Front A/F Sensor P0031 Troubleshooting Help

[carnutdave]

I've just recently gotten my 818 driveable and legal and am supposed to get it tuned next week. The check engine light is on and my Accessport says it's a P0031 code. My research shows there's some issue with the front A/F sensor heater circuit. I can't seem to find the issue despite a week of troubleshooting.

818 Particulars:
-Dieted donor harness
-Factory exhaust
-Running both the front A/F sensor and rear O2 sensor; A/F sensor is NEW (Denso)
-Tune is a base tune from the tuner for my combination (2.5L hybrid, dark blue injectors, VF39, Walbro fuel pump)

What I've Done:
-I have the factory manual and followed the P0031 troubleshooting
--At sensor, voltage of yellow-red wire reads 12-14V
--At ECM, grounds checked good on B84, Pins 8 & 9
--At ECM, output signal on B84, Pins 4 & 5 showed a few tenths of a volt (very hard to check with the connector attached to the ECM; I stuck small wires in the back of the connector as pigtails to my multimeter)
--Resistance across heater circuit on A/F sensor is 1.5 ohms

Everything seems to check out following the troubleshooting. I checked voltage of the white wire at the sensor, which should be the signal wire, and got 3.5V. Sensor was unplugged. This is different than the <1V I got at the ECM on pins 4 & 5. I was expecting to see the same voltage at both places.

All of the wiring is completely covered by firewall, center console, etc so before I start taking things apart, I was hoping someone on the forum had experience and could lend some advice. Thank you for any insight anyone can give!

[RPGs818SNA]

Before you take your 818 apart, you might check the continuity and insulation on each of the sensor wires. If you unplug the connectors from the sensor and the ECM, you can measure the resistance of each wire from end to end and the resistance of each wire to ground. The wires should be much less than an ohm, and wire to ground resistance should be greater than 100 k ohms. Also check the resistance of each wire to the other wires, which should be greater than 100 k ohms. If those values check out, the problem probably isn’t in the wires. If you find a bad wire, it may be easier to run another one rather than dig the bad one out. Hope this helps.

[carnutdave]

Thank you RPGs818SNA. I'll ring out the wires tomorrow and let you know how it goes. Not sure why I haven't done that yet but the garage is 30 degrees and I was getting frustrated!

[carnutdave]

I continuity checked the wires this morning and everything checked out.

- Thick White wire - PWM ground for heater circuit - 2 wires from ECU, 1 at sensor - less than 1 ohm
- Thin White and Black wires - sensor signal circuit - both less than 1 ohm
- Checked each wire to each other and to ground - no continuity

I also rechecked the Yellow/Red power wire with ignition on just for the heck of it and got 12V. Thank you again RPGs818SNA!

At this point, I'm pretty confident the wiring isn't the issue. The only things left are the ECU and the sensor.

I called my tuner to explain the issue to him and he said if the sensor isn't an OEM Subaru sensor, that may be the issue. The sensor I put in is a Denso sensor but it's wasn't "Subaru". I've seen things like this at my job where some companies use commercially available sensors in their equipment but for their equipment to work properly, they need to have stricter tolerances on inputs, outputs, etc. to meet their system requirements. They "filter" the sensors so that only the ones that meet their tolerances work with the equipment. That my be the case with the off the shelf Denso sensor and the Subaru (also Denso) sensor. If it's not the sensor, it could be the ECU.

At this point, I've made an appointment with the shop for them to install an OEM Subaru A/F sensor later this week so we'll see how that goes. I'll update the post with the outcome.

[RPGs818SNA]

I checked voltage of the white wire at the sensor, which should be the signal wire, and got 3.5V. Sensor was unplugged. This is different than the <1V I got at the ECM on pins 4 & 5. I was expecting to see the same voltage at both places.

Your expectation is correct. Both ends of a wire should be pretty much the same voltage. From the 2002 schematic, it looks like the sensor has 2 white wires (confusing), one a signal wire (pin 4) and the other a heater current return wire (pin 1, probably thicker wire). Pin D4 and D5 at the ECM are the heater return pins and should be <1V with the heater on and around 12V when the heater is off. ECM pin D19 should be the same voltage as sensor pin 4. 3.5V sounds reasonable.

Too bad the wires are OK. Fixing them would be simpler than sensor swapping or ECM issues. I hope your new sensor solves the problem. Good luck.

[Sgt.Gator]

This is a really common issue, NASIOC is full of threads about it. One possible fix I've read:
1) Use Accessport to clear the code.
2) Disconnect the battery.
3) Step on the brake pedal to get any power still in the system to drain thru the brake lights. (This may be Internet trash but it can't hurt.)
4) Reconnect the battery.
5) Use Accessport to check for codes.
6) See if it runs ok!

