Batboard

I am trying to determine what sort of filtration we will need for a proposed repeater project. Ordinarily we are used to sites where the transmit frequencies are in the 150-154 range, recieve frequencies from about 157-160, and so on. But in this case we will have an existing telemetry station that transmits/recieves on one frequency and our transmitter would be 360 KC higher than that. So obviously it will be well outside of any adjacent channel interaction with the reciever but it will require some tight filtering to get it down to an acceptable level. What I am trying to do is figure what exactly an acceptable level at the reciever would be.

If I transmit say 100W of power out the antenna, that is +50 dBm. Say 30' vertical separation between antennas is 55 db, so I have -5 dBm coming back down the reciever line. This is where I don't know how to proceed.

What spec should I use from the reciever? I am thinking the spurious/image rejection one is the one to use. Usually 90 dB. But 90 dB of what? Does that mean 90 dB of the ultimate sensitivity of the reciever? Say the reciever sensitivity is -116 dBm, does that mean that I subtract another 90 from the -5 mentioned earlier to give me a -95 meaning I have to take out another 21 dB with the filters? If not would somebody steer me in the right direction here.

Birken

Birken,

You need to speak to chip at angle, http://www.anglelinear.com/.

The man is a super brain when it comes to RX filtration. He makes some awesome products.

There are two issues to consider. Transmitter sideband noise and recevier de-sensitization. Together these issues are known as TNRD.

Here is a link to many technical articles at various levels:

http://www.repeater-builder.com/rbtip/a ... index.html

My two bits worth:
This will be a rather expensive project.

The problem that you will have is "Gross RF" getting into the receiver front end of the telemetry station.

Spurious responses:
Any receiver has "spurious responses".
The image frequency is predictable, the other responses often are not.
There are also "Illegal frequencies" which the manufacturer may publish, or consider "propritary". These are spurious responses that are known. Often the receiver IF must be changed, or another receiver used.

Transmitters also have spurious emmissions, usuially specified as 60 or mord DB from carrier. This can be a large signal on a 100 Watt transmitter. +50-60=-10DBM. A large signal if on a site any where near your receiver.

Then there is "side band noise" All amplifiers have noise.
This is the purpose of the filter in the transmit transmission line.


The manufacturer of the telemitry station may have a chart known as "Duplex curves" which specifies the required isolation in DB versus frequency seperation.

Another important issue is what sort of signal levels are used on the telemetry station.
Perhaps it can tolerate some degraditon from your transmitter.
Then the telemetry might tolerate interruptions, it will re try until it can communicate with its slave stations when your repeater is not keyed.
In that case dont worry about interference.

I would guess about 110DB or so for for 360KC.
With this sort of seperation you will have to be very careful about shielding of the equipment.

The solution is notch filters, perhaps a crystal filter in the receive line.
as well as a notch filter in the transmit line.
Pass filters are not appropriate for such close spacing.

One possible solution is to obtain a duplexer suitable for such close spacing, which will be about the size of two file cabinets, use one port for the telemetry TX\RX, the other port for your transmitter.
The "T" cable on the output of the duplexer could be removed and seperate antennas for both systems connected.

Alternativly two filters which would be the equivalent of the duplexer could be ordered.

The choice may be determined by the space available on the site.

OK, it all makes sense so far. Indeed the telemetry station does seem to have a low duty cycle so desensing it when transmitting might be an acceptable state of affairs. That will be up to the site owners, they also own all the existing equipment. Additionally, as stated the recieved signal strength on the telemetry station might be strong enough that it could tolerate desense anyway, and I suspect that it is.

So I gather that notching my transmitter out of the telemetry reciever line is pretty much assurred. And that it might also be necessary to put another notch on my transmitter output to take care of wideband transmitter noise that would make it to the telemetry station.

What I am really getting at, though, is what general site principles apply to this situation. It would be nice to be able to say that such and such signal needs to be attenuated to a certain level and be done with it that way. And learn how to use the recivever specs to arrive at such a number. Angle linear's web site says it should be -50 dBm or less. Probably a "good enough" number to work with I guess.

