WiFi dish mod to 3.6GHz?

[chris]

Hi SEDers!

At the industrial premises where my office is, the company next door was
throwing this out:
http://i13.photobucket.com/albums/a258/ChrisTH01/IMGP1813.jpg. It's
alleged to be a WiFi (2.54GHz) dish.

I'm more interested in finding out how much work is involved in adapting
it to the local wireless internet ISP's frequency of 3.5-3.6GHz.

The innards of the item at the dish focus are shown here:
http://i13.photobucket.com/albums/a258/ChrisTH01/IMGP1815-2.jpg along
with some approximate measurements (+/- 1mm).

The last time I dealt with any RF theory was in University (20 years
ago) and I've never used it since, so I'm hoping I can get some pointers
as to whether it's worth bothering with. The local price for
manufactured aerials for this ISP are in the AUD 150-200 range as far as
I've found so far.

So, my guesses so far:
It seems like the aluminium part is a secondary reflector?

The dipole part is quite close to the theoretical value for 1/2
wavelength at 2.54 GHz (I calculate ~59mm for 1/2 wavelength).

Is it just a matter of trimming the dipole to the equivalent 1/2
wavelength for 3.6GHz (~42mm) ?

The reason I ask is that it looks like there's some sort of balun
incorporated by the way both brass shims are actually soldered to the
braid, but the center conductor is tied to the 'center' (approx) of the
dipole. Am I close?

If so, does that mean the length of the brass 'feeders' that are at
right-angles to the dipole is critical too? If so, I presume the fold in
the shim should be roughly 1/4 wavelength?

Thanks in advance for any advice offered.

Chris.

 

[rex]

Hi SEDers!

At the industrial premises where my office is, the company next door was
throwing this out:
http://i13.photobucket.com/albums/a258/ChrisTH01/IMGP1813.jpg. It's
alleged to be a WiFi (2.54GHz) dish.

I'm more interested in finding out how much work is involved in adapting
it to the local wireless internet ISP's frequency of 3.5-3.6GHz.

The innards of the item at the dish focus are shown here:
http://i13.photobucket.com/albums/a258/ChrisTH01/IMGP1815-2.jpg along
with some approximate measurements (+/- 1mm).

The last time I dealt with any RF theory was in University (20 years
ago) and I've never used it since, so I'm hoping I can get some pointers
as to whether it's worth bothering with. The local price for
manufactured aerials for this ISP are in the AUD 150-200 range as far as
I've found so far.

So, my guesses so far:
It seems like the aluminium part is a secondary reflector?

The dipole part is quite close to the theoretical value for 1/2
wavelength at 2.54 GHz (I calculate ~59mm for 1/2 wavelength).

Is it just a matter of trimming the dipole to the equivalent 1/2
wavelength for 3.6GHz (~42mm) ?

The reason I ask is that it looks like there's some sort of balun
incorporated by the way both brass shims are actually soldered to the
braid, but the center conductor is tied to the 'center' (approx) of the
dipole. Am I close?

If so, does that mean the length of the brass 'feeders' that are at
right-angles to the dipole is critical too? If so, I presume the fold in
the shim should be roughly 1/4 wavelength?

Thanks in advance for any advice offered.

Chris.

I think everything you said is about right. Getting the dipole/reflector
working good at exactly the new frequency may be difficult without a way
to measure return loss at the new frequency. Not to say you couldn't get
lucky.

One thing you didn't mention is the mesh nature of the dish reflector.
As you go up in frequency it will work less and less good.

 

[[email protected]]

I think everything you said is about right.

Really! Wow

Getting the dipole/reflector working good at exactly the new frequency
may be difficult without a way to measure return loss at the new frequency.

Yes, I'm not sure how to get around this yet. I do have access to an HP
8485A power sensor and power meter but I don't see how to get the
return loss using this gear.

Not to say you couldn't get lucky.

Well, I'd like to maximize my 'luck' if possible ;-)

One thing you didn't mention is the mesh nature of the dish reflector.
As you go up in frequency it will work less and less good.

Yes, I was thinking about this too. I thought I'd cover the existing
reflector with a finer mesh like brass flyscreen or similar.

Thanks for your reply!

Chris.

 

[Mark]

the dish itself should work fine

the dipole can easily be trimmed smaller and the exact size is not
really that critical

it may be very difficult to modify the electronics to the new frequency
and modulation and any other phy layer changes needed.

Mark

 

[Joerg]

Hello Chris,

So, my guesses so far:
It seems like the aluminium part is a secondary reflector?

Yes, and it needs to be shortened as well.

The dipole part is quite close to the theoretical value for 1/2
wavelength at 2.54 GHz (I calculate ~59mm for 1/2 wavelength).

There are some other factors (vicinity to plastic etc.). Just shortening
it proportionately usually works. Same with the aluminum reflector.

