RomRaider

Got the RomRaider logger working and logged some runs. It appears that I am making much less boost than I should be, is TD pulling boost or trying to add it? It looks like it is adding very little which is weird as manifold pressure is extremely far off from target boost. WGDC doesn't seem to be moving past 70%. I don't think I have any boost leaks but I can check, I literally just put everything back together. It could also be my boost control solenoid, and hopefully not a failing output from the ECU. IAMs are 8 when they should ideally be 16?

Here are 3 full throttle runs, 2nd gear though as I was in the city. Some pulls boost higher than others. Tomorrow I will get some good 3rd gear logs.

These are the parameters I was logging to try and get a holistic view of what is going on. Any suggestions for parameters to omit or include?

A/F Sensor #1 (AFR)
Engine Speed (rpm)
IAM* (raw ecu value)
Manifold Relative Pressure (Direct)* (psi relative)
Mass Airflow (g/s)
Primary Wastegate Duty Cycle (%)
Turbo Dynamics Continuous* (absolute %)
Feedback Knock Correction* (degrees)
CL/OL Fueling* (status)
A/F Correction #1 (Direct)* (%)
Closed Loop Fueling Target* (estimated AFR)
Engine Load (Calculated) (g/rev)
Fine Learning Knock Correction* (degrees)
Ignition Total Timing (degrees)
Intake Air Temperature (F)
Throttle Opening Angle (%)
Target Boost* (psi relative sea level)

agreed - your quite far below target boost.


TD is adding duty in an attempt to meet target boost, you didnt say which rom these logs came from but i'd assume your pegged on whatever value is in the wastegate duty MAX tables.
These tables out of the box should be able to achieve target boost without issue - going by the shape of your boost plot i'd assume the boost solenoid isn't hooked up properly so check your electrical wiring too it and check the plumbing to/from it as most likely something is amiss.


On a standard map there is an initial IAM value defined - usually 8 - so the car doesn't start at maximum timing (incase the fuel quality doesn't support it). However you need to achieve a certain amount of engine load before the rough correction can take over to increase the IAM value to 16 and give you maximum timing.

Check your solenoid setup, check for boost leaks and then log again in 3rd gear and see what's what.

I pulled the boost control solenoid out and verified its operation on the bench. It opens and closes perfectly fine with 12V, I blew through it while doing this to confirm it opens fully.

I could not verify operation in the car as test mode does not exercise the boost control solenoid. I have confirmed this through some research as well. Connector is hot with +12 on one side and it is floating on the other pin, I imagine the ECU grounds this pin to open the valve.

I instead hooked the boost lines up as though the BCS was at 100% duty cycle, I connected the two lines that go into and out of the BCS together. Then I took the car on a drive to see if it would hit full boost. I had tried this in the past and it overboosted massively. This time it topped out exactly where it topped out with the BCS in place.

I am going to do a thorough leak testing of exhaust and intake, but my guess is that my 285,000 mile turbo is done. I have another TD04 I could swap in to check to see if it is the turbo's fault.

I did some 3rd gear logs today but without making full boost I don't know how useful they are so I will hold off on posting them.

Also I went into the ROM and checked what the max WGDC is, it is where it should be, though I guess I am surprised that TD isn't increasing the value? I would imagine that if the WGDC is sitting at 69.9, but boost is well below the 13.61psi specified that TD would add to the WGDC to try and increase boost. This is a screenshot from my ROM.



My guess (and hope) though is that the turbo is tired. I had the oil feel line go on me and I didn't realize it immediately. I have since upgraded to a braided line, but maybe it was too late for the turbo.

Was looking at new CHRAs to rebuild my spare turbo to throw in, this one actually seems really nice, made right in UK at their location https://www.ebay.com/itm/322419478352#rwid

But I wanted to test a few more things first just for sanity's sake before spending the money.

I disconnected the line that goes directly to the wastegate this time, (plugged the line) boom instant boost, hit 17 PSI really quickly before I let off. So looks like no boost leaks, no exhaust leaks, and no broken turbo, that's great.

I then unplugged lines in various combinations, but the most telling of all was unplugging the line from the wastegate that goes to the BCS. So the WG was plugged into the T, which was connected to the turbo boost nipple, and the T was also connected to the line going to the BCS, but the BCS was not attached to the line. It only made 9 pounds of boost again. So I think I have found the culprit - a dirty T and line.

I will report back if everything is working after I clean the bits. Still a chance the BCS is not working properly, no way to know until it's all back together.

Full boost is back

After cleaning out the vacuum lines and T with some wire and isopropyl, I configured it so that the WG goes to the perpendicular portion of the T, as there is no actual flow to and from the diaphragm (very very little) and there was no gunk in the line going to the WG. Originally I had the perpendicular portion of the T going to the BCS, but I think that lends itself to getting clogged over time as the edge of the 90* junction picks up gunk.

Went for a test run and it hits 14 psi in 3rd no problem, spools way faster too. Will get some new 3rd gear logs posted and start to log for MAF scaling.

Really ended up being a best case scenario, nothing was broken, just dirty!

make sure you don't lose the restriction pill when cleaning with a bit of wire as losing that will cost you lots of boost.

