I have been trying to get to the bottom of this problem for four months now. . . I have yet to find any post that has come up with an adequate explanation for this very annoying and dangerous behavior. Lets be the first. My next course of action is to ring the tech department in Alfa Ireland and to find a Bosch agent (they are responsibe for the fuel and electrical systems) who may have a better understanding of whats happening.
The 150 is very different to the 136bhp ricer has.
The 150bhp (providing it has not been chipped) should boost to 1.2/3bar absolute max (standard) on 'overboost' then drop to 1/1.1 bar after overboost.
Even allowing a little more, say 1.4 then to 1.2bar.
The map sensor on the 10v measures up to 2.5bar, 1 bar atmospheric so 1.5 bar boost.
Yes it may see 1.6 bar, but its odd that your reading is 1.9bar from a sensor that can't go that high? need a 3bar sensor for that.
If you have changed the overboost module, similar to the N75 module on Vags, and you have changed the MAP sensor on the inlet manifold (30quid) then you may have sticking VNT vaines?
The MAP sensor outputs 2-5volts, 2 on idle, 5volts on full boost.
The ECU reads the voltage, it allows near 5 volts for a couple of seconds, this is overboost, then activates the overboost module which thus alters the angle of the vaines via the small vacuum pipe which runs fron the module to the turbo, thus lowers boost pressure.
If the vaines stick closed, the turbo will run full boost, over 2.5bar reducing to 1.9 bar at 4900rpm (all it can sustain). In this situation the engine management will usually cut the fuel supply as its so serious, if it sees a little more than standard paramters, it will illuminate the engine management light, dropping the clutch, allow idle revs for a second will normally cure it.
Of coarse, you may have a leaking boost pipe, which can cause a similar effect with little smoke at all.