Knock sensor code. cant clear it. bad knock sensor?

[cj91legss]

Yeah those connectors can be tough. You gotta press the tab and wiggle them sideways.

Sent from my SGH-T889 using Tapatalk

[aaronllewellyn]

It should read 0 ohms

Sent from my SGH-T889 using Tapatalk

Alright wire's in, no check engine light on the dash but when I check the codes by connecting the black connectors 22 and 33 still show up. nomatter how hard I try, the battery could be disconnected for hours upon hours and the codes will still not clear, they exist before the engine is even ran, yet will not cause the CEL to come on while running.

there must actually be NO codes yet the ECU memory will not clear :C any Ideas? why would it secretly store codes but not illuminate the CEL!?

still stuck at 5psi fo no real reason, seems to be less "boggy" when you get on it but still missing 30hp none the less. I havent found any vacuum lines around the turbo that could cause it to operate on only wastegate pressure.. but I'll look again.

[aaronllewellyn]

Yeah those connectors can be tough. You gotta press the tab and wiggle them sideways.

Sent from my SGH-T889 using Tapatalk

no engine codes, drove it for three or so miles to make sure it wouldn't come right back on. if I check the codes with the black connector still shows the same 22 and 33 code, no way to clear it.

well this is frustrating

[aaronllewellyn]

would a bad BPV cause it to not boost over 5psi? what led up to this ordeal was my BPV hose that goes into the intake coming off and letting un-metered air in... could that potentially kill one? I have an extra, but I just snapped my nipple off the intake for it while un-doing the hose :C JB weld it is


EDIT: changed the BPV with a spare and no luck still stuck in limp mode or just cant get over 5psi for some reason.

left the battery disconnected for four hours and no improvement.

Pressure surges to 6-6.5 psi in between shifts from 5psi. (it's an auto)

[aaronllewellyn]

The other day I was driving my 1991 legacy SS down the road, unknown to me the BPV hose going back into the intake had come off. knock sensor code 22 and VSS code 33 appear, I drive the car home in limp mode, cant seems to build any boost stuck at 5psi, I have a boost gauge.


I replace the knock sensor, and the code still exists, then I replaced the knock sensor wire with a new shielded wire and the code still exists.

my ECU will not clear, code 22 and 33 will be on after supposedly "Clearing" the computer I've tried everything from the "right way" by connecting red and green connectors the whole process blah blah to disconnecting the battery for nearly four hours, the check engine light is not on.

The codes still exist when you check them with the black connectors even after "Clearing" them

This car has never worked long enough for me to even consider having any fun driving it, only seems to run for a few weeks before needing something

Have only been able to drive it for a combined three months before needing a transmission and now whatever this is... and it only has 130K

Do I have to drive this thing for 100 miles or something before it can come out of limp mode..?

EDIT: I ran the car in "D check" mode with the green connectors only to discover a code 51 (neutral inhibitor switch?) but if you check the codes via the black connectors 22 and 33 still exist.

[James614]

Is this car completely stock?

[aaronllewellyn]

Is this car completely stock?

Only thing that's not stock is the 3" TB exhaust.

[robertpaige]

Pull the battery over night for at least 12 hours, and try again.

[aaronllewellyn]

Pull the battery over night for at least 12 hours, and try again.

Also, I ran D Check mode on the ECU with the green connectors and it only showed code 51, neutral inhibitor blah blah no knock sensor or VSS code.... if the code doesn't show in D check mode there's no way I imagine it can exist.

[robertpaige]

Right. Pulling power will reset those codes. Give it a shot, and report back.

[aaronllewellyn]

Right. Pulling power will reset those codes. Give it a shot, and report back.

I had it disconnected for about four hours today... is that not long enough to reset the computer? If not it seems to hold codes for a very long time, even without power.

[aaronllewellyn]

Right. Pulling power will reset those codes. Give it a shot, and report back.

connected the battery and drove it around the block, doesn't seem to be boggy anymore... but still 5psi. I dont have an intercooler so it's not like there's anywhere for boost to leak from. I unbolted and cleaned the knock sensor mating surface also. no improvement.

I did notice that normal driving conditions kept it at 5-6psi but when you floored it with power mode on it would spike to 8-9 psi for a half second at like 6000 RPM in between gears.

cant find any leaks for the life of me.

I myself cannot figure this out, there's nowhere for boost to leak, in between the turbo outlet and intake it's one solid tube, no intercooler. I rewired the knock sensor but I don't know if it's actually "working" since the code shows up when you check them but otherwise there's NO CEL.

I will try:

MAF cleaner (then maybe a new MAF if it improves)

A new set of spark plugs.

I'll check the continuity of my knock sensor wire,

Then go from there.

[robertpaige]

I don't think you have an issue with boost leak. Sounds like your Boost Control Solenoid (passenger side strut tower) is not working properly. They tend to stick sometimes and not allow full boost, mine did and would only allow 5psi-ish on my old turbo wagon.

