[snake2332]

When under any sort of boost, my downpipe is reading leaner than my header. In my brainstorming and initial investigation, this has to be a leak at the uppipe/turbo/downpipe, right? I don't see any blown out gaskets, but I haven't removed any parts yet to take a closer look. It's a bit of a pain to get that heat-wrapped stuff removed, so I wanted to get some community input before I start taking parts off.

I did replace the Bosch sensor since it only worked about half the time anyway and I read that they will start reading leaner and leaner as they "die", but of course my sensor was fine so that was a waste.

FWIW, the header to uppipe gasket was previous oblong after just 5k miles, so I replaced it. My header and uppipe are Tomei Expreme. I've heard that basically every header uppipe blows out gaskets except the Grimmspeed one. If that's true, anyone have an idea to keep my Tomei stuff from blowing out gaskets?

My STI is nothing crazy, 300WHP and 18psi on the OEM turbo, upgraded fuel system, intake, exhaust, TMIC, IAG AOS, etc., so it's not like it's under extreme conditions.

Thanks!

 

[yamahaSHO]

This is normal. O2's are inaccurate when drive pressure is high. This is why you see any real turbo setup with the O2 after the turbo.

If this wasn't a problem, you'd have no need for a redundant sensor.

 

[GK1707]

I had this issue and it was my exhaust gasket between the header and up pipe that had a leak on it. This was less than 500miles after installing an oem gasket there. I installed the Grimmspeed double thick gasket with a thin layer of copper rtv and my AFRs on my wb02 were back in check.

I'd say check your gaskets, if you can smoke test that's the best to check for leaks.

 

[killerbmotorsport]

This is normal. O2's are inaccurate when drive pressure is high.

Is there a chart on this? I would assume then that AFR response would need to be scaled as that pressure changes then (with mods, increased boost, etc...)?

 

[mheyman]

I've heard that basically every header uppipe blows out gaskets except the Grimmspeed one.

Not true in the least. "we" don't replace header gaskets "regularly" Personally I've never replaced mine except when it had to come apart for other reasons, and that has not been often.

 

[snake2332]

Not true in the least. "we" don't replace header gaskets "regularly" Personally I've never replaced mine except when it had to come apart for other reasons, and that has not been often.

The info I heard was from one older Subaru guy that's worked on quite a few 04-05 FXT's and STI's. He has some strong opinions on things, some of which I didn't agree with. But since I'd already seen the Tomei gasket blow out, it sounded legit to me. I didn't replace it with a Grimmspeed so I am going to dive in and use one of those instead of this CNS one I got in a set of gaskets.
Also, your use of quotes is invalid. I didn't use the words "we" or "regularly", so what other meaning are you trying to imply with the quotes? I don't really care, but you might want to tone it down on the air quotes, haha!

This is normal. O2's are inaccurate when drive pressure is high. This is why you see any real turbo setup with the O2 after the turbo.

If this wasn't a problem, you'd have no need for a redundant sensor.

There was a sudden change though, where the OEM sensor reads the same and is hitting 11.03 (I had the car hitting low 11's last winter but it's running richer right now) but my AEM is reading lean, like 12.8 at WOT where it should be more like 10.5 or something rich like that. If high pressure can throw it off, it must be higher than what I'm doing. It's just a VF48 pushing 18psi.

I had this issue and it was my exhaust gasket between the header and up pipe that had a leak on it. This was less than 500miles after installing an oem gasket there. I installed the Grimmspeed double thick gasket with a thin layer of copper rtv and my AFRs on my wb02 were back in check.

I'd say check your gaskets, if you can smoke test that's the best to check for leaks.

Thanks, I am going to do this, too. I did spray copper RTV on the cheaper MLS gasket, brand CNS(?), but maybe I didn't on the uppipe-to-turbo gasket.

Time to bust out the ramps and creeper! Thanks for the replies all.

 

[snake2332]

Is there a chart on this? I would assume then that AFR response would need to be scaled as that pressure changes then (with mods, increased boost, etc...)?

Too bad the previous owner didn't buy EL Holy Header from you guys! This Tomei UEL unit, even with heat-wrap, melted my clear timing cover a bit, too, until I added heat tape to the timing cover itself.

 

[yamahaSHO]

The info I heard was from one older Subaru guy that's worked on quite a few 04-05 FXT's and STI's. He has some strong opinions on things, some of which I didn't agree with. But since I'd already seen the Tomei gasket blow out, it sounded legit to me. I didn't replace it with a Grimmspeed so I am going to dive in and use one of those instead of this CNS one I got in a set of gaskets.
Also, your use of quotes is invalid. I didn't use the words "we" or "regularly", so what other meaning are you trying to imply with the quotes? I don't really care, but you might want to tone it down on the air quotes, haha!



