New Member
homerwsmith
Posts: 19
Registered: ‎05-12-2011
Posts: 19
Registered: 05-12-2011

Comments and corrections solicited II

RECEIVE AMPS

In the 1990's it was fashionable to use Hyperlink amps, which
gave an automatically adjusted 1 watt out, and a 17db receive gain.

They have their use but are not necessarily what you might think
they are for.

Consider the following theoretical scenario, one AP and one client
out in space with a background noise level of -90db

Construct a plot with two vertical lines one on the left hand side
of the paper and one on the right. Then draw in a horizontal line
connecting both of them at the botom.

The two y axes are in db and the x axis is distance between
them, say 10 miles from left to right.

The left Y axis represents an AP and the right one represents a client,
and both have an identical parabolic that sends out 100mw from each
towards the other.

Each radio has a receiver sensitive down to -100db, but the noise floor
is -90db and the radios are not able to decode a signal less than the
noise floor.

Thus the signal will start out at 100mw on the left and fall down
as it propagates to the right until it hits the noise floor at -90db
at a distance of say 9 miles to the right.

The same thing happens with the right hand client, which sends out
starting at 100mw which falls off as the signal travels to the left until
it too hits the drink of -90db at 9 miles to the left.

Clearly since these two radios are 10 miles apart, they can't
talk to each other.

1.) If we put a transmit amp on the left side so it puts out
1 watt instead of 100 mw, the signal will hit the drink at 20 miles,
and thus be clear and strong at 10 miles so the client can read it.

However the client is still putting out 100mw, so it can't talk back,
so the link does not work.

If we put a transmit amp on the client also, then the client's signal
will hit the drink at 20 miles to the left, and thus be loud and clear to
the AP and so the link will work again.

Moral: having a 1 watt AP, with 100mw laptop clients will only work
at 100mw ranges. The 1 watt is essentially wasted. The lattops will
show strong signal from the AP, but not be able to talk anyhow, causing
customer complaints.

2.) Now let's say we remove the amp from the client side, but put
a receive amp on the AP side along with its transmit amp

Thus the client is transmitting at 100mw as before, the AP is
transmitting at 1 watt, but the AP is amplifying the recevied signal
by 20 db say.

Does this help? No.

The client signal at the AP is say -100db lost in a noise floor of -90db.

The receive amp adds 20 db to both, so the AP sees a client signal
of -80db lost in a noise floor of -70db, and the operative word is still LOST.

Moral: Receive amps don't do what you think they are going to do.
particularly if your existing sensitivity is below the noise floor, and
the signal is lost in the noise floor anyhow. Every wifi radio today has
a sensivity below the noise floor, even if not by much, and in any case a receive amp
will not help you at the receiving end EXCEPT if 1.) the receive amp is more
sensitive than your radio already is, and 2.) the signal at the amp is above
the noise floor.


3.) Now let's say we get rid of both transmit and receive amps and
have both AP and client talking at 100mw across 10 miles as before,
with the signal hittting the drink at 1 mile away from each, so they can't
talk.

What happens if we put an amp in the middle of the link
at 5 miles? Does this help?

Yes.

At 5 miles the signal from each side is down from 100mw to say -80db, still
above the noise floor of -90db by 10 db.

The amp receives the signal and boosts it back up 20 db, so the
signal is now -60db with a noise floor of -70db. Remember the amp can't tell signal from noise, so EVERYTHING is amped.

Notice the amplified noise floor of -70db will now fall off as it moves to the end point, at the
same rate as the signal.

Over the next 5 miles from the amp to either end, the signal/noise of
-60/-70 will fall off to say -80/-90 and the AP or client Will be able to pick up
the signal and the link will work.

Moral: receive amps are only useful when placed BETWEEN the
two ends of a link, not AT either end.

4.) Real work application.

Say you have a tower that is 100 feet tall, and your radio
is at the bottom, separated by 200 feet of LMR 400 cable which
runs to the top where the antenna is.

