Connecting to a weak router signal (again)

[Loose rivets]

It's been quite good for a few weeks, 2 or 3 bars. But suddenly it reverted to its old poor signal. My friend's NETGEAR is still hanging in a plastic back on the garage wall, on the other side of four brick walls.

How could it have been okay for a while?

The cheapest option seemed to be to buy a 5 or 7 dB high gain antenna. Worth a try. The 7 db just lost the signal altogether from the 3G unit. A borrowed 5 db of the same design did the same. I'm suspicious that despite assurances from several people, the gold pin is not reaching the new connection. The knurled wheel seems to be getting in the way of it going all the way in.

The signal type/frequency is correct for matching the router.

So, I borrowed a socking great external wall unit 50 or 60 cm long, and the wire seemed to make a more satisfactory connection. :mad: all signal.

Is there anything glaringly obvious I'm missing, like having to chant some incantations over the bigger units?

I have an identical router to play around with, and even with it just 50 feet away, the 'better' aerials will kill the signal from that. Just totally bewildered.

[Loose rivets]

Calling on old electronics training, I suspected a double or even a triple fault. So, using the spare router, I put a tiny piece of steel wire in the 5db aerial connector so that it reached the router pin. It worked! but still not quite as good as they tiny kit one.

Okay, so now why the big one not working? It's sealed into a rigid 'glass' tube, but when I invert it, it clonks from within. So, maybe fault 2.

Wire to this one is okay measured with a test meter. Pins well engineered - but I did the trick of extending the connection. Poor signal, but now it was at least there.

Tried different channels on the basis hi-gain aerials might be finely tuned to one frequency. Slight changes, but nothing that would account for such a low signal. Put diddy one back, and 5 bars again. Still flummoxed - as are the computer company owner than loaned the kit and my ex IBM pal oooop north. "It/they should just work."

[Mr Optimistic]

Confused of Bedford here.
'Put diddy one back, and 5 bars again.'
So solved ?
Are you testing with receiving unit in the garage to see if its a range or 'local' issue ? Wonder what the impedance of a bent bit of wire is.
Have you eliminated issues at the receiving end ? Interference from networks ? Nothing else varying (ie car in or out of garage......)

[Loose rivets]

Got up this morning determined to find a solution.

The last part of this post is a mystery to me - but first:

Having to assume huge aerial is faulty. 5db one not as good. Back to the little one.

To test the local signal, I put the second NETGEAR v 3 unit in the plastic conservatory that's fixed to the back of the garage. To my astonishment, I'm now getting the strongest signal for a long time from the FIRST and active unit. Coincidence? Probably.

The borrowed second unit giving a modest two bars.

Haven't co-related car in or out of garage - good thinking Bat Man.

Mystery bit: My pal once put a yellow wire into the conservatory. He says he simply connected the Linksys one to his NETGEAR via a yellow wire. OUTPUT TO OUTPUT. This can't be right . . . can it?

It's my intention today to take a 'snapshot' of his settings, then use mine as a relay station. It means pal and family must not need the net during my messing about time as it seems to be a lengthy procedure. Well, not if things go well, but in the real world...

[Mr Optimistic]

Don't forget that higher gain antenna only means more directional, so get the geometry between Tx and Rx wrong and it will make matters worse. Gain directionality will also be more sensitive to near by metallic objects. Don't know about polarisation either.

From memory, the antenna will have a doughnut shaped pattern with max strength orthogonal to antenna axis (though if I am wrong no doubt someone will chirp up ), so antenna should be perpendicular to line of sight from computer to router.

[green granite]

You also need to take into account whether the signal is horizontally or vertically polarized.

[Loose rivets]

Aerials OMNI, but apart from the flat doughnut they have a lobe out of the top which can be pointed at the target.


There seems to be so many variables. The spare NETGEAR unit suddenly disappeared from the list of local routers. The Rivetess' laptop is showing differing stations available, and she's two feet away.

Signal from pal's NETGEAR is consistently swinging between 1 and 3 bars, though some days we get 2 to 3 bars and everything is fine - at least for our temporary needs.

I can only assume the signal is being affected by other transmissions, unless these routers try to make assessments of the signal from what they receive, then turn down their output. But, I've never read of an agc function.

I need pal and his family out of the way so as not to inconvenience them, then I'll have a go at the relay procedure. If that fails, it'll be the BT 'approved' through the house wiring deal - BUT, pal seems to have bought several of these houses and had 3phase put in for his workshops. 50/50 it'll work for normal people, but for me in this phase (ho ho) of my life, 99 to 1 against.:uhoh:

[green granite]

BT 'approved' through the house wiring deal

Soon, hopefully, to be made illegal :=

[Loose rivets]

Yes, following on from the last discussion, I said to a computer company owner about the prospect of them being made illegal. He was noncommittal, but I'm sure once his stock was sold out he'd tell people they should buy something else.

Said pal and I are prone to nattering, so I got as far as putting second unit under the perspex roof of a nearby workshop and I get three bars from there.

Most of the settings on his router seem to already be set to allow the relay procedure, but any further experimenting will have to be done when he won't need the net.

Using the dg834g as a repeater (http://www.unix.ms/netgear/index2.html)

[Mr Optimistic]

Yeras ago I had no end of trouble trying to set up a WDS repeater (or something like that). Found the computer which was supposed to benefit often tried to lock on to the weak source signal rather than the amplified local one. Are the 'roaming' settings OK ?

[Loose rivets]

just as I was downloading the above, and planning a session getting the relay working, my pal comes round and says his brother has a dongle device that is a receiver as well. It is highly directional.

He uses it on his boat, so Zoomed off to said ship which is moored at the local marina and took charge of a PheeNet Tech device and CD.

Loaded the Vista soft, but there things started to go wrong. (no real surprise.) It never presented itself as an installed program, and the danger of Windows claiming the device seemed to be becoming a reality.

Indeed, the only evidence I can find of it being on the computer at all is in the uninstall section. Uninstalled it once and tried again, this time puting the USB plug in on an unidentified command. (might have been the OS or the install soft. Just nothing to say.)

However, now it loaded as a name I'd not heard of and seems to be going quite well. Yes, I did leave the thing unplugged until the soft was in, but it didn't follow the pdf with a reboot etc., indeed, it never asked where the files should be put and other small items as per the pdf.

I hate when that happens.:ugh:

Although the number of bars is not improved, the general performance is undoubtedly better. One will persevere on the morrow.

The above testing software would be a very definite boon to assessing the benefits, but I've yet to make it run. Just sits there with empty blocks, and talks of GPS. Will look at that on the morrow as well.

[Loose rivets]

Put laptop back on its own aerial and system, and the inSSIDer worked straight away.

Fascinating!


My signal on the active router seemed very steady, while my nearer and quiescent unit had its output bobbing about over several db. Odd. It's almost line of sight in a closed down workshop with plastic roof.

Despite few other users showing, the signal was switching between poor-fair-good in sharp square-wave jumps. Yes I understand sampling rates, but this just had the feel of some external influence.

Again, inSSIDer did not function with the external USB devices being a boosted receiver/aerial.

[Loose rivets]

Credit where credit's due.

When the inSSIDer program wouldn't work with the external PheeNet's WLU 805 G, I was disappointed, as the unit is directional and I wanted to see the ratio of other signals in the area compared to mine. I e'd MetaGeek Support <[email protected]> and they suggested not only turning off the laptop's wireless device, but also disabling its drivers in Device Manager. It worked.

The little receiver/aerial is certainly picking up a lot more signals, and now I can search for not only the best angle for power, but as a ratio of the other signals.

Thanks NotGettingAnyYounger for a good link.