Recommendation - use Notch or Bandpass for desense

[brushfire21]

Good Evening All!

I need a little input on a small issue with desense occuring in a VHF repeater system at a site with no other equipment other than a 900mhz SS link radio for telemetry control at a tank site on a 1600'ft hill. Here are the spec's:

-Vertex VXR-7000 running at 50-watts out
-Telewave TPRD-1554 4-can/5" BpBr duplexer with ~77dB isolation (per the spec's) at 1.5dB loss
-RG-214 double braided coax between repeater/duplexer
-38'ft of 1/2" LDF4 Heliax
-15'ft of 1/2" Superflex (between duplexer-lightening arrestor and jumper to LDF4)
-Lightening Arrestor mounted to cabinet (don't remember brand)
-Telewave 2.5dB VHF (First Antenna)
-Commscope DB-222a 2-element 3dB/6dB exposed array dipole (2nd antenna)

My issue is that with the original Telewave antenna, no desense was occurring that I could perceive enabling/disabling the tx on the repeater with a weak station, but the antenna was not receiving very well and kinda deaf. I installed the dipole in an omni configuration and I started getting desense on the input. I then switch it to a directional antenna with the null of the antenna to the repeater, and almost all of the desense went away. In this configuration the repeater is working quite well and I am now a believer in dipoles. I would prefer it in a true omni configuration as HT's in the null are weaker then they used to be, but it's doable in this configuration.

I have a Wacom notch/bandpass filter cavity (can be configured either way), and was wondering what might be best; install it on the rx of the repeater, notching out the tx frequency. Or set it up more as a single pre-selecter in a bandpass configuration on the rx side of the repeater to only allow the rx frequency through. It appears that at 2mhz away, I am getting ~18dB of isolation in a bandpass configuration with a 1/2dB of insertion loss (per my VNA).

So what do you experts think or feel? I may go to the site and test both configurations, but leaning towards the bandpass configuration with the cavity located between the existing duplexer and repeater rx port. Do the cables need to be a specific length like the duplexer's need? Any help would be great!

 

[kayn1n32008]

Wirelessly posted (Mozilla/5.0 (BlackBerry; U; BlackBerry 9900; en-US) AppleWebKit/534.11+ (KHTML, like Gecko) Version/7.1.0.746 Mobile Safari/534.11+)

Notch out the transmitter.

 

[fineshot1]

So what do you experts think or feel? I may go to the site and test both configurations, but leaning towards the bandpass configuration with the cavity located between the existing duplexer and repeater rx port. Do the cables need to be a specific length like the duplexer's need? Any help would be great!

I am not an expert but i have almost the same duplexer and both sides have 2 cans with pass/rejet
and i did just what you are considering with the bandpass can in the rx side and this works very well
so you should not need anymore notching out on the rx side.

 

[kayn1n32008]

Wirelessly posted (Mozilla/5.0 (BlackBerry; U; BlackBerry 9900; en-US) AppleWebKit/534.11+ (KHTML, like Gecko) Version/7.1.0.746 Mobile Safari/534.11+)

What is the TX/RX seperation for the repeater? 4 cans may not provide enough isolation

 

[brushfire21]

Thanks, both of you. In this setup the split is 1.99 mhz On the system i mentioned before. This weekend I am going to the site and will play with the system. Waiting on a 3' superflex jumper to arrive to add the notch/bandpass filter for testing.

 

[prcguy]

The duplexer should have been tuned with a good 50 ohm match on all ports so if your antenna presents a bad enough match the duplexer pass and notch can be pulled around a little. Check the antenna match.

Or the duplexer can be touched up while in circuit but you need a directional coupler to inject a signal generator from the antenna port and tweak for best receive performance.

Or maybe the radios are not really suitable for duplex use and you have direct radiation into the case and circuit board of the receive radio.
prcguy

 

[kayn1n32008]

Wirelessly posted (Mozilla/5.0 (iPad; CPU OS 6_1_3 like Mac OS X) AppleWebKit/536.26 (KHTML, like Gecko) Version/6.0 Mobile/10B329 Safari/8536.25)

Thanks, both of you. In this setup the split is 1.99 mhz On the system i mentioned before. This weekend I am going to the site and will play with the system. Waiting on a 3' superflex jumper to arrive to add the notch/bandpass filter for testing.

At 77db isolation, what is the duplexer minimum separation rating?

