theprf

First: my car is a stock K16 no tune car, 39k miles. When I first got it, 5 years and 4000 miles ago it would easily boost to 0.8 bar on the dash gauge - not spike to 0.8, it would show a steady 0.8 during 2nd & 3rd gear pulls. Then, after one winter's hibernation it would only boost to 0.6 bar on the dash gauge. Circumstances kept me from driving it much, and last year I did a lot of preventative maintenance which took most of the summer.

The maintenance was to pull the engine, weld the 8 coolant fittings, replace (1) every coolant hose in the engine bay; (2) the coolant reservoir; (3) every O-ring and crush washer external to the engine; and (4) the water pump, modify to GT2 clutch actuation including a GT2 PS pump & tank, install 7.2GT2RS intercoolers, and probably more than I care to write about. I also washed the engine and all the parts that were removed before I put it back together.
While that was in progress I tested every vacuum and pressure hose, every check valve, every solenoid valve, both DV's, and they all were fine. I also ziptied every vacuum and pressure hose that wasn't clamped.
When it was back together I did a boost leak check at the turbo outlet to IC hoses, and it held 1 bar for 10 minutes with the air shut off. So it's tight.
However, still only 0.6 bar....

So next I adjusted the wastegate actuators. My dial indicators read inches so bear with me. Spec is 4.0 +/- 0.5mm @ 0.5 bar - that's 0.157 +/- 0.020 inches.
They were both set to move 0.128 inches @ 0.5 bar so I adjusted them to 0.157 inches. Man, the passenger's side is not easy to reach. And for good measure another boost leak check with held 1.0 bar for 10+ minutes.
However, still 0.6 bar...

My next move was to change the "cycle valve" that the DME uses to control boost (the N75 valve? can't remember it's name). For $22 it's a cheap part and seems to be a likely candidate. The valve I removed tested fine, so I doubt that will change things. I just finished this and I ran out of time to do a test drive, so the N75 valve test is not complete. I don't think it's going to change though.

There seem to be only the DV's and the DV solenoid ("change over valve") left as suspects. I'd think that if the DV's were leaking that would show on the boost leak test?
I have a Forge F-pipe and a pair of diverters which I intended to install while the engine was out - but it's SO TIGHT in that little space there wasn't any way I could fit the Forge F-pipe along with all the other stuff that is in the same space. Since the original F-pipe was fine I reused it.

Suggestions as to what should I look at next would be appreciated. I do have a Durametric.
Thanks, Aaron

Kevin

Grab a Mityvac and pull vacuum on your DV's.. That will tell you if they are good.

Fire up your Durametric and do a 3rd to 4th gear pull

Measure, RPM, Engine load, requested boost, actual boost, injector duty and N75 Don't add any other sensors or you samples rate per second will suffer.

theprf

Thanks, I will do this. What should I look for with vacuum on the DV's - just to hold vacuum? I have a Mityvac. (I have every tool known to man it seems, but never the right one...)

Is there a how-to on logging with Durametric. My attempts with it were not so successful. And I have been a software developer since the 80's.

I can't wait to get the car out and try this. It would be nice to have those HP's back.

pancing

https://www.6speedonline.com/forums/...fterwards.html

Here you go.

manimal

Thanks. I need to do this, as well.

2fcknfst

Very helpful - totally explains how I screwed it up.

theprf

Well, it took a few tries to get this. Finally had a shot when the traffic diminished. I've attached the trimmed spreadsheet. I believe that I got the inputs although Durametric has other names for some of them. I used the formulas that were very helpfully linked in the above post to compute boost in bar and injector duty cycle - which I believe I even did correctly.
This road is uphill, outside temp was about 72F, ambient pressure 1003 mb. I managed two pulls.

Based on my understanding of actual vs. desired boost, it looks like actual closely matches desired - which seems to rule out any induction leaks or funny business in that direction.

I should mention that this car got all new plugs, coilpacks, and fuel & air filters at the same time as the other preventative maint. That's about 500 miles ago.

theprf

Looking over this again, I plotted DME boost demand vs. boost measured. They follow pretty well once the turbos spool up. At low RPM the DME wants 0.75 bar but the turbos aren't spooled so can only give 0.2 bar. As the turbos spool the DME continually lowers the boost demand so that by 3500 RPM, turbos spooled, DME only wants 0.65 bar. The as RPM rises even less boost is demanded, at 4000 RPM the DME wants 0.6 bar and that's what the turbos are giving.
Injector duty cycle also looks sort of high - assuming I have the proper formula: injection time * RPM / 1200 (looks right for a 6 cylinder engine with 3 injection events per revolution).

