[chittychittybangbang]

The VNT actuator works off vacuum. This is the opposite of wastegate acuators which work off pressure. At rest there is no vacuum so the spring inside the can is the only force acting on it. Your understanding of it is correct. When it's at rest the vanes are wide open and the rod is longest, when the actuator gets full vacuum the vanes are closed and the rod is shortest. When it's running it's somewhere in between. The car modulates vacuum using the N75 solenoid.

http://www.myturbodiesel.com/1000q_how_to/a4/VNTadjust.htm see here for some videos

Have you checked for camshaft wear? There's an article in the mk5 FAQ for this linked below. It was a problem area on the US 1.9L pumpe duses. Worn camshaft lobes would create a lumpy idle or strange acceleration problems. The misfire symptom seems like it might be caused by this. There's a video of a car with a strongly worn camshaft idling too. See this: http://www.myturbodiesel.com/1000q_how_to/multi/camshaftInspection.htm

 

[chittychittybangbang]

I said..."when engine off rod is longest, but with engine idle, rod is shortest".
You said " when it's at rest.....rod is longest"...is this with engine off or at idle ?? Sorry to be so pedantic here but obvously there's a big difference here.

Sorry, when the engine is off the rod is longest because only the spring is acting on it. When you turn the engine on it cycles short/long. At a certain engine range the rod gradually gets shorter because vacuum is working against the spring and shortening it.