CovertRussian's 05 LGT Build Thread

I would think if turbo seals were shot, it would eat oil constantly. I think the oil sheering and becoming thinner is the best explanation on why it only starts to eat oil after X miles.

In my case, I did oil change, no consumption for 500 miles, drove ~500 miles highway (could different trips) it consumed about 16 (I topped it off at 4-8 oz increments weekly). I now drove about 500 miles since then, city only, and it hasn't consumed any at all. Thus the added miles increasing consumption doesn't apply it seems.

Also my city drive is fully of hills, I do a lot of deceleration, and consumption is still good.

Are you running a stock or modified (in some way) PCV system? It may be you're getting more consistent blowby at highway speeds pushing oil out. Which could be due to age, or YNANSB.

I will say coming form a car that used to consistently eat oil that a new block, valve guides and turbo will fix it for sure

PCV is slightly modified, actually going to an aftermarket inlet reduced consumption, since it moved the PCV port from right off the turbo to further up the pipe (lower pressure drop further away from turbo, thus creates less vacuum).

I need to try out 10w30 oil from Valvoline, it seems to have much lower evaporative properties.

I do need to get under the car and see if there is another oil leak that I'm missing.

Got an update to the oil consumption and a riddle: For last two months I haven't done a highway trip and my oil consumption went back down to pretty much none!

 

Thus it seems like oil consumption is happening mostly at steady state highway driving (~3,000 rpm). Could it be an oil seal leak that only opens at that RPM's oil pressure? I would think piston rings would seal better at that RPM.

It's possible that you do have a seal issue - that's seeping more under higher engine RPM pressures. Post up what you find from your under the car exploration.

Another data point, I did about 500 miles city, not a drop of oil was consumed. Then this weekend went on a highway MPG test, 160miles, ate 8oz.

Highway MPG test was for testing LED's, I'm still working on getting enough data for city tests, but so far looks promising!

Been thinking about this a little more and finally looked it up. It seems like the reason city level doesn't drop is due to fuel/water getting trapped in oil and increasing the level. But when you drive on the highway, the prolonged warm engine will burn off the fuel/water, which makes it seem like highway driving is eating oil.

https://forums.nasioc.com/forums/showthread.php?t=1101570

Typically you will notice the most consumption after doing a highway trip after a lot of city driving which involves a lot of short trips.

 

You will notice it more because city driving with short trips will let the oil accumulate fuel/water contamination which gives the appearance of full oil levels, when it's actually being offset by the fuel/water in the sump.

 

Do an extended high speed run and you'll notice the oil level drop quickly as the fuel/water gets burned off, but then steady off which gives the appearance of highway driving burning more oil.

 

All else being equal, legal freeway cruising will burn off less oil per mile than city driving because the engine had time to fully warm up and is running under light load. City driving will burn more oil per mile because you do a lot more acceleration.

BITOG thread mentioning the same thing:

https://www.bobistheoilguy.com/forums/ubbthreads.php/topics/1019951/Re:_Car_burns_more_oil_during_

Overall my PCV system should be fairly clean, I cleaned up a lot of pipes last year, when building the inlet. Guess I'll have to do the PCV catch can setup that UTC_Pyro mentioned sooner then later .

You guys might remember my MAF connector contact issue from about two years ago, just like with my OBD2 port, repeated unplugging simply wore it out. Well it happened again, one day, the car just started being pig rich. It happened about a week or two after I hand to unplug it to replace the headlight/fog light bulbs.

Now, when I first did the MAF connector fix, I didn't understand how they worked so I probably made it worse. I was planning on doing a new connector eventually, but this failure happened spontaneously and I didn't have the parts on hand yet. Also, this time around I was armed with knowledge of how Subaru pins work, thanks to the OBD2 connector fix that I did a few months back.

Here's the problem child, this is the new MAF that I got in 2017

Even though I pull one wire at a time, still a good idea to have a picture of wire color and their respective locations.

In order to take the female receptacles out, the little white tray needs to be pulled out, just don't pull it all the way out.

Inside the connector you'll see a little plastic tab that you need to push up or down and it will free up the female receptacle (sorry didn't get a pic of it). Once removed, it will look like this:

The way this receptacle works is, the metal wraps from the front of the receptacle halfway back. It works like a flat spring, where the inside should constantly push up against the MAF's male pins. Overtime this flat spring compresses too much, resulting in a loose fitting male pin.

You can somewhat see this on the before (left image) and after (right image) here:

Inside the female receptacle, you should be able to push the flat spring up(down in the picture since the receptacle is upside down), which will reduce the receptacle's overall gap.

