iBurst Antennae

[Mze]

I am situated in Blairgowrie and installed my UTD last year in May and experienced the most fantastic transfer rates, day and night.

By September/October last year the performance experienced became very poor.

After a visit by an iBurst expert, I had the uni directional Webb antenna with a 5m extension cable directed at BS 13 in Randburg. Performance better but not as good as originally experienced.

After a further visit by Iburst expert, I put together a real Heath Robinson mast comprised of two 25mm dia swimming pool "poles" overlapped by approx 1 metre and held together by strong cable ties. Total height of the mast is approx 6 metres.

The Webb antenna was hung on the top of the mast by a strategically bent wire coat-hanger and two wire coat-hangers attached to the eaves of my house secured the mast against the rainwater gutter. the bottom of the mast rested on the ground, and was secured by nylon rope to a steel stake driven into the ground. The top of the mast was prevented from swaying excessively by two guy ropes attached to the holes where a pool brush would normally be attached. This marvellous edifice could be rotated by my son while I ran utTraceStar to determine the optimum direction for the Randburg BS. Performance was better but still not what I wanted.

A visit by Renier Smit, who marvelled at the antenna mast, resulted in the Webb antenna being directed at the Illovo BS 27. The performance was now much better but still not good enough.

I learnt of the existance a Yagi antenna. The general manager of the company was kind enought to have it delivered to my house this week. After it was installed and tested and correctly positioned, I was very pleasantly surprised by the increase in performance.

The data below serves as an illustration of how the different configurations worked.

UTD only. I put this in to illustrate what the standard UTD omni directional performance is right now. Note that the second-last column shows a load of 3 which results in poor transfer rates of plus/minus 100 kbps

2006/11/29 12:24:30,13,Randburg (Rand_Pres),-99,10,3,3,4/10
2006/11/29 12:24:40,13,Randburg (Rand_Pres),-96,10,3,3,4/10
2006/11/29 12:24:51,13,Randburg (Rand_Pres),-98,10,3,3,4/10
2006/11/29 12:25:01,13,Randburg (Rand_Pres),-99,11,3,3,4/10
2006/11/29 12:25:11,13,Randburg (Rand_Pres),-97,11,3,3,4/10
2006/11/29 12:25:22,13,Randburg (Rand_Pres),-98,10,3,3,4/10
2006/11/29 12:25:32,13,Randburg (Rand_Pres),-98,10,3,3,4/10
2006/11/29 12:25:52,13,Randburg (Rand_Pres),-97,11,3,3,4/10
2006/11/29 12:26:03,13,Randburg (Rand_Pres),-93,10,3,3,5/10
2006/11/29 12:26:13,13,Randburg (Rand_Pres),-96,11,3,3,4/10
2006/11/29 12:26:34,13,Randburg (Rand_Pres),-97,10,3,3,4/10
2006/11/29 12:26:44,13,Randburg (Rand_Pres),-97,11,3,3,4/10
2006/11/29 12:26:54,13,Randburg (Rand_Pres),-99,10,3,3,4/10
2006/11/29 12:27:05,13,Randburg (Rand_Pres),-100,11,3,3,4/10
2006/11/29 12:27:25,13,Randburg (Rand_Pres),-97,11,3,3,4/10
2006/11/29 12:27:35,13,Randburg (Rand_Pres),-99,10,3,3,4/10
2006/11/29 12:27:46,13,Randburg (Rand_Pres),-96,10,3,3,4/10
2006/11/29 12:28:06,13,Randburg (Rand_Pres),-95,11,3,3,5/10
2006/11/29 12:28:16,13,Randburg (Rand_Pres),-96,11,3,3,4/10
2006/11/29 12:28:27,13,Randburg (Rand_Pres),-90,11,3,3,6/10
2006/11/29 12:28:37,13,Randburg (Rand_Pres),-95,11,3,3,5/10

Webb Uni Directional Antenna. 6m above the ground with 2.5m cable, 5m extension and pig-tail. Antenna directed at BS 27. Note the load of 1 and the good db readings.

