rob, my actuator should be right. I tuned it to start moving at 4inhg when I installed it in the car and it was set there last I checked. I really don't feel like messing with it right now as its cold and wet outside.
From user dieseljohnny from ecuconnections:
I have found this to be accurate with regards to my ECU.
Locate the boost target map and make note of the address. Sort the maps by address and look for the unnamed maps with addresses just before the boost target map and you will locate a small 0 0 map, then the PID X?, PID D, Upper N75 limit, Lower N75 limiter, and PID I maps.
The Lower N75 limiter has flat values of 500 and a Y axis of rpms. The upper N75 limiter has values starting at 7500.
Please note that VCDS reverses the N75 duty cycle numbers in the ECU.
7500 is 75% duty cycle, vag com reports 25% duty cycle. VagSuite has also reversed these numbers to avoid confusion with VCDS.
So the lower limiter actually is saying that it can have a duty cycle of 10000-500 = 9500 = 95% across the board.
First I'm going to try raising my lower limiter values to 3000 (70% max duty cycle) and see if that fixes my boost spikes. This should be equivalent to changing my turbo actuator length. If that does not then I will start digging into the PID controller (first to confirm that those are the PID X?, D, and I maps respectively).
I still think digging into the PID controller would be better as I would still be able to have a quick turbo response but damp out those spikes, whereas limiting the duty cycle is kind of a hamfisted response.
http://en.wikipedia.org/wiki/PID_controller
It looks like really tuning a PID controller is difficult, but for my desire I could simply scale the PID I map in half and that should greatly (hopefully overdampen) dampen the oscillations and overboost.
The simple explanation of the PID controller is that the P factor is proportional. If actual MAP and target MAP are different the difference D is mapped to a proportional response. IE. 100 mbar difference -> 5% change in the N75 DC. I think the Y axis in the PID maps is the difference in mbar from actual MAP and target MAP, and the mapped values are the change in the N75 DC.
The PID I factor is integrative over the Dt factor (length of time it integrates over). Initially if there is a change from actual MAP and target MAP PID I is still 0 or close to 0. As time passes and the MAP difference D is not corrected the PID I factor increases. The table should be the same as the PID P map, look up the PID I factor in mbar on the Y axis to find the duty cycle response.
The PID D factor is derivative and, I think, also uses the same Dt factor (though it doesn't need to).
The PID controller looks up each individual P I D factor and sums them to get the final N75 DC value.
For the record my maps are at:
353372: Upper N75 DC limiter
353304: Lower N75 DC limiter
353236: N75 PID I
353440: N75 PID D
353508: N75 PID X?
354262: N75 PID P (I think)
Also: There seem to be two separate control schemes for the N75 valve. Under a particular RPM, boost, or fueling (I'm still not sure) the static map is used (The N75 duty cycle map). When a condition is met the ECU switches to the PID controller to meet the boost target map and the N75 duty cycle map is ignored. I'm going to try to locate the switch so that it can be changed. I think if you don't do this then simply editing the static map will have no effect on boost spikes because the ECU will be using the PID controller and not the static map.
