PacMan

Sorry Buddyepa, but I can't give you anything specific at the moment. However, I did get some news today from a guy that I've been collaborating with about the problem and I can say that it was very good news indeed. We will still need to do more testing and, since two of my test pilots are away for for a few weeks, this will take some time. I'm still looking for Test Pilots in Sydney, Australia. Come on guys, my cars now running fine - I need another broken one to work on locally. There must be hundred's of Liberty's in Sydney with this problem. PacMan

Yeah Randy, you're right - it's time for an update... OK, we've now got 4 FixKits installed (including my original). We've got two cars fixed (mine is one - no problems for 3 months now). BUT, we've got two vehicles that are, say, 90% fixed. For normal driving the cars are running fine over terrain which they previously would have had the rough running problem. But, if the owners drive them up very long steep hills, they still fail sometimes (but they used to fail always - so we're getting somewhere). At the moment we are not sure whether the kit is not 100% effective or if these two cars have additional problems which the kit cannot fix (e.g. poor fuel pressure or faulty ignition components). Sorting this out is taking a lot longer than I had anticipated as both these vehicles are a long way from where I live (One is in another country). I have also had the pleasure to have been contacted by another EE who has been working on a 1991 Legacy and he's implemented the changes internal to the ECU, similar to those changes which I originally started with, plus one of my changes to the Injector wiring and he emailed me just today to say that his car is now running fine. So it would seem we're on the same path. I'm hoping to work with this guy to sort out some of the remaining problems (if they exist - at the moment I really don't know). Since my car is now working fine it's hard for me to visualise why the kit isn't fixing the problems in other vehicles all the time. What I'd really love is for some people with this problem who live in Sydney Australia (where I live) to email me and we can work together hopefully to nail this problem once and for all. Please email me at xxxxxx@iinet.net.au (replace the xxxxxx with pacman50 - just trying to avoid some of the spammers). One really strange thing is that I was walking down the street a few weeks ago and saw a guy getting out of what looked to be a 1992-1993 Liberty (Legacy) sedan. I asked him if he'd experienced this problem and he said "yes, it was quite annoying". I then offered to help him fix it and he said words to the effect of "Heck, no way! I just pull over, restart it and then it's fine. You might make it worse". I tried to point out that it's actually a serious safety issue when the engine craps out while you're being chased up a hill by a big truck and I would also do the work for FREE - but he didn't give a toss. The answer was NO. Amazing! I'll Be back again when I've got something constructive to say... Cheers, PacMan

Matt, sorry to confuse you with the ECU pin connector numbers I gave you. These were for the 1990 to early 1995 series Legacy's. You obviously have the updated model (introduced in mid 1995 in Australia) and the ECU and the wiring is drammatically changed from what I have to work with (a 1992 model). However, you should still get the wiring diagrams for your car and check the supply voltage to the ECU. ESPECIALLY since you've just found a fried relay! I don't know what you did to repair the damaged relay but I think you need to work out how it got fried and then make sure both the cause and the effects are repaired - fix the obvious things first. I don't have a wiring diagram for your car so I'm gonna butt-out of this thread. Good luck.

I think "PEBBLES" was referring to the Ignition Relay. This thing is known to be a problem. If the contacts on this thing are badly burnt it will starve the ECU of power and the ECU won't run properly. Check the voltage on pins A2 and A13 of the ECU with the ignition switched ON. It should be within 0.5 volts of the battery voltage - if it's not then it's probably a faulty Ignition Relay. Make sure that the battery is fully charged when doing this work. It seems like the car won't even fire although it's got gas in the cylinders - but it should fire even if only cyl 1 & 2 are sparking - unless of course, the timing was way off. The timing is controlled by the ECU. It should be possible to use a timing light to check if cylinder 1 is firing near top dead centre - even though the car won't start.

