Honestly, I am hoping I can connect with cardoc, I am a lurker here but dont have enough posts to message him.

Sometimes it is enough to include their member name in a post, as in @cardoc
Let's see what happens from that, if anything.

 

Thank you! I also am not closed off to any tips or suggestions, I'm pretty comfortable with mechanical systems, I've rebuilt a lot of cars in my life but mainly 60's to 80's. If someone has some pointers to what to check, besides the Autel I have a pretty capable Fluke meter if someone has ohms or pin outs to check on a connector. I consider this one a keeper as it lets me have the most basic of conveniences and has proven itself reliable and durable.

 

I pm'd cardoc for you and provided a link to your thread. We'll see what happens. AFAIK, you cannot receive or send pm's until you have 15 posts, otherwise I'd pm you rather than post this stuff in your thread. No matter, I can always delete later. Note - spamming to get to 15 is usually frowned upon.

 

These may help:

 

The fuel gauge reading is based on the two readings from both level sensors in the tank. The main pump on the passenger side of the car is designed to create a Venturi to pull gas from the left side of the tank to the right side to keep the right side predominantly full. The resistance on the level sensors between left and right is how the computer gauges the level and that information is sent to the instrument cluster. A lot of times the driver side level sensor acts up and you can either take it out and clean it and put it back in and it'll work, or you just have to get a new transfer pump with the level sensor.

I'm working in Round Rock right now. I will send you a PM that I am pretty sure you will get because it has worked in the past with other members who had a short number of post.

 

Thank you all for this help, I will start on the diagnostics today.

 

After talking with Cardoc he suggested besides the ohms of the senders, check the battery connections. I removed and cleaned both terminals, they were not bad to begin with but I sprayed them down with deoxit d5 and brushed them and they are now completely clean and corrosion free. I also removed the main ground on the engine and the approx 10 gauge ground wire to the firewall and cleaned both of those connections well. Battery was made 02/22 and is a AAA battery I bought at NAPA, apparently last year, I only remember I had starting trouble when it was cold one winter and replaced it. With the engine off, I measure 12.61V, with it running I measure 14.40. The fuel tank is today roughly 3/4 full, the primary sender with pump shows 39.4 Ohms, the secondary sender shows 159.2 Ohms these both seem within spec. Yesterday the tank was about 7/8 full and I measured 17.2 Ohm and 124.1 Ohms respectively so they seem to be responsive. I am at a loss at this point. The only thing of note is I distinctly recall the original secondary sender had a red dot, the replacement has a black dot, is this possibly a calibration code to the ECM? Is it possible the secondary sender is good but the ohms are too high for my ECM? I ordered the part online from a Subaru dealer by year/model and engine/body style. Both parts are original Subaru parts. I changed the main sender/pump assembly assuming a 200,000 mile pump and filter could be part of my hot starting issue, this issue started after that, I then checked the secondary assembly and saw the float arm had a great deal of side to side play and just assumed contact was intermittent/poor. FWIW, I only get a fuel tank reading on the dash when the tank is full, it then suddenly goes to a Brake, traction control and check engine light and F refill notice and the gauge shows empty at right around 7/8 tank, once full, it goes away and fully resets. The only code I get is P0463, there is not another single code that comes up if I run a scan. Part of me is wondering if the ECM has an issue simply because I have a hard shift into drive when cold, that has never triggered a code, I also have one bank that tends to cycle in and out of cam advancement at 2400rpm and that is despite all new solenoids, one would figure this would trip a code. All the sender contacts on both end are clean and shiny and the plugs are firmly connected and I see no damaged wires.

 

Well, I think I just figured this out, seems my car has the California emissions sender, so my hunch was correct, that is what the red dot means, off to order the correct unit.

 

Interesting . . . I've never seen a distinction in the senders based on the emissions system. The FSM is silent on this, as is the Subaru on-line parts catalog ( Subaru Parts Online from Subaru of America ), as far as I can tell. But I might be missing something in my search.

Are there two different part numbers for the sub-sensor depending on emissions-spec?

 

California 42081AJ021
Standard 42081AJ140

Apparently the pass through voltage influences the tank vapor purge and the fuel temperature sensor reading. The standard has a black dot, California is a red/orange. This time I searched parts by vin and that was where I found this issue

 

When I get the California one I’ll post back, fwiw the hard starting seems to have been resolved by all the contact cleaning but I’d also unplugged and replugged and cleaned the sender/ pump harness connections so hard to say which one actually did it. First time in month I didn’t have to try 2-4 times to get it start.

 

When I get the California one I’ll post back, . . .

When you have the CA one, would it be possible to compare the two (the CA and the FED in there now) side-by-side?

There might be mechanical differences, e.g., different seals/gaskets or the way the top "cap" part is shaped. Electrically, there's only the fuel level variable resistance -- could that be different? According to the FSM pages posted above, the sub level sensor resistance ranges from about 9 Ohms when the tank is full (float high) to about 271 Ohms when the tank is empty (float low). Perhaps compare the resistances of the two units to see if the CA one has a significantly different resistance range.

I plugged the CA part number into the catalog I linked to above. It recognized it for 2010 and 2011, but not 2012. A Google search of the part number itself led to numerous other parts suppliers, but all indicated this part number is not applicable to 2012. Where (website?) did you use the VIN that led to finding there's a different part specifically for the CA 2012? (It might be useful for others.)