There was a sudden change though, where the OEM sensor reads the same and is hitting 11.03 (I had the car hitting low 11's last winter but it's running richer right now) but my AEM is reading lean, like 12.8 at WOT where it should be more like 10.5 or something rich like that. If high pressure can throw it off, it must be higher than what I'm doing. It's just a VF48 pushing 18psi.



Thanks, I am going to do this, too. I did spray copper RTV on the cheaper MLS gasket, brand CNS(?), but maybe I didn't on the uppipe-to-turbo gasket.

Time to bust out the ramps and creeper! Thanks for the replies all.

The OEM will flatline at 11.03 as there is a hard limit in the tune. That said, 11.03 on the OEM sensor can be 12.x:1 or 5:1. Unless you suspect an issue with your WBO2, follow the WBO2.

 

[snake2332]

The OEM will flatline at 11.03 as there is a hard limit in the tune. That said, 11.03 on the OEM sensor can be 12.x:1 or 5:1. Unless you suspect an issue with your WBO2, follow the WBO2.

One thing I considered was the OEM sensor failing or misreading, but that seems less likely to me than a post-uppipe leak. Have you seen the OEM sensor fail on multiple occasions or something? I mean, the sensor doesn't inherently misread 12:1 or 5:1 as 11.03:1, so I'm curious what makes you say that.

Thanks!

 

[yamahaSHO]

It's is unable to take an accurate reading with high drive pressures, so yes, it will inherently flatline with high drive pressure.

Once it enters open loop, consider it inaccurate.

 

[snake2332]

It's is unable to take an accurate reading with high drive pressures, so yes, it will inherently flatline with high drive pressure.

Once it enters open loop, consider it inaccurate.

OK thanks. What's your threshold for high pressure?
I'm confident mine is still accurate because it was always spot on with my wideband prior to suddenly not being even close except at stoich/lambda.

 

[enoxard]

Timely thread.. I plan to install a Killer B holy header and up pipe [ and cobb downpipe ] this weekend.. I also have an extra bosch wideband O2 that was going to use as the rear sensor.

Finally, I thought the header bung was for EGT, not O2?!?! The downpipe and mid-pipe bungs are O2.

 

[snake2332]

Timely thread.. I plan to install a Killer B holy header and up pipe [ and cobb downpipe ] this weekend.. I also have an extra bosch wideband O2 that was going to use as the rear sensor.

Finally, I thought the header bung was for EGT, not O2?!?! The downpipe and mid-pipe bungs are O2.

On my Subarus, the EGT is in the uppipe, OEM AFR in the header/exhaust mani, and wideband goes anywhere in the exhaust post-turbo, usually in bellmouth of the downpipe.

EDIT: Midpipe has a bung for the rear O2 sensor that just tells you if your cat is performing (P0420 code anyone?), right?

 

[yamahaSHO]

OK thanks. What's your threshold for high pressure?
I'm confident mine is still accurate because it was always spot on with my wideband prior to suddenly not being even close except at stoich/lambda.

I only pay attention to the OEM sensor in closed loop.

 

[snake2332]

I only pay attention to the OEM sensor in closed loop.

Makes sense since it is definitely inaccurate near the 11:1 compared to the LSU4.9 wideband that reads down to 8:1, and you're probably tuning a lot of cars that run E85 so you need those lower values. That's what I was using when tuning as well, with the analog-out wire from the AEM gauge going to TGV-Right and logged on my Cobb AP. I did tune a pair of FXT's with open-source and the built in AFR, gradually getting it close to 11:1. We only have mediocre 91 octane gas here in Montana - no E85 anywhere, not even 92 octane Shell.

 

[yamahaSHO]

These reference lambda and are converted for gas, so you actually need 'leaner' values when that is factored in. In other words, instead of targeting 10.8-11.0:1 like you would on gas, you'd be looking at doing more like 12:1 on E85, again gas referenced.

If you have a gauge that will allow you to display other fuels, then you'll see lower numbers, but lambde will not change.

 

[a5m]

Finally, I thought the header bung was for EGT, not O2?!?! The downpipe and mid-pipe bungs are O2.

The header bung is O2, as is the downpipe one post Cat. The pre Cat is for WB.

 

[snake2332]

These reference lambda and are converted for gas, so you actually need 'leaner' values when that is factored in. In other words, instead of targeting 10.8-11.0:1 like you would on gas, you'd be looking at doing more like 12:1 on E85, again gas referenced.

If you have a gauge that will allow you to display other fuels, then you'll see lower numbers, but lambde will not change.

Oh nice, didn't know that. I have given up hopes of getting E85 here. We'll probably get EV charging stations instead.