The reason I ask is that it looks like there's some sort of balun
incorporated by the way both brass shims are actually soldered to the
braid, but the center conductor is tied to the 'center' (approx) of the
dipole. Am I close?

The connection of the "hot" element side (the one fed by the center
conductor) to the coax sleeve may have to be dropped.

You'll lose the holding pegs because both elements will now be shorter.
You could glue in a plastic flange to screw the center of the aluminum
reflector to. For the main element maybe some potting compound can hold
the center. Try to avoid affixing the ends. Any load, even dielectric,
can mess up the performance because they are high impedance there.

If they "threw" it into the dumpster followed by other heavy stuff the
reflector might be bent which degrades performance. Also, as Rex said,
the mesh size may be a bit on the coarse side for the higher frequency.
But heck, even if this antenna only gets you a few dB that would be
better than nothing. Considering that it didn't cost you anything.

On a side note, sometimes higher gain antennas aren't allowed by the
authorities.

Regards, Joerg

 

[chris]

the dish itself should work fine

the dipole can easily be trimmed smaller and the exact size is not
really that critical

it may be very difficult to modify the electronics to the new frequency
and modulation and any other phy layer changes needed.

That part shouldn't be a problem because I want to try it with a
Wireless broadband 'modem' (for want of a better word) which is designed
to work at the 3.6GHz frequency. It also has external antenna connections.

Mark

Thanks for your reply.

Chris.

 

[chris]

Hello Chris,



Yes, and it needs to be shortened as well.



There are some other factors (vicinity to plastic etc.). Just shortening
it proportionately usually works. Same with the aluminum reflector.



The connection of the "hot" element side (the one fed by the center
conductor) to the coax sleeve may have to be dropped.

You'll lose the holding pegs because both elements will now be shorter.
You could glue in a plastic flange to screw the center of the aluminum
reflector to. For the main element maybe some potting compound can hold
the center. Try to avoid affixing the ends. Any load, even dielectric,
can mess up the performance because they are high impedance there.

If they "threw" it into the dumpster followed by other heavy stuff the
reflector might be bent which degrades performance. Also, as Rex said,
the mesh size may be a bit on the coarse side for the higher frequency.
But heck, even if this antenna only gets you a few dB that would be
better than nothing. Considering that it didn't cost you anything.

Well, I salvaged it before anything else was thrown on top of it so I'm
reasonably confident the reflector's OK ;-)

I was also thinking about the mesh size and I thought I'd add a layer of
brass flyscreen or similar just to be on the safe side.

On a side note, sometimes higher gain antennas aren't allowed by the
authorities.

I'll look up the local rules, but for now it's only a temporary test on
borrowed equipment.

I live on the outskirts of town where there's no wired broadband and
here in Oz, the sattelite broadband (my only retail option) is quite
expensive (relatively) especially two-way. I don't rate my chances
success very highly as the nearest wireless broadband tower is ~15km
away and I don't have a clear line-of-sight.

Thanks for your reply! It's encouraging to think that some of that
20-year-old knowledge stuck!

Regards, Joerg

Chris.

 

[Ken Taylor]

Well, I salvaged it before anything else was thrown on top of it so I'm
reasonably confident the reflector's OK ;-)

I was also thinking about the mesh size and I thought I'd add a layer of
brass flyscreen or similar just to be on the safe side.


I'll look up the local rules, but for now it's only a temporary test on
borrowed equipment.

I live on the outskirts of town where there's no wired broadband and here
in Oz, the sattelite broadband (my only retail option) is quite expensive
(relatively) especially two-way. I don't rate my chances success very
highly as the nearest wireless broadband tower is ~15km away and I don't
have a clear line-of-sight.

Thanks for your reply! It's encouraging to think that some of that
20-year-old knowledge stuck!


Chris.

One thing to contemplate as you point 3.6G around the place is that it's
*awful* close to C-band satellite downlinks. I don't know where in Oz you
are but you need to be mindful, as those services will be protected.

FWIW, since you're playing, you could try 'just' shortening the lengths and
distances involved in the various bits of copper/Al. You might look to
applying the same 'fudge-factor' that the original manufacturer's came up
with when they got their lengths to be slightly off 1/4-wave, etc. The ARRL
Antenna Handbook (or was it the RSGB VHF/UHF Manual?) gave a pretty
reasonable run-down on working at these freq's. Both books at Tech-Books in
Melbourne.

Cheers.

Ken

 

[Joerg]

Hello Chris,

I live on the outskirts of town where there's no wired broadband and
here in Oz, the sattelite broadband (my only retail option) is quite
expensive (relatively) especially two-way. ...

Same here in California. About $80 AFAIR. DSL is $50 and web over TV
cable is in the same range but only if you also take cable TV. No way
we'd do that.

...I don't rate my chances
success very highly as the nearest wireless broadband tower is ~15km
away and I don't have a clear line-of-sight.