LMR is lossy, say -10 db, a signal that comes in to the antenna at
-80/-90 SNR
will be degraded to -90/-100 SNR by the time it gets to the bottom.

If the -90 signal is less than or on the edge of the sensitivity of the radio it may not be decodable,
even though the signal hit the antenna at a decodable level.
Notice the LMR antennuates both the signal AND the received noise floor!

If you place a receive amp at the top, the incoming -80/-90 will
be boosted to -60/-70, the LMR will degrade it to -70/-80 at the radio
and the radio will do just fine.

Notice the receive amp is not where the radio is, it is someplace
earlier in the link where is can pickup a usable signal, make it MORE usable,
so that the degradation between the amp and its final destinatin will not render it
totally useless.

Moral: receive amps never go at the ends of a link, but somewhere
earlier.

And put your radios up where the antenna is with ethernet/POE
and throw all your LMR out :manhappy: LMR gets wet so easy. And if lightning is going to hit your tower, your radio is going to be toast whether it is at the
top or the bottom. The only downside, is you will need to climb the tower
to replace your toast, but it just isn't worth all the amps and nonsense necessary to get a dim signal from the antenna to the radio at the bottom.
Let ethernet packets descend your tower, not analogue signals.


Homer Wilson Smith
Lightlink CEO
SuperUser
WHT
Posts: 29,546
Registered: ‎03-17-2008
Posts: 29546
Kudos: 5533
Solutions: 116
Registered: 03-17-2008

Re: Comments and corrections solicited II

You haven't been in the wireless game very long, have you?
Receiver amplifiers also amplify the ambient noise, so your net gain is zero.
WHT = Short Form Acronym for "You couldn't handle me even if I came with instructions!"
Well engineered projects are indistinguishable from crazy ideas.

Speed, distance, reliability, cost...Pick three.
...World's First Ubiquiti AirMax WISP....
New Member
homerwsmith
Posts: 19
Registered: ‎05-12-2011
Posts: 19
Registered: 05-12-2011

Re: Comments and corrections solicited II

Isn't that what I said?

As for the game, I am Cornell EE 1973, a ham KC2ITF general, and I been deploying
wireless since 1999.

Homer
SuperUser
WHT
Posts: 29,546
Registered: ‎03-17-2008
Posts: 29546
Kudos: 5533
Solutions: 116
Registered: 03-17-2008

Re: Comments and corrections solicited II

No...You're still advocating a receiver amplifier.
WHT = Short Form Acronym for "You couldn't handle me even if I came with instructions!"
Well engineered projects are indistinguishable from crazy ideas.

Speed, distance, reliability, cost...Pick three.
...World's First Ubiquiti AirMax WISP....
SuperUser
Dave-D
Posts: 25,359
Registered: ‎06-23-2009
Posts: 25359
Kudos: 3721
Solutions: 419
Registered: 06-23-2009

Re: Comments and corrections solicited II

Why is this even a thread? Dave
No disclaimer. Nothing to sell. I need to fix that.
Established Member
Idea
Posts: 1,145
Registered: ‎10-28-2010
Posts: 1145
Kudos: 418
Solutions: 10
Registered: 10-28-2010

Re: Comments and corrections solicited II

I thought it was a spam comment at first. Especially when it got into the "imagine a plot" part. Just draw it and post it.
Regular Member
brom42
Posts: 379
Registered: ‎10-29-2011
Posts: 379
Kudos: 130
Solutions: 1
Registered: 10-29-2011

Re: Comments and corrections solicited II

TL;DR.

Cliff notes?
SuperUser
Josh_SPITwSPOTS
Posts: 17,500
Registered: ‎11-20-2011
Posts: 17500
Kudos: 5183
Solutions: 124
Registered: 11-20-2011

Re: Comments and corrections solicited II

TL;DR.

Cliff notes?


If it were possible to remove ears like legos and put bigger ones on, you'd hear things just as well as you did before.... the background noise will rise just as much as the stuff you'd like to pay attention to.

