When to Use 2.4 GHz Bandpass Filter

[Guest]

Archived from groups: alt.internet.wireless (More info?)

I am wondering why or when something like this
http://www.teletronics.com/Accessories.html#24ghzbandpassfilter

is needed. I would have thought that an access point/bridge
(e.g., connected to an outdoor antenna)
would have a receiver tuned to the 2.4 (or 5.x) GHz range, and so
would ignore or be immune to signals at other frequencies such as
VHF/UHF, AM/FM, etc. So why the need for an external
bandpass filter?

Are there any rules of thumb when to put one in? For example,
"Use when WiFi antenna is within x feet of VHF antenna on
same tower radiating at y watts..."?

Or, maybe "Use when WiFi antenna is in Fresnel zone of
gigantic telco solid dish less than x miles away..."?

 

[Guest]

Archived from groups: alt.internet.wireless (More info?)

On 30 Sep 2004 09:14:26 -0700, carhore@yahoo.com (c hore) wrote:

>I am wondering why or when something like this
>http://www.teletronics.com/Accessories.html#24ghzbandpassfilter
>is needed.

That's a bandpass filter for the entire 2.4GHz ISM band, 83.5Mhz wide.
You would need one if you have interference from a souce adjacent to
the the 2.4GHz band, such as MMDS wireless, or ITFS television. It
does NOTHING for any interference in the 2.4GHz wireless band.

There are band pass filters that are only 26MHz wide, which are used
to seperate the 3ea non-overlapping channels (1, 6, 11) and are used
to eliminate interference from co-located wireless users on other
channels.

>I would have thought that an access point/bridge
>(e.g., connected to an outdoor antenna)
>would have a receiver tuned to the 2.4 (or 5.x) GHz range, and so
>would ignore or be immune to signals at other frequencies such as
>VHF/UHF, AM/FM, etc. So why the need for an external
>bandpass filter?

Most radios include ceramic bandpass filters on the board. The
Teletronics data sheet says that it's a 3 pole ceramic filter, which
is no better than what's already on your wireless card. My guess(tm)
is that all you're getting is the same filter in a nice box with
connectors. Inside is something like this:
http://www.toko.co.jp/products/pdf/filters_dielectric/tdf-24lan.pdf
Note the -2 to -3dB loss. Add another -2dB of loss for the connectors
and internal circuit board wiring. Can you afford the loss?

>Are there any rules of thumb when to put one in?

No. Additional filtering is only required if there is some type of
interference situation.

>For example,
>"Use when WiFi antenna is within x feet of VHF antenna on
>same tower radiating at y watts..."?

Well, ok... let's do it like that. Since the antennas are unknown,
it's easier to make a measurement. Install a spectrum analyzer on
YOUR antenna. Look around (wide sweep) for strong signals. If there
are any signals near the 2.4Ghz band, and that are strong enough to
overload your receiver. My guess(tm) is that most cheap access points
will overload at about -10dBm. If the signal is a spread spectrum
smear, you'll need to compensate for the bandwidth (observed bw
divided by spectrum analyzer dispersion) to get the actual despread
power.

>Or, maybe "Use when WiFi antenna is in Fresnel zone of
>gigantic telco solid dish less than x miles away..."?

You'll find that the bigger dishes are NOT major sources of
interference. They use big dishes because the various coordinating
organizations demand very narrow beamwidths with few side lobes. The
actual power used is very small and often less than 802.11. The idea
is to protect other users on the same frequency by keeping the RF away
from useless areas. You would NEVER see a commerical microwave
installation with an omnidirectional antenna. Unless you're exactly
in the path of the dish, I wouldn't bother.


--
Jeff Liebermann jeffl@comix.santa-cruz.ca.us
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

 

[Guest]

Archived from groups: alt.internet.wireless (More info?)

On Thu, 30 Sep 2004 10:04:30 -0700, Jeff Liebermann
<jeffl@comix.santa-cruz.ca.us> wrote:

>There are band pass filters that are only 26MHz wide, which are used
>to seperate the 3ea non-overlapping channels (1, 6, 11) and are used
>to eliminate interference from co-located wireless users on other
>channels.

