Log1call

Rough!! Take the cable terminal off the battery, clean it inside and the battery post outside, then put it back on and try without the ugly screw poking our of your Subaru. If you don't I'm going to tell Subaru on you!

Sitting at the lights with the clutch in is bad for them. Starting the motor with the clutch in is even worse due to no oil pressure at that stage.

It might be the brake band needs adjusting. Go here and download "transbrakebandadj", have a read... http://cid-4ca3c3459aaa7f7f.skydrive.live.com/browse.aspx/.Public?uc=1

Did you get the circlip in?

Inside there are some sliding hubs that lock the gears to the shaft to engage a gear, they sort of slide over the gears. There are forked/fingered sort of castings that are fitted to shifter shafts, and those fingers slip over the hubs to slide them along to engage gear. The gearchange lever you move goes inside and chooses one of three shift shafts when you move the lever left and right, then when you push forwards it slides the shaft which carries the fingers which push the hub onto the gear. The cast fingers are a bit deicate and they are only spring/roll pinned to the shift shaft, so, they are vunerable to breakage on most makes and models. It could be the pin, the finger, or the end of the shaft where the gear lever chooses which shaft. Try and find a manual online for it and it will have all the pictures.

The clutch is working isn't it? Sounds like inside the box to me, probably a shift fork.

My pump has never played up but apparently they can sometimes make noises before they die, other times not. Best test for the fuel pump is to test it's flow and it's pressure.

Since there are no codes, other than a 420? which I have never heard of, I would be inspecting the main relay, or, since it is intermittent, taking it out and putting it in another car to try or try a known good relay in your car.

Hi, I have a description of a code 1446 with the same description as your 1445. 1100 I have no record of but there is a 1110. They are both described in detail in a PDF file called (04_DTC Criteria" which you can down load from.... http://cid-4ca3c3459aaa7f7f.skydrive.live.com/browse.aspx/.Public?wa=wsignin1.0&sa=90828925 Might be worth a read.

Greetings Olnick. Thanks for the welcome. I have done as you suggest and put my location in.

Agreed, the axles are very sensitive to throttle, which wheel balance isn't. Axles and wheels all turn at the same speed. The driveshaft to the rear can also vibrate but it is a lot faster than the wheel/axle. If you stick a large piece of sticky tape to a tyre you will probably be able to hear it hit the ground as the wheel turns. If the vibration is at that speed, suspect wheels or axles. If the vibration is at about four times the frequency of the tape, it could be driveshaft.

I'm in New Zealand. There is always a lot of confusion(for me), the same model over here can sometimes be older, and yet sometimes more sophisticated than the American models. If I ask questions, it's not being critical, it's because I need to know if I'm to help people. I may not be up to the play with your American models, but, my logic and diagnostic skills are normally very good. I used to be the highest paid toyota mechanic in New Zealand a few years ago. I got to that position after three months with toyota, they said to name my price.... based on my diagnostic skills. There is more to diagnostics than guesses based on statistical probability(which is also a valid line of investigation).

Well it's not obvious to me apart from the trouble code(which was what I was assuming he had read). Over here, that car wouldn't have been OBD2. I didn't read that he hadn't had it on an oscilloscope, so that seemed like a possibility. No matter. I will assume then that it's coils feed either both front or both rear cylinders? In which case it won't be coil. Cam belts are easy to check. Headgasket would be the next thing to check after that.

First thing, since the light is on, is to check for trouble codes. You can check for trouble codes by connecting two plugs under the dash. It might be a good idea to check what the vacuum is at idle. Vacuum leaks or worn motors give those symptoms too. You might like to go here... http://cid-4ca3c3459aaa7f7f.skydrive.live.com/browse.aspx/.Public?sa=12810571 and read "diagnostic aids for".

There are specific cylinder missfiring trouble codes in OBD2. I know how the OBD2 computers calculate a specific cyinders missfire. As I understood it, poster one had those trouble codes. If he didn't have a trouble code that specifically stated it was cylinder three and one then he has diagnosed three and one cylinders as missing some other way... an oscilloscpe for instance. If he got a missfire on a specific cylinder trouble code then he has OBD2. If he has OBD2 then it will also record a cam sensor code if the belt jumps.

