exiting from inside for a pigtail going into the deck surface. Some gymbals also have similar slots. So that's no real worry.

My idea about using antennae with N-style connectors is to allow unscrewing them without pulling cable. Antennae with center coax pigtails built into them would be impossible to swap easily. Something like this:

| The N-style antenna | /-[|] A z-shaped bracket collared around an extension pipe | / +-- -0- The usual marine gymbal mount |

-------------

Or turn the bracket back in a U-shape over the gymbal to keep the weight centered.

| The N-style antenna | [|]-+ A U-shaped bracket collared around an extension pipe | +--+ -0- The usual marine gymbal mount |

-------------

I'd just need to get the bracket fabricated. Really not much more than some sheet metal with the right size holes in it and bended appropriately. And then powder or otherwise coated to ward of corrosion. And it'd also allow some 'fine tuning' of the angles by slightly bending the bracket.

Well, if I go with the above ideas on bracketry there ought to be 'enough' bend radius to accomodate LMR-400.

Heh, nothing worth doing ever is.

-Bill Kearney

"Bill Kearney" hath wroth:

Yep. I did some work on the problem. It looks quite easy. See photos at: |

Antenna is an Antenna Specialists ASPT2976 which methinks is no longer manufactured. | gain. 11 degree vertical beamwidth. Ouch.

The pipe is PVC 1/2" water pipe. I didn't have the proper adapter to mate the pipe to the 1"-14 thread on the base, so I built one out of two couplers.

The base is kinda nifty. It's a Shakespear 4186 ratchet mount made from nylon. Also available in stainless as the 4187. |

The nice thing about 1/2" water pipe is that the inside diamter is only a tiny bit larger than the outside diameter of the Type N male connector. A layer of tape around the N male will give a rather snug fit for the antenna base. If you want something stronger, the big nut that comes with the antenna can be heat swaged into the 1/2" pipe.

The base conveniently has a hollow center and an outlet for 0.25" dia coax cable. Perfect for LMR-240. |

a pigtail from the antenna, through the pipe, and through the base. Terminate it with whatever will go through the bulkhead.

I'm not quite sure what to do about vibration. I guess don't make the extension pipe too tall. The inherent flexibility of the water pipe should help somewhat.

I'll look later, but I don't recall seeing anything similar or suitable. Perhaps a searchlight mount would be sufficiently strong.

One of the projects I worked on was the Intech AN/SRD-21 direction finder (homer) for the USCG. They were installed on 41 and 44 footers. There were 2ea 30" VHF fiberglass antennas mounted on the top of the mast. The system was expected to survive a mast high wave impact. They were so-so and were much later replaced by more flexible stainless whip antennas (at the expense of DF accuracy).

What you'll find is that when the mount, pipe, and antenna are sufficiently strong to survive fairly high forces, the weak spot will be the mounting base to vessel connection. Just drilling the fiberglass arch and installing some washers isn't going to work. The leverage from the extension pipe (if stiff) will be sufficient to rip apart the fiberglass. The connection needs to be made to something much stronger.

The alternative is to make the antenna system flop around like a wet noodle. That will drastically reduce the stresses, but will also cause alignment problems. With an 11 degree beamwidth, having the antenna deformed by the wind or wave pressure is not a great idea.

What I was going to do is put a large piece of sheet metal between the two antennas. Maybe two corner reflectors. I already have a test point on the diversity switch chip to see which antenna is active. In theory, as I move the other end of the link from one side to the other, the antennas should switch. My guess is that it's not going to be a very graceful switch-over. Be sure to do it in client mode on the WRT54G with DD-WRT as it might act differently in other modes.

For signal strength, I prefer Kismet (Linux). For Windoze, there are various client managers (Boingo) that give signal strength. Also:

also does signal strength graphs and plots. There are probably others, but I don't recall the names.

Ummmm.... just how high were you thinking of making the extension? The

1/2" pipe isn't that heavy.

The copper center conductor of LMR-400 is 0.100" diameter. I don't think it's going to break.

Methinks the bracket needs to be a box shape or perhaps tubular. The way you drew them, I suspect they'll vibrate.

