EJ25 running rough

[mad max]

After many attempts I got on to the fly-soob site and asked my question. The next morning I attempted to get on the site and it gave me the run around. What was wrong with the 1960"s -- No F#*king computers and loverly carburettors on top of all engines?

[birdy]

Sound like you need a rotax fix Max.

[mad max]

I like the sound of the big heart throbbing ej25"s Birdy, the same madness that attracts Harley owners ;D

[birdy]

Solong as its not relyn ona computer controled pace maker ay.

[mad max]

Russ, you said you saw my post on flysoob and a reply. It won"t let me back on. same password ect. Did they say anything interesting? I"m still trying to get back on it.

[Tim]

G"day Max,You probably already know this but I will state the obvious anyway. It has probably already been said but I haven"t been keeping up with the threads very well lately. Most original Subaru computers (and many other brands) have "live" injectors, that is to say that they are permanently "live", some models even with the ignition switched off. The computer is told by the crank angle sensor (and sometimes also the cam angle sensor) exactly which injector to fire and when. All the computer does then is earth that injector wire to complete the circuit.If that particular return wire from that injector back to the computer happens to be rubbed through or bared somewhere in the loom and intermittently touching earth, it will activate the injector each time it touches, creating an over rich mixture in that cylinder. Like I said, this has probably already been suggested, but I am too lazy to go back and read the whole thread.Regards Tim.

[chook]

Thanks Tim. This is a common design method of safety. If you short to ground the return wire it will only activate the injector. If you short the active or live wire coming from an internal semiconductor then you end up causing damage to the semiconductors within the ECU box which is expensive! This way the live comes from a fused bus - fuses are cheap to replace.I have suggested that Max puts his CRO leads onto each end of that feed wire to injector #4.He has at this stage convinced me that whatever the problem is, it is associated with that #4 injector circuit.At this stage he has replaced the injector and he has replaced the ECU (which I am assuming is connected directly to injector #4 via connector pins)This ECU has been further tested by Russ (?) back into his own machine.So that leaves only one last link - the wiring between the two.And as a tech I would be looking at both ends of that feed simultaneously on the CRO. The problem is that actually connecting to the ends is particularly difficult because they are well insulated. To be fair, Max has already had to perform a significant amount of work just to get to this stage - like shifting those injectors would not have been a 5 minute job.This next course of action will either confirm

[mad max]

Good to hear you on the forum Tim. I"ve just spoken to Russ and he has read my Fly-soob answer. Apparently the suggestion is I have a fault with the timer which tells the ECU that the wheels are turning and it"s causing the engine to go into limp mode. My frequency meter tells me it"s working but I shall follow up on that angle for now.

[mad max]

Yes Chook, I"ve rewired the 4 injectors with nice new coax cable. The props still on as I"ve got used to hanging close to them after working in the 70"s checking charging systems on GA light aircraft hanging on to the fuselage trying to read an Amps/ Volt meter while the pilot revs the engine. Being in front of the prop of course is more dangerous. I make sure the test cables are very secure and stay at the controls.

[chook]

Max that also makes sense - but why would they richen just cylinder # 4? That would place extra stress on the crankshaft with uneven power pulses.It would make more sense to just rev limit the engine.I have failed to ask you 3 questions:-1) What was the RPM when you captured those CRO readings?2) Is the rough running over the whole rev range?3) At what RPM does the fault appear?You can check that pulse generator again with your cro. You are looking for nice even square pulses. I believe they use an integrated circuit called a 555 timer. Very easy to work with (read repair).Chook.

[mad max]

I agree Chook, it would be a silly way to control the revs. The tests were done at 2000 rpm. The rough running is at the whole rev. range. The timer produces 1.84 volts at 65.5 htz. I made it from J-Car bits using a 555 chip. I"m just clasping at straws by rechecking this area :"(

[birdy]

Maybe theres no piston in the No. 4 hole.?

[mad max]

That only happens to Rotax"s Birdy ;D

[mad max]

OK I"ve made a decision to do something radical as I"m getting no where with my EJ25 problem. I"ve concluded that because of the extensions I"ve added to the leads on my oscilloscope they are picking up HT interference and giving me false readings and causing me to bark up the wrong tree. I picked up some info. on the web which hinted that some EJ25s had a problem with a "leaking" head gasket. So I"m about to remove the LH head and inspect. Could be leakage between no. 2 and 4. Aparantly it is not an engine design fault but an inferior gasket. Time will tell weather I"ve made the right decision :-

[Russell]

Hang on mate.........if your gasket is suss..........would"nt your radiator header tank be pumping out ( over flowing ) ???give waddles a call.........he did have a gasket issue.....now fixed.