An Anritsu Reflection Bridge: a few questions.

[Skywave]

Intro.
1. This thread has arisen as a consequence of another thread on this forum which is currently W-I-P.

2. This item has been sitting in my 'round-to-it' pile for longer than I can remember; I can't even recall how - or from where - it came into my possession, but it was many years ago.
Attached are three photos. to aid my questions about this item. Photo 1 is simply to supply its size. The other two are the really relevant ones.

3. I must admit that devices like this are on the edge of my understanding of RF measurement devices and methods, but I do believe that I have the basic concepts of its use understood. However, searching the 'Net for info. on this specific item has revealed nothing (probably due to its age). What I did unearth - which was of a general nature on similar devices - has left me confused, and after some initial playing around with it I have become more confused.
Hence, it's time to ask!

Main text.
First, I have assumed that the BNC connectors are designed to be used in the following way:
(a) The 'INPUT' conn. is where the signal source is connected (in my case an HP 8640B);
(b) The 'REFERENCE' conn. is for connecting a 50 Ohm dummy load;
(c) The 'TEST' conn. is for connecting the measurement device (which in my case is a non-sweeping spectrum analyzer);
(d) The 'X' conn. is for connecting the DUT.
► Can someone confirm all of that - or otherwise - please?

Second, the insertion loss, from 'INPUT' (as input) to 'X' (as output), is 6 dB (at 10 MHz): that sounds about right to me √. However, the insertion loss, 'X' (as input) to 'TEST' (as output) is 26 dB (at 10 MHz). I was expecting another 6 dB, not 26 dB.
► Is that difference (26 instead of 6) a sign of a malfunction, or is that 26 dB to be expected? I ask that since general info. from the 'Net about similar devices states 6 dB, not 26 dB. And if it is correct, why the extra -20 dB? Surely that will have the effect of desensitizing the device for the purpose of measuring the return loss?

And finally, any general remarks of how to gain useful info. about the DUT using this particular item will be appreciated.

Thank you.

Al.

[Skywave]

Using the time-honoured method of experiment and test, plus studying various non-specific product text-books, I feel that I have gathered enough data about this device to be able to put it to good and meaningful use.

So on account of those remarks and the zero response to date, I am happy for this thread to be closed.

Al. / Aug. 29, 2015 //

[G8HQP Dave]

My guess is that the extra -20dB is so the receiver or spectrum analyser is not overloaded.

[Radio Wrangler]

Yes, that should do the job for you.

You can use a sweeping spectrum analyser and a noise generator or else a comb generator or else a tracking generator.

David

[Pamphonica]

There's a brief spec at the end of this document:
https://www.google.co.uk/url?sa=t&rc...Y_j3j8u9FxYrHQ

Jeremy

[Radio Wrangler]

6dB would have been what you would have got from a maximally-coupled hybrid. This one is either set up for lower coupling by the transformer turns ratios, or else there is an attenuator in there to buffer the impedance presented by the analyser/receiver/detector so that it doesn't unbalance things and spoil the directivity.

Essentially there's a transformer coupled Wheatstone bridge in there. To get wide bandwidth in a measurement transformer one trick is to use the attenuation of a minimum loss pad to bring the impedance level from 50 Ohms down to as low as feasible, the transformer magnetising L sets a lower 3dB point for the low frequency rolloff.

Wes Heyward put some info on transformer hybrids in "Intro to RF design" and his follow-up book. I put a bit in Sprat 61 configured as a dual directional power meter.

David

[Skywave]

Thank you, Jeremy, re: post #5. (What does puzzle me how come I didn't find that link! )

Al.

[Skywave]

Originally Posted by Radio Wrangler

You can use a sweeping spectrum analyser and a noise generator or else a comb generator or else a tracking generator.

Yes, I probably could - but, unfortunately, I don't own such expensive R.F. exotica!

However, my thanks go to David (Radio Wrangler) and Dave (G8HQP) for their respective remarks.

Al.