With that number, if the signal is +50 dBm off the transmit antenna, with 55 dB of isolation it comes down the line at -5 dBm, and hits the recieve pass cavity dropping another 10 dB or so (depending on how that cavity is set up) so that leaves me with 35 dB to go...which off the cuff looks like it could be taken care of by a single 8" band reject cavity...lots of variables of course....

The only thing left is transmitter wideband noise supression. I guess it might take another band reject cavity on the transmitter line to reject the recivever freq. But the whole thing seems doable I guess.

Birken

My post got clipped off, I forgot to add.

VERY IMPORTANT.

You need to take a service monitor up to the site, roof, etc. wherever the TX units are, or where your equipment is going to be located.

Chip will tell you this if you speak to him, I am not an RF tech head, I am far from even amateur, so speak to chip to clarify this stuff.

You need to determine your sites noise floor, and identify any offending large signals present.

Grab a monitor, a vhf yagi, a rubber ducky, attenuators or a selectable attenuator, about 10 feet of military grade coax (jumper between monitor and antenna).

Go up to your site with the monitor set on your RX freq. with 0 attenuation and with the rubber ducky on the monitor and see what your noise floor is. In my case on the roof where my box is at the noise for is close to -70db, it is a crowded, and nasty RF environment.

Start searching up and down from your RX freq in small increments and see where you get some big signals. You also want to check the image freqs in the other bands.

Record all your data with 0 attenuation. Pad the input with 10db and start searching around. See if the signal is coming from your site. Try to make a list of all the freqs that are going to cause your input a problem. Add more attenuation to see what you are dealing with, and exactly how nasty it is.

This interference may also not be from your site if your are going to be on a roof top. Grab your yagi with 0 attenuation and with yagi in hand sweep slowly around in a 360 scanning for interference. See what you have and record the signals that will be above the noise floor. Keep the yagi with the elements when generally scanning with the elements in a horizontal position.

Have all of your findings ready when you order your filters, this will help determine how much filtering, and let the manufacturer custom tune the filters.

Here is a excerpt from chips website:
Interference Data Collection:

The best way to assure interference free operation of your repeater is to prove what interference you need to eliminate. This interference information or data can be obtained with a spectrum analyzer. Measure all signals from at least 50 MHz to 1000 MHz and record all signals greater than -50 dbm. The objective of the filtering is to reduce all out of band signals to below -50 dbm. With the correct amount of filtering you can do this and have virtually interference free repeater system. The minimum filtering we recommend is two (2) band pass resonators on the receive port of any duplexer regardless of who makes it or what fancy name it has. They all need band pass filtering to get the out of band unwanted signals down to a tolerable level. Some locations may require three (3) resonators on the receive port of the duplexer to survive; it all depends on your location and circumstances. Remember: no guessing.......hoping you won't have any interference won't work either...prove everything

Good luck, Hopefully all else can add in.

My two bits worth:

I consider it to be extremely bad practice to have a transmitter and receiver only 390 kc apart on the same site, assuming that very little degradiation is allowed.

Normal filters are far too broad to have any effect on signals so close together.

The chassis filtering is also very critical.
I have had the frustrating experience of having to optimize a couple of these "Kludges".

You will not find isolation recommendations in the manufacturer's manuals for such close spacing, since it is such bad practice in repeaters.

Assuming that degradiation of the receiver can be tolerated
then there is no problem.

There are certainly instances where this is un avoidable.
Public Safety Dispatch points being an example: when many simplex frequincies only a couple of channels apart are in use.

If the stations cannot be physically seperated, cross muting is used being understood that reception is not possible on other channels when dispatch is talking.

Since small departments have only one dispatcher this is of no great concern.

It all depends on your application.

only 390 kc apart

Are we dating ourselves?

(I once went on a tour of a WWII Naval Auxiliary Landing Field, which had been abandoned since the end of the war. At the end of the weed-dominated blacktop was the old woodframe "tower" (only 2 stories tall), all of the glass of which was long since gone. And there painted on the side, in 6-foot tall letters, extremely faded but still readable, was the terminal freq: "396 KC." Sort of felt like you had fallen in with Rod Serling and ended up in the Twilight Zone.)