Naw. Use LMR-240 from the antenna to the arch and LMR-400 the rest of the way. Better yet, put the radio in a waterproof box near the base of the anntenna to reduce coax losses.

I lied. It's apparently quite easy with plumbing parts and pieces.

Wow Jeff, that's great! Thanks for doing that.

Sure, work with what you've got, makes sense. As I get closer to actually implementing this I'll be sure to get the right couplers or cut the right sort of threads into whatever works right.

I'd be inclined to go with stainless only if the antenna ends up being more than 20" or so. The existing VHF antenna on the boat is a 6 footer and is on a stainless base. But if I end up going with a pair of 5db units there'd be little point in bothering with stainless bases. Just one more thing to corrode, eh?

Hmm, so you're not actually making a secure connection of the antenna to the coupler? "Rather snug" is good enough for stationary installs but on a moving boat I'd want to be pretty sure about it staying put. Also, how does this lend itself to swapping antennae? Push the wire 'up a bit' and remove it manually? Perhaps a threaded collar down around the antenna, with a washer, threaded onto the extension would work.

Or cut a slot in the coupler assembly and run the cable out the side of there, instead of through the base. Either would work.

Good point. I'm only really worried about it working loose enough to be a maintenance issue.

Yeah, while some sort of tripod mount would provide multi-axis movement I don't think it'd be the right solution for this particular installation attempt.

Now THAT would likely be overkill. Still, coupled with a USB controller it might be a novel way to adjust the antenna automagically! I'll leave that idea for the NEXT project!

Heh, if I'm taking THAT much water over the boat I've got bigger problems than WiFi connectivity!

I'm most certainly going to be using a backer plate underneath the mounts. I'm sure it'll be a pain in the ass to get them installed but I'd want to have the added security.

Yeah, marine environments do present their range of challenges, don't they?

Yes, I'm going to try it this weekend with a WRT54G running DD-WRT in client mode and a pair of 9db rubber ducks.

I have as yet not picked up a WiFi card for the laptop that'll handle an external antenna. I'm hoping to set things up such that I've got an SSID for the boat so my laptops never have to change their configuration. The boat's equipment will handle making the connection to the shore access points. It would incur a second level of NAT redirection but I can probably live with it for the limited amount of stuff I'd be doing on the boat. Stuff like mail, http, ssh and the like work fine this way. Trying to VPN and such would get messier but since I know that it's not like I can't plan around it.

I'll have to measure it but I think the radome is about 10" tall. So if I went with a single antenna I'd want to raise it's base above that point. I'd have to take the ratchet mount into account. But it'd likely be no more than a 12" extension, at maximum.

Yes, you're probably right.

Sure, if I end up having something made it'd be worth going with a box of some sort.

I'm figuring the WRT54g (or other client device) will end up residing inside the radar arch itself. The pigtail from the antenna to the device would probably be no more than about 3 feet long. I may end up using a Soekris or other type of device to better handle dual radios. One for making the shore connection and another for providing the boat's network (presumably on a different channel).

I think with some sort of compression fitting it might be possible to slide a collar down around the antenna and secure it tightly 'enough' to the extension. Plumbing stuff has a wealth of options for this sort of thing. Some sort of silicone or rubber washer around the antenna could fill the gap enough to mate up with the collar and make a reasonably secure connection. Then the LMR-240 would be flexible enough to let me pull it up far enough out of the base to get a connection to the N-style connector and unscrew it. Yep, this is sounding like a good way to try it.

THANKS JEFF!

I'll be sure to post pictures once I get this mess working.

-Bill Kearney

"Bill Kearney" hath wroth:

Well, I goofed. It's 3/4" PVC pipe, not 1/2". Sorry. (It was late, I was tired, etc).

The 5.5dBi omni antennas have a vertical beamwidth of 32 degrees, which is largely what you're looking for in the way of sway resistance.

maybe not this one. There's no N connector at the base.