 

[rescuecomm]

The duplexer should have been tuned with a good 50 ohm match on all ports so if your antenna presents a bad enough match the duplexer pass and notch can be pulled around a little. Check the antenna match.

THIS!

I am running a VXR-7000 VHF with a Sinclair Q2220 at 25 watts with no problem. The frequency split is 3.025 mhz with the antenna match is almost 1 to 1 on a Stationmaster.

If the desense changes with the antenna orientation, then your antenna match definitely needs to be checked.

Bob

 

[brushfire21]

Minimum split on the duplexer is .6mhz, which is at the 77dB isolation and 1.5 dB loss per the spec's. noise floor is low as well. The setup was put on an analyzer at the site and it checked out fine when first put online with the 2.5 dB Telewave omni and no further tuning was needed. Were getting about 34-36 watts out the duplexer with either antenna.

My superflex cables arrived and am trying to squeeze in some time tomorrow to had to the site.

 

[WB5ITT]

Minimum split on the duplexer is .6mhz, which is at the 77dB isolation and 1.5 dB loss per the spec's. noise floor is low as well. The setup was put on an analyzer at the site and it checked out fine when first put online with the 2.5 dB Telewave omni and no further tuning was needed. Were getting about 34-36 watts out the duplexer with either antenna.

My superflex cables arrived and am trying to squeeze in some time tomorrow to had to the site.

Your repeater at 1.9MHz split will require more like 82db or more isolation..Also HOW deep are the nulls?
I think the duplexer may be underrated for this split.....to make it work I would add a notch cavity in the xmtr side notching the RCV freq and put another notch can in the front of the rcv path notching the XMT signal..
That would kill the noise from the xmtr from getting to the T and the added isolation from the xmtr on the rcvr side should improve its operation.

IMHPO, I think you went with a too small of a duplexer....always strive for 80-88db isolation....especially on VHF narrow splits....If the split had been `10 MHz, this duplexer MAY have worked,,,,
Try the two new cans as mentioned above and I think the desense will be gone....

 

[brushfire21]

Thankyou for the input. Still having a little desense but have reduced power to around 20 watts out of duplexer and the desense is almost gone. I put in a notch on the rx to block out the tx and it was hardly noticable but seemed to do a little bit of good.

 

[beachmark]

This is waaay late, Brushfire, but I have to wonder if:

1> Your problem could be direct radiation from the transmitter getting into the antenna, and you had a bad Telewave antenna...the classic 2-problem situation. The poor Telewave antenna did not respond to enough of the nearby TX radiation to casue any desense, but the 2nd antenna did, and then changing the 2nd antenna to directional reduced the pickup some.

The TX notch in the RX line should have helped this, but did you use a 1/4 coax between that and the rest of the duplexer? Not knowing the duplexer config makes it a bit harder to advise you on that aspect. And you could have direct TX radiation into the RX lines/duplexer, somewhere, as previously noted.

2> Have you checked to see if you have a good VSWR on the 2nd antenna, measured into the antenna line right at the duplexer port? A poor VSWR will refect more TX energy back to the duplexer and depending on the locations of the voltage peaks on the line, the duplexer could have more to deal with than with a good antenna load.

3> Is the 2nd antenna actually new? Or used? the reason to ask is if there is something corroded in the antenna or in the lines and connections, or in the actual metals in the coax shield and connectors, that can cause what is know as spectral re-growth or phase noise spreading. It is a form of passive inermodulation and only shows up when real RF power is applied. This will create phase noise in the antenna line at the RX frequency, and there is nothing in the RX side of the duplexer to stop this so any tiny amount will desense the RX. (All of the RX frequency phase noise filtering is in the TX side of the duplexer.) The desense may have dissappeared when you reduced power to 20W due to this issue. It would also exactly explain why adding the TX notch after the RX side of the duplexer did not make much difference. The more I think of it, the more I think this is the first place to investigate.

4> The 77 dB isolation is marginal for most repeaters as noted. You can sometimes sacrifice some insertion loss, maybe 0.5 to 0.7 dB, and get 5 dB or more added isolation of the TX phase noise on the TX side of the duplexer, and do the same on the RX side to get more suppression of the TX carrier with a small loss of RX sensitivity. That would be better on the TX side than reducing 50W to 20W which is a 4 dB loss. You just have to watch the VSWR back to the TX input when you do this.