2006/11/28 10:35:56,27,Illovo (Laudergate),-77,16,5,1,8/10 Very Good
2006/11/28 10:36:06,27,Illovo (Laudergate),-76,15,4,1,8/10 Very Good
2006/11/28 10:36:16,27,Illovo (Laudergate),-78,15,4,1,8/10 Very Good
2006/11/28 10:36:26,27,Illovo (Laudergate),-77,15,4,1,8/10 Very Good
2006/11/28 10:36:37,27,Illovo (Laudergate),-77,15,4,1,8/10 Very Good
2006/11/28 10:36:47,27,Illovo (Laudergate),-77,15,4,2,8/10 Very Good
2006/11/28 10:36:58,27,Illovo (Laudergate),-76,15,4,2,8/10 Very Good
2006/11/28 11:10:56,27,Illovo (Laudergate),-78,16,5,1,8/10 Very Good
2006/11/28 11:11:07,27,Illovo (Laudergate),-78,15,4,1,8/10 Very Good
2006/11/28 11:11:17,27,Illovo (Laudergate),-77,15,4,1,8/10 Very Good
2006/11/28 11:11:27,27,Illovo (Laudergate),-78,15,4,1,8/10 Very Good
2006/11/28 11:11:37,27,Illovo (Laudergate),-78,15,4,1,8/10 Very Good
2006/11/28 11:11:48,27,Illovo (Laudergate),-77,15,4,1,8/10 Very Good
2006/11/28 11:11:58,27,Illovo (Laudergate),-77,16,5,1,8/10 Very Good
2006/11/28 11:12:08,27,Illovo (Laudergate),-78,15,4,1,8/10 Very Good
2006/11/28 11:12:18,27,Illovo (Laudergate),-77,16,5,1,8/10 Very Good
2006/11/28 11:12:29,27,Illovo (Laudergate),-76,15,4,1,8/10 Very Good
2006/11/28 11:12:39,27,Illovo (Laudergate),-78,15,4,1,8/10 Very Good
2006/11/28 11:12:49,27,Illovo (Laudergate),-76,15,4,1,8/10 Very Good
2006/11/28 11:13:00,27,Illovo (Laudergate),-78,15,4,1,8/10 Very Good
2006/11/28 11:13:10,27,Illovo (Laudergate),-78,15,4,1,8/10 Very Good

Yagi Uni Directional Antenna. 6m above the ground, plus 10 m extension cable, and pig-tail. These readings were obtained before fine tuning direction. Almost a perfect fluke the way I positioned it.

2006/11/28 13:32:46,27,Illovo (Laudergate),-71,16,5,2,9/10 Excellent
2006/11/28 13:32:56,27,Illovo (Laudergate),-72,15,4,2,9/10 Excellent
2006/11/28 13:33:06,27,Illovo (Laudergate),-72,15,4,2,9/10 Excellent
2006/11/28 13:33:16,27,Illovo (Laudergate),-72,15,4,1,9/10 Excellent
2006/11/28 13:33:27,27,Illovo (Laudergate),-72,15,4,2,9/10 Excellent
2006/11/28 13:33:37,27,Illovo (Laudergate),-72,16,5,2,9/10 Excellent
2006/11/28 13:33:47,27,Illovo (Laudergate),-72,15,4,2,9/10 Excellent
2006/11/28 13:33:58,27,Illovo (Laudergate),-71,15,4,1,9/10 Excellent
2006/11/28 13:34:08,27,Illovo (Laudergate),-73,15,4,1,9/10 Excellent
2006/11/28 13:34:18,27,Illovo (Laudergate),-71,15,4,2,9/10 Excellent
2006/11/28 13:34:28,27,Illovo (Laudergate),-72,16,5,2,9/10 Excellent
2006/11/28 13:34:39,27,Illovo (Laudergate),-71,15,4,1,9/10 Excellent
2006/11/28 13:34:49,27,Illovo (Laudergate),-72,15,4,2,9/10 Excellent
2006/11/28 13:35:00,27,Illovo (Laudergate),-71,15,4,2,9/10 Excellent
2006/11/28 13:35:10,27,Illovo (Laudergate),-72,15,4,1,9/10 Excellent
2006/11/28 13:35:20,27,Illovo (Laudergate),-72,15,4,1,9/10 Excellent
2006/11/28 13:35:30,27,Illovo (Laudergate),-72,15,4,1,9/10 Excellent
2006/11/28 13:35:41,27,Illovo (Laudergate),-71,15,4,1,9/10 Excellent
2006/11/28 13:35:51,27,Illovo (Laudergate),-72,15,4,2,9/10 Excellent
2006/11/28 13:36:01,27,Illovo (Laudergate),-71,16,5,1,9/10 Excellent
2006/11/28 13:36:11,27,Illovo (Laudergate),-72,15,4,2,9/10 Excellent