Yes, I have replaced all those line with silicone and the pill is very securely held in there.

Here is a clean csv of a 3rd gear run, my IAM is reading 16 now which is cool and surprising since my MAF needs scaling. Unless I am misunderstanding IAM.

Car feels great right now. Planning to do a MAF scaling run later today.

Started recording MAF data, and driving in constant closed loop I realized that my idle is hunting, and during constant low throttle cruising the speed fluctuates the same way the idle hunts. I can hear the Idle valve open every time that the revs then increase. I replaced the vacuum line from the fuel rail that goes to the evap stuff at the tank, and thought that maybe I introduced a leak there, but pinching the line closed had no affect on the hunting idle.

In regards to the MAF data I was collecting, I can see that the MAF values are constantly bouncing with this idle/cruise hunting. You can see from the screen shot that the dense clusters during driving are spread out above and below the interpolated line. When idling the green dot was bouncing from the top to the bottom constantly. I have to do some investigating to see if I can find out why this is happening. Going to log at idle and see what I can see.

I did not apply any changes to my MAF scaling because I figured that this phenomenon would throw it off. Any suggestions welcome.

Did a log at idle and my AF correction maxes out over and over as the RPMs fluctuate. The idle speed is definitely changing because of the Idle Valve, as I can watch the duty cycle change along with the MAF reading.

The question is whether the O2 sensor is bad and is making it seem overly lean which then the injectors dump in a bunch of fuel and then the revs drop so the idle valve opens up and it starts all over.

Or whether it is actually going lean because I have a vacuum leak and this hysteresis is the result of that. I will definitely be investigating this option first, as it is mostly free. I did remove the majority of vacuum lines when I rebuilt the motor so it could definitely have introduced a minor leak.

Attached is CSV and image is the plotted data, I divided RPM by 10 so it would be in the same-ish scale.

This ECU business is fun

Been doing more reading and more logging, I think that my fluctuations are occurring between -20 and -18inHg, whether idling or cruising at low throttle inputs like when scaling for MAF. Still have not had a chance to check for intake leaks but I will be doing that soon. I am very tempted to buy another O2 sensor also to make sure that is not the issue as the one I have no was shipped in an envelope with no box and I worry it got damaged. Not really a waste of money to have a spare O2 sensor. I should replace the rear O2 also as I never have. If it is not the O2 sensor and there are no leaks, then I will start playing with AFR Corrections table/limits.

My 3port BCS came in and I am going to hook that up.

My understanding is that I will drop all the WGDC by 30% (multiply values by 0.7) including the Turbo dynamics WGDC adjustments.

Car is running really great in open loop, and real snappy in closed loop asides from the idle, low throttle input cruise, and sometimes a casual 1-2 shift has a little bit of a stumble and lurch (probably because vacuum ends up around where it is for idle and low throttle cruise).

what happens if you try with the maf or o2 sensor unplugged, also if your not running the standard air box / maf tube setup then you'll definitely need to do some MAF scaling to get it running nice. Unplugging the maf will put the ECU into limp mode but because its based on MAP pressure then it's easier to see if you have a leak anywhere on the intake side that's messing with things. Unplugging the o2 instead helps rule out if the sensor feedback is making things worse. Use your wideband to see what the true AFR is then and you can make some tweaks to your MAF scale based on the wideband readings.

You've not got your wideband on the logs, make sure to get it added in so you can make accurate changes to your tune.

I will try both of those things in addition to leak testing the intake. I am rewiring in my wideband controller with a dedicated relay and mini fuse box with power and ground directly from the battery. I have new the wiring run, just need to get a relay. I have it running pretty rich for safety's sake until I get the wideband in.

I have spent probably over 30 hours logging, graphing, tuning, repeat. My main focus is boost control, I was making good boost, but I would overboost when it got cold at night, and I would overboost if I stabbed the throttle on the highway in 4th or 5th. I have played around a lot with the turbo dynamics to prevent boost fluctuation, and to control the ramp of boost as it gets close to target boost. I have found around 3500rpms the turbo really wants to spool, so I have pulled WGDC from this area and it results in a really flat boost curve. I am still a bit under my target boost, but I will be able to finesse the WGDC until it gets there. It is fun to spool really hard all the way to target boost, but I think it is much easier to control overboosting by having it slightly undershoot target boost. From a performance drivability standpoint it feels really good. I am much less concerned with max boost as early as possible, and much more concerned with being able to control the car in a corner with the throttle, and if you aren't starting at 2500rpms, it spools really linearly.

Here is a log from a 5th gear passing situation simulation, boost is very well controlled. Key is at bottom.

Recently forced the car into permanent open loop via tuning and there is 0 idling or cruising issues. I bought a new O2 sensor and will be installing it.