Here's one channel filter and diplexer:
http://www.rflinx.com/Filters/2400-UHQ-TRIPLXR.htm
Only $1500.

--
Jeff Liebermann jeffl@comix.santa-cruz.ca.us
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

 

[Guest]

Archived from groups: alt.internet.wireless (More info?)

> would have a receiver tuned to the 2.4 (or 5.x) GHz range, and so
> would ignore or be immune to signals at other frequencies such as
> VHF/UHF, AM/FM, etc. So why the need for an external
> bandpass filter?

I don't really mean to be blunt, but have you ever studied receiver design?
A look through an Amateur Radio Handbook will prove illuminating.
While the frequencies are different (we do have an allocation at 2.4G where
we are allowed 1500W) the general principles still apply.

Out of band signals, even WAY out of band, can cause a significant loss of
sensitivity in your receiver. Also, there are frequencies that will mix
with your first IF, and give you what's called "image", where the receiver
picks up the desired frequency, and a few others, as if they were all on the
same frequency.

I see desense in my IC-R8500 receiver, over most of the band, (0.1 - 1999
MHz) due to a local FM broadcaster on 104.1 MHz. I use a simple stub filter
to eliminate this problem. The attenuation that I get through the filter on
the FM broadcast signal is enough to take the receiver's front end out of
saturation.

> Are there any rules of thumb when to put one in? For example,
> "Use when WiFi antenna is within x feet of VHF antenna on
> same tower radiating at y watts..."?

How will you know?
The most common "symptom" of desense, is no symptom at all, just
unexplainably short range.


> Or, maybe "Use when WiFi antenna is in Fresnel zone of
> gigantic telco solid dish less than x miles away..."?

--
KC6ETE Dave's Engineering Page, www.dvanhorn.org
Microcontroller Consultant, specializing in Atmel AVR

 

[Guest]

Archived from groups: alt.internet.wireless (More info?)

On Thu, 30 Sep 2004 12:15:16 -0500, "Dave VanHorn"
<dvanhorn@cedar.net> wrote:

>I don't really mean to be blunt, but have you ever studied receiver design?

I have. A few comments and hair splitting.

>A look through an Amateur Radio Handbook will prove illuminating.

Yeah. Like one must suffer before enlightenment. I suggest something
more relevant like:
Complete Wireless Design by Sayre

http://www.amazon.com/exec/obidos/tg/detail/-/0071370161/102-7783866-6131366

>While the frequencies are different (we do have an allocation at 2.4G where
>we are allowed 1500W) the general principles still apply.

Shhh. If the FCC ever noticed, we're likely to lose that band. Who's
gonna win in case of conflict? A few hundred hams that do EME and
AO-51, or a few million 802.11 users backed by some rather large
companies?

>Out of band signals, even WAY out of band, can cause a significant loss of
>sensitivity in your receiver.

Yep. That's called blocking. You have to have a fairly strong signal
to do that. The traditional 2.4GHz band pass filter in the front end
of the 802.11 receiver should get rid of most of the out of band
stuff. Only the stuff near the band edges will be a problem.

>Also, there are frequencies that will mix
>with your first IF, and give you what's called "image", where the receiver
>picks up the desired frequency, and a few others, as if they were all on the
>same frequency.

802.11 radios come in many flavours. The current incantations are
direct conversion, where all the demodulation happens at 2.4GHz. No
local oscillator and no IF amplifier.
http://www.maxim-ic.com/quick_view2.cfm/qv_pk/3556
Therefore, no image. Of course, older receivers will have an image
problem. As I recall, the Prism I chipset used a 280MHz first IF with
low side injection.

>I see desense in my IC-R8500 receiver, over most of the band, (0.1 - 1999
>MHz) due to a local FM broadcaster on 104.1 MHz. I use a simple stub filter
>to eliminate this problem. The attenuation that I get through the filter on
>the FM broadcast signal is enough to take the receiver's front end out of
>saturation.