Tuned Yagi. This is as it is working today.

2006/11/29 09:03:08,27,Illovo (Laudergate),-69,16,5,1,9/10 Excellent
2006/11/29 09:03:18,27,Illovo (Laudergate),-71,15,4,1,9/10 Excellent
2006/11/29 09:03:28,27,Illovo (Laudergate),-71,15,4,1,9/10 Excellent
2006/11/29 09:03:39,27,Illovo (Laudergate),-72,15,4,1,9/10 Excellent
2006/11/29 09:03:49,27,Illovo (Laudergate),-72,15,4,1,9/10 Excellent
2006/11/29 09:03:59,27,Illovo (Laudergate),-71,15,4,1,9/10 Excellent
2006/11/29 09:04:10,27,Illovo (Laudergate),-71,16,5,1,9/10 Excellent
2006/11/29 09:04:20,27,Illovo (Laudergate),-69,15,4,2,9/10 Excellent
2006/11/29 09:04:30,27,Illovo (Laudergate),-71,15,4,1,9/10 Excellent
2006/11/29 09:04:40,27,Illovo (Laudergate),-71,16,5,1,9/10 Excellent
2006/11/29 09:04:51,27,Illovo (Laudergate),-71,16,5,1,9/10 Excellent
2006/11/29 09:05:01,27,Illovo (Laudergate),-69,16,5,2,9/10 Excellent
2006/11/29 09:05:11,27,Illovo (Laudergate),-72,16,5,2,9/10 Excellent
2006/11/29 09:05:22,27,Illovo (Laudergate),-72,15,4,2,9/10 Excellent
2006/11/29 09:05:32,27,Illovo (Laudergate),-71,16,5,2,9/10 Excellent
2006/11/29 09:05:42,27,Illovo (Laudergate),-72,16,5,2,9/10 Excellent
2006/11/29 09:05:52,27,Illovo (Laudergate),-72,15,4,2,9/10 Excellent
2006/11/29 09:06:03,27,Illovo (Laudergate),-72,16,5,2,9/10 Excellent
2006/11/29 09:06:13,27,Illovo (Laudergate),-71,15,4,2,9/10 Excellent
2006/11/29 09:06:23,27,Illovo (Laudergate),-72,16,5,2,9/10 Excellent
2006/11/29 09:06:34,27,Illovo (Laudergate),-71,16,5,2,9/10 Excellent

What about transfer rates? I hear you ask. Right now, 11:32 I am getting these results using iBurst's speed test. (417/140)(431/136)(424/121). At night I measure the transfer rate as being up to and over 1Mbps.

Under normal operating conditions I estimate that the Yagi performs 30 to 50% better than the Webb.

What does it cost?

The pricing: All prices exclude VAT
Antenna – R325.00
Extension Cable – R10 per metre
Connections x 2 – R20 each.
Cost of my installation = 465 + VAT = R530.10
6m x 50mm Mast = ±R427 from Ellies (I must still buy the mast)

Where from?
From: Graeme-Radiant Antennas [mailto:[email protected]]
Sent: 24 November 2006 11:39
To: [email protected]
Subject: Antenna Picture
Dear Mze
Many thanks for the enquiry.
Attached is a picture of the antenna.
See you Tuesday.

