73 replies to this topic

[duncanm]

probably gunked up with oil -- isn't this where blowby usually collects in the TMIC -- try giving it a good clean.

[neggles]

I cleaned it out and that just made it move to halfway even more quickly - I think the return spring must have been permanently compressed. The actual valve moves very freely, the actuator just doesn't push it open when you remove vacuum like it should.

 

SPRV is functional, assuming that it should behave like a regular BOV (open with vacuum, closed with pressure). No cracks in anything that I can detect.

At this stage I'm assuming the late changeover is just a tuning issue or another issue with ECV actuation in positive pressure mode.

 

BBOD has been modified slightly to apply boost pressure to the ICV instead of just atmo (swapped the tiny little filter on the solenoid for a line to a manifold tee), which should function as a workaround until I can get a replacement ICV.

[duncanm]

you could get away with blocking the SPRV boost bypass for diagnosis -- at least to help rule it in our out.

[neggles]

EDIT: OH GOD NEVER MIND I'M AN IDIOT AND I FORGOT TO REATTACH THE VACUUM RESERVOIR

It works much better when it doesn't have a catastrophic vacuum leak and no ability to control half the valves.

 

Holy f%&$ing s#!t this thing is FAST in twin turbo mode with a functional ICV. IT MADE 20PSI OF BOOST AND BLEW THE OUTLET HOSE OFF THE SECONDARY. I need a new hose clamp for the secondary outlet.

 

For future reference, you can totally replace the little air filter on the ICV solenoid with a boost line if you have a similar problem to me and your ICV doesn't open the whole way without a little bit of help.

 

Right! We've sorted the weird behavior when it stages into twin turbo mode. Now it's just the "why are you still not staging until 6000rpm" problem, which I think is going to be down to ECV duty mapping or ECV rod length. I've confirmed that both solenoids are functional.

 

I was very confused for a few hours and it was all my own stupid fault.

[duncanm]

might be worth getting in touch with Matt (kido tune) to help sort out tune problems or questions -- didn't he do sbv's revA-C TT outback with eculabs ? (link to build thread)
 
 
Good to hear your making progress -- it certainly sounds like fundamental problems -- not tuning tweaks.

[LukeFranky]

I'm just wondering if the mod you made to add boost to the ICV solenoid is one worth doing regardless. It seems like good idea to help the valve open more forcefully.

Thoughts?

[neggles]

The only thing I can think of is that the diaphragm might not be super happy about having boost pressure behind it (since it's not designed for that) which means it might cause damage to the diaphragm over time, so I probably wouldn't do it to a still-functional valve.

 

I've tried to contact Matt from KiDo multiple times and never received a reply to my emails or SMSes, unfortunately. He must be rather busy. Also, it appears the revD uses a somewhat different approach to mapping to the A-C, though I'm sure he'd be able to figure it out.

 

The most recent map I have suggests that the car just isn't dumping enough gas into the secondary to keep up with the primary's boost output during changeover. One way or another, it's an issue with ECV +ve control, and all the solenoids/lines are operating correctly, so it's got to be either the actuator (unlikely, -ve solenoid still works) or the mapping. I also haven't had a chance to properly wind it out from single mode in 2nd or 3rd, it's possible it just didn't sequence 'till 6000 in first because it straight up didn't have enough time for the secondary to catch up before shifting.

[duncanm]

yeh - I wouldn't be trying to diagnose anything in 1st..

Mine will switch over pretty obviously in 2nd and above. It's pretty obvious if you boot it in 2nd to about 5k, then change to third into the VOD. Anything over 1/2 throttle will do it.

[neggles]

Took the car for another test drive and tried out a few things.

 

You make a good point, I'll summarize my findings thus far (and in doing so hopefully explain why I think it's the ECU, not a mechanical issue) - I will admit that I am somewhat biased towards electrical issues rather than mechanical as fixing electrical problems is my specialty.

 

First of all, here's a reference image of how the lines are set up on my BBOD (looking at the BBOD end-on);

 

 

Alright, so;

• ICV was not opening fully. This has been resolved with a workaround and I have verified that in test mode it now moves fully open/closed if I supply some pressure to the relevant line for it to operate with.


• BBOD has been checked and cleaned. All solenoids direct air through the relevant ports in their respective energized/de-energized states.


• All vacuum lines have been checked more times than I can count, and correspond with the diagram above (internally they correspond with the service manual and my spare BBOD, save for my minor modification to supply boost to the top of the ICV solenoid instead of atmo


• ECU does not command ICV open until 6000rpm, even if DPS is reporting less than 2psi difference.


• I do not currently have a stock map loaded.

 

Map details:

• The ignition and fueling maps are from Got It Rex. I can't use the ones that came with the car as it had stock injectors and now has ID1000s, and I don't trust myself enough to get the fueling right.


• IAT is a potential issue as the IAT sensor is now located in the intercooler end-tank instead of at the MAF, but the IATs I've been seeing on the logs are reasonable (mid-20s) so I'm not super worried there.


• The boost target maps are what was loaded into the car when I got it; I don't think these are stock as the boost target gets up to 2 bar on both maps!


• The primary wastegate duty cycle maps are mostly just set to be nice and high; the car mostly ignores these when it starts prespooling at 2500ish anyway, using the ECV to control boost.


• Secondary duty cycle (aka ECV +ve duty) is set to what came with the car.

 

Again, I'm unsure that the ECU hadn't been touched prior to my taking ownership of this car, so if anyone has a stock ECUlabs map for an MY02 B4 (ECU P/N 22611AG850) please let me know (or PM me a link if you would be so kind).