Different animal. Your IC-8500 has a broad band RF amplifier, with no
tracking filter. It wasn't made to operate on one frequency but
rather on a wide range of frequencies. That requires a compromise
between overload and interference performance, and bandwidth. If
you're IC-8500 had a front end designed to tune only a small range of
frequencies, the necessary front end filter would have eliminated the
FM interference.

>> Are there any rules of thumb when to put one in? For example,
>> "Use when WiFi antenna is within x feet of VHF antenna on
>> same tower radiating at y watts..."?

>How will you know?
>The most common "symptom" of desense, is no symptom at all, just
>unexplainably short range.

Oh, you can find out easily enough. The same method is used to
determine if the junk on the spectrum analyzer screen is real or
generated by mixing in the spectrum analyzer. Install a switchable
attenuator between the antenna and the receiver. Generate a weak
signal, mix through a combiner, and measure it's level (RSSI is fine).
Add in 6dB of attenuation. Does the signal level (RSSI) change 6dB?
If yes, then there's no blocking in the front end. If the receiver
suddently changes MORE than 6dB, then there's blocking. Blocking and
intermod are threshold phenomenon. Drop the levels below the
threshold and they don't change linearly, they disappear suddenly.

>> Or, maybe "Use when WiFi antenna is in Fresnel zone of
>> gigantic telco solid dish less than x miles away..."?


--
Jeff Liebermann jeffl@comix.santa-cruz.ca.us
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

 

[Guest]

Archived from groups: alt.internet.wireless (More info?)

>>A look through an Amateur Radio Handbook will prove illuminating.
>
> Yeah. Like one must suffer before enlightenment. I suggest something
> more relevant like:
> Complete Wireless Design by Sayre
> http://www.amazon.com/exec/obidos/tg/detail/-/0071370161/102-7783866-6131366

I'd still reccomend the AR manual for a first cut, to get the basic
principles in hand.


> Shhh. If the FCC ever noticed, we're likely to lose that band. Who's
> gonna win in case of conflict? A few hundred hams that do EME and
> AO-51, or a few million 802.11 users backed by some rather large
> companies?

I think the popularity of Wi-Fi and such has been WAY beyond what the FCC
anticipated.
My microwave oven, cordless phone, Wi-Fi, and ATV all operate pretty much on
top of each other. OTOH, 902-928 is getting quieter every day

I'm sure they haven't forgotten, but then again, I don't know anyone
personally, who runs more than 20W up there.


> 802.11 radios come in many flavours. The current incantations are
> direct conversion, where all the demodulation happens at 2.4GHz. No
> local oscillator and no IF amplifier.

I'm aware, wasn't sure what the OP is using. It's good to be aware of these
issues.
I would expect that the direct conversion radios still have desense problems
possible, and maybe some new sorts of problems.

> Different animal. Your IC-8500 has a broad band RF amplifier, with no
> tracking filter. It wasn't made to operate on one frequency but
> rather on a wide range of frequencies. That requires a compromise
> between overload and interference performance, and bandwidth. If
> you're IC-8500 had a front end designed to tune only a small range of
> frequencies, the necessary front end filter would have eliminated the
> FM interference.

True, I guess I didn't make my point very well there.
The 8500 is a broadbanded, but high performance receiver, with lots of
shielding, and a lot of effort spent in it's design, and yet it's not immune
from these sorts of issues.

> If the receiver
> suddently changes MORE than 6dB, then there's blocking. Blocking and
> intermod are threshold phenomenon. Drop the levels below the
> threshold and they don't change linearly, they disappear suddenly.

But, you need to know these techniques, in order to apply them.
Hence the point to the books.

 

[Guest]

Archived from groups: alt.internet.wireless (More info?)

On Thu, 30 Sep 2004 13:36:39 -0500, "Dave VanHorn"
<dvanhorn@cedar.net> wrote:

>I'd still reccomend the AR manual for a first cut, to get the basic
>principles in hand.