 

I would also suggest that you maybe go down to Napa or O'Reilly's and get a 4 gauge battery cable with eyelets on both ends, approximately 36 in long, and run it between the negative post clamp over to the upper AC bracket bolt. This will increase available ground flow to the engine block and especially the alternator.

As for the cam phase differences - Before you drain the oil for your next oil change, put in 1/2 bottle of Sea Foam and idle it for 10 minutes, then drive it a few miles keeping it under 45 mph, then do the oil change. The Sea Foam will clean out the engine and with driving it clean the cam phasers. This might help with equalizing the cam timing.

 

The only physical differences that are obvious is a black dot on the top of the fed, and red dot on top of the CA. Obviously the resistance of the sender is different as well, all connections are identical. I used Subaru Parts Exchange, I would recommend them hesitantly, primarily for price, I am replying so slowly as it seems to take a good while to get parts, the first time I searched by year and model and the second by VIN as this dot discrepancy jumped out at me and while I never expected this to be emissions related, I wondered if certain VIN's had a 2013 or 2011 version that was different. The part only arrived Monday afternoon. Anyway, all the issues are solved, from what I can gather online and from the scan tool the resistance of this secondary sender has an effect on the evap canister purge and the fuel temperature sensor signal, hence the two models for emissions. Since changing the sender, the codes have not recurred (granted it hasn't even been a day and about 40 miles) where in the last run on the old sender, they recurred immediately upon being cleared, you really could not clear them. Besides P0463 I had a recurring code for fuel temperature sensor that took longer to reappear, about 10 miles, this is also seemingly resolved. The fuel gauge now reads normally. It seems cleaning the grounds as instructed by cardoc resolved the hard starting when hot. I had not expected such an issue, I am used to dealing with ground issues on more typical vehicles for me, I tend to buy 80's trucks and Mercedes as I just don't want computers and knick knacks on my cars, about as much luxury as I like is power steering, ABS and a radio, I like to stay as simple, rebuildable and analog as possible. On the Subaru, my hunch is that the issue is that the ground is to aluminum and it oxidizes over time, I do plan to add the second cable as Cardoc described, I will try to find the braided strap type though in a suitable length as I am not a huge fan of a cable sticking out over things and I prefer the way the straps can be kind of contoured out of the way. I do plan to get to Cardoc in the future as I would like to keep this car long term, to me it would be more practical to rebuild the engine and transmission someday and stick to what I know vs diving into the headaches of GDI and CVT's.

 

I never expected this to be emissions related, I wondered if certain VIN's had a 2013 or 2011 version that was different.

Yes, it looks as if the 2012 CA-spec uses the fuel tank sub-sensor/fuel pickup that is used in the previous years.

As I mentioned earlier, your CA-spec part number appeared to be relevant to the 2010 and 2011, but not the 2012. With your update, I looked at the 2011 FSM pages for the sub-sensor. Whereas the 2012 FSM indicates the sub-sensor resistance should range from about 9 Ohms (when tank full) to 270 Ohms (tank empty), as shown in the pages posted above, for the 2011 the range is 2 Ohms to 62 Ohms. That's a significant difference, especially as the tank level changes from full to empty. (Did you measure the resistances of the CA and FED sub-sensors?)

Some of the fuel tank-related emissions functions are dependent on the fuel level in the tank, typically being done only when the tank is between 1/4 to 3/4 full.

However, there is only one fuel temperature sensor, and it's on the main fuel pump assembly in the passenger side of the tank, so the sub-sensor assembly would not affect the fuel temperature reading.

Interesting outcome, and a reflection of further deficiencies in Subaru service documentation.

 

I am also an analog guy living in a digital world. Not many of us left these days.

 

I generally choose based on what will last the longest and nothing I have seen shows electronics to last very long, most new cars seem to start having annoying issues in only a few years and when the touch screen runs everything, they get really annoying. I chose this car as my wife wanted me to have something more modern and safer to take the kids around and I read a lot on the 3.6 and it has always made sense to me to overbuild something and not push it so hard, it lasts a lot longer than say squeezing the same power out of a small 4 cylinder. It has been very usable, I can get most anything at the hardware store or nursery and bring it home so I am a fan of the Outback chassis but unless I could get the diesel boxer, it would have to be a 3.6.

 

It may well be that the fuel temp sensor is not in anyway linked to the signal from the secondary sender, but it does seem to trip the fuel temp sensor code if something is wrong with the secondary, I had assumed the circuit resistance influenced the temp reading from the probe on the pump assembly, but it could just be the computer sees a signal conflict, empty but yet the sensor is still detecting a cooler liquid. When I was trying to figure this out on forums, this was a common link that identified a secondary sender failure, a fuel temperature sensor error code.

 

What code was that?

I found three in the FSM: P0181, P0812 and P0183. The latter two are for continuity or short problems in the wiring of the fuel temperature sensor. The first is for sensor "performance".

 

I can't claim to remember perfectly as it is now cleared, however, in my search history, the code I had searched for more info on was P0183, that is where I found a few posts that changing the secondary sender solved this issue and that was what lead me down this path, I was not searching specifically for Outback, search terms were just Subaru P0183. This was the first code issue I had after changing the primary sender/fuel pump assembly and that assembly is a genuine Subaru/Hitachi. I did also have an inaccurate fuel gauge but that tripped no codes or displays until a wrong secondary assembly was installed.