In trying to make a diode ladder filter (failed of course) I got something entirely different...
I was poking around with the differential amplifier that was fed by each side of the ladder and happened to disconnect both inputs. I still got signal. There was a feedback path from the output to the base of the opposite transistor (not the input). As far as I can tell, the signal is coming from the base of the transistor.
Running an LFO into the CV in makes for a kind of phasing sound. I'm not entirely sure what I did. I suppose I like it...
Why would it not come? If proper biasing conditions are met and the base is left floating the base's potential is diode drop higher than emitter's and the other transistor in the pair happens to receive input through it's emitter.
It's the rudimentary princeiple of operation of the LTP.
Of course in a proper design you might wish to ground the base (to make the common base amp operate to it's full potential) or alternatively direct some feedback to the base. ...But leaving the base floating will cause it more or less "track" what is happening at the emitter.