Well, it does have a few chapters on basic principles. After so many
years of re-writing the same stuff for annual revisions, one would
suspect that it should be perfect by now. Yet, I'm still finding and
submitting errors. The Antenna Handbook was even worse, but at least
went through a major rehash about 8 years ago. The various handbooks
also cover topics that are of little interest to someone dealing with
802.11 wireless. HF long wire antennas, construction articles, and DX
are interesting, but of little value in understanding wireless
communications. Methinks the various ARRL microwave handbooks might
be of use, but I have no idea what's in them.

>I think the popularity of Wi-Fi and such has been WAY beyond what the FCC
>anticipated.

That's an understatement. It has grown so fast that I'm hearing
comments from FCC staff such as "out of control" and "unregulated".
It's not the first time this has happened as they also missed with CB
and FRS. However, GMRS got over-regulated, so it went nowhere. MURS
is an unknown at this time.

>My microwave oven, cordless phone, Wi-Fi, and ATV all operate pretty much on
>top of each other. OTOH, 902-928 is getting quieter every day

900MHz was suppose to be totally polluted so overyone went to 2.4GHz.
Now, 900MHz is almost empty. This was not left un-noticed by the
manufactories. Motorola now has Canopy running on 900MHz at
3Mbits/sec. 900MHz also goes through trees much easier than 2.4Ghz.

>I'm sure they haven't forgotten, but then again, I don't know anyone
>personally, who runs more than 20W up there.

Magnetron 2.4GHz
http://lea.hamradio.si/~s57uuu/mischam/magnetr/

SSB Electronics had a 100 watt 2.4Ghz power amp for sale for a while
until someone convinced them to withdraw the product. Here's the
1.2GHz version:
http://www.ssbusa.com/tlapow.html
They still sell a 10watt 2.4Ghz amp:
http://www.ssbusa.com/m13amp.html
There are plenty of military/industrial 100 watt amplifiers available:
http://www.ophirrf.com/pdf/5302019.pdf (50 watts)
http://www.ophirrf.com/pdf/5038B.pdf
You can find almost anything on eBay:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=5722523078 (60w)
Hmmm... I'm tempted. Be the first on your block to dominate the
2.4GHz band.

>I'm aware, wasn't sure what the OP is using. It's good to be aware of these
>issues.
>I would expect that the direct conversion radios still have desense problems
>possible, and maybe some new sorts of problems.

Oh, direct conversion (Zero IF) has all kinds of performance related
problems. In terms of overall receiver performance, they're quite
inferior when compared to single or multiple conversion receivers.
However, improvements in IC manufacturing technology (i.e. SiGe) has
largely compensated for some of the performance issues. Direct
conversion is also becoming popular in cell phones and GPS receivers.
The reasons are far from subtle. Direct conversion is less board real
estate and lots cheaper. With commercial SDR (software defined
radio), direct conversion is almost a requirement. Some clues:
http://www.commsdesign.com/showArticle.jhtml?articleID=16505963
http://www.commsdesign.com/showArticle.jhtml?articleID=16504800
http://www.commsdesign.com/showArticle.jhtml?articleID=16503600

>But, you need to know these techniques, in order to apply them.
>Hence the point to the books.

You won't find those in the ARRL handbooks. Hints and Kinks in QST is
a possible source. You will find it in the instruction manual for
older HP spectrum analyzers, and in books/articles on test equipment
use.

I dunno what to recommend to someone with a computah background and
trying to learn microwave RF. It's been so long since I've read the
basic books, I don't know what's in them. The Wireless book I
previously suggested is fairly comprehensive, but would be useless to
someone without an electronics or circuit design background.


--
Jeff Liebermann jeffl@comix.santa-cruz.ca.us
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

 

[Guest]

Archived from groups: alt.internet.wireless (More info?)

