P0170 on 2000 Subaru Outback happens only after refueling

I enjoyed your post but it's pretty far fetched that an evap issue would cause that code in my experience! I mean it can but It may be just coincidence! IF it was an evap issue I'd think u'd have an evap code as well with your po170 code. I'd start by checking the vaccum lines at the cannister and filler neck and if u want to eliminate the evap as a casue drive the car around after fueling with the cap off, . If the po170 dtc stays gone but an evap code occurs then I'd say you're right, if the po170 still arrives then I'd think it's not related.

jeff compton wrote: ...it's pretty far fetched that an evap issue would cause that code in my experience! . I was guessing that the lines in the diagram I mentioned were part of the evap system. I tried to upload it before but got an error message: "Sorry, the board attachment quota has been reached." So I tried a different way this time. It's #10 valve and the lines by it that I was wondering about, but that's if this is even the correct diagram for my car.

I found another image showing hoses off the filler neck. Could it be possible the fuel is splashing into the these lines, and if that would cause the code.

I don't know if it's the correct diagram or not. If anything I'd doubt it's number ten! If anything on evap issues I find more problems with the other valves and solenoids not the flapper in the neck. Have u considered having subaru look at the live data and see what they can see if there is any correlation or not?

jeff compton wrote: Have u considered having subaru look at the live data and see what they can see if there is any correlation or not? Considered it, yes. Afford to have them chase it down, no. The code will not trigger at any time other than driving it for a couple miles down the road immediately after than after refueling to over half-way mark or so. I had some solenoid replaced (by a shop) on it in about 2004 or 2005 that I was told it was located above the gas tank, so I let them do it. I cannot remember which one it was for sure, but I do recall it threw a different code and it would not stay cleared like this one. It would pop back up immediately. The image of the filler neck and related hoses is from a pdf manual for Subaru. The other diagram I cannot be positive on application.

Update - The mystery continues... With no codes registering, I refilled the gas tank yesterday. The CEL came again on after 2.2 miles. I stopped by A-zone to have them read and confirm it was the P0170 code again, then clear it. It hasn't came back in subsequent driving around on errands, and likely will not reappear until I refill the tank again. Any suggestions of which sensor or valve to test would be greatly appreciated!

Update - It happened again!! After several months and a few fill-ups with the code NOT popping up, it reappeared again but within a few startups after refueling. This time it was after adding fuel from a an old 5-gal. Jeep can and using a long narrow funnel to get past the filler neck flap (I was empyting all my gas cans generally reserved for lawnmowing to refill them cheap with a certain store discount). I still wonder about the valve in the filler neck being the problem. I had to replace the PS pump (bad bearing) so it sat for a few weeks until I got the part then with only a few startups to bleed the PS lines and recheck the belt tension. Then a few days later when I drove it again, the code popped up about 5 miles down the road. I found a diagnostic procedure for this code and the list includes testing; Intake and Manifold Leaks, Exhaust Leaks, Fuel Pressure / & Regulator, Coolant Temperature Sensor, and Manifold Pressure Sensor. It also mentions replacing the Intake Air Temperature Sensor if the combined Manifold Pressure sensor is out of specs, but does not reference testing the Air Temp Sensor part separetely. I tested the fuel pressure with and without the vacuum line connected to the pressure regulator. Specs are 41-46 psi with hose off. I got 44psi. And though the pressure with the vac hose on is at 35 when the spec is 30-34psi. It may be my gauge or it may actually be slightly high indicating clogged return line or to replace the regulator. Pinching the return line several times immediately shot the pressure up to 65-ish and right back to the exact same lower readings when released. So the regulator is low on the suspicion list. I have a scan tool, and the Coolant reached normal temperatures. The Manifold Pressure Sensor signal was within specs both at idle (7.09-12.2 inHg) and with key on / eng off (21.65-31.5 inHg). I did see that one O2 sensor reading was at .75 or so while the other was at 1.5 or so. I could not find specs to verify if this was okay or not. I cannot see anything obviously loose broken or cracked to cause a vacuum leak. The next step will to carefully spray some carb cleaner around the engine to see it varies the idle rpm indicating the possible problem in the engine area. Otherwise, I'm looking at perhaps replacing exhaust gaskets as a try-n-see if any little leak there may cause this. I have not found a definate answer one way or the other on which direction the fuel trim is in error; rich or lean for this code. Could it possibly be a leaky injector causing a "rich condition" as mentioned in another article as being the main trigger for this code? Does anyone have a spec for any allowable fuel pressure bleed-down on 2000 Subaru Outback? I'm also wondering about fuel pump volume discharge. What if the pump is producing the pressure but not the volume of supply needed? Is that possible and perhaps causing the code if engine is starved for fuel at times? I cannot find a spec for pump volume output to test this.

