SDK 5

[L&LD]

I did not correct any claim I made. I corrected the assumption you made about me.

Positioning antennae is also important too. But the standard \|/ orientation as recommended by Asus serves most very well. And when needed in a particular environment, I do test different antennae orientation too.

I am glad to see that you have corrected your previous claim on this issue.

I'll offer another "real world" experience to this thread. I live in a 113-year-old house with very thick brick walls. About 60 feet behind the house is a similarly-built carriage house whose second floor I wished to serve with WiFi.

My first attempt was an RT-N66U in a ground floor window of the main building, facing the carriage house. Like most installers I oriented the antennas vertically. Connectivity in the carriage house was spotty.

So I tipped the top of the antennas back about 10 degrees so a line drawn at right angles to their length intercepted the carriage house second floor. The signal inside the carriage house, 60 feet north, through its brick walls, rose to -50dB -- a nice solid connection. By tipping the antennas 10 degrees I tipped the "doughnut" by the same amount.

 

[dlandiss]

I did not correct any claim I made. I corrected the assumption you made about me.

Please review Message #8.

 

[L&LD]

That is not my post nor yours.

My statements stand by themselves and the support for hggomes was directly related to what I write in my own posts and not what others write in theirs.

Please review Message #8.

 

[dlandiss]

That is not my post nor yours.

My statements stand by themselves and the support for hggomes was directly related to what I write in my own posts and not what others write in theirs.

My error. Yes, for some reason I thought I was replying to gomes, not to you. It is clear that you have a much better understanding of antennas than most installers, and your experience and mine do not contradict each other.

 

[L&LD]

One small correction. I am not an installer, but I do that too.

My error. Yes, for some reason I thought I was replying to gomes, not to you. It is clear that you have a much better understanding of antennas than most installers, and your experience and mine do not contradict each other.

 

[ilgeor]

Hi
Uhm, I see, my small question resulted in some general discussion, showing that this question really matters.

Thanks for all the advice given.
There is, however, an important restrictions:
I simply cannot re-position the router - it should stay where it is.

So, the problem is, my "room of interest" is separated from the router by three concrete (or brick, idk) walls. The house was built some 50 years ago, so you could guess they are pretty thick. You should add furnishing, built-in wardrobe and other nice stuff. And don't forget, my neighbors also love using WiFi. All in all, that's called bad conditions

Anyway, 3.0.0.4.374.35_4-sdk5 firmware is providing me with the signal, which could be called bearable.

By the way, I'm using higher gain antennas (dbi +9, i presume http://www.ebay.com/itm/3x-Lot-9-db...ain-antenna-/141020367766?hash=item20d5780b96) . The router stands vertically (well, almost).

I'll try adjusting the position and (thanks to Rmerlin - John's fork). I'll let you know the results.

 

[L&LD]

You might want to try to pm hggomes and ask for the link to his latest RMerlin fork with unlocked power options. Works very well for a few customers.

Even with an extra power boost, the three concrete / brick walls may still win in the end. But only testing will let you know for sure.

I would download john9527's nvram save/restore utility and also do a gui backup of the settings with the version you now have installed. Make sure you have the current firmware downloaded and checked for errors too on your computer before continuing.

Download the appropriate version of hggomes fork and after it has loaded, do a full reset to factory defaults and another reboot too.

Use the nvram save/restore utility to setup the basic settings of the router again, but I would suggest leaving things as close to default as possible and also change the ssid's of both bands. Go to the options and set the power to full or any level you want to start testing at.

With the power level settings adjusted, I would now start testing the channels individually on the 2.4GHz band (1, 6, 11 and 13, if that is available for your devices) and see if one works better than another.

Test for maximum download speed from Ookla speedtest, for example. Test for wireless performance copying a large 1GB+ file from a wired computer to a wireless computer and also test how fast or responsive the internet is on each channel tested. Do all these tests at a variety of locations inside your home and you'll see one is better than all the others. You can also do this for the 5GHz band, but you don't need to skip channels because they don't overlap.

Here, the best channels are 11 for the 2.4GHz band and the lower channels for the 5GHz band (usually 48 or so). The 2.4GHz band should be on a 20 MHz channel width and the 5GHz band 20/40/80.

If all is working good at this point, create another backup for the router settings with both the gui and the nvram save/restore utility too.

If you do decide to try this, I will be very interested in your results, so keep us posted.

 

[dlandiss]

L&LD, I have not found it necessary to increase the transmitter power. John's fork gives me very good range. 160mW would only be a 3dB increase anyway. I do use 9dBi antennas and they did help, raising the throughput from 7Mbps to 33Mbps on the 5GHz band. If we needed more I would put an ASUS USB-N66 antenna back in the carriage house, as it is tons better than the ones built into laptops.

You are spot on about testing for throughput instead of signal strength. There are so many other factors that influence the throughput that relying on signal strength alone is deceptive. It is just a starting point, and it varies too much from second to second during testing to trust as a final result.

 

[hggomes]

Have you measured your signal after attaching the 9dBi antennas vs the stock antennas on the top and down floors?

How the measurement went? Since high gain antennas will have less vertical angle...

 

[dlandiss]

Have you measured your signal after attaching the 9dBi antennas vs the stock atennas on the top and down floors?

How the measurement went?

Yes. Since the stock antennas are said to be about 2dBi I might have gotten as much as 7dB increase. I'd say on the average it was more like 4-5dB increase.

Since high gain antennas will have less vertical angle...

Exactly, which is why I had to tip the tops back about 10 degrees.

