|
May 30th, 2010 at 11:27am
Ideally you need the same source and load impedences as well as connectors and cable type.
If you have a mixture, the effect is partially reflected signals which results in amplitude ripple due to partial signal cancellations/additions across the frequency band. This will add errors to measurement accuracy - but it depends what you are trying to measure or observe - if you are simply looking to see a particular small carrier on a spectrum analyser then maybe the ripple errors and absolute levels don't matter. The partial reflections also cause mulitpath delay effects, like shadow images on analogue TV and inter-symbol interference in digital systems. Similarly don't crush coax cables at regular intervals (e.g. using wall clips, clamps or cable ties).
If you are connecting a C band BUC with a coax output via coax to the feed OMT you should must use 50 ohm Andrew Heliax cable (or similar) and Andrew 50 ohm N type connectors at both ends.
Be aware that 50 ohm N type and 75 ohm N type plugs exist. Never put a 75 ohm N type plug to a 50 ohm N type socket - the centre pin won't connect properly as the 75 ohm centre pin is very slightly smaller diameter.
I agree with the desirability of using 50 ohm for all C band RF measurements.
Best regards, Eric.
|