What might be useful it to epoxy or RTV the male end of the N connector to the 3/4" pipe at the antenna end. You can then screw in any antenna with an N connector. Some kind of an outer sleeve around this connection should keep the water out of the connection. If you want to experiment, I like to use hot melt glue for assembling my prototypes before I ossify the connector with a permanent glue.

Measuring: Pipe ID = 0.809" N Male OD = 0.786" The N connector is conveniently knurreled on the outside. Glue it in place.

Easy. With the N connector an integral part of the pipe assembly, just screw in whatever antenna is convenient. As long as it has an N female connector on the base, it should fit. If you want to try a rubber ducky antenna, find an N to RP-TNC female adapter, and install it in between.

If you want the coax joint totally waterproof, find some 1" wide PTFE plumbers tape. Embalm the connection with the PTFE tape. Then protect the PTFE wrap with whatever looks good (duct tape, stainless flashing an hose clamps, aluminum vent tape, etc). The PTFE will cold flow and keep the water out. The tape protects the PTFE from mechanical and UV damage.

Those will both work. If you want real lateral strength without gluing, just add an N connector male barrel into the coax connection. The idea is to lengthen the contact area with the 3/4" pipe. I think you'll find that it's adequate with just one connector, but a few more won't hurt.

No. That would seriously weaken the base. Because of the long lever arm involved, the strength at the base must be many times the strength at the antenna end. A slot would be a stress concentration point. Use the hole provided in the base mount.

One model of the fiberglass antennas we used for the USCG had an interesting problem. The joint between the fiberglass tube and the chromed brass base was filled with epoxy. The epoxy used was originally rather hard and brittle. A dash of UV exposure and some severe vibration on top of the 30ft(?) mast, and the joint would crack. The first indication of a problem was the entire outer fiberglass tube would fall off but leaving the antenna internals largely intact. The glue was replaced with a urethane composition and that problem disappeared. Be careful with anything that is brittle.

The original specification called for surviving a roll over while making about 5 knots of headway. This was specifically for service at the mouth of the Columbia River. Eventually, someone remembered that the operating water depth in the surf line was less than the height of the mast. It was also impossible to use the direction finder (homer) in the surf as the vessel has to be doing donuts in order to get the direction. That spec was mercifully dropped.

If it were easy, it would be no fun.

As you may have noticed, my knowledge of everything except marine electronics is rather limited. That allows me to make some really dumb design mistakes. The AN/SRD-22 automagic DF mechanics was a good one. The antenna consisted of two parts. A base and the electronics head. However, these were screwed together from the bottom with 8 screws. After we started seeing installations with large numbers of missing screws, I realized that it's almost impossible to sit in a bosons chair and assemble the antenna. I replaced the screws with captive screws, which was a big help.

No need. The WRT54G with DD-WRT can handle multiple client connections. Just plug in the various computers and get your DHCP assigned IP address from the marina WISP.

For traffic on the vessel, just get another wireless router, but use it as an access point (on a different channel). The laptop connects to the access point, which goes to the WRT54G, which goes to the marina WISP, which delivers the DHCP IP address.

What's nice about this derangement is that you can change your mind and run double NAT if you want. Just move the ethernet cable from one of the LAN ports on the 2nd wireless router, to the WAN port, and you have instant double NAT.

VPN is no problem unless you're running IPSec with AH encapsulation, which encapsulates the entire packets. Any header re-writes, such as with NAT, will break this form of VPN. IPSec with ESP encapsulation works just fine.

No problem. The 3/4" PVC pipe is quite rigid for such short lengths. I was visualizing something that was 5 or more feet long.

If you're going for an outside install, I would suggest a pressurized box. These are a pain to build and deal with but solve many problems. I use a bicycle pump (with a compressed air dryer) to raise the pressure about 0.2 ATM. That's all that's needed as it only has to keep changes in atmospheric and internal pressure from pumping water in and out. The big gotcha is that you can't easily open the box while underway.

If that is too much trouble, at least mount the internal electronics vertically so that condensed water will drain off the boards and not puddle on the circuitry. I suspect you'll find this difficult to do on the arch.