Here's another update. I have re-verified all the steps (to the best of my ability) again in the diagnostic procedure for this code down to where it finally says to contact Subaru of America. Which to me, translates to handing my local dealership an open check for however long it takes to figure out what else may cause the code not mentioned on the diagnostic procedure... or maybe not on the one that I got for it. The exhaust system looks okay, ...great even for 101K miles and without any obvious holes or dark blow-by stains at the gaskets. Beyond visual inspection of the hoses, I sprayed short bursts of carb cleaner around the hoses, throttle body and intake manifold gaskets to check for vacuum leaks and that did not produce any change in rpm, or it was compensated for so quickly by the computer that I could not notice any change. If there is a better way for testing vacuum leaks on newer cars, or I missed a step to avoid computer compensation (if/as needed), please let me know!! This is the simple way I knew how to test them all at once on older cars. Shy of replacing every little section of hose and exhaust gasket just for the fun of it to try-n-see, I am presuming they are okay at this point. I'm now looking at other possible things that may mimic this condition. I cleaned the stock fuel filler cap o-ring and springed vent inside it. I tried to check the crossover tube at the fuel tank for the "Jet Pump" but without dropping the tank, I could only access the outer ends that appeared okay, so (hopefully) no suction loss or air cavitation there in drawing the fuel over to the pump side. The "Drain Filter" also seems like a possibly if it is clogged and/or the "Drain Valve" not working properly and not allowing proper flow functions the vapor canister. Thoughts on this posibility? Has anyone here ever had to change theirs due to problems? * I have recently, and finally, found mention of the ON-BOARD REFUELING VAPOR RECOVERY (ORVR) SYSTEM. It describes the process of vapor circulation during driving and refueling. The main components mentioned here are the Pressure Control Solenoid Valve, Purge Control Solenoid Valve and the Vent Valve. Here's what I understand of the description- The Drain Valve allows fresh air into the canister through the Drain Filter as needed during driving. And then as the fuel in the tank evaporates, sloshes around or heats up creating vapor pressure above atmpspheric pressure, the Pressure Control Solenoid Valve diaphragm is pushed open to allow these tank vapors to also enter the canister. The Purge Control Solenoid Valve is then supposed to open to allow these vapors present in the canister to be added to the filtered clean air flow at the throttle body. During refueling, as the pump-nozzle is inserted, the Shut Valve in the filler neck is supposed to close to prevent liquid fuel from entering the EVAP lines. With the increased vapor pressure created during refueling from the displacement of volume in the tank, the Vent Valve (located on top of the tank and is the most difficult to access to test for sticky float) is supposed to open and flow these vapors directly to the canister where it is purified by the charcoal and vented to the atmosphere through the Drain Valve and Drain Filter. So with this, I'm back to looking at the Drain Filter & Valve here again as a possible problem because of the commonality of the code happening right after refueling. Being a 'filter' in it's purpose and the placement under the car subject to road grime and schtuff, I would be leaning towards replacement of this part for general principle anyway and I will get that part soon. I might get lucky and never see this P0170 code again with just the Drain Filter replacement. I won't know until I replace it, drive it around some more and refuel again at least a couple more times. Anyone have testing specs for the two solenoid valves mentioned above to go ahead and test them to eliminate the possibility there? I cannot find the specs, so far. I really hate to just throw parts at anything for a try-n-see. If the Drain Filter doesn't resolve this, then I'd be looking at testing these other parts. BTW... With the code cleared, the O2 sensor readings on my scantool were; B1S1= 1.53 and B1S2= 0.76 at idle. Are these okay or within range? Again, I have not found these specs for this vehicle. I think they're probably mentioned somewhere in the DTC procedure lists, but which one? Any further information here would be greatly appreciated on this mystery as I have been spending numerous hours, as I can, on the computer trying to find the relevant information, and I simply cannot afford to start throwing parts at it nor to just take it in to the dealership or local independent shop because on top of purchasing any needed parts, they would not only want to charge to repeat my work but also charge for any further unknown hours of searching for the real problem. ...back out to the garage... --->