The graph shows the first 3rd gear pull. RPM on the bottom axis. I couldn't get libreoffice to plot RPM as an actual axis... Notice the spike when I came off the throttle and the slow drop in boost as the turbos unspool. I'd have expected a quicker drop if the DV's opened on time. Need to check that.



I suspect more datalogging is called for, I'm not exactly sure what is best to log: Lambda values? Ignition angle? IAT?

With these six logged values I got about 10 Hz samples. I've got a fairly new laptop with a SSD.

theprf

The DV's held 26 inches of vacuum on the vacuum inputs, so I'm thinking the diaphragms are good.

Suggestions of where to look next, or what to log?

Kevin

I have attached a Stock log screenshot. You can look at the boost, and how the ECU adjusts it to maintain target "load"
If your engine is making the rated 420HP with timing it will lower boost because it had reached the limiter.
Your load is a tad low.
I'd wind in two turns (clockwise) adjustment to your wastegates.

T Kono

Forgive my sophomoric question.......I am totally new to data logging, etc. The graph above looks to me like the engine is supplying the requested boost, yes? Does that mean we need to look for why it's not asking for more?

TK

theprf

That indeed is my question. When I got this car the dash would show 0.8 bar boost from 4000-6500 in 3rd gear at WOT. Now it only shows 0.6 bar during a WOT pull in 3rd.
I had a thought that there was a boost leak which would decrease the actual boost.
From my graphs we can see that the boost the DME is demanding and the boost the turbos are creating are actually the same. Thus the DME is intentionally reducing boost. And there is no boost leak.

Kevin's explanation goes a long way to solving that. On all drive-by-wire cars the "gas" pedal is really a TORQUE pedal. The driver demands an amount of TORQUE by pressing the pedal and the DME attempts to deliver that TORQUE. (Drive by cable cars, the "gas" pedal is actually an AIR pedal and the carb or fuel injection attempts to match the fuel delivery to the air the driver is admitting.)

The Engine Load parameter must be what the DME is using as a target value. When the engine's load matches what the DME wants the DME lowers boost & fiddles with timing to maintain that. My engine load maxes out at 152%. (100% would be maximum load for a NA engine). The graph Kevin loaded shows 165% max engine load and also shows a boost sag similar to mine.

Why it had different behavior before, I cannot tell.

theprf

The spreadsheet you attached looks very similar to the data that I collected, except you have 165 max engine load and I have 152. I don't know if it's making the full 420HP and I don't have an easy way to check. I presume that if the engine load reaches 165-ish that is the best indication of reaching the rated HP short of a dyno run.

I will adjust the wastegates this morning. Of course the passenger's side is a bit tricky to get to so this might take a while.

Dock

The U.S. specification is 309 kW or 414 SAE HP for a stock 996 Turbo.

theprf

I "tightened" both WG actuators 2 turns (12 flats) and went to do a few more pulls. It's a lot hotter today, around 87F.
Before, the max engine load seen was 152. After, max engine load is 157. I did see 0.7 bar flicker onto the screen for a moment. According to the spreadsheet actual boost got well into the 0.7's a couple times.
Three spreadsheets attached... from three consecutive pulls on the same road. It's uphill followed by a bend and just long enough to touch 6500 RPM in 3rd. I'm sure you all know but these are pretty fast cars.

The sheet that has _05 in it's name: RPM/Load/InjDuty/ActBoost/SetptBoost/N75 (same data as before)
The sheet that has _52 in it's name: RPM/Load/IAT/IgnitionAngle/MAF/Lambda1/Lambda2/ActualBoost
The sheet that has _37 in it's name: the same data as _05 plus CorrFactorForBPCWith IAT and CorrFactorForBPCWithKnock

In the _37 sheet the correction factors stay at 1.0 all the time. This indicates to me there is no issue with knock or charge air temp.
In the _52 sheet my ignition angles are similar, perhaps a bit lower than the stock sheet Kevin loaded. Mass air flow is lower, lambdas are similar.

I can't help but think I'm missing something here. I haven't any idea what it might be.