Best regards
Graeme Davis
General Manager
Tel: +27 12 804 6750
Fax: +27 12 804 3691
Cel: +27 82 377 5669

 

[Fudzy]

Interesting findings, does this improve your ping at all? Aren't you worried about lightning strikes?

 

[Ekhaatvensters]

And the price isnt bad either. Thanks a lot for all the testing and info.

People here have talked about Yagi antennas, but im not sure how many have actually tried them.

 

[who.is.michael]

With iBurst ant antenna is essential...

Good to see it helped...

If you have spare cash and want to further improve your signal either purchase another Yagi of the same type from the same supplier and mount them side by side about 16 cm apart (boom-to-boom) and combine them to the main feedline using a “combiner” module, this will narrow the horizontal beamwidth and increase the gain by approximately another 3dB or purchase two higher gain Yagi’s .

However the best cellular antennas can be found in Ireland, reasonably priced (60 EUR) however the shipping cost (125 EUR) is rather high. The shipping cost would be the same for ten antennae, so if you can convince a few iBurst users to purchase as well and split the shipping costs it can be well worth your while.

Insulated Yagi’s

Parabolic antennae

Hope this helps.

Regards

Michael

EDIT: To reply to your post below; as it is not completely accurate. Narrowing the beamwidth would allow for less “noise” from unwanted towers and other sources, therefore an improved signal (> -3dB) allowing for a higher throughput.

Would appreciate your feedback once you have tested it – it is especially relevant where there are several base stations with a difference in signal of around 10dB.

 

[Radiant Antennas]

Two antennas

Hi All
Nice idea to double up on antennas, but doing so will only increase your signal by 3db's, but you will loose it again through your splitter (-3dB's for every split)....so you will not improve the signal strength...just narrow your beamwidth.

Graeme

 

[Mze]

Two Antennae

Good one Graeme. What do you suggest to increase db gain?

 

[Radiant Antennas]

No Miracles

Hi Folks
Unfortunately, as far as antennas are concerned, there have been no technologically miraculous discoveries/designs which, as with the rest of the electronics world, has produced a miniture antenna, that out performs anything before it.
Unfortunately, it's true.........size does matter!

The bigger the antenna, in length, height, diameter.....generally overall volume, the better the performance.....and usually the cost.
Providing the design is sound, and it has been built well......the bigger the antenna...the better.

There are a number of iburst antennas available in South Africa....and those that come from the reputable suppliers/manufacturers, all work very well....I have tested most of them. What is claimed in performance (ie dB gain) is true.......except for one very high gain antenna, which is no where near what is claimed.

Therefore, a few simple pointers (again..no miracles) that could help increase signal strength:
1. Buy the best antenna you can afford.
2. Use low loss cable...also the best you can afford. 3. Keep your antenna cable as short as possible...ie, if the cable is too long, don't coil it up under then table......ask a reputable manufacturer to cut it to length, and re-fit the connector.
4. Put the antenna up as high as you can...or as high as your neigbours will let you.The down side of this is that the higher your antenna, the longer the cable, the greater the loss through the cable....so there will be a point where going higher does not increase signal that significantly.....one has to be practical. (a higher gain antenna can make up for losses in a cable that has to be very long)


If anyone is interested...I will from time to time post a few antenna 'tutorials',
just the basic kind of stuff that will aid in choosing what antenna will be suitable for you, and generally how they work, and why there are so many different designs.....cheers for now.

Graeme

 

[Mze]

Pings and Lightning Strikes

The figures shown below are reasonably representative of continuous pinging.
The average latency shown is on the high side - I usually get between 70ms and 90ms.

Although my antenna is 6m plus high, there are trees and double storey houses in the immediate vicinity which are a lot higher. This, I hope, will make them more attractive to lightning than my antenna. BTW, there is line of site between my antenna and BS 27 at Illovo.

