Long-Distance WiFi Device

[thatboatguy]

We are very happy with the setup we have. I will say that the 15dbi antenna at the masthead of our monohull boat goes against the grain of the majority opinion. Lots of pros out there (on whom I relay heavily for help) will try and talk you out of it. There is all kinds of good theory to support this. However, we use the antenna to good effect. We keep a solid directional antenna hooked up to a second Bullet which we use at deck level and this can also be hoisted with the 8dbi omni if I have to. I've never been anywhere where the 8dbi hoisted outperformed the 15dbi at the masthead.

OK, that's one cruisers experience. I'll defer to the pros for solid technical advice.

George

[Auspicious]

The problem with mounting a high-gain omni antenna too high is that the path between your antenna and the shore-side access point may lead below the 3 dB point of your antenna and/or the shore-side antenna, giving up a lot of the gain you paid for.

The problem with mounting a high-gain omni antenna too low is that the water surface is within the Fresnel zone of the transmission path, a fancy way of quantifying the impact of reflections and refractions of the signal.

Like the third law of thermodynamics it is an indication that physics keeps us from winning. *sigh*

Spreader height may be a reasonable compromise. My Air802 antenna ( http://www.air802.com/2.4-ghz-outdoo...i-antenna.html ) is mounted on an aft radar pole that works quite well. I tried hoisting the antenna to various heights and measuring the signal strength of near and distant access points. I would have been better off a bit higher than I could get on the radar pole but the pole was so much easier an install than the lower spreader that I gave up a bit for that ease.

[s/v Jedi]

Although in theory it's true that using a high gain omni can put the signal path outside it's primary radiation pattern, at some distance between the two stations this is not an issue anymore. Heeling and pitching can be a problem but Wifi is the last thing on our mind when we do that in an anchorage

It's easy to test this or do some math. Check the specs of the antenna for it's vertical opening angle; now make a drawing in scale of the mast height with that antenna angle radiating in both the upwards slope and downwards slope. Now you can see the dead-zone but remember that the antenna will have side-lobes in it's pattern meaning that close-by always works. You can also see how much you need to heel to get into trouble this way.

Some (more expensive) high gain omni's have a built-in downwards radiation angle. These are better for short distances but worse for the real long distance work.

ciao!
Nick.

[Auspicious]

Originally Posted by s/v Jedi

Now you can see the dead-zone

Presumably you mean the nulls in pattern that are going to be yet further out than the 3 dB points and likely not relevant in a real-world scenario. If one is close enough to the access point for nulls to be a factor then you'll be able to make a connection with a wet noodle.

My point is that if using an antenna with a narrow vertical beamwidth you will be well off maximum gain almost all the time and often beyond the 3 dB points (which defines the "beam").

It would be interesting to establish a set of 1/2 mile, 1 mile, 2 mile, and 4 mile scenarios and look at the probable gain from 5, 8, 12, and 15 (if attainable) dBi omnis on a boat at various heights ASL. Maybe 58 combinations? A little busy here to do that.

[thatboatguy]

Seems to all make perfect sense in the way we end up using our antennas. For long distance where there is not a lot of noise, nothing beats that 15dbi antenna at the masthead. We had an 8dbi up there before and it simply did not work as well. In places where the access point is closer and/or there is more noise, the cantenna aimed right at the access point works best.

George

[s/v Jedi]

Originally Posted by Auspicious

Presumably you mean the nulls in pattern that are going to be yet further out than the 3 dB points

No, I mean the area that gets more than -3dB attenuation in the radiation pattern of the antenna. Every decent antenna has specs with the opening angles and good ones even come with a graph showing the pattern, incl. side lobes.

Let me do the math:

if the vertical angle is 25 degrees (typical for 14dBi gain omni) and our antenna is 50 feet high and the AP antenna is 10 feet high:

then the height difference is 40'. In order to get the gain-boost from our high gain antenna, we need to stay within it's 25 degree opening angle.

tan(angle) = height / distance

tan(30) = 40 / distance

0.58 = 40 / distance

distance = 40 / 0.58

distance = 69'

So, as long as we keep a MINIMUM distance of 69 feet between the two antenna's, we are within the radiation pattern of that hi-gain antenna. When we get closer than 69 feet, we get into trouble... but not really because we get so close that it will always work because of the side-lobes.

Of course the difference in height between the two antenna's can be much more than in my example, but now you can calculate it easily by changing the numbers in the math above

ciao!
Nick.

[Auspicious]

Hi Nick,

Originally Posted by s/v Jedi

if the vertical angle is 25 degrees (typical for 14dBi gain omni)

Can you source that information? The E-plane 3 dB points for my 12 dBi omni is well under 10 degrees (+/- 5 degrees). The largest sidelobes are 25 dB down from the peak. See the plots on the link I posted above.

Remember the vertical angle you talk about IS the angle between either 0 and a 3 dB point or between the 3 dB points depending on the geometry being analyzed.

I haven't had a chance to check your arithmetic.

Most of the APs I know of around Annapolis are higher than 10' ASL which makes the minimum distance smaller (a good thing) and increases the distance at which Fresnel effects become significant (another good thing).