If you want conformal coating protection, there are various compounds but all of them ruin connectors. You'll need some removable goo (conformal coating silicon mask) to protect the connectors, and a suitable protective spray to cover the boards. I prefer the wax type as it's cheap and easy to use, but doesn't survive well in high heat. Incidentally, forget about your warranty if you do this.

Sure, but be a bit careful here. The threads on the white PVC plumbing parts are tapered to give a good waterproof seal. However, the almost gray electrical identical parts have straight threads because they're not expected to carry any pressure. The gray electrical parts are more expensive for some unknown reason, but are much easier to work with.

Yeah, that's what I was thinking until I decided that glue would do as well. Because the connectors will need to be crimped onto the ends of the LMR-240 coax, there is no way this can easily be disassembled after construction. Might as well make it permanent.

Yep. However, be careful with the LMR-240. There are different types, some of which are not suitable for marine use. See:

looks like a good choice, although the white LMR-240-PVC-W will probably look better. (I haven't seen this stuff).

However, since you're only going to have about 2ft of coax cable involved, you can probably use just about any type of reasonable coax. Ordinary marine RG-58c/u (with non-wicking braid) should work just fine. Coax 2ft atten (dB) LMR-240 = 12dB/100ft 0.24dB RG-58c/u = 23dB/100ft 0.46dB For a difference of only 0.22dB, it's not worth messing with quality coax.

Good luck.

"Wolfgang S. Rupprecht" hath wroth:

Yep. They're "stylized", also known as faked. Perhaps generated from a computer model in free space and without the influence of the connected wireless router. In addition, the proximity of the 2nd antenna in a diversity derrangment is a big problem with high gain antennas. There's minimal interaction between antennas when the gain is fairly low (i.e. 2dBi), but it really gets weird looking when the gain increases. In other words, the plots are not real (or even close to real world).

Nope. It's a hand drawn model of what one antenna would look like in free space and without any connections to equipment or reality. I do those all the time to "beautify" the antenna patterns. When I throw in the effects of equipment, other antennas, cables, mounting, and even the table, the computer generated pattern gets really weird.

The Hyperlinktech patterns are weird in another way. Look carefully at the scale. The outer ring is 0dB. Then there's -3dB, -10dB, -20dB, and so on. However, a proper antenna pattern should have the log scale marked with equal distances between equal number of decibels. These graphs are double logarithmic for some stupid reason, which makes the beamwidth look very wide. Plotted on a real log scale, the

-3dB beamwidth would be pencil thin or about 12 degrees for a 9dBi antenna. However, they didn't even bother faking that part correctly. The plot shows a beamwidth of 30 degrees, which is correct for a 5dBi omni. However, the 5.5dBi omni has an even weirder plot:

proclaims a beamwidth of 120 degrees, which is pure baloney.

I would say the plots are total hand drawn fabrications performed by someone with no clue as to what they should look like for the alleged antenna gains.

Why me?

Yes, that sounds about right.

Indeed, the bare N-male connector 'affixed' inside the end of the extension should be sufficient to hold it. I'd go with an N-male to RP-TNC pigtail leading down into the radar arch and to the WiFi device. Using an RP-TNC pigtail will mean a slightly bigger hole in the arch but that can be covered and sealed easily enough. I'd prefer to avoid the db loss connectors tend to incur. I'll have to double-check what connectors are usually found on

2.4gHz amplifiers first.

Hehe, ossify... embalm... ever watch Princess Bride? "Inconceivable!"

No, I think you're right in going with just the single N-male connector on the end of the pigtail and glued into the extension.

Ok, good point. I've seen some ratcheting mounts that have a slot cut into their threaded base. What I'd be worried about is whether or not an RP-TNC connector can fit through the base. Any idea if the typical marine mount has a wide enough hole?

Yes, I've had past experience using epoxies and I'd have eventually stopped to ponder that very question. So whose glue would be more suitable?

Ha! Like I said, if I've got that much water trouble then WiFi's the last thing on my mind!

Got that right, I've made it a habit to bring along about four times as many fasteners anytime I go above decks. I can guarantee that at least three will go overboard before the first one gets connected.

I've not begun experimenting with dd-wrt in client mode. Nice that it has this feature.