 

[hggomes]

But the vertical angle % of your antenna will not increase by setting it those 10 degrees, you are simply pointing/adjusting the antennas in a better vertical angle position to your clients due to having a worst vertical angle compared to stock antennas, but at the end the vertical angle will still be less compared to default antennas, specially at small distances which is alot noticable, and that can be a problem if you have several clients spread over the floors.

You cannot have the best of the two, you will always have to choose between having a antenna with less dBi but more vertical angle or having a higher dBi antenna but less vertical angle, which will have a big impact on your coverage on top and down floors.

Most of the times when upgrading to high gain antennas, what happens is that the signal will be alot better on the same floor (horizontal angle) but it will be alot worse and can be barelly seen (black spots) on the lower and higher floors (vertical angle), and that also happens due to low distance, the more the distance the better the vertical angle gets spread/opened.

Having a 20º vertical angle on a 9dBi antenna is not the same of having a > 60º angle on a 2dBi antenna.

Resuming, high gain antennas to low distance on lower and upper floors (vertical angle) is non-sense, you will be alot better with a low gain antenna.

 

[L&LD]

If you had the original antennae to test with the higher power that hggomes firmware fork offers, it would make for an interesting comparison to your 9 dB antennae you have now. Especially in light of hggomes post too.

I know it would be a lot of work, resetting to factory defaults and all. But think of the information you can share.

And, you might find a better solution than what you have now.

L&LD, I have not found it necessary to increase the transmitter power. John's fork gives me very good range. 160mW would only be a 3dB increase anyway. I do use 9dBi antennas and they did help, raising the throughput from 7Mbps to 33Mbps on the 5GHz band. If we needed more I would put an ASUS USB-N66 antenna back in the carriage house, as it is tons better than the ones built into laptops.

You are spot on about testing for throughput instead of signal strength. There are so many other factors that influence the throughput that relying on signal strength alone is deceptive. It is just a starting point, and it varies too much from second to second during testing to trust as a final result.

 

[fiasko]

When indoors, you should not rely too much on the theoretical RF signal pattern models, since the RF signal behavior is quite difficult to predict in indoor environment. Take a look at one example in here http://cnlohr.github.io/voxeltastic/

 

[kvic]

@hggomes With due respect to the info you provided in #31, I don't see what you're trying to refute

Use the same illustration u provided. Let the carriage house be 60 pixels to the right of the antennas. And let the carriage house be 5 pixels tall. Across the three story carriage house, which of the following has the strongest signal strength: A) 9dbi antenna, B) 7dbi antenna, C) 5 dbi antenna, D) 2dbi antenna; and E) none of the above.

The answer is obvious to me.

 

[hggomes]

I think you are missing the picture here, based on your query a 15dBi antenna with a 9º vertical angle would be even better from your point of view, right? Wrong, it doesnt matter if you have the strongest signal in the world when it doesnt covers the upper and lower floors (vertical angle) on your house, im talking about vertical angle here, not horizontal.

The neigbour houses will pick better signal on those floors with a high gain antenna than you inside your house, that's not the idea i think, is it?

 

[dlandiss]

But the vertical angle % of your antenna will not increase by setting it those 10 degrees

Vertical angle (the direction of strongest signal) is not the same as vertical beamwidth (the 25 degree angle in the drawing you attached). Tipping the antennas results in tipping the vertical angle of the "doughnut" but does not change the beamwidth.

Referring again to the drawing you attached, if the top of the antenna was tipped to the left it would rotate the entire diagram CCW. which aims the portion labelled "effective signal coverage" exactly at the second story of the carriage house.

The fact that the left side of the drawing tips down at the same time is of no consequence since the devices that use WiFi on that floor of the main house are in the same room as the access point.

 

[dlandiss]

If you had the original antennae to test with the higher power that hggomes firmware fork offers, it would make for an interesting comparison to your 9 dB antennae you have now. Especially in light of hggomes post too.

I know it would be a lot of work, resetting to factory defaults and all. But think of the information you can share.

And, you might find a better solution than what you have now.

I tested the effects of firmware and antennas on throughput at long range separately. John's fork alone added about 40% to the throughput. The 9dBi antennas nearly quintupled the throughput (message #28). Even though the signal strength only changed by a few dB, the improvement in S/N ratio was enough to cause an impressive speed difference.

 

[dlandiss]

When indoors, you should not rely too much on the theoretical RF signal pattern models, since the RF signal behavior is quite difficult to predict in indoor environment. Take a look at one example in here http://cnlohr.github.io/voxeltastic/

But theory does provide a valid starting point. It would be counterproductive, for example, if one wishing to cover a laptop two floors higher and exactly above the access point, to start with the antennas oriented vertically. That would put the laptop right in the "doughnut hole" and any useful signal it sees would be from random reflections.

 

[kvic]

Vertical angle (the direction of strongest signal) is not the same as vertical beamwidth (the 25 degree angle in the drawing you attached). Tipping the antennas results in tipping the vertical angle of the "doughnut" but does not change the beamwidth.

Referring again to the drawing you attached, if the top of the antenna was tipped to the left it would rotate the entire diagram CCW. which aims the portion labelled "effective signal coverage" exactly at the second story of the carriage house.

The fact that the left side of the drawing tips down at the same time is of no consequence since the devices that use WiFi on that floor of the main house are in the same room as the access point.

This is quite an elegant engineering I must say.

 

[dlandiss]

This is quite an elegant engineering I must say.

I wish! Just a straight-forward application of well-documented antenna principles.

I'll bet thiggins is getting some good smiles out this thread!