John,
Helpful post!
In summary of this thread, I guess the message I'm getting is, there is no
substitution for the right tool for the job.
And its clear that an analyzer is the correct tool, to truely learn the
characteristics of your RF colocated neighbor, without risk of false
assumptions.
Tom DeReggi
RapidDSL & Wireless, Inc
IntAirNet- Fixed Wireless Broadband
----- Original Message -----
From: "John Scrivner" <[EMAIL PROTECTED]>
To: "WISPA General List" <wireless@wispa.org>
Sent: Saturday, October 28, 2006 12:12 PM
Subject: Re: [WISPA] Sprint / Nextel to use 900mz for iDen
Good info Jack. In a past life I was a headend tech in the cable
television industry. and I also performed signal egress and ingress
troubleshooting using a Hewlett Packard 8591B analyzer. I spent a few
thousand hours on this tool and learned much about spectrum analysis at
that time. Here is some info for some of those out there who may be new to
spectrum analysis:
There is something that can make these paging transmitters appear to be
bleeding over into the ISM bands when in fact they may not be. The setting
on the analyzer is called "resolution bandwidth". This setting takes all
power within a given bandspace and averages it together as it sweeps
across the screen. The wider this setting is the fewer bumps you see on
the screen. The trace will smooth out as you increase this setting because
it is averaging power within a wider space of spectrum. This has the added
effect of making a "loud" carrier appear to cover a wider space than it
actually does and can cause you to believe that a paging or other carrier
is bleeding over into the ISM band. On the contrary, narrowing the
resolution bandwidth will show more accurate representation of actual
power in a given bandspace but is slower to scan on most analyzers and
produces a very sporadic display. If you are looking for narrowband or
adjacent channel interference into your band then a narrow resolution
bandwidth will be required. If you are wanting to take a RSSI reading of
your own carrier then a wider resolution bandwidth will be required.
Resolution bandwidth is something you should learn to use and understand
if you want to get more from your work. It is an important part of
spectrum analysis. If you want to see how good an analyzer is then look at
how low the resolution bandwidth setting will allow. For our work a
minimum resolution bandwidth of about 100kHz is probably all you will ever
need. Also run it at its lowest resolution bandwidth and see how long it
takes to scan across the screen. If you are comparing multiple analyzers
make sure you always use the same span setting (difference between upper
and lower frequency on display). A narrower span will display a narrow
resolution bandwidth much faster. Better analyzers will have a wide range
of resolution bandwidth settings and will show a sharp, clean display in
any setting.
Learning to use a spectrum analyzer can seem daunting at first glance. Do
not let this intimidate you. You can learn to use this and get meaningful
information from it if you give it a try. You will not break the analyzer
by experimenting with it. If the unit you are using has knobs and you had
it set by someone previously then just take notes of where they are set
and then experiment with the unit. The most important things to master are
start frequency, stop frequency, span, center frequency, reference level,
attenuation, resolution bandwidth. Anything else you learn is good to know
but not as much as what I just outlined here.
If anyone here is working with an analyzer and does not know what any of
those things mean then feel free to ask here onlist (or offlist if you
would prefer to not tell others you do not know) :-)
Scriv
Jack Unger wrote:
Tom,
Yes, their gear (the paging stuff) not only costs more but their
transmitters spurious emissions have to remain low or the paging company
risks being fined by the FCC. Sure, a transmitter can malfunction once in
a while and cause interference to the ISM band but this is not a common
occurance. Our gear has receivers where the manufacturing cost is quite
low. There may be $50 worth of parts in the receiver section of an AP.
The vendors typically do not spend a lot of money on components that
would raise the cost of their equipment and make it non-competitive such
as adding expensive filters to reduce the overloading problems that only
a minority of WISPs may ever experience. Similarly, the new cars that
people buy don't come with the most expensive tires as standard equipment
because most people would never notice a difference or be willing to pay
more for the premium tires.
I started deploying 900 MHz bridges in 1993 and 900 MHz APs (yes, for
WISP service) in 1995. I used Lucent "Wavelan" cards in those systems.
Whenever I was located within about 1/3 of a mile from a cell site (with
colocated 929 MHz and 930 MHz paging) I had to add an external bandpass
filter between the antenna and the antenna connector on the Wavelan card.
Until I did this, I could not get full throughput (which was about 1.3
Mbps in those days) through the card. The bandpass filter would clear up
the problem every time. Those filters weren't even that strong - only
about 6 dB of attenuation at 900 MHz and at 930 MHz (even less - maybe 5
dB at 929 MHz) but it was enough to protect the Wavelan card's receiver
from being overloaded. These bandpass filters were made by a 3rd-party
source and custom tuned by me in a calibration lab. My filter cost was
$125 each and they were not weatherproof so I mounted them indoors. The
inband attenuation was aboat 1 or 1.5 dB which was insignificant in light
of the fact that the filters worked to eliminate the overloading and
allow the AP to receive client signals up to 10 or 12 miles away.
