[TowerTalk] Andrew CNT-240-FR

jimlux jimlux at earthlink.net
Wed Nov 3 06:56:17 PDT 2010

Jim Brown wrote:
>> Add 0.25-0.50 dB for any connectors/adaptors if used above 50Mhz
> This is another one of those old wives' tales that has no basis in fact 
> -- they are the big lie that is believed simply because it is so often 
> repeated. 

> This spring, VE1RAC sent me documentation of some very good experimental 
> work he had done showing that the loss in decent connectors at VHF is 
> FAR lower than that -- a few HUNDREDTHS of a dB, as I recall.

Indeed.. It shows up even in professional work where someone does a 
gain/loss budget and stacks up the "worst case" loss of the connectors 
in the chain, using the "spec sheet" loss as the worst case. Typically, 
the spec sheet claims something like "dB maximum = .03 sqrt[f(GHz)]", 
which does NOT mean that the loss is likely to be anywhere near that.. 
it just means that the mfr doesn't want to try to measure any lower, and 
0.05dB is as good as "zero".  Same with VSWR specs, where you'll see 
something like 1.15 + .011 f (GHz).  That is the spec is more a 
reflection of the production line test measurement uncertainty than the 
actual connector performance.

Real connectors have very much lower loss, but it's tough to measure. 
Empirical methods like Jim describes are a good way (e.g. run a kW 
through it and see if it gets hot).

In fact, if the connector has loss that *is* measureable, that's a bad sign.

More information about the TowerTalk mailing list