Hi Brama, this is a fairly well known "feature" of the 1990-1995 liberty 2.2 litre Liberty/Legacy models. There are quite a few threads on USMB (and other Suby support sites) regarding this problem - these are the most recent on USMB, I think. http://www.ultimatesubaru.org/forum/showthread.php?t=19616&highlight=gremlins http://www.ultimatesubaru.org/forum/showthread.php?t=22907&highlight=leggy+trouble If you read these threads you'll see that I have fixed this problem with my Liberty and was trying to develop a kit to allow other Subyholics to also fix this problem. The current status is that I have just shipped two of my new design fix-kits to two guys who are having this problem - they are my fix-kit test pilots. One lives in Australia (93 Liberty auto AWD wagon - non-turbo) and the other lives in New Zealand (1990 Legacy 2000 GT turbo). Nobody from the US responded to my requests for volunteers. I'm waiting for my testers to confirm that they've received their fix kits. Then they have to install the kits and do some testing. I'm hoping that they will get back to me in about 2 weeks (hopefully less). Brama, stop throwing money at the car for a while - the MAF is expensive and, whilst a dodgy MAF could cause your problem, I don't think it is your problem (the MAF sensor seems to be pretty darn tough). One thing though, I'm starting to believe that the pinging you describe is another part of the problem. This could explain why these cars don't seem to show the symptoms until they get about 150,000 Kms on the clock. RichieRich has definitely fixed cars by using Moto-Vac (see posts from T.J.) which I think possibly works for some cars as it probably reduces the engine pinging under load. My kit also reduces the pinging on my car when going up hills (but it does not entirely eliminate it - my engine has 205,000Km on it). Therefore I think that continuing to use fuel treatments that remove valve and combustion chamber deposits can help reduce the problem, but again, it isn't the whole story. Cheers, PacMan

Well my last post seems to have "lit the blue touchpaper"! Firstly, I'd like to give credit to sea#3 for determining that the Jideco relay was a source of problems in this area. I've now spent many hours upside down under the dashboard of a Subaru and I've got to admire the fault finding tenacity of a person who found this one. I'd be happy to take my car to such a person - they are few and far between down this way. As for "calling ya out" - not so fast buddy! Secondly, I'd like to point out that I didn't resolve this problem "yesterday" and then test it with a "quick blip around the block". I fixed the problem 2 months ago and because it's such an intermittent problem I've done as much testing as possible to ensure that my fix is as close to 100% as possible. My first fix was to actually modify the guts of the ECU computer so it wasn't so susceptible to "electrical noise". This fixed the problem, but I figured that telling the Subaru world to rip out their ECU computer and start hacking at the circuit board was probably going to cause more problems than it fixed. But it caused me to go looking for where the electrical noise was coming from - this is when I discovered the wiring fault. I then designed the MKII fix which was a combined electrical filter and relay replacement box which plugged into the Jideco Ignition relay socket. This plug-in fix also needed additional changes to the wiring loom under the bonnet (hood) to fix the wiring fault. I then went and spoke to my local Suby wrecker (a nice guy and very helpful). Together we decided that asking someone to climb under the dashboard and remove the Ignition relay and then plug a cable into the now empty socket would be like asking your mother to replace a gudgeon pin on the Space Shuttle - it wasn't going to happen (but, then, maybe your mother is into that sort of stuff). Current plan is:- My Suby wrecker is currently cutting a wiring loom out of a wreck and once I've got this loom in my workshop I will be able to work out the easiest place to tap into the wiring to accomplish all these things... 1) Fix the wiring error. 2) Filter the power to the ECU computer 3) Bypass the troublesome Jideco relay. Once I've figured this out, in the words of that famous Austrian guy (ooops, I meant American guy) "I'll be back".

Thanks for clearing up the identity thing TJ. Well my last post seems to have lit the "blue touchpaper"! Firstly, I'd like to give credit to sea#3 for determining that the Jideco relay was a source of problems in this area. I've now spent many hours upside down under the dashboard of a Subaru and I've got to admire the fault finding tenacity of a person who found this one. I'd be happy to take my car to such a person - they are few and far between down this way. As for "calling ya out" - not so fast buddy! Secondly, I'd like to point out that I didn't resolve this problem "yesterday" and test it with a "quick blip around the block". I fixed the problem 2 months ago and because it's such an intermittent problem I've done as much testing as possible to ensure that my fix is as close to 100% as possible. My first fix was to actually modify the guts of the ECU computer so it wasn't so susceptible to "electrical noise". This fixed the problem but I figured that telling the Subaru world to rip out their ECU computer and start hacking at the circuit board was probably going to cause more problems than it fixed. I then went looking for where the electrical noise was coming from - this is when I discovered the wiring fault. I then designed the MKII fix which was a combined electrical filter and relay replacement box which plugged into the Jideco Ignition relay socket. This plug-in fix also needed additional changes to the wiring loom under the bonnet (hood) to fix the wiring fault. I then went and spoke to my local Suby wrecker (a nice guy and very helpful). Together we decided that asking someone to climb under the dashboard and remove the Ignition relay and then plug a cable into the now empty socket would be like asking your mother to replace a gudgeon pin on the Space Shuttle - it wasn't going to happen (but, then, maybe your mother is into that sort of stuff). Current plan is:- My Suby wrecker is currently cutting a wiring loom out of a wreck and once I've got this loom in my workshop I will be able to work out the easiest place to tap into the wiring to accomplish all these things... 1) Fix the wiring error. 2) Filter the power to the ECU computer 3) Bypass the troublesome Jideco relay. Once I've figured this out, in the words of that famous Austrian guy (ooops, I meant American guy) "I'll be back".