Here is the before and after again, sorry for the after being kind of fuzzy. You can see slightly smaller gaps in the receptacles. When I went to install it, it refused to go in at first! It really needed some force to get it back onto the MAF!

Technical Details

Unfortunately I didn't get a before and after log, since I accidentally fixed it by touching the MAF connector (which temporarily re-established a connection and car ran better already). But as I mentioned earlier, the loose connections made the ECU run pig rich when going into slight boost/open loop. I did see the short and long term fuel trims being maxed out in the negative (removing fuel).

This is interesting because the loose connection was causing the MAF to read a higher higher load. Basically, the MAF made the ECU think that I'm running 6psi when in reality I was only running 1psi. This causes enriched fuel and reduced timing, which both would lead to rich fuel and sluggish performance.

utc_pyro has a more technical description in this post.

Surprising Aftermath

I was in a middle of an MPG test (for headlights) when this happened. While I'm mad that it created another variable, nullifying weeks of headlight testing, the MPG benefits that this made made up for it!

Before the fix I saw the following city MPG:

20.51mpg (Gauge: 21.3)

20.78mpg (Gauge: 21.3) * Failure and fix happened about halfway through this run

After the contact fix my MPG instantly jumped to:

21.44mpg (Gauge: 22.5)

21.38mpg (Gauge: 22.2)

Real world gains (Averaged): 0.765mpg or 3.71%

Also, it's nice to see that gauge MPG saw the difference too, which also makes sense since it's based on mass airflow readings.

Thank you for this post... I’ve been fighting the same issue for the last few weeks, and actually just ordered a new MAF connector and enough shielded wired to replace thing all the way back to the ECU. This seems faster than redoing everything, may try this first.

Heck yeah, I did this over my lunch break, so shouldn't take long at all.

Ok I found my learning view log screenshot. A, B, C was maxed out at -15%. This is ECU removing fuel, which means the engine load is higher when the MAF connection is loose. This also means the MAF is reading higher voltages too, I'm still brainstorming on why it would read higher voltages in this case...

Ok I found my learning view log screenshot. A, B, C was maxed out at -15%. This is ECU removing fuel, which means the engine load is higher when the MAF connection is loose. This also means the MAF is reading higher voltages too, I'm still brainstorming on why it would read higher voltages in this case...

That’s exactly want mine was doing, but at -12.x%. Leading theory is high ground resistance, creating a fairly consistent voltage to offset.. The maf’s Output signal generator (opamp or digital electronics) will be referenced to it’s own ground and thinks everything is fine. But it’s ground is maybe 1/4 Volt higher than the ECU ground (which the ECU uses as reference), creating a bad reading. Compounding this the power draw of heater circuit in the MAF is proportional to the output voltage and thus the ground voltage offset through the ground resistance will rise causing the final scaled difference to the ECU be close to a fixed percentage.

Guess this is why some domestic manufacturers use PWM signals from their maf’s vs the analog signals most Japanese and Euro manufacturers use. As long as the signal is still making it and the MAF input voltage is high through for it to operate these sort of issues don’t matter.

That’s exactly want mine was doing, but at -12.x%. Leading theory is high ground resistance, creating a fairly consistent voltage to offset.. The maf’s Output signal generator (opamp or digital electronics) will be referenced to it’s own ground and thinks everything is fine. But it’s ground is maybe 1/4 Volt higher than the ECU ground (which it uses as reference), creating a bad reading. Compounding this the power draw of heater circuit in the MAF is proportional to the output voltage and thus the ground voltage offset will be close to a fixed percentage.

 

Guess this is why some domestic manufacturers use PWM signals from their maf’s vs the analog signals most Japanese and Euro manufacturers use. As long as the signal is still making it and the MAF input voltage is high through for it to operate these sort of issues don’t matter.

Thank you very much for this, I definitely wasn't considering the ground. I was thinking it was related to hot wire MAF design; where the hot wire wasn't being heated as much, thus ECU thought that it's being overcooled, as it would be with higher airflow, thus seeing higher load.

Also I've never seen such MAF funkyness with 1990's Nissan ECU's. Both are similar design (hot wire), but for some reason Nissan MAF's were much more resilient to airflow and intake changes. MAF V wouldn't jump down, like it does on Subaru, when I would install a cone filter instead of stock airbox.

Seems like maybe something I should do to (all of) the connectors in my engine bay to try to address my misfire.

Definitely! I'm thinking about going through and re-crimping all of the injector and spark plug pins too, fortunately those don't get unplugged nearly as often so they don't seem to fatigue as quickly.

