Silent Bob

I think you need to put a proper set of gauges on it to see what the system is actually doing. Just because the gauge on your can tap is in the green, does not mean the system is correctly charged. I think at this point you should have the system properly evacuated, vacuumed and charged by weight. That way you have a good baseline to start diagnosing the problem.

This^ It's more than 5, but not quite 5 1/2 But don't over-fill it. That can cause it to set codes for the variable valve timing.

There is a pipe connection at the evaporator core. I suppose it could be leaking there, but the catch is you still have to remove the dash to get to it. At that point you a probably farther ahead to put a new core in it just in case, rather than having to pay someone to do the job twice. I have replaced a couple of cores here and there, so they do fail occasionally. Another thing to think about: I don't think there are a lot of techs who would jump for joy at the thought of replacing one, it's kind of a time consuming, PITA job. If it was leaking somewhere else, I have a feeling that they would tell you, since most any other leak would be more of a "gravy " job.

Mitchell flat rate time to replace evaporator core is 6 hours, includes remove & replace instrument panel. (whole dash has to come out to replace core). If there is dye coming out of the evaporator drain, the core is about the only place it can be coming from. So, they are probably being fairly straight with you.

Run the wires through the hole under the spare, zip tie excess wire to hitch. That's the intended route.

If there is fuel in the tank, and the pump is running but nothing coming out, pull the unit out of the tank and make sure the pump itself is still connected to the outlet tube. Also, some models (not sure if 96 is one) have a "pulse damper" on the fuel pump module. The cap can come loose causing the fuel to flow back into the tank instead of out the line.

I have seen one with 150k. '10 Outback owned by someone who had a newspaper route or something similar. Car was beat to death, but the trans seemed to act ok. Fluid was pretty black looking though.

Oil changes at most shops are a loss-leader $29.95 (or so) special just to get people in the door to upsell profitable jobs. They typically pay the tech .3 to .5 hr. We get .4. I'm not sure there is an official time.

This may sound a little off, but if you can find a spare, swap out the MAF with a known good one. My 95 had a MAF problem with the characteristic random stalling. BUT- a time or two it threw trans codes in the process. Why, I don't know. Maybe just the sputtering, stalling, and erratic signals caused the TCM to flip out.

Communication problem: first thing I would do is start checking fuses, particularly anything 7.5a Option 2, is that during the "update" something went awry and it bricked the TCM. May be possible to retry and recover it (never had that happen myself, so not absolutely sure) R&R valve body looks to be 1.8hours according to Mitchell, so not out of line. Sounds harder than it is. Another, more experienced dealer sounds like a good option here. Most are only too happy to point out what someone else has screwed up...

Before things get totally out of hand, look under the dash, just left of the column, and make sure the TCM is fully plugged in. Unplug the connectors to the TCM and reconnect them, ensuring that they "click" into place. I can envision a scenario where the TCM was unplugged somewhere along the line and one or more of the connectors is not fully seated.

Check the ignition switch. Make sure the "ignition" circuit is not dropping out and losing power when you turn the key to the "start" position. Should be a red wire at the ign. switch connector.

Guess I should have been more specific. I was referring to serpentine belt idlers & tensioner- Might potentially make some nasty noises when the bearings come apart, which they do regularly. Quick and easy thing to check. Some people assume the worst when they start hearing funny noises- thus the "don't panic". Of course if it's not that, It probably is something internal, complicated, and expensive...

A rod knock will not usually turn into a grinding noise. Check the idler pulleys and belt tensioner before you panic.

Check the bulb with a meter, even if it looks ok. New bulbs can be bad too. That said, the lamp harness you have pictured is a fairly common failure, usually the bulb connector gets burned/melted.

Funny, I put a switch in an Impreza last week. I think there were at least 3 or 4 failed aftermarket switches sitting in the console.

Fairly common on '11 FB engine. It is sealed to the head with RTV, and the earlier ones like to leak. The engine has to come out to reseal.

MAF fixed mine. No codes, random stalling, would restart and run fine for days/weeks.

This. The low beam bulbs on a '10 are a pain to change with the lamps in place. Very easy to misalign the bulb and have it be very dim and aimed way off. Make sure they are correctly seated in the housing. I have seen this happen many times.

Update: Clicking noise turned out to be the trip meter which coincidentally got stuck at the same time. So, total red herring there, but at least not driving me nuts anymore... Fuel pump& main relay & connectors look nice and shiny. Dug up a spare MAF out of the pile-o-parts. Figure I'll swap that out after lunch & see if is makes a difference.

I could probably write a book on this, but I'll try to keep it as concise as I can. The victim: 95 legacy wagon, ej22, auto, awd, 250k miles The symptoms: Intermittently stalls/dies/sometimes just hiccups. - Happens every couple of days most often after a hot soak. So far, will restart immediately. There is a relay repeatedly cycling on and off under the dash- only when in gear, seems to cycle faster as speed increases. Fairly certain it is not the fuel pump relay or main relay, clicking. No codes in ECM. On one occasion, TCM set code 23 (engine speed signal). Driving along with a code reader connected, reading live data, it will sometimes loose communication with the ECM. This is sometimes, but not always, followed by a hiccup/stall. Code reader does not seem to lose power. Tried so far: Swapped coil, threw a crank sensor at it.(cheap, why not...) Cleaned up as many grounds as i can find. Fuel pressure +-32psi-( need to rig up the gauge so I can monitor while driving). Battery & alternator test good. I have a hunch that the ECM is beginning to get flaky. I say this because even if I start unplugging sensors, it will not set any codes. (except for the TPS) Just wondering if I'm missing something obvious. Any suggestions/advice greatly appreciated. -Bob

Rear tone ring: Remove caliper& rotor Remove parking brake shoes. Unbolt tone ring from hub. You can then move the tone ring aside and have enough clearance to remove and install the studs.

Not mine, but pretty cool. http://roanoke.craigslist.org/mcy/1644511602.html

Chased the same problem for a long time on my 91. Check/replace every vacuum line you can find, same with pcv hoses. Replace the oil filler cap and the o-ring on the dipstick- I think that helped mine more than anything.

My advice would be go with the Audiovox electronic cruise kit (mod# 250-1316) I've installed them on a lot of vehicles, generally pretty easy. Hardest part is usually just finding a good place to put the brain. Connect to throttle cable, just requires electrical connections to brake switch,vss, and switched 12v (&tach if so desired). The control switch looks a little cheesy, but overall they seem to work pretty well.