> The Antenna Handbook was even worse, but at least
> went through a major rehash about 8 years ago. The various handbooks
> also cover topics that are of little interest to someone dealing with
> 802.11 wireless. HF long wire antennas, construction articles, and DX
> are interesting, but of little value in understanding wireless
> communications. Methinks the various ARRL microwave handbooks might
> be of use, but I have no idea what's in them.

Pet peeve there, is not exactly defining where the measurements are taken.
Dosen't matter much for 20m, but matters rather a lot at 2.4 GHz

> That's an understatement. It has grown so fast that I'm hearing
> comments from FCC staff such as "out of control" and "unregulated".
> It's not the first time this has happened as they also missed with CB
> and FRS. However, GMRS got over-regulated, so it went nowhere. MURS
> is an unknown at this time.

Around here, the CB crowd has discovered FRS, and is using modded 440 rigs.
😛
At least their garbage is pretty much localized.

> Magnetron 2.4GHz
> http://lea.hamradio.si/~s57uuu/mischam/magnetr/

I still have that article for an FSTV transmitter using a microwave oven.

> I dunno what to recommend to someone with a computah background and
> trying to learn microwave RF. It's been so long since I've read the
> basic books, I don't know what's in them. The Wireless book I
> previously suggested is fairly comprehensive, but would be useless to
> someone without an electronics or circuit design background.

In my last 2.4G venture, a custom BT module, I had to defend against
someone's "miracle antenna".. This was a "high gain" antenna the size of a
1206 SMD part.. They wanted something like $7 each, and I had a hell of a
time convincing mgt that this was NOT workable for us.
In the end, my $0.00 etched track on the PCB won out.. The mental process
was something like this: "Well, they sell for $7, so they must be pretty
good! Much better than a free wire."

Reminds me of the comedy zone at the local truck stop, with their 2' long CB
antennas, and "10,000W" antennas with built-in RG-58.



--
KC6ETE Dave's Engineering Page, www.dvanhorn.org
Microcontroller Consultant, specializing in Atmel AVR

 

[Guest]

Archived from groups: alt.internet.wireless (More info?)

On Fri, 1 Oct 2004 15:10:41 -0500, "Dave VanHorn" <dvanhorn@cedar.net>
wrote:

>In my last 2.4G venture, a custom BT module, I had to defend against
>someone's "miracle antenna".. This was a "high gain" antenna the size of a
>1206 SMD part.. They wanted something like $7 each, and I had a hell of a
>time convincing mgt that this was NOT workable for us.

Such companies and products are still around:
http://www.palowireless.com/bluetooth/antenna.asp
However, I don't think anyone makes one that small (1.6mm x 3.2mm) and
calls it an antenna. The smallest I could find was 3.9mm x 6.2mm) at:
http://www.gigaant.com/?id=535
If you're building Bluetooth headsets, that's about the only antenna
that will work. If you try to do it on a G10 circuit board, your
antenna will be about 10mm for a 1/4 wave which is difficult to fit if
you don't much available "ground free" board space. G10 is also quite
lossy. The only thing good about PCB trace antennas is that it's
cheap.

>In the end, my $0.00 etched track on the PCB won out.. The mental process
>was something like this: "Well, they sell for $7, so they must be pretty
>good! Much better than a free wire."

Oh, it's a common effect. Bigger numbers are better. I kinda enjoy
reading the cordless phone commercials, where the vendors try to
convince the customer that since 5.6GHz is much "bigger" than 2.4GHz,
it must be better. Most customer don't have a clue what it all means,
much less the difference in free space loss.

>Reminds me of the comedy zone at the local truck stop, with their 2' long CB
>antennas, and "10,000W" antennas with built-in RG-58.

Well, yes. There's hype in almost every product area and market. A
technology isn't considered mature until it's been abused and
polluted. 108Mbits/sec performance, at perhaps 3 feet away. Metro
WLAN's but only if everyone abandons 802.11b/g. Nothing new here.


--
Jeff Liebermann jeffl@comix.santa-cruz.ca.us
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558