Agreed - not easy to fix. However, should be easy enough to mask. Gian, your 2-transistor 'clamp', applied to U1 pin 3, takes care of Rx-Tx transition, and I'm working on Tx-Rx with 10uF (what's to hand) between +12 and junction of 10k/1N4148. If that 'hang' is not long enough (T=CR, uF*k = ms, 10*10=100ms - will that tenth of a second be too much? Lessee, have I got a smaller cap?) I'm sure the 10k could go to 47 or even 100k and still clamp effectively - do we really need (10V/10k=)1mA into either base? There's a couple of 150k still on that scrapped PC PSU .... We shall see ...
73