Reply from 196.30.31.120: bytes=32 time=137ms TTL=125
Reply from 196.30.31.120: bytes=32 time=74ms TTL=125
Reply from 196.30.31.120: bytes=32 time=121ms TTL=125
Reply from 196.30.31.120: bytes=32 time=152ms TTL=125
Reply from 196.30.31.120: bytes=32 time=90ms TTL=125
Reply from 196.30.31.120: bytes=32 time=58ms TTL=125
Reply from 196.30.31.120: bytes=32 time=74ms TTL=125
Reply from 196.30.31.120: bytes=32 time=58ms TTL=125
Reply from 196.30.31.120: bytes=32 time=74ms TTL=125
Reply from 196.30.31.120: bytes=32 time=74ms TTL=125
Reply from 196.30.31.120: bytes=32 time=90ms TTL=125
Reply from 196.30.31.120: bytes=32 time=73ms TTL=125
Reply from 196.30.31.120: bytes=32 time=74ms TTL=125
Reply from 196.30.31.120: bytes=32 time=58ms TTL=125
Reply from 196.30.31.120: bytes=32 time=74ms TTL=125

Ping statistics for 196.30.31.120:
Packets: Sent = 110, Received = 109, Lost = 1 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 57ms, Maximum = 574ms, Average = 102ms

 

[Mze]

Know your facts and reduce frustration

Hi Ed,

I tried to enter this into the blog, but it did not want to know. You should have this fixed.

Please enter it – In fairness, something positive has to be said about iBurst.
Firstly, every user must know the best base-station in his/her area from which to get service - through the advice of either your own techies or those of your resellers.
Following this advice, the appropriate antenna must be correctly installed.
You, at iBurst, do not have antennae of the performance criteria required. Admit this, and advise installation of superior, and proven equipment. This does not include pushing omni-directional antennae for the simple reason that uni-directional antennae will always give the best performance when one has to select a less loaded BS from amongst the heavily loaded BSs in the area.
Having said this, one has to be concerned at the fact that one’s best BS can go down, leaving one in limbo. In my case, there are several base-stations which are visible, and which will still give very good back-up performance because the alternate BS db readings are still in the vicinity of -80 to -85 db. This despite the fact that my Yagi antenna has a narrow “field of vision”.

Whilst the forgoing is intended as a positive contribution and perhaps a measure of encouragement to those who are presently highly frustrated at their iBurst’s performance, you have no idea as to how thoroughly pissed off and frustrated I have been in the last 12 months. It is only by taking a positive approach and finding out as much as possible about base-stations and antennae, that I have been able to achieve at least 60 to 70% (and often more) of the 1Mbps that iBurst offers under the optimum operating conditions. This despite the fact that that I am located in an area where the coverage map indicates that I will achieve 300kbps. (Consider my 600 to 900+kbps vs 300kbps). Having achieved this performance improvement, my connection no longer drops on an unplanned basis the way it used to. It is now rare that I am dropped overnight.

A very hackneyed cliche - To do things right, you must do the right things.

 

[fishfly]

Great post there Mze

 

[Ekhaatvensters]

This despite the fact that that I am located in an area where the coverage map indicates that I will achieve 300kbps. (Consider my 600 to 900+kbps vs 300kbps). Having achieved this performance improvement, my connection no longer drops on an unplanned basis the way it used to. It is now rare that I am dropped overnight.

Well just to note that the coverage map does not include antennae for its measurement. So you would expect to get better results that what the map tells you using an antennae.

But what you are quite right about is that your connection is definitely not just about speed. The lower your speed the worse your connection and therefore the more you will get dropped and the worse your "lag" or latency spikes and general latency will be. So an iBurst connection that only gets about 384kbps (ADSL reference) most of the time will be miles worse than a connection which gets 1mb most of the time, although you would not notice such a big difference or no difference in gaming or browsing on the respective ADSL services.

Just posting that because often people just want to know what kind of speed they will get, and they think living with 500kbps is fine, because for the same price they could only afford 384kb ADSL. However, the ADSL line at that speed would browse faster and be much more consistent in general never mind gaming.

 

[Ed Hall]

MZE, thank you for the comments and the stats on the Yagi. Based on this, I am going to forward your results to our Product Development team.