 

If I set the secondary duty cycle to 0% (commands the ECV to maximum opening in positive pressure mode, the logic is reversed to what you'd expect) manually, the car makes significantly less boost off the primary (as the exhaust gas is all going through the secondary) - around 4-6psi instead of the target - but will happily report that the secondary is making the same or slightly more boost than the primary all the way from 4000 to 6000 where it decides that it's time to open the ICV, even though it should have been time to open the ICV quite a bit sooner based off DPS readings.

 

I suspect at this point my weirdly high boost pressure targets are confusing the car (as I've not got enough time/road space to properly tune the secondary duty map) - with the manifold pressure at only 6psi, but the pressure target at that RPM for twin turbo mode at 20psi, I could see how that might confuse it. The 6k rpm ICV opening definitely seems to be some sort of failsafe.

 

It would be nice if I could see the ECU's list of conditions that need to be met to open the ICV, so that I knew what data to look at and what maps to tweak to make it happier.

 

Edit:

 

Additional info!

 

Here's a log trace showing the 0% secondary duty cycle run (click for full-rez). As you can see, it's not making at lot of boost, but the secondary does keep up, until 6k RPM where it finally decides to open the ICV and then it's SUDDENLY, BOOST! and before you know it you're at 7k.

 

And here's a link to the log file if you'd care to peruse it yourself; https://www.dropbox....145028.csv?dl=0

 

I should also note that if I restore the came-with-car secondary duty map it makes 1.5bar of boost on the primary all the way up to around 4500, but it takes until 6000 for the secondary to catch up; I might try a run with the as-delivered map reduced by 10 or 15% across the board to aid the secondary in spooling.

Edited by neggles, 17 July 2016 - 03:45 PM.

[neggles]

It does, yes - the yellow line is the DPS measurement, the pale green line is the MAP measurement. As in that screenshot, at 4483 rpm the primary was making just under 4psi of boost and the DPS shows 0, indicating the secondary was also making just under 4psi of boost. Something other than the DPS prevented the ECU from staging until 6000, where the orange line (ECV +ve) goes to 0 and the IAC output (bright green line) goes to 1.

 

I suspect in this particular log's case the car wasn't happy that the MAP (4psi above atmo) wasn't anywhere near the boost target for that RPM, but I have no data to back that up.

 

I see the same thing on regular duty cycle runs, however the DPS doesn't get above -4 psi (with a manifold pressure of around 14psi) until after the IAC opens and the ECV switches to wide-open mode. That still indicates 10psi of boost coming from the secondary at that point...

 

Uh... Do I have the ECV lines backwards? Positive pressure should go to the top of the ECV, yes? Or is that backwards on the revD?

E: I can't really have them backwards as if I did then it'd never open, right?

[Jimbo]

Are you sure the DPS is plumbed correctly? 

Spot on with the ECV lines, +ive to the top and -ive at the bottom.

To me it sounds like a sensor or solenoid isn't "doing". Like you have said, something is preventing the ICV from opening until late which means something isn't getting the signal it's supposed to be getting, thus delaying the whole shebang.

This leads me to believe the DPS is the culprit as it is supposed to pull the trigger when the conditions are met. 

If you can, post a pic of your engine bay, one of us might spot something.

Edit, if you had the ECV lines backwards it'd never open as it would be applying vacuum to the top port, plus with the spring tension and +ive pressure under the diaphragm, it would well and truly stay shut!

[duncanm]

Do you have access to the DPS ('open the IACV') voltage threshold in the map ? What's it set to?

Also, I'm not sure the seconday should be 'keeping up' - the SPRV should be open and keeping boost low, unless the IACV is leaking alot when shut (which would explain both the low boost and 'equal' pressure in primary and secondary)

[neggles]

Alright, cool. I was briefly very confused. Will adjust ECV rod length now.

 

DPS is definitely working properly. I've taken it out of the car and tested it manually, plus it correctly reported my ICV-not-quite-opening problem.

 

To me it sounds like a sensor or solenoid isn't "doing". Like you have said, something is preventing the ICV from opening until late which means something isn't getting the signal it's supposed to be getting, thus delaying the whole shebang.

 

Wholeheartedly agree, except IMO it's got to be something sensor-side or mapping-side as the ECU isn't commanding ICV opening; if the ECU was commanding open and it wasn't opening (which was kind of happening before, but isn't now) then I'd see that on the logs, but it just never commands opening until 6k, so there's some sensor or something that it's not liking the input from.

 

It better not be that bloody hidden P0037 code.

 

Edit:

 

Do you have access to the DPS ('open the IACV') voltage threshold in the map ? What's it set to?

 

I don't think so. Emmanuel said that not all of the maps available in the ECU are broken out in ECULabs Studio as he's not sure what the remaining ones do - I've asked for access to more maps for experimenting. However, it should be stock, which iirc waits for the DPS to be above -2psi.

 

I've actually never seen the ECU command the SPRV open (SPRV solenoid #1) on any log, ever. It only ever energizes SPRV solenoid #2, which is the positive pressure solenoid.

The SPRV spends the entire time with its control line connected to atmo pressure, until the ICV opens at 6k rpm and it energizes solenoid 2.

Edited by neggles, 17 July 2016 - 04:39 PM.

[Jimbo]

I thought the same thing and possibly a blocked line filter also. But from memory the DPS throws a code if the lines are wrong.

I had it plumbed incorectly/line fall off it and it never threw a code. But, being AC vs DE, the JECS ECU is probably alot smarter than my tamagotchi based ECU..