That would work too. I was giving consideration to using an NSLU2 device and hanging a pair of USB WiFi dongles off it it. But using a plain-jane access point connected via CAT5 to the WRT54G might be considerably less hassle.

I'm sure I'll have some questions about getting the WRT54G configured properly. One main goal is to avoid doing any reselecting of the SSID on the laptops. I'd prefer to let my wife consistently connect to the boat's SSID and not have to 'fiddle' with anything else beyond that. So I'll have to look into how to tell DD-WRT to jump to other available SSIDs as they appear. For the home port marina it'd be easy. But when at a new destination it'd certainly require some sort of configuration effort, wouldn't it? But I suppose a script running in DD-WRT should be able to work through doing it automagically.

Well, since the latest DD-WRT supports OpenSWAN as both a client and server I may well go with that. I've run a VPN back here at home for ages and find it very useful to tunnel back into it whenever I'm on the road. The traffic stays encrypted that way, not as clear packets on an unsecured SSID. Sure, there's a slight hit, performance-wise, by going all the way back to the house and then out to the Internet but the security makes it worthwhile. Nothing's completely secure but this is at least one added layer that keeps script kiddies at bay.

Nope, pretty short.

I already have a pump onboard to deal with floats, tubes and fenders so repressurizing such a box wouldn't be a terribly big deal.

Dunno, I'll be in there later this week presumably so I'll report back how the install options look.

Warranty? Heh, these are $20 specials off eBay, I have no fantasies about defrauding anyone trying to claim warranty on them. But yeah, for new stuff it'd certainly be a deal breaker. Although I suppose you could just stuff 'em back inside their normal case and the clerk at the store would never know. By the time it got to someone that might actually service it the trail would be long cold. Not that I'd suggest doing that of course.

Well, they're both DAMNED cheap compared to "marine" stuff!

connection.

Yep.

Cripes, there's just no end to the specialization, is there?

Well, it'll have everything to do with whether or not I can find ready-made pigtails in the lengths I'd need. Once First I'll figure out where to put everything and then hunt down the cabling.

Jeff, you've been a tremendous help here. Many thanks!

-Bill Kearney Syndic8.com

"Bill Kearney" hath wroth:

Cancer of the vocabulary. It's a non-fatal disease caused by excessive exposure to education.

It won't fit and probably cannot be made to fit. The hole in the top of the Shakespear 4186 is 0.540" ID. The TNC male is 0.570" OD. Won't fit. Worse, the coax exit hole at the bottom is 0.298" dia, which is just large enough to accomidate the 0.240" of LMR-240. Looking at the construction, I don't think it possible to enlarge the coax exit hole.

White rubberized bathroom caulk (RTV) or rain gutter glue. It was ok, but was a problem in a production environment having to wait an hour for it to harden. We later switched to a moisture activated urathane glue (forgot the name) for unknown reasons. Todays version is probably:

would use RTV.

The client mode also has a site survey feature which includes a "join" button. It's on the: Status -> Wireless page. It should make connecting to random WISP's fairly easy from a local computer. However, if you put another router in between the WRT54G client mode radio, and your computer, you can't easily get to the WRT54G config pages without juggling cables.

Groan. Why me?

You only have to do that once and it can be done from any client computah. Connect to the WRT54G client radio. Find the SSID. Hit "join". Save. Y'er done.

It can be done with the "wl" command and a shell script. See:

I don't wanna write it.

For example, some useful commands: wl ap 0 (turn off access point mode) wl scan (scan for access points) wl scanresults (display results) wl ap 1 (turn access point mode back on)

SSID: "Mariposa's Art" Mode: Managed RSSI: -89 dBm noise: -99 dBm Channel: 6 BSSID: 00:0D:88:BF:5A:97 Capability: ESS WEP ShortPre ShortSlot Supported Rates: [ 1(b) 2(b) 5.5(b) 11(b) 6 12 24 36 9 18 48 54 ]