It can be a symptom of a crook axle, although in my experiance, the axles tend to be set off by something like a pothole and then they vibrate until you either change the throttle setting, speed changes, or the axle just aligns itself again after a few seconds. Wheels out of balance tend to be much more regular and vibrate once you hit the critical speed. If you suspect an axle then have a look at their cv boots, grab the axle in the cantre and check for play. It is, in my experiance, the inner axle where it fits onto the gearbox that causes the vibration so check there too. There is always some play where the axles join the gearbox so compare the two sides and if possible check some other cars that don't have vibrations. To test the axle on a road test try and hit a pothole with one wheel at a time and see if the vibration is set off. Often it only happens at certain throttle settings so try it with light throttle, heavy throttle and backed off throttle.

Hmm. Poster one says it is a cyl 1 and cyl 3 missfire code, which OBD2 can post. Older subarus I agree don't set cylinder missfire codes. Perhaps poster one needs to verify what the trouble codes were?

Hi Gary, all due respect(your suggestions are always excellent), but..... If it's late model enough to have a missfire code then it should have a cam angle sensor code that will set if the timing doesn't coincide with the ecu's expectations wouldn't it? Apart from that, yeah good suggestion and easy thing to check too.

Hi, I'm in New Zealand so bear with me if I sound a bit stupid.... Our cars are often different to yours. Does that car have one coil for the front cylinders and one coil for the rears(mounted on top of the motor with long wires), or is there one coil for left cylinders and one for right cylinders? If it's a left/right coil setup, I'd suspect the coil or ignitor. If it's a front/rear cylinder setup I'd suspect a mechanical fault, like head gasket.

Oh, that was for an older model by the way, just thought it might be worth investigating.

There was a service bulletin a few years back about wind noise from the drivers wing mirror. It was caused by the mirrors mount being too far in against the door and the test was to press against the inside where the mirror mounts and see if the noise got better, or even different, when you pressed it out a fraction of an inch as you were driving. If it made a difference then you fitted the thin plastic gasket subaru supplied and you put it under the mirror mount to pack it out from the car a fraction. A piece of cereal packet thick plastic it was from memory.

That site was perhaps a little biased... they were selling sensors! Their description was that the anti-freeze would eat away the sensor, they didn't say it would interfere with it's operation exactly. I still think that antifreeze will not set a O2 code or cause your ecu to missbehave. I'm a mechanic and have done dozens of head gaskets and never had to change a O2 sensor afterwards. Anything that can make the motor missfire, or run a bit too lean or rich, can set that trouble code. Since it only happens at highway speed I would doubt it is an air leak. Have you checked your air and fuel filters?

Sounds like you have already decided to do it but.... The two parts both fail with different symptoms. If the solenoid fails the transfer clutch locks up and you have full time drive to the rear. If the clutches fail you have no drive to the rear. What is it that makes you think there is a problem with the transfer system ?

First off you need to be sure you put enough in to get picked up by the pump. If the pump has run dry it may need fuel almost around it to prime, so put plenty in. Second thing is that the fuel pump can be primed by turning the key on and off. Every time you turn it on the fuel pump runs f about two seconds. Most cars will reprime their fuel lines with about ten turns off and on again. That way you save battery cranking while the fuel gets there. if yu don't hear the fuel pump run back in the fuel tank for about two seconds when you turn the key on, there is something wrong like a fuse or connectro.

I don't think coolant is rated as O2 sensor safe but i don't think it causes sensor failure either, not in my experiance. The problem that causes coolant loss can cause O2 sensor readings to be out and therby cause a O2 code however. Anything that can cause a missfire or a lean or a rich mixture can set that O2 code, it doesn't mean the sensor is faulty, just that the reading it was feeding into the ecu was out of acceptable range. Since it only does it on the highway I'd suspect it was just a missfire that was going unnoticed.