Interesting Astor compressor circuit

I'm in the process of acquiring and sorting through a bunch of old Astor Australian made broadcast gear, some of which I have schematics for very closely related modules, some of which I don't. One is an 'intercom' module with a mic input, speaker output and a built-in compressor. Still tracing it out, but it's very closely related to the schematic for the talkback section of an Astor mixer of the same era.
    This section has a little compressor, (schematic attached, compressor section down the bottom) which is using a transistor to drive a diode into conduction to pull the signal down. Although similar, it's not a diode bridge. I've seen some theoretical discussions on this type of compressor, but nothing comes to mind of actual products using this.

  Will be a while before I have time to get these modules functioning and actually listen to them.  Any thoughts, examples, explanations or other opinions on this kind of compression circuit?

NOON said:

Although similar, it's not a diode bridge.

Click to expand...

A diode bridge has a self-cancelling effect on distortion; a single diode has not. The crude performance may be acceptable for TB, but not for any kind of "noble" audio application. Maybe for a telephone effect...
This crude and cheap approach is in contrast with the rest of the circuit, which is rather well designed (xfmr in/out, nice power amp...); they may have deemed it sufficient for the purpose... I guess German engineers would have had a different approach.

I don't really expect it to be 'hi-fi'. They certainly don't write manuals like they used to, see the attached circuit description.

And tech specs.

I was browsing through the Mullard Transistor Audio and Radio Circuits Second Edition and ran across pretty much this exact circuit, described as AGC for a tape machine.
It also has a very good description of the circuit operation. It explains that the B-E junction of the transistor acts as the second diode to reduce distortion, thereby saving one whole diode.
I've attached a scan of the relevant section. (If it's still under copyright it should count as fair use for educational purposes.)

Also some interesting charts at the end analysing the circuit behaviour. Looks reasonable (for the time) as long as signal levels are kept low, hence attenuating right at the input.