RF isolator loss due to abuse

[EE4all]

I have noticed a trend for almost all RF isolators that I come across, where all the specifications are on point when I tune a new isolator, but for a used isolator, insertion loss is often around double the factory specification, despite multiple tuning rounds. All other specifications will still be pretty much on point. For example, a dual isolator with a .45dB specification now will read 1dB. This is not for every used one I have worked on, but multiple.

I believe these used isolators have normally seen years of heat and often abuse, sometimes perhaps getting close enough in range to the Curie temperature, that the magnetic components degrade and thus impacts signal flow. If you have ever been in a non air-conditioned rooftop repeater shack like I have, you know about that added heat factor as well. The increased loss ends up as heat in the isolator and its loads, compounding heating and damage more as time goes on. Older isolators may also have oxidation to internal components I suppose as a cause. Hmm.

What say you? Is there some other reason I am not thinking of?

[radiolistener]

I think that old connectors have a lot of metal dust from metal contacts due to a friction. That metal dust leads to isolator contamination with a thin layer of metalization and it leads to additional power loss.

[EE4all]

When I say all isolators I come across, I should qualify the statement to say this is not my day job, so my sample size is pretty small compared to someone who perhaps does this commercially a lot more often. I figure the collective experience here might give me more insight.

[EE4all]

I think that old connectors have a lot of metal dust from metal contacts due to a friction. That metal dust leads to isolator contamination with a thin layer of metalization and it leads to additional power loss.

Could be, but I have one EMR dual isolator here that is very clean, no corrosion outside, adjustment plugs were in place with no rust, connectors look flawless, and the unit is well sealed, but having a 1dB insertion loss even with multiple retuning rounds. On the other hand I have one that looks well worn, and has half that. 

[coppercone2]

im pretty sure that magnet material in there ages and changes the parameters of it

oh yeah and I am pretty sure another factor is if you take one apart they seem to glue the connectors on with conductive glue IIRC. The one I Had was built in such a way that it was not physically possible to solder the new connector on, so I used conductive epoxy, but inside IIRC there was a glue already, I did not rip a solder joint taking the dented connector off. Or perhaps it was just regular glue then press fit into place.

To solder it you would need like a laser to get in there through the tiniest of gaps and then push it in when its molten and hope for the best

and since it has glue in there, maybe stuff shifted on high temperatures.

I have not seen a tunable one though?

[EE4all]

I have had one old nontunable Telewave double isolator across my desk. Yeah, lots of glue, and it was really worn out looking. It also had a bit more loss than it should have had, but at least it was nice and big, and could handle the extra heat. This EMR on the other hand is about half the size and mass. These are a bit like tuning a duplexer, in that you need to make several passes to get the absolute best performance out of them.

[coppercone2]

well they are fascinating I need to get some tunable ones or ones that you can open up non destructively

Also, I have been meaning to stick a magnet on one when its plugged into VNA

[EE4all]

Well, just look for an eBay special. That's what this EMR was, and probably why is has half a decibel more loss than it should. 

As they say, you get what you pay for.