FIXED FIXED FIXED FIXED FIXED FIXED !!!!!!! Oh, by the way, I fixed the problem!!!! I spent nearly $500 on parts chasing this problem and a couple of hundred hours with my head under the bonnet and the dashboard (yeah, we like the car a lot - when it's running well). After replacing EVERYTHING (MAF sensor, coil pack, Fuel pump, pressure regulator, knock sensor, ECU, putting heaps of valve cleaner through it and a heap of other stuff) I'd run out of ideas and was looking to buy another car. I then decided to build a data logger to monitor the ECU inputs while I was driving the car. This is basically a device that monitored the inputs from the MAF sensor, the oxygen sensor, throttle sensor and also the outputs to the fuel injectors. I also hooked up a fuel pressure guage. It's amazing to watch all the sensors and the ECU at work as the ECU constantly adjusts the fuel mixture for optimal engine performance and efficiency. I found it hard to drive whilst also watching all the lights and displays but, in the end, it paid off. What I found was really bizarre. And I can tell you guys right now "stop swapping parts" because you will NOT fix the problem. Firstly, the problem is a design problem (yes, even Subaru get it wrong sometimes) which is causing the ECU computer to "crash". This is why switching it off and restarting makes the problem "go away". Secondly there is a wiring error in these cars - my wiring diagram does not match the wiring that is actually in the car. If my wiring diagram is wrong then Subaru have two design problems. The problem will show up with different symptoms too. Usually no engine faults are recorded (probably because the ECU has crashed) but if a fault is recorded - Oxygen sensor faults are the usual ones but also knock sensor faults will show as well. Other symptoms are lack of power, misfiring and engine pinging. I'll give you a clue - the engine is running very lean but the ECU computer cannot compensate for this and it cannot work out why. It does seem strange that the problem only seems to appear as these cars get older - the Jideco ignition relay contacts deteriorating definitely play a part in this but the relay is only exacerbating the other underlying problems. I have DEFINITELY fixed my car and I can actually reproduce the problem at will. However, fixing the problem is not easy unless you're an electronic engineer (lucky for me, I am an electronics engineer, heh, heh, heh). I'm trying to work out whether I can make a kit of parts to fix the problem which can be easily installed by a handyperson. My current fix is fairly complex and would confuse the heck out of non-nerdy electronic type people. How many people are still interested in getting a fix for this problem?

FIXED FIXED FIXED FIXED FIXED FIXED !!!!!!! Oh, by the way, I fixed the problem!!!! I spent nearly $500 on parts chasing this problem and a couple of hundred hours with my head under the bonnet and the dashboard (yeah, we like the car a lot - when it's running well). After replacing EVERYTHING (MAF sensor, coil pack, Fuel pump, pressure regulator, knock sensor, ECU, putting heaps of valve cleaner through it and a heap of other stuff) I'd run out of ideas and was looking to buy another car. I then decided to build a data logger to monitor the ECU inputs while I was driving the car. This is basically a device that monitored the inputs from the MAF sensor, the oxygen sensor, throttle sensor and also the outputs to the fuel injectors. I also hooked up a fuel pressure guage. It's amazing to watch all the sensors and the ECU at work as the ECU constantly adjusts the fuel mixture for optimal engine performance and efficiency. I found it hard to drive whilst also watching all the lights and displays but, in the end, it paid off. What I found was really bizarre. And I can tell you guys right now "stop swapping parts" because you will NOT fix the problem. Firstly, the problem is a design problem (yes, even Subaru get it wrong sometimes) which is causing the ECU computer to "crash". This is why switching it off and restarting makes the problem "go away". Secondly there is a wiring error in these cars - my wiring diagram does not match the wiring that is actually in the car. If my wiring diagram is wrong then Subaru have two design problems. The problem will show up with different symptoms too. Usually no engine faults are recorded (probably because the ECU has crashed) but if a fault is recorded - Oxygen sensor faults are the usual ones but also knock sensor faults will show as well. Other symptoms are lack of power, misfiring and engine pinging. I'll give you a clue - the engine is running very lean but the ECU computer cannot compensate for this and it cannot work out why. It does seem strange that the problem only seems to appear as these cars get older - the Jideco ignition relay contacts deteriorating definitely play a part in this but the relay is only exacerbating the other underlying problems. I have DEFINITELY fixed my car and I can actually reproduce the problem at will. However, fixing the problem is not easy unless you're an electronic engineer (lucky for me, I am an electronics engineer, heh, heh, heh). I'm trying to work out whether I can make a kit of parts to fix the problem which can be easily installed by a handyperson. My current fix is fairly complex and would confuse the heck out of non-nerdy electronic type people. How many people are still interested in getting a fix for this problem?