SSID: "CMS" Mode: Managed RSSI: -75 dBm noise: -93 dBm Channel: 11 BSSID: 00:11:50:0C:92:F7 Capability: ESS WEP ShortSlot Supported Rates: [ 1(b) 2(b) 5.5(b) 11(b) 18 24 36 54 6 9 12 48 ] WPA: multicast cipher: TKIP unicast ciphers(1): TKIP AKM Suites(1): WPA-PSK No WPA Capabilities advertised

The standard version of DD-WRT supports Microsloth PPTP VPN termination. I use the stock Microsoft PPTP VPN client to connect. It's not a secure as IPSec, but good enough. DD-WRT also now supports WOL (wake on LAN) pass through, so I can turn on my office computer, grab the files I forgot in the office, and shutdown (hope it doesn't hang).

Make sure you're pumping dry air. I have a pneumatic air dryer in line with my bicycle pump. I also install guages on some of my boxes. Also, you might wanna pickup some kids bubble mix to help check for leaks. Other soaps work for gross leaks, but bubble mix seems to find the tiny ones. I also have an ultransonic leak detector, but that's overkill for this project.

Yep. There are also some suprises. I ordered a roll of "super-flex" LMR-400 and attempted to crimp N connectors from my pile. The rubberized outer jacket is about 0.010" larger than the common PVC jacket. I made it fit with the usual application brute force and 4 letter incantations. Be careful with connector selection.

You won't. The problem with this derrangement is that once the connectors are attached, it can't be disassembled. The problem is the tilt base is made to only pass the coax cable, not the connectors. You might be able to sneak an SMA female connector through the hole, but that's about it. I don't have one handy so I can't measure it.

Good luck.

On Fri, 23 Jun 2006 10:00:38 -0700, Jeff Liebermann wrote in :

"You broke the code." :)

Hmmm, it looks like hypertech sells a base with a large enough hole:

It's mentioned on their marine antenna page (a unit with an RP-SMA pigtail): So I'll pickup one of these, I suppose and see if it's possible to get an RP-TNC pigtail through there with "some persuasion". Probably a drill.

But if I can't get it through then I'd be faced with using an RP-SMA to RP-TNC adapter. How much loss might I expect through one?

Yeah, I've some past experience with higher-end RF cabling. I did some

900mHz packet data programming (Ardis) some years ago. I had a yagi on a 60' mast outside the house fed down to a handheld PDA. Talk about adapters, N-style *big-ass* cable down to SMA into a PCMCIA card. Talk about 'non-portable'. But I lived in the sticks and the nearest tower couldn't be picked up with the dinky antenna on the PCMCIA card. The end result of trying to fabricate my own cabling was calling up Tessco (right down the road in Hunt Valley) and getting some $400 worth of cabling made. I'll choose to avoid that hassle again, if possible.

So I'll pickup their mount and see what can be done with it.

-Bill Kearney

I may just consider getting one. This unit looks interesting:

It might take some clever mounting alternatives if I go with the idea of an N-male stubbed inside an upright extension from a marine ratcheting base.

I'd really like to keep things 'clean looking'. So going with a ratcheting base with extension with an N-female stubbed down into it really holds some appeal. It'd let me switch between 5db and 8db omni sticks by doing little more than unscrewing them. Putting a directional on there might be as simple as making a collar of some sort that slid down over the upright extension. One with an elbow or short pigtail to make the angled connection to the directional antenna. Then just mount the panel's pole mount onto this collar extension. I'd never be leaving this antenna connected while underway so it having a little bit of rotational slip might help with aiming. It'd be less attractive than a plain upright stick but not terribly so. Certainly less ugly than a NEMA box.

I'll have to see what sort of results I get with omni sticks first. If they work 'well enough' then going with a directional won't be necessary.

Meanwhile, what're the opinions of various 2.4gHz bi-directional amps for use in the boat to help boost on-shore connectivity?