I don't want to spoil the broth here but at 3.5GHz and without a line of
sight the chances are pretty slim. If trees are blocking the path,
maybe. But if it's building probably not.

Thanks for your reply! It's encouraging to think that some of that
20-year-old knowledge stuck!

Once you have practiced RF for a while it is like riding a bike. You may
get a bit wobbly but you won't forget the basics.

Regards, Joerg

 

[Joerg]

Hello Chris,

That part shouldn't be a problem because I want to try it with a
Wireless broadband 'modem' (for want of a better word) which is designed
to work at the 3.6GHz frequency. It also has external antenna connections.

Coax losses are huge at that frequency. So if you detect a faint signal
you can improve things by placing the modem really close to the antenna.
Or you could then try to find one of those very large dishes at a yard sale.

Regards, Joerg

 

[chris]

Hello Chris,



Coax losses are huge at that frequency. So if you detect a faint signal
you can improve things by placing the modem really close to the antenna.
Or you could then try to find one of those very large dishes at a yard
sale.

Yes, I'm just now looking around at what to use to feed the signal back
into the house and it's a little disturbing (both the cost and the db
loss per m)!

Regards, Joerg

Chris.

 

[chris]

One thing to contemplate as you point 3.6G around the place is that it's
*awful* close to C-band satellite downlinks. I don't know where in Oz you
are but you need to be mindful, as those services will be protected.

I'm in NW Sydney. Who are the primary users of C-Band? The only likely
installation in the area I can think of that could have sat downlinks is
a RAAF base but it's ~90deg to where I want to aim.

FWIW, since you're playing, you could try 'just' shortening the lengths and
distances involved in the various bits of copper/Al. You might look to
applying the same 'fudge-factor' that the original manufacturer's came up
with when they got their lengths to be slightly off 1/4-wave, etc.

That was basically going to be my approach.

The ARRL
Antenna Handbook (or was it the RSGB VHF/UHF Manual?) gave a pretty
reasonable run-down on working at these freq's. Both books at Tech-Books in
Melbourne.

I'll check the libray first, I think, if I need to. It's not something
that's likely to crop up that much in my case ;-)

Cheers.

Ken

Thanks for your suggestions!

Chris.

 

[Keith Williams]

Hello Chris,


Coax losses are huge at that frequency. So if you detect a faint signal
you can improve things by placing the modem really close to the antenna.
Or you could then try to find one of those very large dishes at a yard sale.

Or you could find some parabolic Chineese cookware.

http://www.usbwifi.orcon.net.nz/

 

[Joerg]

Hello Chris,

Yes, I'm just now looking around at what to use to feed the signal back
into the house and it's a little disturbing (both the cost and the db
loss per m)!

I'd first have the laptop in the attic or something, maybe even on the
roof with you if that is safe. Once it works you could place the modem
close by. CAT-5 is a whole lot cheaper than mil grade coax.

Regards, Joerg

 

[Joerg]

Hello Keith,

Or you could find some parabolic Chineese cookware.

No joke: When I was young UK ham radio operators discovered that trash
can lids in London were the perfect reflector.

Regards, Joerg

 

[chris]

Or you could find some parabolic Chineese cookware.

http://www.usbwifi.orcon.net.nz/

Saw this on Slashdot a while back - quite a cheap solution for WiFi in
any case ;-)

Cheers,
Chris.

 

[chris]

Hello Keith,



No joke: When I was young UK ham radio operators discovered that trash
can lids in London were the perfect reflector.

Regards, Joerg

LOL! Our bins now all have square hinged lids so... but maybe if I
forget the lid & line the entire bin with tin foil... I'm thinking
Pringles can on steroids! ;-)

Cheers
Chris.

 

[chris]

Hello Chris,



I'd first have the laptop in the attic or something, maybe even on the
roof with you if that is safe. Once it works you could place the modem
close by. CAT-5 is a whole lot cheaper than mil grade coax.

Regards, Joerg

Yep, I've got an upstairs window that's facing in the roughly the right
direction so it'll definitely be a case of minimal length coax to start
with...

Cheers,
Chris.

 

[Joerg]

Hello Chris,

LOL! Our bins now all have square hinged lids so... but maybe if I
forget the lid & line the entire bin with tin foil... I'm thinking
Pringles can on steroids! ;-)

It's all plastic out here now. That makes for interesting scenarios if
people don't let their wood stove ashes cool down long enough. We bought
an extra couple of steel cans that we can rotate but most people have
just one.

Regards, Joerg

 

[Rich Grise]

And moved closer to the dipole, by the same factor

Exactly, but I'd find a piece of braid so that you could extened that
solder-soaked part of the braid such that the "open" part of the balun
is also shortened by the same factor.

Essentially, just shrink the whole assembly. Everything everybody else
said is spot-on.

Good Luck!
Rich