Regarding Trango - I have not verified the accuracy of their spectrum
analysis tool but what you're seeing can be explained by one observation
and one guestimation. The -20 dBm to -30 dBm signal indications above 929
MHz are likely fairly accurate. Nearby paging transmitters could easily
be that loud. The fact that you're seeing signals down to 924 MHz or so
could be explained by the Trango receiver "front-end" (the first stage
connected to the antenna) being overloaded by one or more nearby paging
transmitters. When a receiver is overloaded, it generates "spurious"
signals that are not really being transmitted on the frequency where they
show up. The "spurs" are being generated inside the receiver itself as a
consequence of the overloading. It's fairly easy to test to see if this
is the case. Just insert a bandpass filter between the antenna and the
antenna connector (assuming a connectorized AP). If the AP receiving
distance and/or the throughput increases, you have just proved that
overloading was a problem. You can also re-run the spectrum analysis tool
and see if it no longer reports signals down to 924 MHz. It should now
report that the non-WISP signals start around 929 MHz.
I hope this explanation helps.
jack
Tom DeReggi wrote:
Jack,
That all sounds good, and it brings up a good point, that we are just as
probable to be the culprit, not just the other guy.
Besides, their gear costs more, right :-)
However, what specific gear do you have experience with, on this issue,
to support your comment?
I'm not sure that I am knowledgable enough on the topic, to know for
sure which side is the flaw, how would we tell?
I use Trango 900. Trango's have a built-in specrum site survey tool,
that also scans a bit lower and higher than the ISM edge. My comment
was based on the fact that, when I do the site survey, I see signals in
the neg 20-30 range, spanning from significantly above 930 down to mid
portion of ISM channel 4 (924 or so).
Have you verified the accuracy of the Trango tool, and how it reacts to
this situation?
Tom DeReggi
RapidDSL & Wireless, Inc
IntAirNet- Fixed Wireless Broadband
----- Original Message ----- From: "Jack Unger" <[EMAIL PROTECTED]>
To: "WISPA General List" <wireless@wispa.org>
Sent: Friday, October 27, 2006 1:07 PM
Subject: Re: [WISPA] Sprint / Nextel to use 900mz for iDen
"Bleed over" implies that the paging system is transmitting a signal
that is too wide. This is typically NOT the case. Our rather
inexpensive WISP AP receivers do not have adequate selectivity to
reject strong nearby signals. In other words, it's our equipment
problem not their equipment problem.
Also, WISP subscriber sites, unless located right under a
paging/cellular tower aren't close enough to be overloaded by
paging/cellular so they would not need the bandpass filter. Only our
APs which are located near paging/cellular towers should need the
bandpass filters.
jack
Larry Yunker wrote:
While filters can help, the problem that I see is that filters are: 1)
expensive and 2) bulky. Last time I checked, a cavity filter for the
902-928 range was roughly $300-$400. I don't see it being practical
to install one of these at every customer site!
Cavity filters are fine for your broadcast sites, but that is of
little help when the 900Mhz paging systems bleed over so much that
they "deafen" the subscriber radios.
- Larry
----- Original Message ----- From: "Mike Cowan"
<[EMAIL PROTECTED]>
To: "WISPA General List" <wireless@wispa.org>
Sent: Thursday, October 26, 2006 7:32 PM
Subject: Re: [WISPA] Sprint / Nextel to use 900mz for iDen
Filters fix this problem quite handily. We recommend one on every
system needed or not. I don't see an issue here.
Mike
At 07:07 PM 10/26/2006, you wrote:
ISM 902-928.
Exact band and Power limit is relevant. Currently, the top 25% of
ISM 900 bandwidth (channel 4) is unusable, in MANY areas, due to
blead over from 930 Licensed high power gear (500W). If the same
thing were to occur at the lower portion of 900 ISM bandwdith, it
could kill Channel 1 also, horribly effecting WISPs using
unlicenced. They also may be requesting to use higher power on the
actual ISM bands, argueing Public Safety is more important than
unlicensed use. Iftheir request is granted, specifics should be
lsited on how they are going to prevent interference with existing
unlicensed band users. Remember that the goal may not only be to
use the spectrum. They have benefit in killing off all the 900Mhz
WISPs, that could compete with Sprint/Nextel Next generation WiMax
type Licensed 700M-900M solutions.
Tom DeReggi
RapidDSL & Wireless, Inc
IntAirNet- Fixed Wireless Broadband
Mike Cowan
Wireless Connections
A Division of ACC
166 Milan Ave
Norwalk, OH 44857
419-660-6100
[EMAIL PROTECTED]
www.wirelessconnections.net
--
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/
--
Jack Unger ([EMAIL PROTECTED]) - President, Ask-Wi.Com, Inc.
Serving the License-Free Wireless Industry Since 1993
Author of the WISP Handbook - "Deploying License-Free Wireless WANs"
True Vendor-Neutral WISP Consulting-Training-Troubleshooting
Newsletters Downloadable from http://ask-wi.com/newsletters.html
Phone (VoIP Over Broadband Wireless) 818-227-4220 www.ask-wi.com
--
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/
--
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/
--
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/