It's a placebo.
Josh Reynolds :: Chief Information Officer :: www.spitwspots.com
Ubiquiti Carrier Wireless Admin, Trainer
SuperUser
WHT
Posts: 29,546
Registered: ‎03-17-2008
Posts: 29546
Kudos: 5533
Solutions: 116
Registered: 03-17-2008

Re: Comments and corrections solicited II

No matter what color of Calvin Klein underwear you dress up a goat...it still looks like a goat.
www.youtube.com/watch?v=qJ7lh52J8sg
WHT = Short Form Acronym for "You couldn't handle me even if I came with instructions!"
Well engineered projects are indistinguishable from crazy ideas.

Speed, distance, reliability, cost...Pick three.
...World's First Ubiquiti AirMax WISP....
Emerging Member
theory240
Posts: 70
Registered: ‎10-02-2008
Posts: 70
Kudos: 17
Registered: 10-02-2008

Re: Comments and corrections solicited II

What was the point of this?

Amps are useful is some conditions, but with low cost high power CPE and AP's, not so much as before.

Design your links.
SuperUser
Bean_Planter
Posts: 5,427
Registered: ‎08-09-2011
Posts: 5427
Kudos: 2265
Solutions: 164
Contributions: 10
Registered: 08-09-2011

Re: Comments and corrections solicited II


And put your radios up where the antenna is with ethernet/POE
and throw all your LMR out :manhappy: ... Let ethernet packets descend your tower, not analogue signals.

Hi Homer,
Welcome to the UBNT forum.
Your statements are completely sound. You likely received the reaction that you did because what you described (radios on top of the tower, ethernet carrying data from there) is why people abandon other LMR-type solutions and adopt UBNT solutions.
In effect, you "restated the obvious" (at least, to those who have been here for a while). I don't think that is bad.
Your background obviously includes LMR-savvy. Your Ham Radio background is helpful in that regard. In fact, if you were active on the uWave bands, you would likewise be placing your transverters at the top of your tower, feeding only the much lower Intermediate Frequency down to "the shack" for the same reason that you mention.
It is nice to see that you have a WISP business going, too!
On the off-chance that you already know this anyway, here's some information that may be helpful to you as you continue to explore UBNT solutions...
You mention placing "intermediate" stations along a pathway that is otherwise not good enough to sustain direct communication. That sounds like it may have been grounded in LMR's "repeater" or "voting receiver" concept to some degree.
UBNT equipment offers several strategies that provide that level of funcationality, too.
Repeating: You can establish another AP between two points and invoke AP-AP-Station WDS repeaters. The over-the-air throughput drops by half each time one is invoked...but sometimes that is the only way.
Relaying: If you have an intermediate location for a repeater, you certainly could install a "relayer" instead (and maintain throughput speed). An intermediate relay combines a Station AND an AP in the same location. One "side" is on one channel and the other "side" is on another channel. Example:
AP(Channel-1)<~~><~~>Station
All you need at the intermediate is two UBNT radios and the electricity to power them as data will pass from one to the other at Ethernet speed (there is no RX-Buffer-TX delay that occurs with the WDS Repeater described above).
Again, welcome to the forum!
Cheers,
Gene
NEW FORUM MEMBERS:
Seek knowledge. Press the [Knowledge Base] link at the top.
Established Member
Idea
Posts: 1,145
Registered: ‎10-28-2010
Posts: 1145
Kudos: 418
Solutions: 10
Registered: 10-28-2010

Re: Comments and corrections solicited II

Wow Gene, that was very cordial. Nicely written.
New Member
homerwsmith
Posts: 19
Registered: ‎05-12-2011
Posts: 19
Registered: 05-12-2011

Re: Comments and corrections solicited II

No...You're still advocating a receiver amplifier.


Yes, under two conditions.

1.) In the middle of a link between two stations to far too
talk to each other, a repeater receiver/amp.

2.) At the top of towers with long lossy cable between antenna
at top and radio receiver at bottom.