Thanks again.
Ed

 

[bdt]

DAMN but I hope I don't come off as pompous as this when I get going..!

I am situated in Blairgowrie and installed my UTD last year in May and experienced the most fantastic transfer rates, day and night.

By September/October last year the performance experienced became very poor.

After a visit by an iBurst expert, I had the uni directional Webb antenna with a 5m extension cable directed at BS 13 in Randburg. Performance better but not as good as originally experienced.

After a further visit by Iburst expert, I put together a real Heath Robinson mast comprised of two 25mm dia swimming pool "poles" overlapped by approx 1 metre and held together by strong cable ties. Total height of the mast is approx 6 metres.

*sigh* .. the things done to the public/consumers by "experts" - it's little wonder people get so hostile at technicians after the treatment they get at the hands of "experts"!

The Webb antenna was hung on the top of the mast by a strategically bent wire coat-hanger and two wire coat-hangers attached to the eaves of my house secured the mast against the rainwater gutter. the bottom of the mast rested on the ground, and was secured by nylon rope to a steel stake driven into the ground. The top of the mast was prevented from swaying excessively by two guy ropes attached to the holes where a pool brush would normally be attached. This marvellous edifice could be rotated by my son while I ran utTraceStar to determine the optimum direction for the Randburg BS. Performance was better but still not what I wanted.

Thanks for a great chuckle, this is definitely not what I would have done, but then I have the right tools for the job that are properly set-up AND I know how to use ..would that make me an "expert" too?

A visit by Renier Smit, who marvelled at the antenna mast, resulted in the Webb antenna being directed at the Illovo BS 27. The performance was now much better but still not good enough.

I learnt of the existance a Yagi antenna. The general manager of the company was kind enought to have it delivered to my house this week. After it was installed and tested and correctly positioned, I was very pleasantly surprised by the increase in performance.

Ahh yes, our old friend the Yagi (actually, once upon a twice it was a Yagi-Uda after the inventors' names, but that form is long since obsolete). Truth be told, there ARE other antenna configurations, but there aren't many off-the-shelf products in the 1700MHz band that iBurst use, so for directional HGA (high gain antenna, the de-facto abbrev. here, get used to it ) action, we're pretty much restricted to the not very exciting (in terms of gain and beamwidth) Yagi or the somewhat better flat panel, with 40deg horiz. & vert. beamwidth, (which are still a LONG way off the characteristics, at 12- & 16deg. respectively, of a 19dBi grid) but, short of cooking up a home-brew job, it is what it is.

Oh, and creepy crawly things that scuttle in the dark have antennae ..what WE use are antennas. Yes, I'm being pedantic, but it IS technically correct and as a tech. I can't abide things being done incorrectly.

 

[bdt]

Hi All
Nice idea to double up on antennas, but doing so will only increase your signal by 3db's, but you will loose it again through your splitter (-3dB's for every split)....so you will not improve the signal strength...just narrow your beamwidth.

Graeme

Ahh, not exactly. By my calcs (and for the sake of simplicity I'm excluding the cost of the rest of the signal path) if I take two of the aforementioned 12dBi Poynting quads and feed them into a splitter that costs me 3dB insertion loss I'm still sitting with 9dBi, so I don't "loose" *cringe* (hint: it's "lose"!) enough to make it not worth the effort.

Actually, I've got a site in Hyde Park that sits on the edge of a signal black-hole so I have to mount his UTD in a box up on the top of the house and in the name of NOT being at the mercy of just one tower (like we were with Stinktech) I'm seriously thinking about aiming two panels at two different towers ..I can let y'all know how it turns out if I end up having to go that route.

 

[Radiant Antennas]

Yagi and panel antennas

Hmmm....had to think about that alittle....you have clearly formulated an opinion from what you have read on product spec sheets...however, somewhat correct...not fully so.

Does a Mazda out perform a Toyota? Which models you ask.....exactly!!

For the most....a yagi of 10dBi gain and f/b of say 15dB, and a panel of 10dBi gain and f/b of 15dBi, will perform exactly the same.