Been meaning to do this for a while, but my other car was occupying this sensor. Once I sold it I finally could use it on the Legacy.

 

First, the easiest part if your downpipe has a bunch already...

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/CIMG1252.jpg~original

 

UEGO has two main wires, one that comes from the O2 sensor and second one is the power delivery + serial out put wire (which sadly you have to wire in yourself). I did this a long time ago so will dig up the steps on how to add serial cable later.

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/CIMG1253.jpg~original

 

I like to use quick disconnects, especially when I tend to move sensors around different cars often. I also like using the fuse splitters, got sick and tired of splicing and re-splicing in car wires.

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/CIMG1258.jpg~original

 

Next thing I like to do is use different ends for power and ground wires, in this case I ended up using different plugs. This will help you avoid plugging in a positive wire into a negative wire (note sensor end is show that's why you see male and female connections).

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/CIMG1283.jpg~original

 

For ground wire I like to use ECU ground, it keeps readings consistent with what the ECU would see.

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/CIMG1289.jpg~original

 

For power I tapped into the least useful fuse I could find, 10A LAMP IG

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/CIMG1295.jpg~original

 

UEGO comes with a bunch of wires, have to tuck it under the radio

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/CIMG1282.jpg~original

 

Wired in the gauge and tested it before putting it all together

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/CIMG1285.jpg~original

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/CIMG1296.jpg~original

 

Heat shield fits well with the wideband in place

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/CIMG1340.jpg~original

 

Now here is why you want a wideband over the stock sensor. Stock sensor doesn't read below 11 AFR, wideband is showing it gets as rich as 10.5:1 afr!

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Interior/Gauges/Wideband/WBOStockvsUEGO.png~original

what was the color wire that you tapped in for .5v on the 02 for signal. looking at the FSM I'm seeing white/green.

Update: It looks like the 2008 WRX radiator (Denso 221-9235) is the same dimensions as the STI radiator, but also has the bigger turbo tank drain, which would save you some headaches...

I've been just daily driving the Legacy testing the LED's still. This summer we had a couple hot weeks (at ~95F) which started taking a toll on the cooling system, even though water temps themselves would not get above 205*F.

I would smell coolant on those hotter days, until finally it started gushing from the turbo tank drain line. I had to cut the line slightly shorter and put a new clamp on it:

A couple weeks go by and I smell coolant again! This time I found a crack in the upper radiator, not far from the previous leak. Not wanting to replace the radiator just yet, I JBwelded it...

Another couple weeks go by, I decide to try new radiator caps out and all the sudden I got a major coolant leak. Now I'm worried that something is wrong with the thermostat/waterpump and is over-pressuring the system, which would mean a new radiator would crack too. But I think the real reason for the failure to properly bleed the system, since I was trying to reduce idle times (since I was doing MPG tests heh).

Either way now it's time for a new radiator... I went with 2008 STI Denso radiator (Part 221-9236). Specs made it sound like it would be slightly thicker/wider, plus didn't have AT cooler like all the LGT specific ones did.

Started removing the stock radiator:

Radiator had to come out with fans, since the bolts were a bit too rusted on. Overall real easy on the Legacy due to cavernous space:

2008 STI Denso 221-9236 vs Stock 2005 LGT specs:

2005 LGT (Denso 221-3605): link

Core Thickness: 0.6250 in.

Core Height: 13.3750 in.

Core Width: 27.0620 in.

Inlet Diameter: 1.5620 in.

Outlet Diameter: 1.5620 in.

2008 STI (Denso 221-9236): link

Core Thickness: 0.6300 in.

Core Height: 13.3900 in.

Core Width: 27.8000 in.

Inlet Diameter: 1.5630 in.

Outlet Diameter: 1.5700 in.

Interestingly enough the WRX one is identical to the STI AND it has the same sized turbo tank drain. If I did it again, I would go with the WRX model!

2008 WRX (Denso 221-9235): link

Core Thickness: 0.6300 in.

Core Height: 13.3900 in.

Core Width: 27.8000 in.

Inlet Diameter: 1.5630 in.

Outlet Diameter: 1.5700 in.

Even though the specs say they should be bigger, they look pretty identical to me:

Unfortunately the same width too:

Same length too:

New radiator fit right in:

The lower hose fit perfectly:

Upper fit perfectly too:

First real difference/issue is the turbo tank drain, the STI is MUCH smaller, which required a downsizing adapter to work:

The radiator coolant cap overflow was really loose too, this was easily fixed by cutting the line a tad shorter

All done:

There is now a gap between radiator and rad support, since the rubber gasket was not reusable. I'm not sure it makes that big of a difference though.

