pcmhacking.net

kph and rpm are in the first graph, no spike. Im wondering if there is anything else that could affect it, like air going to the wrong place until something is full, brake booster, or something like that. The map sensor should be trustable, so the question is why the turbo drops pressure at that point, or where the air might be going that could reduce pressure at that point. The sample rate on the dyno looks slower and rounded compared to the ecm. Loosing 40kpa over half a second, then back again in the next half a second. Is there any kind of pcv or similar that might open when it shouldnt when boost hits that point?

The only valve in the intake path is the brake booster. It's probably worth plugging it on the dyno for testing. One less variable. What is more interesting is that in steady state at 4000rpm the boost oscillates between those two points.

I'm thinking that maybe the head is lifting and leaking between adjacent cylinders. This could explain the loss of boost pressure (less exhaust energy) and no change to AFR's or VE while it happens. The simpler explanation could be the wastegate arm is binding up as the turbo heats up. I'm not sure on a way to verify this

I'm assuming you have actually observed the motor when this happens? No hoses in front of turbo collapsing?
Definitely do an intake boost leak test. Plug up the intake with a pvc pipe cap or something and stick some compressed air into the manifold via the fuel reg reference hose. Get out the soapy water if she leaks...

Its possible it could be something to do with the wastegate too.

Charlay86, haven't seen the motor when this happens but there's nothing flexible on the turbo inlet, just a 3" flex pipe with steel coil inside up to a pod filter. I'll grab some plumbing this weekend and pressure test it.

I think it's going to end up being wastegate related...

Holden202T, finished up my ALDL LCD.

Primarily to replace my broken dash temp gauge, it shows temp when off boost and will show boost pressure when positive MAP is read. Using Mode1 Message 0 so most values are easily accessible. Should work fine while tunerpro logging and I could always add an SD socket for logging without the laptop. Ideal for boost control too.

I'll probably get a few pcb's made up for these, and maybe a couple for the larger 3.5" IPS panels. Let me know if you're interested in one and I'll get extras made up. I'll release the source too so you can mix it up though you can just connect a USB cable to it and it'll give you a config text file that you can adjust (not yet implemented but the framework is there)

https://www.youtube.com/watch?v=0AJSLYTQKBs

yeah i'd love to give one a try!

Small update on the boost fluctuation. Turns out it was a leaking manifold gasket which eventually blew out. Didn't leak at idle, only under boost and couldn't be heard over the dyno. After replacing the gasket, we get a graph which looks like this:
I would have thought a turbo designed for a 2.2L saab would have a bit more power in it on a 1.8L but its clear we're seeing boost creep past 4k and diminishing returns on boost past 7psi, plus full boost by 1900rpm. All these together I think indicates we're running out of exhaust turbine and back pressure is now the bottleneck. This is with the cat-back disconnected.

So, now begins the GT2871 (china spec) build!

It could also be your timing, if your way low in a 'safe' tune it'll gain a lot with more advance. But of course you dont want too much, and how much is always the question.

It was tuned for peak torque across the boost map and 2 degrees pulled out of it just to be safe. Still a whole lot of fun to drive at 85kw, I just need more

2 cylinders are running a lot richer than the other two so I'll be off to the wreckers this weekend to find a genuine set of injectors. 440cc are overkill for this thing, 220cc were getting close to 90%. Wonder what has ~330cc's in them from factory.