I also got around to remaking my gauge setup. Originally I had them zip tied where the clock would be and said it was just temporary, and then years went by and I never re-did it. I got a cluster surround gauge pod because I miss my clock. Also I just tapped into random power for stuff and I wanted to do it cleaner so I have a dedicated ground and power cable coming off the battery. Now I have a fuse block with a bunch of hot outputs that turns on with ACC, I can use these for other gauges, or for my radar detector too. I used a fuse tap in the #11 fuse to turn the relay on. Here is a picture of my setup on the bench, the plexiglass mounting plate is temporary unless it fits somewhere well. I just wanted everything contained. If I did it again I would have used a fuse block that accepted ATM fuses to keep everything physically smaller. The main power cable is fused to 25 amps but the 12 gauge wire should be good for over 50, and each of the individual circuits are fused to 5 amps. Using an M6 bolt as a ground lug, wish I had a spade terminal setup though for easy install and removal of wiring.



Power and ground lines not shown here as they are in car, orange line goes to fuse tap

Got my gauges installed and calibrated the LC-2 wideband. My LC-2 would not calibrate, continually blinking green which meant it was heating up. I could feel the O2 sensor and it was pretty hot. After some research I hooked a battery charger up to the battery and tried again and it calibrated in under a minute. So if your battery is below 13v and your LC-2 won't calibrate, that is why.



Logging my wideband it seems that the car is running really lean at idle, idling around 17:1. When cruising it is around 15.5:1. I am still running forced open loop. Under load the AFRs are what they should be, 10:1 and dips as low as 9.5:1. If you go back to my MAF scaling plot, you can see how it is a little lean at cruise but quite lean at idle. This led me to think that it was definitely a vacuum leak to the intake manifold which would manifest itself when the car is pulling vacuum when idling or cruising. If it was a leak post MAF but pre intake manifold, I would imagine it would make it leaner everywhere. My friend made me a smoke tester for Xmas so I put it to use.



I had it set to a few PSI and when I would open the oil cap it would clearly release pressure and smoke would billow out. The only leak I found was between BPV hose and the turbo inlet. Tightened it up and then checked again and no leaks anywhere. I started the car up and the idle was fluctuating while it was warming up. I unplugged the Idle Valve and while it stopped fluctuating as badly, it would still fluctuate a little. I plugged it back in and unplugged the MAF and it idled perfectly and right around 14.7:1. Normally while in open loop it does not fluctuate at idle, so I am not sure why it was doing it now, possibly in response to just having everything full of smoke.

But unplugging the MAF and having it idle where it should be means that the MAF is either reading correctly and there is a leak at idle, or the MAF is going bad and reading poorly at the low end.

I remembered that while I did replace my MAF, I think I did so with a used unit. I am just confused because my car had no issues with idle or cruise until I rebuilt the motor, so I would be surprised if I had a sudden sensor failure. I did clean the MAF with MAF cleaner before reinstalling it, possibly I damaged it during this process though I clean it regularly.

I think I am going to do two things. One is retest it with the smoke tester but at night to better see the smoke with a flashlight, and two is do a boost leak test and spray everything with soapy water and see if I get any leaks from my TGV deletes, TGV to manifold, or manifold to heads as I think those are the most likely culprits if something is leaking that does not show up under a low pressure smoke testing.

If both of those turn up no results, then I will re-clean the MAF as well as maybe swap it out with a spare I think I have to see if that changes anything.

Any other advice is welcome, I just want this resolved so I can continue on tuning.

Bit the bullet and bought a new Denso MAF for the WRX, I figured even if it was not the issue that it would be a good idea to have a new MAF to make sure it is accurate, as my old one has over 100k miles on it for sure.

The MAF helped a lot, the car is tuned to right around 14.6AFR for cruise to allow me to force open loop, and it nails that while driving around and on the highway. It still goes lean during engine braking which I believe is normal behavior but correct me if I am wrong. Idle is still a little lean, idling around 15.5:1. I don't think I have any leaks, I could definitely see that idle AFR being caused by the COBB SF intake not lining up with the stock MAF scaling. If it is normal to idle lean, let me know, but I didn't think so.

I am going to stop forcing the car into open loop and see if it runs better or worse with the O2 sensor. Then I am going to replace the O2 sensor with the new one I bought as I suspect the O2 sensor is bad. My theory is that the current O2 sensor will still make it fluctuate at idle and cruise, and then the new one will fix the issue.

At that point I feel confident to rescale my MAF now that I have a working MAF, working O2, and no leaks.

We will see if my theory is true. I will make some logs today with the LC-2 output included.

Car is running better than it has in a long time right now, boost is only at like 16 psi and I am close to maxing out the injectors. Once I get my MAF scaling dialed in I will replace the injectors with the pinks and scale for those, then I can turn the boost up some more safely. I will install the 3 port, and nail down the fueling table. Mine still hits 9.5:1. Then I will play with the timing table and getting that nailed down, currently my knocking at WOT is very minimal, but I have some knock at partial throttle enough to the point where there is fine learning applied. I have to take lots of logs trying to replicate the scenarios so I know what cells to adjust.

I will definitely say though that with the numbers I see on the wideband with the stock MAF scaling applied right now, the SF intake is not too far off, none of the AFRs are unacceptable. I would think it could be run on at least my stock untuned 02 WRX without issue (how the car was for 6 years).