Just in case, I did reinstall the older radiator caps. I'll take a look at the waterpump/thermostat later when I go to install a warmer thermostat (another MPG test idea heh). Meanwhile I reused the previous coolant, since it was fairly new anyway.

There is now a gap between radiator and rad support, since the rubber gasket was not reusable. I'm not sure it makes that big of a difference though.

On a street car, maybe not much difference. On track cars, ducting & sealing around the radiator makes a huge difference, and in any case there's no benefit in allowing air to bypass the core. I took my old rotted trim to the weather stripping section at Home Depot and picked up new stuff of comparable size.

On a street car, maybe not much difference. On track cars, ducting & sealing around the radiator makes a huge difference, and in any case there's no benefit in allowing air to bypass the core. I took my old rotted trim to the weather stripping section at Home Depot and picked up new stuff of comparable size.

I always try to get the seals back. My job deals with lots of electronic components so it's easy to come by relatively firm packaging foams that boards come in. I just cut to size and use 3M spray adhesive to glue in place. HD is not a bad idea too.

So CR STI rad ended up being pretty much same as LGT stock? I know OEM Subaru radiators for diesel engines are much thicker but sourcing them here is too costly of an option...

On a street car, maybe not much difference. On track cars, ducting & sealing around the radiator makes a huge difference, and in any case there's no benefit in allowing air to bypass the core. I took my old rotted trim to the weather stripping section at Home Depot and picked up new stuff of comparable size.

That's a good idea, I was thinking of doing that but figured it might not be worth the effort. Might do that on my "racecar" first, since it sees all the track duty.

So CR STI rad ended up being pretty much same as LGT stock? I know OEM Subaru radiators for diesel engines are much thicker but sourcing them here is too costly of an option...

Yup pretty much identical, I was looking at Denso specs at RockAuto and they were showing up as slightly different:

2005 LGT (Denso 221-3605): link

Core Thickness: 0.6250 in.

Core Height: 13.3750 in.

Core Width: 27.0620 in.

Inlet Diameter: 1.5620 in.

Outlet Diameter: 1.5620 in.

2008 STI (Denso 221-9236): link

Core Thickness: 0.6300 in.

Core Height: 13.3900 in.

Core Width: 27.8000 in.

Inlet Diameter: 1.5630 in.

Outlet Diameter: 1.5700 in.

Interestingly enough the WRX one is identical to the STI AND it has the same sized turbo tank drain, darn wish I saw that before!

2008 WRX (Denso 221-9235): link

Core Thickness: 0.6300 in.

Core Height: 13.3900 in.

Core Width: 27.8000 in.

Inlet Diameter: 1.5630 in.

Outlet Diameter: 1.5700 in.

Anyway, I'm not convinced that going to thicker radiator would help, there are a few threads that discuss the pitfalls of thicker cores (takes much longer to cool them down).

Anyway, I'm not convinced that going to thicker radiator would help, there are a few threads that discuss the pitfalls of thicker cores (takes much longer to cool them down).

yeah, for MPG purposes, i think a correctly sized core (OEM) is better than an oversized core (2 row aluminum Koyorad). extends warmup time among other things.

Not exclusively to MPG even. My G20, with a dual core all aluminum radiator, is getting pretty damn good MPG heh.

It's similar to why stock Intercooler works so damn well, while aftermarket thicker ones can be worse. While a lower mass intercooler/radiator heats up quicker, it's able to shed the heat much quicker too.

Now with the Intercoolers, I noticed that it gets to a certain heat point and it simply doesn't go above that point. The LGT FMIC would then start shedding that heat VERY quickly. When I compared it to my G20's much denser FMIC, it would get to that peak heat point a little slower, but it would also take much more driving around to get it cooled down again.

Anyway back to radiators mwiener2 explains it real well in this post:

Heat and Temperature are not the same thing. Lets say the stock radiator is capable of storing 30,000 BTUs and can reject at 10,000 BTUs/hr. The aftermarket radiator may be capable of storing 50,000 BTUs, but can only reject it at 9,000 BTUs/hr.

 

In simpler car terms, the stock radiator can store the heat from only 2 back-to-back 1/4 miles runs and then needs at least 15min to cool down. The aftermarket one can handle 5 back-to-back runs, but then needs almost 40min to cool down.

This is also partly why cooler thermostats don't necessarily help, they slow down the warming of the car on cold starts but at full operational temps it doesn't matter. Now it's possible that a cooler thermostat will allow you to supercool the coolant 10-20* below the warmer thermostat, giving you slightly more time before coolant gets hot again, but shortly after you're back to needing a more efficient cooling setup (radiator/fans).