I got your post before, and since then I have acquired a 2m circulator, and a 50 ohm dummy load for an antenna. My network analyzer says that it does exhibit the 30db isolation that I'm looking for.. I can insure that the isolator sees that load with a 2m antenna tuner which I already have, or a custom built loading circuit if needed - the frequency range that needs to be covered is very narrow - less than 1mhz wide.. so I should be able to get a good match across the entire range.
You say "makeshift" but it gets me about the same isolation that splitting the rx and tx antennas does for the space I have available to split the antenna, or the same amount of isolation as a single cavity (which would be multiplied by 3 for my 3 receivers).. so I'm just as well off and the circulator is as cheap if not cheaper than the 2nd antenna, and definitely cheaper than 3 cavities in terms of both cost and space. it also gets my RX and TX at the same height and the same gain. definite benefits. And even if I don't achieve full 30db isolation, I can make up the difference elsewhere in the system - trading off cavity bandwidth and insertion loss to get the additional isolation needed To date I have acquired the repeater controller, 4 cavity filters, 1 2m circulator, and 2 of the 4 radios I'll be needing.. still a bit more to go before I can do a full systems check - at least one more circulator and one more radio before I can setup a minimal test system (1 simplex radio + the repeater). I've got 2 more cavities on the way that should get me all I need for the receiver filters. So, until I have a few more parts, it would help to have some sides of how receiver specs will impact the required isolation needed. This isn't guessing - its called RESEARCH _____ From: [email protected] [mailto:[EMAIL PROTECTED] On Behalf Of Jeff DePolo Sent: Monday, October 22, 2007 3:11 PM To: [email protected] Subject: RE: RE: [Repeater-Builder] Re: duplexer isolation and reciever noise budget > Transmit and receive paths will be separate with 30db of > isolation (either split antennas, or a 30db isolation > circulator just before the feed to the antenna) I posted a followup to one of your previous messages regarding using isolators as a makeshift way of getting more Tx to Rx isolation, using the reject port to feed to a receiver, and the problems therein. For some reason, that post never showed up. Anyway, the 30 dB of theoretical isolation that you refer to is never going to be realized in your application. The isolator's spec of 30 dB that you're quoting is from antenna to transmitter, assuming an ideally-matched reject load. You're using the isolator differently; that spec doesn't apply. In your configuration, the Tx to "Rx" (reject port) isolation is only as good as the return loss of the antenna; consider that a 1.5:1 VSWR is only 14 dB return loss. Maybe you'll do a little better than that, but you'll probably never get the full 30 dB. And there are a whole slew of other issues related to IM and harmonic generation within the isolator. > That brings the transmit power down to +17dbm (55mw) before > any other splitting and filtering to get to the 3 receivers What about noise supression? At close T/R spacings, the noise supression requirement is likely going to be higher than the carrier supression required. > Am I blowing smoke ?? Is there a way to figure from the usual > receiver specs just how much signal can be handled before > de-sense occurs ?? A typical 50 watt solid state 2m repeater typically needs 80 dB or more of noise supression at 600 kHz split, and about 80 dB of carrier supression (both figures plus or minus maybe 10 dB depending on the radios used). If your Tx is closer than 600 kHz to your Rx, you're going to need proportionally more noise supression; 100 dB at 400 kHz spacing would probably be a good guess. Carrier attenuation likely won't need to increase substantially until you get appreciably close to the Rx frequency, like maybe within 200 kHz as a rough ballpark guess (it will likely vary quite a bit depending on the quality of IF filtering in the radio). Instead of guessing at all of these numbers, why not quantify them if you already have the equipment you're planning on using. --- Jeff