-Bill Kearney

Yep, looks like I'll have to go N-male from the antenna to RP-SMA through the mount. This works OK as it'll also mean the hole I have to put through the radar arch will be smaller. The question then is how to deal with the cabling from there to the WiFi device. It may be impractical to put the WiFi client up in the arch. Between limited space and environmental issues it might be better to bring a cable down the arch to a cabinet on the deck. So should I just use an N-male to RP-SMA pigtail using thin cable short enough to get inside the arch and then switch to something better? Like a

195 pigtail to an RP-SMA to N-style using 400? Then down the arch to the radio and convert from N-style to RP-TNC. 2' N-male to RP-SMA pigtail using 195 RP-SMA to N-style adapter N-style cable using 400, probably about 10' N-style to RP-TNC adapter to radio

Seeing as how it'll be about 12' total distance I figure that'll be the 'least lossy' of the options, that sound about right?

-Bill Kearney

"Bill Kearney" hath wroth:

That's EXACTLY what I used. It's a Shakespear 4186.

That would be too easy. It lacks the added height of the extension we discussed. Vertical beamwidth of 15 degrees is typical for 8.5dBi gain. Just remember that most of the RF belched from vertical end-fed colinear antennas comes out the bottom of the antenna, not the top. (Detailed explanation on request).

No. I previous measured out the hole sizes. The hole in the top of the mount is too small. The hole in the bottom is far too small and cannot be enlarged. Even if they were enlargd to the right diameter, I don't think it would be possible to make the 45 degree bend very gracefully. The holes are made to pass the cable, not the connector. Actually, I don't think you can even get an SMA Male connector through the hole. A female SMA might fit. Yeah, that's what Hyperlinktech does. They specify a "3ft coax lead with RP/SMA jack" on the web page.

0.25dB or so if you can keep it dry. Avoid dissimilar metals. Most SMA's are gold plated. Mixing a gold SMA connector with a nickel plated TNC or N is going to be a problem. I'm not sure how much the loss will increase, but I know that it will.

Tessco is VERY expensive. There are places that will make cables for you for much less. I bought several hundred dollars worth of crimp tools and RF Industries connectors because I kept getting crappy crimp connections from allegedly reputeable vendors. They were apparently all learning how build cables and I was getting garbage. Hopefully it's better these days. At the time, the only vendor that didn't mangle my custom cables was Electro-Comm:

These daze, you can get a set of 3 crimp tools for about $40/ea which will do everything from MCX to LMR-400 size N connectors. If you're going to make your own antennas, you might as well make your own cables and pigtails.

Same as Shakespear 4186 but almost exactly the same price.

Incidentally, if you don't like the PVC water pipe extension idea, Shakespear makes 1"-14 extensions: |

"Bill Kearney" hath wroth:

Well, you could also do SMA connectors instead of TNC. No need to go RP (reverse polarized) for a connection that will only be a splice or extension cable between the antenna pigtail and the radio.

Without seeing the installation, I can only guess(tm).

2.5' N-male to SMA-female pigtail using LMR-240 (SMA might fit through the hole in the mount). 10ft of LMR-240 with SMA-male and RP-TNC-male. (1.3dB loss) No pigtail required. **** OR **** 2.5' N-male to SMA-female pigtail using LMR-240 (SMA might fit through the hole in the mount). 10ft of LMR-400 with SMA-male and N-male. (0.7dB loss) N-female to RP-TNC pigtail with LMR240. **** OR **** 12.5' of LMR-240 with N-male to RP-TNC-male. Crimp it yourself. No intermediate connectors (that need waterproofing).

Don't forget to leave a service loop at the antenna base to lessen the stresses on the coax of tilting the antenna.

LMR-240 is 0.24" OD.

Buy the tool: |

are other vendors, but I'm too lazy to go out to my truck and get the numbers off my tools.

Cheap: |

I like numbers instead of "least lossy" conclusions. The difference in coax loss of 0.6dB between 10ft of LMR-400 and 10ft of LMR240 is not worth the effort. The additional pigtail required for LMR-400 will reduce the differential and raise the cost. I would use LMR-240 throughout and in one long run.

The only reason I'd stick with RP would be to allow plugging actual devices in at that point. For testing and such.

Thanks for running the numbers.

I have a ratcheting crimper already. I'll have to see what jaws it can be changed to use. The tool is, of course, all the way out at the boat in Annapolis while I'm here in Bethesda. All the more reason to dodge the idea of crimping my own connectors...