Love your call RichieRich! Hey, what have I got to lose here? I'm darned if I can figure out why it doesn't wanna go up certain hills. I'll give your idea a shot - but can you clarify what is "Motor Vac"? I live in Australia so if you can tell me who make this stuff and what it's called on the label I'm sure I can find it. I've been thinking about this a bit more and it occurs to me that it can't be a bug in the Engine Management Computer program otherwise all these Lagacy/Liberty models would have had the problem right from day one and Subaru would have gone bankrupt by now. Therefore it's probably something to do with the fact that these 2.2 litre (pardon my "Queen's English" spelling) engines are getting pretty old now and the concept of the valve gear mis-behaving makes sense. I'm also starting to wonder if the problem is not caused just by the fact that I'm driving the car up a (fairly) steep hill. But the problem may be started by first letting the car roll down a long steep hill with no throttle first (high engine manifold vacuum for a couple of minutes). This may be sucking oil down worn valve guides and causing the valves to stick? Maybe it's something to do with the hydraulic lifters? Anyways, I'll give Moto-Vac a shot and see what happens.

Thanks for the Injector Cleaner tip TJ. The thing that troubles me is that I can't see how Injector Cleaner will "fix" the problem. The reason for my concern is that after the engine runs rough, I switch it off and it restarts fine. Then the car will drive up the exact same hill without a problem. So if it were clogged/dirty injectors then just restarting the car would not fix the problem (for a while). I did find that the Ignition relay was faulty as described in this site I whacked the relay with the engine idling and, sure enough, the engine cut out. Wow, I thought, this is it. But I cleaned the relay contacts and put it back in the car and now you can whack the relay all you like the engine does not miss a beat. I was sure this fixed the problem until a few weeks later when my Wife reported that "it's up to it's old tricks". I checked the relay again and it's fine. So it's something else. Another thing I've noticed is that the problem has occurred three times at almost exactly the same place on the same hill on the way to my Wife's folks house. The last time it did this I went back down the bottom of the hill and drove up again - the second time I was giving it some serious gas - no problems. Then, to my Wife's annoyance, I went back down and drove up the hill again gentle-like and I was really expecting it to fail - but the darn thing worked fine. Grrrrrr! I've had it in the shop and the guy said he couldn't find anything wrong but cleaned the injectors anyways. The thing sure idles nicely now but the problem still occurs. I'm starting to wonder if there's a bug in the Engine Management computer program. I can't see that it could be a crank/cam sensor because it's failed at the same place on the same hill three times - hills don't change sensors (do they?).

Did you ever resolve this problem? I have a '92 Legacy (actually a Liberty - I'm in Australia) with 200,00Km on the clock - and it's been doing this for a many months now. I've replaced the O2 sensor, Fuel Pressure regulator, Ignition relay and I even tried swearing at it but no luck. One thing that I have worked out is that it mostly happens when starting to climb a steep hill and slowly pressing the throttle to the floor. Before the auto kicks back a gear you can feel the engine start to run really rough and the power drops to almost nothing. Pull over, switch off, say a prayer or two, restart and it's fine - maybe for weeks. I dearly love the car but this is driving me nuts. I find many reports of this problem on the web but no concrete solution (that I haven't tried). Someone MUST have worked out what the heck is causing this problem by now - please respond!