Can you try and get ahold of a manual boost controller? That would rule out the BCS completely.

The things listed above that you are going to try would have no affect on underboost.

[aaronllewellyn]

I don't think you have an issue with boost leak. Sounds like your Boost Control Solenoid (passenger side strut tower) is not working properly. They tend to stick sometimes and not allow full boost, mine did and would only allow 5psi-ish on my old turbo wagon.

Can you try and get ahold of a manual boost controller? That would rule out the BCS completely.

The things listed above that you are going to try would have no affect on underboost.

i'll try and get my hands on a mbc then.. maybe make one? then slowly dial up boost... is there a way to take apart the solenoid to maybe see it sticking partly open... sorry im throwing so many questions at you i'm pretty desparate to get the car working again.

is the d-check mode not throwing codes be verification it's not in limp mode? i unbolted my knock sensor again and started the car just to verify it would activate a CEL, i couldnt tell you why they exist in memory after clears...

[aaronllewellyn]

Alright I removed the solenoid and I can blow air through it! through the top nipple which I shouldn't be able to do so maybe that settles it, the solenoid has stuck partially open and is not holding my waste gate closed.

[Legacy777]

I merged your two topics since they are related.

Have you tried clearing the ECU memory with both the green & black connectors? If not, I'd try that. However, I'm not sure a code 51 is causing your problem. Instructions for clearing the ECU using the two connectors is on my site.

http://www.surrealmirage.com/subaru/engine.html

[aaronllewellyn]

I merged your two topics since they are related.

Have you tried clearing the ECU memory with both the green & black connectors? If not, I'd try that. However, I'm not sure a code 51 is causing your problem. Instructions for clearing the ECU using the two connectors is on my site.

http://www.surrealmirage.com/subaru/engine.html

I have cleared them VIA the battery method, with both pairs of connectors (the "right" way) and used D check mode, only showed code 51...

I have installed a new knock sensor and a new shielded wire, continuity is good between the sensor and ECU.

I removed my boost control solenoid and found that it was partially stuck open (I could blow trough the top tube through to the the waste gate vaccume line, I'm cleaning it with carb cleaner but I believe I will have to order a cheap MBC to solve the issue.

EDIT: Ordered a torbologic MBC to hopefully solve the issue.... car seems to be running fine, doesn't seem to be pulling timing as far as I can tell... so hopefully I wont be adding a MBC in limp mode.

[aaronllewellyn]

I merged your two topics since they are related.

Have you tried clearing the ECU memory with both the green & black connectors? If not, I'd try that. However, I'm not sure a code 51 is causing your problem. Instructions for clearing the ECU using the two connectors is on my site.

http://www.surrealmirage.com/subaru/engine.html

Alright well, even with a boost controller it sticks at 5psi then surges to 8~ on shifts, car has no pickup, laggy.

no CEL, but when you check the codes it says knock sensor 22 and VSS 33 D check mode is useless as it doent say there are any codes, I cant clear them and nothing I do will fix them either, it's probably pulling timing in some BS Subaru limp mode. I love owning a car that every morning you wake up and something else is broken this happened overnight after a rain storm, so it probably shorted out or something but im ready to give up on this car.

I would invest in a fuel pump or filter change or ECU but I'm done spending money if it's never going to fix the problem.

It's getting sold. this is overall the worst car experience I've ever had.

anyone in NY want a rustless 130k turbo legacy?

[cj91legss]

Do you have access to another ecu? Maybe just barrow one for a day and see if that helps.

Sent from my SGH-T889 using Tapatalk

[aaronllewellyn]

Do you have access to another ecu? Maybe just barrow one for a day and see if that helps.

Sent from my SGH-T889 using Tapatalk

There are no turbo legacy in Binghamton NY, I've seen one other up here in my life and it had out of state plates.

could leaving the stock boost solenoid connected prevent it from running properly.. even if there are no hoses attached..? either way it has to be still in some sort of limp mode...

Well, maybe I've connected the knock sensor wire incorrectly, I have this poorly drawn diagram of the pinout I'm just curious to know if this is the right side of the plug to be looking at when you wire it, way it's wired right now is opposite to the diagram, maybe me assuming the black wire was the ground was incorrect?

[aaronllewellyn]

Have I lost everyone's interest...

[cj91legss]

Black should be ground I believe

Sent from my SGH-T889 using Tapatalk

[aaronllewellyn]

Black should be ground I believe

Sent from my SGH-T889 using Tapatalk

I suppose I'll try re-wiring it again.

[aaronllewellyn]

Do you have access to another ecu? Maybe just barrow one for a day and see if that helps.

Sent from my SGH-T889 using Tapatalk

I'm going to check the continuity of my "new" $10 ebay knock sensor and see if it
s DOA

[aaronllewellyn]

BUmp update, today I finally got a code for the MAF code 23 and 49

abnormal voltage and wrong MAF code!

WOOOHHOoo you guys know what MAF's would be interchangeable with mine? I have a local PnP with tons of subarus