Well, I'll have real distances once I get a better feel for where I can cram the WRT54GS. If there's room in the radar arch then I'll go with N-male to RP-SMA and an RP-TNC adapter on the back of the WRT54GS. If that doesn't prove to be a reliable location then I'll use an RP-SMA to RP-TNC cable to reach to whatever location works better.

-Bill Kearney

Figured as much. That'll mean spec'ing the correct gender connections unless I want to add to my already abundant supply of mismatched adapters.

Yeah, good point about mixing metals. Marine environments aren't very forgiving.

Sure, but over the years all work I've had done by them has been high quality. Granted, most of it was like 600 series cable with soldered (silver?) connections. For some stuff I'm not opposed to paying to get quality work. Especially if it keeps me away from the crimping tools! Me and coax just never got along. Dunno, just never quite got the hang of it. It's one thing to have a bunch of half-ass connections on a cable TV line, but for a marine environment pushing 2.4gHz signals I'm just more comfortable letting someone else do it for me.

I may end up with some of their extensions along with PVC. It'll depend on how the various pieces start fittng together. That and some testing to see whether or not I need to get it up above the existing radome and by how much, or whether using a pair of the on either side of the arch would work better. Won't know until I get an actual antenna here and rigged up into the mount. Hopefully that'll be sometime next week.

-Bill Kearney

Went with a Maxrad 8db unit. It's mounted on top of a shakespeare 12" extension. The extension's stands about 2" above the Raymarine radome. At the top of the extension I added a mylon 3/4" pipe thread female to 1" male adapter. I then cut a hole in the middle of a 1" threaded cap and installed the N-female end of the antenna through it.

The 3/4-1" adapter leaves enough room inside it to allow the N-male connector to fit onto the antenna. The N-male connector cable is simply tightened on to the base of the antenna directly. In order to minimize twisting of the cable the adapter is spun onto the base first and then backed up into the cap as it's tightened. I can just as easily keep the pigtail cable loose enough to turn while tightening.

It's on a regular marine antenna base.

Signal strength on a nearby connection DOUBLED with this positioning.

The hassle was finding the right sort of way to attach the antenna to the extension shaft. I'd looked into using regular PVC but each extra coupler adds weight. That and the inside diameter of many fittings wouldn't accomodate the N-Male connector. I was tempted to bore out the hole in the extension shaft. It'd only need to be about 1/8" larger. But fearing reduction of the mount's integrity sent me back out to the local hardware store's plumbing section. I poked around in the parts drawers and discovered the 3/4" to 1" adapter and it had more than enough room for the N-male connector. I had to go to home despot to find a 1" threaded cap for it though. Two local plumbing supply places had NONE of these parts with threads, go figure.

Anyway, if you're looking to put an omni antenna on a marine mount this might be a solution worth considering.

My next step will be to work out a way to easily swap the omni for a directional. But I'm only going to bother with it if I discover the omni giving me less-than-satisfactory results. I'd probably go with just a small PVC extension using a 3/4" male to slip connection and a U-bolt style clamp. I'd back a nut onto the extension to allow controlling the rotational angle. The regular omni is just tightened down fully since rotation doesn't matter.

From the antenna it's an N-male to RP-SMA pigtail, with a 1 meter extension. This leads to an RP-SMA to RP-TNC adapter on the back of the WRT54G that's mounted up in the radar arch. This feeds an ethernet twisted pair to another WRT54G in the lower cabin. I didn't use power-over-ethernet, I just ran 12v power to them.

The only problem seems to be related to the firmware. If the shore SSID comes and goes "too much" the firmware *sometimes* seems to get confused. The symptom is it refuses to show anything but one or two SSIDs. If I soft reboot the router it immediately shows all detected SSIDs during it's Site Survey. This isn't easily repeatable, but it does seem to coincide with a lot of wave action rocking the boat. I haven't yet dug into how the firmware manages intermittent connections and reconnections.

Anyway, save for getting a 'sealed' enclosure I've now got wifi working quite well on the boat.

-Bill Kearney