9pec

I have been reading this thread with great interest. I fried the ignitor pack once on my Mitsu V6 after a spark-plug-change, even with the batter negative disconnected. Human budy carry static charge that is capable of frying transistor. Static might have fried something half-dead during the timing belt change. For reference, this is a Mitsu ignition system info, but Subaru has very similar waste-spark system :- http://www.stealth316.com/0-frames.htm:D It seems one channel of the ignition still fires, but the other won't. While I have seen 3-cylinders engine runs rough with only 1 cylinder's ignition firing, at this juncture it should be reasonable to assume the problem is due to one ignition channel not firing. Crank sensor sends weak pules of current that the ECU converts into voltage pulses. That's the first "trigger". These signal pulses trigger a pair of Mosfet-gate-drivers that in turn, drive/trigger the Ignitor pack, in an inverted fashion, that is, it switches off the ignitors to fire sparks. Ignitor pack has 2 high current mosfet transistors that act as swicthes which connect the ground lead of the primary coils to ground. When the ignitor is at "on" stage, high current (probably 6-8 amps) goes from battery/fuse to primary coil, via the ignitor and then to ground. When you turn the key to "on", this coil-energising stage takes place but if you do not crank the car a timer will turn off the ignitor after a pre-set time. You can actually hear the coil energising once the key is switched to "on" position before cranking if you stick your ears closed to the coils. When ECU trigger the ignitor to fire, the mosfet-gate-driver pulses from the ECU turn off the mosfet swicth in the ignitor and cut the primary coil circuit from ground. Primary coil voltage shoots to few hundred volts in a split second, and this get magnetically coupled to secondary coil and shoots to few kilo-volts. Electrons now jumps across 2 spark plugs' tips to ground/and ground to tips respectively in 2 cylinders. One cylinder is near its compression stroke, and one cylinder is in exhaust stroke ("wasted spark"). I have always confused myself with EJ20's cylinder numbering system, so I must make it clear that I am assuming that when we are talking about cylinder 1-2 and 3-4 we are accepting that the cylinder pairs are the right pair. I say this because in a waste spark system, if a particular cylinder has no spark the other cylinder will have no spark as well since they are electricly in the same loop. Seeing that we have one channel still firing, I suggest pulling out the ignitor-to-coil connector and use jumper wires to swap the 2 channels. If we have spark in the other pars of cylinder after the swap, at least we know the ignitor, coils, spark cable, spark plugs are all fine. (note :- strong spark in free air doesn't mean good spark in compression-chamber, pressure and fuel make it harder to form spark). Again, we can try swapping the channels at ignitor-to-ECU connector to see if it is ECU having one channel fried, or the ignitor having one channel fried, or both having one channel fried! Perhaps one of the ignitor or ECU's triggering is dead. I am no expert but my understanding is, if the crank sensor is faulty, then we will have no spark at all in all cylinders. My Forester has so far not been giving me any electrical problem, only tons of suspension problem. I have no acess to any Subaru service manuals so please correct me if I said anything not applicable to Subaru's system. Just hoping to see Mjez nail the problem and get the car up and running soon.:cool:

Great! You have found the cause! Has the head cracked? Hmm....probably just a blown head gasket?