Try not to compare a 17dBi panel against a 10dBi Yagi, and then make the statement that a panel is better than a yagi. If you do compare two antennas of similar spec's, and one does not match up......the manufacturer is lying about his gain.....simple as that.
By the way...wind loading on a yagi is way less than a panel.

Cheers
Graeme

 

[Radiant Antennas]

Doubling up

Ahh, not exactly. By my calcs (and for the sake of simplicity I'm excluding the cost of the rest of the signal path) if I take two of the aforementioned 12dBi Poynting quads and feed them into a splitter that costs me 3dB insertion loss I'm still sitting with 9dBi, so I don't "loose" *cringe* (hint: it's "lose"!) enough to make it not worth the effort.

Forgive me but maybe I don't understand.....you started with a 12dBi Yagi....you added another, through a splitter, and have now got a 9dBi antenna.....how did you not lose? (except you are now a few hundred rand lighter)
Cheers

 

[bdt]

Interesting findings, does this improve your ping at all? Aren't you worried about lightning strikes?

Anytime you put a nice big metal spike in the sky you're saying 'look at me' to lightning but we counter that by putting a lightning protector in the line. While it's no guarantee of safety, it's like chicken soup: it may not help but it doesn't hurt!

 

[bdt]

Forgive me but maybe I don't understand.....you started with a 12dBi Yagi....you added another, through a splitter, and have now got a 9dBi antenna.....how did you not lose? (except you are now a few hundred rand lighter)
Cheers

Heh, this can take a bit to get your head around. OK, take a look at the splitter we (have available to) use: its *intended* usage is to put the antennas on the side with two connectors and the CPE (customer premises equipment, in this case the UTD) on the side with one (tho' I've done it the other way round and it worked a treat). So, we plug the 12dBi Yagis in on the one side -and they don't 'see'/interfere with each other- so, for all intents and purposes we calculate our link budget with only one: a starting point of 12dBi. Now they SAY it has an insertion loss of only 0.6dB, but it's never that kind to you - hence the generally accepted claim of 3dBi ..which is why we do: 12-3=9dBi resulting potential gain THAT far (at the single-sided end of the splitter - you still have to take into account the cost of your co-ax, but that's outside the scope of this reply).

 

[Radiant Antennas]

One of us is confused

Firstly about splitters....if you have a box that has one input connector and two outputs (a splitter), and you put in a signal of 100mW....you get a theoretical 50mW from each connector....each one -3dB's.
No splitter is perfect, so insertion loss is added...lets use your -0.6dBs, thus your loss at each connector is -3dBs + (-0.6dBs)....total -3.6dB....understand so far??

Now take your 12dBi yagi....add another, get another 3dB's......total 15dBi.
Connect each one into the splitter........-3dB (and the -0.6, but this is negligable)....and you are back to 12dB......where you get 9 from beats me, and why you think this 9 is better than your starting 12 is a mystery.

 

[bdt]

Hmmm....had to think about that alittle....you have clearly formulated an opinion from what you have read on product spec sheets...however, somewhat correct...not fully so.

No-o-o-o, I'm clearly formulating my opinion on having been working with HGAs right up and down the performance spectrum for oh, the last two and a half years or so worth of being out in the field ..well, more like up on top of many, many houses, all around greater northern Jozi. Disclaimer: I did HGA installations for MyWireless, but all that means is I worked in the 2.5GHz spectrum and had fewer towers to aim at.

Try not to compare a 17dBi panel against a 10dBi Yagi, and then make the statement that a panel is better than a yagi. If you do compare two antennas of similar spec's, and one does not match up......the manufacturer is lying about his gain.....simple as that.

Tell you what: I'll go with my field experience of the last little while coupled with the familiarity of the products I have at-hand from suppliers whose product specifications I've learnt I can believe - that sound reasonable to you?

By the way...wind loading on a yagi is way less than a panel.

By the way, the crappy plastic Yagi we have doesn't have the world's greatest retention mechanism compared with the panel. Moreover the panel has better (enough) beamwidth characteristics that I'm prepared to prefer it over the Yagi ..but that's just me.