Topologically they are almost the same thing.

If that is wrong, a little intelligent direction would be of help.

Homer
SuperUser
WHT
Posts: 29,546
Registered: ‎03-17-2008
Posts: 29546
Kudos: 5533
Solutions: 116
Registered: 03-17-2008

Re: Comments and corrections solicited II

Already explained your fallacy.
WHT = Short Form Acronym for "You couldn't handle me even if I came with instructions!"
Well engineered projects are indistinguishable from crazy ideas.

Speed, distance, reliability, cost...Pick three.
...World's First Ubiquiti AirMax WISP....
New Member
homerwsmith
Posts: 19
Registered: ‎05-12-2011
Posts: 19
Registered: 05-12-2011

Re: Comments and corrections solicited II

No...You're still advocating a receiver amplifier.



Yes, under two conditions.

1.) In the middle of long distance links where the two ends can not
talk to each other.

2.) At the top of high towers using lossy cables between antenna
at top and radio receiver at bottom.

Both are topologically equivalent.

Homer
SuperUser
WHT
Posts: 29,546
Registered: ‎03-17-2008
Posts: 29546
Kudos: 5533
Solutions: 116
Registered: 03-17-2008

Re: Comments and corrections solicited II

Topology has nothing to do with ambient noise.
Please insert another quarter to play again.
WHT = Short Form Acronym for "You couldn't handle me even if I came with instructions!"
Well engineered projects are indistinguishable from crazy ideas.

Speed, distance, reliability, cost...Pick three.
...World's First Ubiquiti AirMax WISP....
Regular Member
marwood0
Posts: 439
Registered: ‎12-03-2010
Posts: 439
Kudos: 43
Solutions: 1
Registered: 12-03-2010

Re: Comments and corrections solicited II

I think your post is generally correct though probably clearer with illustrations. And maybe some editing.
I didn't see amplifier noise figure and intermod specs referenced; those can be important.
About 2 years ago I asked Hyperlink for their amplifier specs so that I could actually design a system since they are not published. I got no response.
ATTN UBNT: in the USA, telecom systems commonly run at -50VDC (that means "positive" ground). Why is my toughswitch using +24V (negative ground)?
Senior Member
walrus0
Posts: 3,082
Registered: ‎09-28-2010
Posts: 3082
Kudos: 485
Registered: 09-28-2010

Re: Comments and corrections solicited II

TL;DR.

Cliff notes?


The original poster has an idea which involves generating electricity through the use of over-sized hamster wheels and several hundred dachshunds. It is estimated that each dachshund will be good for about 10W, but the food supply is a constraint.
If you have a particularly unique or challenging project, professional consulting services are available in central Asia, south Asia (AF, PK, IN, BD) and the Vancouver, BC area. Please contact me for more details.
SuperUser
Josh_SPITwSPOTS
Posts: 17,500
Registered: ‎11-20-2011
Posts: 17500
Kudos: 5183
Solutions: 124
Registered: 11-20-2011

Re: Comments and corrections solicited II

The original poster has an idea which involves generating electricity through the use of over-sized hamster wheels and several hundred dachshunds. It is estimated that each dachshund will be good for about 10W, but the food supply is a constraint.


^^^ This
Josh Reynolds :: Chief Information Officer :: www.spitwspots.com
Ubiquiti Carrier Wireless Admin, Trainer
SuperUser
WHT
Posts: 29,546
Registered: ‎03-17-2008
Posts: 29546
Kudos: 5533
Solutions: 116
Registered: 03-17-2008

Re: Comments and corrections solicited II

The original poster has an idea which involves generating electricity through the use of over-sized hamster wheels
How did I miss the obvious? :manhappy:
WHT = Short Form Acronym for "You couldn't handle me even if I came with instructions!"
Well engineered projects are indistinguishable from crazy ideas.

Speed, distance, reliability, cost...Pick three.
...World's First Ubiquiti AirMax WISP....