2005 Legacy GT wagon: a bone stock rebuild

[birkhoff]

That's kind of where I was heading. A leak is a leak.

 

Under positive m'fold pressure, it is a boost leak and you run rich without trims.

 

Under negative pressure, it is a vacuum leak and you run lean without trims

 

But wait. What you just said . . .

 

I was assuming that under open loop the output of the O2 sensor is ignored. Maybe that is true for instantaneous fueling, but it would makes sense to continue monitoring the O2 for input into the learning values. And that would partly explain how I can get negative values for trims with what is 99% of the time a vacuum leak, but 1% of the time a significant boost leak.

 

Thanks, I think this may be starting to make sense.

[birkhoff]

Swapped in a different injector from my stock of old parts. World of difference.

 

Zero misfires now. Idles smooth (within the limitations of engine design) and generally smoother over the whole power range. There was a strange running mode where it seemed like the timing would go seriously late -- like pistons trying to catch up to the crank. That's gone too. This is the first time since I owned the car (well, since I apparently broke it on day 2 of ownership) that the thing has felt like it is running properly.

 

I'm at 1500 miles on the rebuild. I'll pull a log, but for then next 1500, unless something comes up, I think I'll just run it in and consider the job done.

 

Thanks to the members who helped out with this tuning stuff.

[StkmltS]

All right! Time to put some kilometers on that baby!

[xt2005bonbon]

But wait, your trims are still off no?

[birkhoff]

I don't know. How fast should they respond to having fixed the injector problem?

 

Is there a short refresh time, or do they just take a long average since the last ECU reset?

[xt2005bonbon]

I personally view my long term fuel trims live when I drive and I can tell you they can change very quickly. In my case though, fuel trim D takes forever to get adjusted. But the others change a bit every day.

 

You did not reset your ecu did you?

[birkhoff]

 

You did not reset your ecu did you?

 

On your advice, no. I'll pull another LV and post it.

[birkhoff]

Aug 10, 2016 fuel trims look like below. Swapping to a new injector has shifted the trim issue to the lowest load column. Looks like I still have a vacuum leak, although it is not causing any serious running issues at the moment.

 

Next up: do a full pressure test on the intake. I could have something going on at the manifold to tumbler assembly gaskets, or upstream of that since none of those items were touched at the time of rebuild. The tumbler to head intake gaskets are new and properly torqued.

 

I have a setup here now that I can use pressurize from the turbo intake tube (intake box end) all the way to the cylinders. What should be disconnected from the manifold in order to not overly pressurize the crankcase or otherwise leak, just to be safe?

 

My primitive test was to block the intake snout and blow by mouth into the BOV pressure line. That test was a pass.

[StkmltS]

As long as you only apply a reasonable amount of pressure to the motor (<18 psi?) I think you shouldn't need to remove anything. Everything post-turbo already regularly sees positive pressure, so if you find a large leak at 18 psi, it may also be a small (hard to find) leak at lower pressures. Because our cars are boosted from the factory we can smoke/pressure test at a higher pressure than what's usually recommended for N/A vehicles.

 

Sent from inner space.