Originally Posted by pmugabi
Maybe the one way valve is leaking?
I used a similar setup for a few months at one time and it worked perfectly. Although it was not a one way valve, it was simply a T inline with the MAP sensor with one side left to the atmosphere (I know unmetered air etc, but it worked fine).
Another option is to put the 97+ MAP sensor, which is what I'm using now, by default it will cut boost at 1.5bar. It has a different electrical socket, so there is a bit of re-wiring to do.
If the valve was leaking I would think replacing it would have done the trick.
As to the 97+ MAP sensor. I would think that using Sensor that max's out at a higher value would require the ecu to have a new or rewritten table of MAP sensor values in the same way that a giant MAF sensor cannot simply be dropped in without a new tune that correctly interprets MAF values from a sensor with a higher max value.
As to the OP question, Do you have any datalogs with MAP sensor readings from before and now?
Is it possible that you have a vacuum leak somwhere else that is effecting the functionality of your one way valve. I dont think the ECU would typically let something doesnt run for months and stop working. If the ECU somehow relearned its usually in 2 weeks not 2 months.