Error codes in respect of "air control valve" and "purge control solenoid valve"? Hmm.....these codes can be something else altogether. Not related to the coolant/head gasket issue but nonetheless maybe one of the factor contributing towards what's happening. On my 97 Forester the purge solenoid is beneath the intake manifold near the throttlebody. There is vacuum hose going from the manifold to the solenoid and then coming out from the solenoid going to somewhere else (the solenoid is a vacuum/pressure switch). At one time my engine will cough and choked and die and throw a code every now and then but not always. I eventually discovered that the vacuum hose going from the manifold to the solenoid has cracked. When I tried replacing it I then discovered that the one of the hose-connecion-nipple on the solenoid itself has actually broken-off! I got that fixed and the engine runs smooth as hell after that. My boost/vaccum gauge reading wasn't very stable at that tome so that has helped me to identify the problem quickly. Now if your car has similar problem (maybe the purge solenoid is leaking, stucked or failed entirely) this seems to explain the 2 codes. I think the idle air control valve will attempt to compensate the manifold vacuum leak and eventually the out-of-range compensation will cause an error code as well. On my Forester the idle air control valve ("IAT") is right next to the throttle body, light brown in color. Meanwhile, just some thoughts reading through the posts:- On boxer engines isn't there a coolant cross-over-pipe running across underneath the intake manifold? Perhaps there is leak on this pipe or it's gasket? Isn't the coolant temp sensor located on this piece? Since there is a head gasket work done earlier, and if things aren't put-back together properly.........well, I just thought pulling the heads will require taking off the manifold anyway so if the problem can be identified without actually having to take apart the heads would surely be good! I know, still it doesn't explain the exhaust smell in the radiator, so keep the finger crossed!

Check if the nut that is holding the wiper arm down isn't loose?

Sometime brake "shudder" isn't caused by by warped rotors, but uneven brake pad deposits on the rotors' surface. I would confirm that the rotors are actually warped before going straight to turning them. Turned rotors have less mass and will have less ability to disperse heat. Also, hot, thinner rotors have more tendency to crack/warp when come into contact with cold water splash. If the rotors are not warped, scrubbing the surface clean and put in new pads and follow proper bedding-in procedure will help. Do not use ordinary sanding paper! The abrasive materials are mostly alum-oxide and they will make things worst! I used 3M-Garnet Paper for rubbing and followed with Synko's Rotten-Stone for final polishing on rotors before each pads-change. Good results.

counterweight?

Don't give up. My experience with "weird" issue like this is that usually it'll turn out to be something very basic and simple....that's why it is difficult to find it in the first place. (Guys, if I say something wrong, please correct me. Not trying to be a smart boy here, but I really think he needs a structured approach to resolve this rather than hitting a moving target randomly. It is obviously driving him nuts:eek: ) Ok, let's trust the ECU for what it did do. The code tells us it's a faulty temp sensor or circuit. The aircon is functioning well, no overheating actually takes place, so radiator, cap, hose, pump, thermostat must be working. Hence, I am guessing it's electrical. Temp sensor is usually a thermister, resistance drops when temp goes up. ECU sends current through the sensor via 2 wires, and takes the voltage across the sensor as signal. Somehow the ECU has been getting voltage readings that are out-of-range, hence it stores an error code, and tries to runs a "safe" mode, ratarding timing, runs higher injector duty cycle etc. coz it thinks the engine may be overheating, which is actually fine. We'll need:- 1. Multimeter/Ohmmeter: a Taiwan-made al-cheapo one will do. To meassure resistance of the circuit and components involved. 2. An accurate circuit diagram of your car. (sorry I do not have one) You'll need to know exactly how the whole circuit is wired up in order to zero-in. Let's walk through how the electrical part of a cooling system usually work. Key swithed on. ECU checks signal from all sensos. Status check ok, it starts. Runs close loop warm up map, rich and advance, to warm things up. Waiting for temp signal to come into the prescribed range, meanwhile engine is really warming up. ECU now ready to cease warm up mode and go open loop. ECU receives signal from Temp sensor and............ohoh.... From this point onwards, I am guessing a few possible scenarios or a combinition of all of them:- 1) Voltage too high (resistance too low). That is, ECU thinks coolant is toooooo hot. Ground and run the fan circuit. Throw a CEL. Stop open loop. Goes into "save" mode. This may explain the low rpm (max duty cycle and full timing retard) If temp sensor gives good reading after a while, ECU goes open loop but stores the error code. OR 2) No voltage, i.e. no connection (open circuit). ECU thinks either the wires has snapped, or the sensor is dead as hell. Stop open loop. Throw CEL. Run "save" mode. Fan will not run in this case. OR 3) Voltage too low, current draw too high (the circuit is grounded somewhere). ECU cant keep up with the current draw and something will overheat and possibly a breaker circuit steps-in before something start burning. Throw CEL. Kill ignition and injection kill everything. Engine stop running. So this is what I propose:- a) Meassure the Temp Sensor's resistance (ohms) alone, out of the car. Record the reading down and remember this. Dip the sensor in hot water. meassure again. It should have a lower resistance now. If not, sensor is screwed. c) Find the 2 pins that link the sensor wires to ECU, on the ECU's side. Unplug that plug from ECU. This is where you need the ECU diagrame. One of the pin is supposed to be grounded in the ECU itself, or somewhere along the harness, or near the sensor on the block, etc. You'll need the diagram to idetify where to meassure. The diagram will tell you this. d) Plug the sensor into the circuit (without installing it to the engine). Take ohm readings between the 2 pins on the ECU plug's side. If the wires are fine, it should give you the same resistance reading of the sensor. Very,very slight increase of ohmage is ok (like 0.1ohm more) ...it's the wires' resistance adding up. If no readings (open circuit), the wires are faulty. If the resistance shoots very high....connectors are dead. Clean them, or replace them. If the resistance drops through the floor, the wires' skin has probably shreaded somewhere and is touching the chassis, i.e. the circuit is grounded, which is not supposed to. e) Repeat this with the sensor installed to the engine. Note if there is any difference after the sensor is installed. If the ohm drops, the sensor body is grounded to engine which is not supposed to. Sensor is at fault. e) if all things are within spec, suspect the ECU itself. Open the casing. If visually you can see anything burned, snapped etc....well you'll probably should hunt for a replacement component or replace the ECU unit. I am guessing there is nothing wrong with the fan system because the fan runs once the ECU decides to ground it and complete the circuit. If the fan system is at fault, should it throw a "temp sensor" code or a seperate code? Good luck.