[s/v Jedi]

Originally Posted by Auspicious

Can you source that information?

sure: WiFi Omni Antenna, 2.4 GHz, Wi-Fi, Omnidirectional, 14dBi, Outdoor, Wi-Fi, Wireless LAN shows a 14 dB with 25 degree vertical angle.

The E-plane 3 dB points for my 12 dBi omni is well under 10 degrees (+/- 5 degrees).

Okay when we put that 5 degrees into the math we get a minimum distance of 440' instead of 69' which is a significant difference but still not a problem at all.

The largest sidelobes are 25 dB down from the peak. See the plots on the link I posted above.

Yes but the 25dB attenuation is not a problem when you are using a side lobe on close distance. Basically, close distance, anything works.

Remember the vertical angle you talk about IS the angle between either 0 and a 3 dB point or between the 3 dB points depending on the geometry being analyzed.

Correct. You can get twice as strong a signal when you are at the center of the radiation pattern as compared to being on the edge of the -3dB point.

I haven't had a chance to check your arithmetic.

I must admit that I had to look it up to be sure

Most of the APs I know of around Annapolis are higher than 10' ASL which makes the minimum distance smaller (a good thing) and increases the distance at which Fresnel effects become significant (another good thing).

The best thing to do with boats on the water is change to horizontal polarization. But when the shoreside AP is a vertical omni, you will have to comply (new advanced APs have MiMo arrays with both horizontal and vertical polarization even with omni-directional antenna's !!). A 15dB gain omni-directional aboard will work better than a medium gain omni as long as the boat is not rolling.

ciao!
Nick.

[thatboatguy]

Here are the actual specs for the Hawking 15dbi antenna we have. The 50 foot figure is a good number for it's mounting elevation.


Network Specifications:

• IIEEE 802.11b/g

Electrical Properties

• Frequency: 2.4~2.4835 Ghz
• Impedence: 50 Ohms Nominal
• Gain: 15dBi
• Radiation: Omni
• Polorization: Vertical
• H-Plan:360º
• E-Plan: 30º

Mechanical Properties

• Connector: RP-SMA and N-Plug

Material Properties

• Color:White & Silver
• Body: Brass & Ni Plating
• Pin: Phosphor Bronze Gold
• Insulator:Teflon or Delrin
• Dimensions: 48 (H) x 1 (r) in (w/out joint)
• Weight: 2.2 lbs

George

[s/v Jedi]

That's a 30 degree and the way they state it I assume you have to use a number of 15 degrees for the formula I posted.

Now that I look at the link I posted before... that antenna's 25 degree angle is also combined with a 360 degree horizontal angle meaning it needs 12.5 degrees in the formula. I just Googled for a 25 degree angle antenna to find a source but should have used 15 degrees or so in my first calculation, making the minimum distance 150'.

The point I'm trying to make though is that it's an urban legend that a high gain omni on a boat doesn't work because of the flat donut shaped radiation pattern. The minimum distance is still small enough to always be just fine. The only problems occur during rolling.

ciao!
Nick.

[thatboatguy]

Exactly Nick! I already know it works, I just leave to someone else to explain why. heh heh

George

[Amgine]

Wow. 57 pages at the moment and still going strong...

You know, it would be really great if some of the successful installations (and pros) could write up a series of articles on the wiki explaining getting wifi on the boat along with a couple of 'recipes' for reasonable deck-level and permanent installations. (Plus an experts article going into the math, etc?)

[Extemporaneous]

Originally Posted by richardhula

Thanks Bob, points noted. Useful info about the bridge mode as well.

I had a look at the Discovery utility but noted it was a Java app, so needs JRE installed on PC.

I'm wondering for anyone starting with Bullet for the first time if it wouldn't be easier to just reconfigure PC LAN adapter for static IP initially.

Final version of instructions here as pdf file too big to attach.

Well I'm hooked up and NO pain what so ever!!

Thank-You Richard! (for the expert article)

I can't say whether my hardware is any good ;-)...... but I AM connected.

Cheers,
Extemp.

[s/v Jedi]

Originally Posted by Extemporaneous

Well I'm hooked up and NO pain what so ever!!

Thank-You Richard! (for the expert article)

I can't say whether my hardware is any good ;-)...... but I AM connected

Just for fun you should connect the picostation to the computer and config it into station mode just like the Bullet. This will show you how well that little antenna works and it gives you a benchmark that must be beat by any other antenna solution. In 90% of all cases the little picostation is all that's needed for a link

ciao!
Nick.

[Viking Sailor]

I see 3 dB used a lot in the above posts with respect to signal strength and antenna patterns. The 3 dB value is used as a reference because it describes the halving or doubling of the effective radiated power. And not coincidently, it is also the minimum change in effective radiated power the human ear can detect.

However, one should keep in mind that when it comes to Digital Signal Processing (DSP) the 3 dB detection limitation does not apply. In the DSP world, being 1 dB above the noise or interference is huge.

So, don't get to hung-up on 3 dB points. It is really all about finding what works best. Which you seem to be doing just find without my extraneous comments.