[carnutdave]

This is a really common issue, NASIOC is full of threads about it. One possible fix I've read:
1) Use Accessport to clear the code.
2) Disconnect the battery.
3) Step on the brake pedal to get any power still in the system to drain thru the brake lights. (This may be Internet trash but it can't hurt.)
4) Reconnect the battery.
5) Use Accessport to check for codes.
6) See if it runs ok!

Thank you for the reply Gator! If the genuine Subaru sensor doesn't work for some reason, I will try this next. I brought the car to the shop yesterday morning but haven't heard anything back yet.

[carnutdave]

Short update. The tuner installed a new OEM A/F sensor but the code still comes up. They're going to troubleshoot a little but if they can't find anything, I'm going to get the car back home and continue trying to figure out what's going on.

[carnutdave]

The saga continues. The OEM Subaru AFR sensor did not solve the problem. I had the tuner troubleshoot the wiring and he came to the same conclusion I did, the wiring looked ok. The only thing left was the ECU so I bought a used one and tried it. Same code pops up at the same time. Something has to be going on with the wiring. I've started disassembly to get to the wiring and going to start with grounds as "bigger" suggested.

It's worth mentioning, one other thing I tried when troubleshooting was to hook a random incandescent tail light bulb to the heater wires to simulate a heating element. It measured about 3 ohms. This can show if the ECU is even trying the heater at all. The light lit up and pulsed for a few seconds after startup which is correct. It then turns off when the ECU throws the code and disables the sensor and heating element.

Thank you all for your comments. I'll post again when (not IF!) I find the issue.

[RPGs818SNA]

With a 1.5 ohm sensor heater, the ECU should be looking for around 8 amps. A fairly good wire that checks out good with a voltmeter may not work so well with that much current. One other helpful test would be to measure the voltage on the sensor YR wire during those few seconds of startup to make sure it's getting full battery voltage from the main relay. Also check the larger white wire on sensor pin 1 for ground during startup. That might tell you which side to investigate first. Good luck. I'm sure your perseverance will pay off.

[aquillen]

I hope you know all this already, but since you are still chasing I'll throw in my 2cents -

I don't recall above, you did verify and trace the polarity of the signal wires to the O2 sensor ? They are particular as it works somewhat like a DC battery (not really but close enough for discussion here). The heater side is not polarity sensitive.

About the grounds:

The correct ground location for pretty much all the grounds associated with the ECU is to the two clusters of smaller black ground wires on the top of the engine in left and right locations, not to some chassis point or elsewhere.This is because (1) the engine block and manifold are a big solid chunk of electrical conductor. (2) the sensors and controls associated with the engine are screwed, bolted or fastened to that chunk of conductor. (3) the ECU and those components need to be very well connected to each other for high quality signal integrity needed by the computer to measure what is happening in the sensors (of course at both the signals side as well as the ground sides). So the circuits need to be common to the best conductor in the "area" hence the engine.

SO - locate the two engine ground clusters, which at the engine bolts they really look physically just like the bits with my red arrow below. Make sure none of those black wires are corroded, half gone from wear/damage/corrosion and all is good and tight. Find both sets and make sure they are good. Trace them back to the related connectors, inspect those connectors - again no corrosion, fully connected and locked together. Note there is more than one Subaru page for the engine grounds, and those engine top ground clusters go through at least 2 connectors (the "big fat" ones), I only posted 2002's page: GND-08 so get your wiring diagram for your year and make sure you know what you are looking for.

I've stripped 3 harnesses off used 02-04 Legacy models (cause I did the H6 motor and got hold of some pretty bad used wiring, etc.,) and all of them had the ground tabs - at the engine - in rather crappy condition. I replaced all those with new connector blocks of my own making. Not everyone can or is going to do that but you can clean up and repair or buy crimp lugs and redo them if they don't look right.

Any of those ground wires - if bad - will introduce a bug almost for certain in the particular ground paths they handle for one or another sensor, actuator, etc. Grounds so often get lower attention but they are, after-all, 50% of the pathway for any loop of electric current flow in the system (not discussing the odds and ends of twin isolated lines of course).

And another item - there is typically a B83 connector which shows up about half way through the engine wiring section, not necessarily on the ground pages at the beginning of the various year manuals. This has to be in good shape and has jumpers inside it that tie a bunch of sensor circuit grounds together. It is typically close to the ECU. Make sure it is intact, all plugged together, no wires pull out, etc. That connector is easy to abuse when handling the ECU connector harness, could get messed up. This is another piece of the sensor ground "world" that you need to trust.