Good to hear about the increased ground clearance! The positive camber issue can be fixed with a camber plate even if it can't be resolved with the stock camber screw adjusted to max negative. In addition, instead of removing the spacer entirely some shims can be swapped-in to limit the effect. But why would the halfshaft's life affected? I would worry about the lower ball joint's wear instead, as on WRX STI Spec C the ball joint seat on the aluminium arms were revised to have an angle instead of vertically positioned, so that seems to suggest that there might be issue that Subaru is trying to resolve.

It is possible that the block did experienced spun bearings before and had since been fully rebuilt. crankshaft could "walked" with severely spun thrust bearing and the marks may have been a result of that. Crankcase with some metal shaved off like this will not be a problem, I guess. I would pay more attention to the crankshaft that may have came into contact with the block instead, to check if it's still balanced and within spec?

Why don't you check the thermostat? It could be stucked opened when it's cold, and stucked closed when it's hot? Radiator clogged? Give the whole system a good flush? It may not be the sensor or fan afterall?

Well not really. Let's leave the handling solution aside for the time being, although I am in the process of doing just that. Rather, what I would like to find out is how these two pieces fit into the Forester's overall suspension geometry in comparison with the sedans sharing the same platform, e.g. WRX/Legacy/etc. It seems to me that by mouting the Foresters' x-member directly to the framerail should result in a higher front roll center. In fact I suspect the reason why Foresters have front-control-arms rear bushings that sit slightly lower (similar to Whiteline's Anti-lift Kit for the WRX) is to work with these spacers to keep the front geometry in tact. I am just trying to see if anyone have done it before, and how the associated challenges are resolved, e.g. the pitch-stop rod needs to be shortened? tranny mount needs spacers? Will downpipe hits the body? Will it screw up the propeller shaft agle? etc.

i do not know what it's called. From tapping them lightly they do not seems like rubber to me (perhaps mine are already hardened from old age! ). I am certain that on WRX/Impreaza chassis the x-member is bolted directy to the framerail without these pieces in between.

In SF5 chassis (not sure about SG) there are 2 "wedges" or spacers sandwiched between the front x-member (engine cradle) and the frame where it is bolted to the frame, about 1 inch thick. Has anyone ever tried raising the front roll center by removing these?