If the cause of your 0170 code was the Evap. system then the fault code would be 0440 - the"loose gas cap" code. I took out the charcoal canister from the LH rear wheel arch of my '99 Outback and checked the solenoids and hoses when looking for the cause of a persistent 0440 code. Eventually found a rust pin hole in the metal upper part of the steel fuel-filler neck, in a position just below where the base of the filler cap would be when installed. (Thank you, 10% Ethanol). Filled the pin hole with two-part epoxy putty - end of problem. Now i'm getting 0170 fault code! Research leads me to suspect engine oil contamination of the O2 sensors' electrical connectors, above the RH front driveshaft in the engine bay. About six months ago the RH exhaust camshaft front oil seal started to leak badly; the oil ran back along the underside of the cam cover and misted the whole area, including the O2 harnesses. At first I thought that it was the cam cover seal, which I changed in vain before realising it was the camshaft seal. Need the car for work so will try degreasing the harnesses and splitting the connectors to degrease the contacts this weekend. Hope this helps.

englenick wrote: ...found a rust pin hole in the metal upper part of the steel fuel-filler neck, in a position just below where the base of the filler cap would be when installed. (Thank you, 10% Ethanol). A pinhole in the filler neck may be a good possibility because P0170 is neither specified as being rich or lean. But I haven't had the time lately over winter to further check mine out yet with other 'do-me-first' repairs on other vehicles in the family fleet taking space in the garage. **Anyone here know about '79 Lincoln EEC-I timing and V-V carbs? I've already heard; 'just replace them.' It's a challenge to me to get this car running again as-is. Anywho, I'd have to look back on my research info about potentially oil contaminated O2 wiring being suspect for this code. My engine is fairly clean with some seapage for the age but no major leaks. My O2 wiring connectors were clean and dry when I replaced the sensors. They were due for replacement anyway according to the book. The front sensor tip had a very slight carbon colored darkening on the front side, but that wasn't the cure here. The code is still randomly happening. I'll go a few stops at the gas station for $10-20 worth and not get a code. Then another time, it will turn the CEL on. I am finding another commonality in where I get fuel, by location not by brand. If I drive away from any gas station and continue at normal driving speeds, 30mph or more, it will not likely come on versus if I pull out into heavy traffic and basically idle along slowly stoplight to stoplight for a while. So this is telling me it may have (or may also have) something to do with the tank venting and vapor recirculation system. I will notice a little stumble in the engine on rare occasions as if it is suddenly not receiving the correct mixture when at a stop or creeping along slowly in traffic. This usually preceeds the CEL coming on, but oddly not immediately. The mechanic at the dealership told me that either of the trim corrections going beyond the specified range in either direction or both would likely set the code. Again, they've never encountered this specific code before at that location. If the fuel trim over-correction in one direction sets the P0171 and the other sets the P0172 as they are specified as rich and lean fuel trim issues... Then, maybe exceeding the trim levels in both directions with extreme rich and lean injector on-time compensations will set the more general P0170. I dunno, yet. I will have to look further at the long-term trim levels with the scantool and try to find what spec is supposed to be a "normal" range compared to my readings. And try to find the max peak allowed for short-term correction when the system opens to add the stored tank vapors to the overall fuel mixture. Perhaps another similar Subaru owner here can check their fuel trim and O2 readings with a scantool and post them for comparison???? Shy of letting the dealership figure it out for the supposed $90 inspection fee to determine the cause, the last thing I am down to physically inspecting is the "shut-off valve" in the filler neck. See #10 in earlier diagram. And perhaps valves 14 or 16 in the diagram that I cannot directly access to inspect without dropping the tank. This is where I may finally let the dealership do the inspection and testing. I will report back again when I find out anything more on my car's specific issue. Also, FYI for those that let Jiffy-Lube or another place change your oil for convenience, the mess you see may be from careless employees and may not be from leaking seals or gaskets. This can also be a trick used to sell you a sealant additive. They tried this on my neice. I cleaned the area and no leaks were evident where they said it was.

check with dealer several TSB's and recalls for emissions parameters... A swag- everything working - FUEL GAUGE? and modofocations?/

I'm chasing down the same P0170 code on my 97 Legacy 2.2. I've also solved the P0440 code, so I have some experience with EVAP too. Do you know what your long term fuel trim value is? Mine long term fuel trim is 27%, so I know it's adding fuel for some reason. You might have to use a higher quality scanner to get the freeze frame report. I'm using the Scantool.net 425801 scanner + it's included software on my laptop to get mine, which was fairly inexpensive. Since you think it's related to your EVAP system, have you verified your EVAP solenoids are working correctly? You can connect the green diagnostic connector under the dash and feel them. You should hear/feel them click. If the click is faint, it's best to pull them and bench test. Your front O2 sensor should be cycling back and forth, not at a constant voltage. The rear shouldn't affect this (it's used to monitor the cat), but should be a steady voltage, around .5V I think. EDIT: I believe any correction over 20% will set the P0170 code.