What no one should do is move these engine grounds to a chassis tie point and then rely on a ground strap to the engine for the "tie - in". First you have poor signal quality into the ECU and next if that battery ground strap fails, your small wiring will be fried at the first starter motor crank... pretty sure I read about 5 years back someone posted "did that"...


.

[carnutdave]

Thank you so much for your input RPGs818SNA, bigger and aquillen! I've done some disassembly to get to the wiring (seats, firewall, etc). The heater ground and signal wires are pretty short and direct and seem to be in good shape. I've been discussing this with my tuner and he mentioned that he measured only about 12.7V with the car running and suggested I check the alternator (it's from the donor). His thinking is it's a "heater low" code and may be caused by low voltage. I checked voltage with the car running at the alternator, battery, main engine relay and the AFR sensor and they were all reading 14.7-14.1V. The strange thing is my Accessport "Battery Voltage" monitor reads 12.7 volts.

My next step, as you've all recommended, is to start checking grounds and making sure everything looks well connected and clean. As RPGs818SNA said, the grounds may check good with the voltmeter but can't handle a multiple amp draw from the heater. aquillen, I didn't change the location of grounds from how they were arranged on the car (engine is still engine, chassis are still chassis, etc). Getting to the two engine grounds are going to be fun since they're under the intake manifold....ugh. :-)

[Bob_n_Cincy]

On my engine, both of the engine ground connectors were easy to access on top of the intake.

ground1.jpg ground 2.jpg

[Bicyclops]

Thank you so much for your input RPGs818SNA, bigger and aquillen! I've done some disassembly to get to the wiring (seats, firewall, etc). The heater ground and signal wires are pretty short and direct and seem to be in good shape. I've been discussing this with my tuner and he mentioned that he measured only about 12.7V with the car running and suggested I check the alternator (it's from the donor). His thinking is it's a "heater low" code and may be caused by low voltage. I checked voltage with the car running at the alternator, battery, main engine relay and the AFR sensor and they were all reading 14.7-14.1V. The strange thing is my Accessport "Battery Voltage" monitor reads 12.7 volts.

My next step, as you've all recommended, is to start checking grounds and making sure everything looks well connected and clean. As RPGs818SNA said, the grounds may check good with the voltmeter but can't handle a multiple amp draw from the heater. aquillen, I didn't change the location of grounds from how they were arranged on the car (engine is still engine, chassis are still chassis, etc). Getting to the two engine grounds are going to be fun since they're under the intake manifold....ugh. :-)

If you have 14.x volts at one point and 12.x v elsewhere, you have a high resistance connection somewhere. The same can also be true in ground paths. Lots of starters get replaced because the battery negative wire has corrosion on it at the tail end. As has been pointed out, an iffy ground can play havoc with signals.

Ed Holyoke

[carnutdave]

On my engine, both of the engine ground connectors were easy to access on top of the intake.

I just ran down to the garage and looked again and my grounds are the same as yours. Not sure what I was thinking when I though it was under the intake manifold! There is a lug underneath I was looking at but it's the mount for the hard fuel line. It's been awhile since I rebuilt the engine, so I guess I forgot where everything went. When I reinstalled everything, all I did was wipe the harness down and put it back on. Didn't make any modifications to it.

[Bob_n_Cincy]

Car Nut Dave,
How are you wiring the battery negative terminal to the engine block?
Where is your battery located?
Bob

[carnutdave]

Car Nut Dave,
How are you wiring the battery negative terminal to the engine block?
Where is your battery located?
Bob

Bob, taking advice from this forum (maybe even one of your posts!), I ran a dedicated 4ga wire from the battery negative terminal to the starter bolt which is also where the frame ground strap is attached. The battery is in the standard, front location and is also grounded to the frame in the front. Here's an older pic of the ground wire:
20200705_093809_cropped.jpg

[carnutdave]

I found it! It was a bad ground. Started tracing grounds from the ECU and ECU connector B134 pins 7 and 15 measured 5 ohms. Pin out diagram says these are "control system" grounds. All others were less than .5 ohms. Wiring diagram also shows this is a shared ground with the OBD connector (if you remember, my Accessport voltage didn't match the actual measured voltage). Isolated it to the E3 engine connector and was able to put dielectric grease on it and get it down to .2 ohms. Started the car and no MIL! Accessport is now showing 14V.

Lesson learned, the Subaru troubleshooting won't always lead you to the solution. In this case, the ground was part of a separate circuit and not directly linked with the AF sensor.

Thank you all for your input and help. This forum is full of great people.

[RPGs818SNA]

Congratulations CarNutDave! Well done. Thanks for sharing your experience, especially the need to check ALL the grounds.