[Radio Wrangler]

I know the 8901A which doesn't have the head with the power sensor.

8901s are fairly expensive instruments, even second hand, even today. They are also very expensive to repair.

The easiest and most popular way od damaging them is to put in too much power.

A dummy load and a sniffer may be OK in general terms, but with so much at stake, you really should be using a good quality attenuator.

I've used 8901As a lot over the past dozen years measuring aviation transceivers wit powere from about 4W to 60W pep. A 20dB RF attenuator would allow the 8901 to work to 100W much of the time I was using a 30dB attenuator because that was what we had.

Go for a good make of 20 or 30dB RF attenuator with N-type connections.

Be aware that almost all attenuators are asymmetric. They may give the same attenuation in either direction, but they don't have the same power rating each way. Usually the shunt resistive element is beefier at one end than the other. So a 100W 20dB attenuator is sized to run 100W at one end and not necessarily much more than 1W at the other end. So pay great attention to which end is the input.

On its own, in default mode, the 8901 will detect some RF power which will trigger a search to find its frequency and lock on. Then you will start to get modulation readings.

You can circumvent the search time and get speedier results if you type in the carrier frequency, it will then lock much faster.

There are lots and lots of special function commands from the front panel which do things you will never need, and some you might like. One of them selects better counter resolution, for example. There should be pull-out plastic cards with listings.

David

Oh, people also blow up attenuators by transmitting into the wrong end of them. They seem to treat RF like software hacking and just give it a go.... Sadly this winds up being expensive. When you buy an attenuator, if you go to one of those "We don't have the equipment to test it, but it sure looks nice!" places, be a bit suspicious. You can test them with a DVM and a bit of calculation, so that sort of statement means that they haven't a clue at best.