ASV: 11mm pump - how to "unmap" boost spikes?

Hello!

My original 10mm IP (ASV engine) died this summer and I replaced it with a 11mm IP from an automatic. A previous owner had the car chipped at ABT (from 110HP to 132HP) but I decided to take it a tiny bit further and started remapping it to 1,25bar @ 2500rpm and a nice torque curve that shouldn't kill the clutch too soon. All that has worked out pretty well, there's just one thing I can't get rid of - boost spikes.

Whenever you increase the throttle pedal position, boost pressure will go to about +40-50% of what is requested and then settles to whatever level it should be at. When slowly applying throttle it doesn't happen...

Now since the change in IP results in a ~20% increase in IQ there's more exhaust gas flow spooling the turbo faster than the ECU anticipates. So I decreased IQ in two maps by 20% (torque limiter, driver's wish) and adjusted the smoke limiter to lambda > 1.3 (taking the extra IQ into account). Still spikes.

Adjusted the N75 map to open the vanes earlier and further. Still spikes.

Now what? I don't know what else to do and I really dont want to kill that tiny GT1749V.

Should I decrease the values in the pump voltage map's IQ-axis according to the added 20% of fuel?

Cheers,
hybridfreak

make the vnt rod 1 a 2 mm longer, ore turn in the stop screw a bit 1 turn so vnt isn't so snappi any more.

Technically that boost spike is a programmed result of the N75's PID control

Lift the high fueling sections of the N75 map by a few %. This will kill the spikes although will create undershoot if you go too far.

Thank you guys! I lengthened the VNT rod (turbo was bought refurbished and propably set incorrectly) and it helped a bit. The spike vanishes faster but it tends to underboost for a while after a spike. So a little tweaking of the N75 map should do the trick!

Since I am already messing with the vnt rod, N75 and so on I thought it'd be a good idea to check if the stop screw is set correctly.

So, using VCDS Lite I tried going into basic settings and "Group 11". But instead of reving to 1400rpm it just stays at idle and the cluster starts beeping and showing an "Engine malfunction - Workshop!" fault. Can anyone tell me if I am doing something wrong or if I have to login to the ECU before using basic settings?

Looks like the vnt rod needs some shortening. But how does that explain the boost spikes I am getting? Is it because the turbo builds boost too late and at that point too quickly for the ECU to react?

The tune by default will spool the turbo like that. The overshoot is due to the PID control parameters, not the N75 map.
https://en.wikipedia.org/wiki/PID_controller

Ok, thank you!

I have adjusted the vnt rod length according to the standard procedure (70-80% @ 4000rpm WOT) because it was way off (rerubish...furbished turbo). Now the overboost spikes are higher and continue for longer...



Can you give me a hint on where to find the P, I and D values in my hex file? VAGEDCSuite doesn't find them automatically and all the information I've found online is for PD engines (find 3 2D maps close to the N75 map). Do the PID-maps for VP engines have the same adresses?

I've had the pleasure of learning about PID controllers and such in University (mech. engineering) but it's been a few semesters since. But I'll figure it out

If you use Garrett OEM actuator or very good refurbished, you can fix that problem most easily from N75 map with opening vanes from the problem area ( about 1400-1700rpm), and possibly raise boost spec a bit if more air needed there. As you can see, vanes are fully closed on 1700rpm, where they should be much more opened. Please also note, N75 map IQ axis follows gas pedal, not total IQ.

Basically how i adjust VNT rod, is that I adjust it sitting towards opening limit of the vanes when vacuum is zero. This will allow maximum injection quantity without overboosting that easily, and you can have full adjustment range to use.

You also must remember, that refurbished turbo actuators work differently than OEM, and it may be very tricky to make it work properly. Sometimes, it just cant. It always spikes and then it underboosts, programming will not help it and very very small duty cycle change affects too much for the boost. Why it happens, is that the trajectory per pressure unit is different, and they are not that precise too. Have found this many, many times making a tune.

Still FMO, i´d say on this case N75 map is the main reason. Either you have edited wrong coding map, or it´s done wrong. You need to open vanes from problem area, and add more fueling or close vanes a bit from underboost area. There is NO need to change PID to fix this very simple "problem".

hybridfreak said:

Ok, thank you!

I have adjusted the vnt rod length according to the standard procedure (70-80% @ 4000rpm WOT) because it was way off (rerubish...furbished turbo). Now the overboost spikes are higher and continue for longer...

Can you give me a hint on where to find the P, I and D values in my hex file? VAGEDCSuite doesn't find them automatically and all the information I've found online is for PD engines (find 3 2D maps close to the N75 map). Do the PID-maps for VP engines have the same adresses?

I've had the pleasure of learning about PID controllers and such in University (mech. engineering) but it's been a few semesters since. But I'll figure it out

Click to expand...

The addresses are specific to the software version and ECU from the car. If memory serves they are a 12x1 or 10x1 map. I'd suggest looking around on ecuconnections.com a bit as they have a wealth of information and you just have to search it out.