Overboost - 17/22 and Malone Stage 5

Long time lurker who hasn't posted in ages. I'm hoping you guys can help me sort out an overboost issue that I'm having on an 02 Golf TDI that I picked up fairly recently.

The car has a 17/22 with supporting mods and a Malone Stage 5 tune. The analog boost gauge is showing spikes/being pinned at 30 PSI if I go WOT (every once in a while it will catch and hold at around 26). I'm tried flashing to the Stage 2 Tune on the Flashzilla and it's still boosting to 30, so there's something mechanical causing the overboost.

Here are my symptoms:
- 30 PSI (or more because gauge stops there) spikes on WOT
- Some oil on the back of the engine and minor drops on the ground (could be valve cover leaking down, or possible intake manifold)
- Some oil in the intercooler to EGR (deleted) piping
- Small amount of oil consumption, but not topping up between changes.
- Turbo is 90,000km old.
- Still getting 5l/100km on the highway
- Small amount of smoke on first start of the day - no noticeable smoke after

Here's what I've tried:
- Pulled VCDS Block 11 logs and sent to Malone - logs show a max boost of 2600 mbar. Malone though that the car probably had a 2.5 bar map that was being maxed out, but I verified the installed map by part number is a 3 bar sensor. I don't have an explanation as to why the sensor is capping at 2600 mbar when it' s boosting over that value and the installed sensor is a 3 bar map.
- Vac tested the actuator (moves freely at 4 and stops around 17/18) **Does anyone know what the range should be for a 17/22?
- N75 tested using VCDS output test and Mityvac (output test hits 25 on Mityvac and then drops in stages while the N75 is cycled.
- Inspected all vacuum hoses for obvious breaks (I'm planning on replacing them all this weekend.
- There's oil on backside of the block, but not around the turbo.

What would you recommend checking next? I'm a bit lost at this point.

So if I am figuring correctly, your boost gauge is saying 30 plus psi of boost and your map is saying approximately 23 psi?

Are you getting overboost codes? If not, I would try another boost gauge, or another map if it is not registering overboost that actually exists.

Take it for a hard drive. Run it to redline in 3rd and 4th at WOT a few times. You can work up to it by using part throttle at first to prevent running at 30 PSI, but after a couple runs the overboost should go away. 17/22s are notorious for building carbon on the vanes, causing erratic operation. This especially true if the tune is overfueling. If you're persistently seeing less than 40 MPG (US) then you're probably overfueled.

But a hard drive usually sorts things out. Oil collecting in your intercooler is an indication you're not running full boost often enough.

"17/22s are notorious for building carbon on the vanes," IBW

I was having trouble with mine overboosting. The stop screw on the actuator was loose, apparently not the case with yours as you've checked
the vacuum. At Jeff of RC's (Rocketchip) recommendation I also installed
a catch can. Over time, with some hard driving, the turbo is much smoother. I can still spike it if I start WOT at 1600rpm, but from
2000 it holds at 24psi. My purely subjective butt dyno tells me I've cleaned up the carbon on the blades. The additional mass of carboned blades when
spooling was running away with the boost IMO. No more puff of smoke
at initial throttle tip in.

From the sounds of your mystery oil leaks, I'm guessing you have a worn motor and probably a fair amount of blowby through the CCV.

Mileage? Nozzles?

Dial the IQ (fuel) back, ie increase the number, check timing and deviation.

Do you have VCDS?

The car has 430,000 KM on it. Nozzles are 520s.

I went though the vaccum lines today and replaced all of them. The car has an EGR delete, but the EGR vaccum lines hadn't been fully removed, just plugged off. I rerouted all of the vacuum lines to delete the lines running to the EGR and ASV.

I took it for a quick drive and it seems slightly improved. It will still peak up to the 30, but it seems to taper and hold around 27/28. That said, it was tough to get any real WOT pulls today - the roads are covered in snow and ice.

I do have access to VCDS. Would dialing back the IQ have any direct interference with the Malone tune?

only thing that would interfere with the tune is over-writing it with another file. If it's visibly smoking at WOT then dialing that back will probably help with boost control.

Here's my internet diagnostic recommendations, (For What It's Worth)

Adjust the actuator to stop a little lower, 15-16"hg.

Put the IQ up 1-2 points.

Install a catch can.



Roll into the WOT at 2000rpm, ie. don't just stomp on it.

Take it into the low 4k rpm on a regular basis. Keep the revs up!

Downshift earlier!

All of the above has worked for me.

Although I ran a 17/22 happily for 120K miles, this is why I don't miss it.

I checked out the IQ last night. It was set around 3.2 and I brought it up to about 6. The car is considerably smoother when starting in 1st and doesn't feel like it wants to stall anymore. From a driveability perspective, it's definitely an improvement.That said, it's still spiking to 30 PSI on WOT, so I'll be checking the actuator too.

What does everyone make of the MAP sensor readings? I pulled a Block 11 log the other day and it shows that the MAP sensor is topping out at 2600mbar and does not read any higher. I double checked and the installed MAP is a 3bar sensor based on the part number. There seems to be a disconnect between the log and boost gauge, as the log shows the car "tops out" at ~23 PSI, but the boost gauge is going to (or over) 30 PSI. In this any chance that the boost gauge is showing absolute vs relative pressure? Or could it be a MAP sensor issue that is preventing the car from reading any higher than 2600 mbar?

Hopefully this link works:

https://log.malonetuning.com/chart/...kw6rCnsOyw7HDgsKLwr3CusOewpsfwoUmw7wEw68AAAA=

map sensor is flat lined, n75 is fully open and still not able to control boost. boost gauge is probably correct.

That's not our stage 5 tune or it'd read the 3 bar sensor properly, normally set for 26 psi. There were a bunch of pirated/botched tunes floating around, this may be one of them, or it could just be a lower power stage 4 file set for a 2.5 bar map. Only way to tell would be to read off the file and send it to us.