2005 turbo forester wont start. Any ideas?

tambox · Post by **tambox** » Fri Oct 25, 2013 5:25 pm

I looked at a 2005 (I think) turbo forester that had an engine swap and now will not start.
The car was driven in, the motor/loom swapped, then it would not start.

It fires on the front two cylinders, but not the rear two.
If left for a while, it will give one spark on the rear cylinders, then nothing till left for at least 15 mins, then one spark again.
After some ECU testing with an old OEM Subaru analyser.
It showed no fault codes, but going through the scope pictures on the analyser, it did not show any pulses for the crank sensor.
This analyser reads from the test connector, it correctly recognises all the Forestors electrics.

All the original sensors and loom, inc crank gear were were then fitted to the new engine, no difference.

With a real scope, I checked there were pulses on the loom plug for the crank sensor, they looked fine.
An autoelec ran a scan tool on the system, tested all conections and it looked fine. All readouts were ok.
The ECU was sent away and bench tested ok.
The one spark on the rears is apparently part of the initialisation and it shows the ECU can drive the rear cylinders, if it want to.
Coil packs have been swaped/changed.

Any ideas????

RSR 555 · Post by **RSR 555** » Fri Oct 25, 2013 7:27 pm

Yep.. wrong crank shaft pulley on the timing belt or different LH intake timing gear pulley

tambox · Post by **tambox** » Sat Oct 26, 2013 7:04 pm

They had tried two different crank pulley's, the new engine, then the original it was runining with.
The cam sensors are one for each cam, non magenetic, but 3 wire piezio or similar.
It has auoto advance on the cams,but basic timing is still the same.
Tried new motor sensors, then original.
Nothing made it fire correctly, except the test bench.
Asked test bench guys to give me low voltage and high voltage trigger points for crank sensor, will compare to car.
All the sensors are happy, except crank.

Scope figures are required here.
What is the low voltage switching point,0.2 ,0.3 ???
What is the high voltage switching point 4.0, 4.1 etc???

Looking for hints on this headache.

RSR 555 · Post by **RSR 555** » Sat Oct 26, 2013 9:59 pm

To me there is still a fault at the crank.. maybe a pickup sensor or reluctor ring.

From memory (haven't got the book at home) 0.5v-0.7v should be low but if you're getting 0.4v then you should be looking at pickup sensor gap?

Now I know I'm being silly for asking this question to you Greg but I just want to know it in my head.. is the timing belt setup correctly (sorry this is not meaning to sound insulting) but I want to cover all bases. The other thing is the coil pack wires are going to the correct coil? Was the 2 engines of the same model? and lastly, was everything from the old engine used on the new one? I see you mentioned all this but it just doesn't seem right??

But I'm still going with pickup sensor. Sorry hard to test from in front of a computer screen

tambox · Post by **tambox** » Sun Oct 27, 2013 8:07 pm

It got to the stage where all original loom and sensors have been fitted, still no go. The motors are identical, well ones stuffed.
Every sensor has been swapped at least once.
The cam timing has been done at least 6 times, due to swapping bits, once with me advising, so I know it was right.

Both myself and the auto elec agreed, the ECU showed a crank angle sensor fault by scope, but no fault code as the sensor looked ok to the ECU, just not receiving any signals from it.

I have asked the guys that bench tested the ECU and said it was fine, what their crank input voltages were for their test bed. I have then asked them to raise the low voltage till it fails, then drop the high voltage till it fails. This was done on Sat morn, will find out on Mon if they can do this or not.

At least that will give me some figures to work with.
As far as I can see the pickup may be in a slightly higher position on the new motor as being a cause of this fault.
I am only advising on this job, have not seen it from start to finish, they are running out of options.
The guys that tested the ECU said it may be borderline, put another ECU in it and it will run fine, bit scary a statement like that.

RSR 555 · Post by **RSR 555** » Sun Oct 27, 2013 11:32 pm

Are both Cam and Crank sensors the same? Have you swapped these 2 around? not old to new but cam to crank and vice versa.

Also check that the main engine harness plugs don't have water in them (say from washing the engine bay when the engine is out) and/or have bent/broken pins.

tambox · Post by **tambox** » Mon Oct 28, 2013 11:09 am

This is one of the newer style electronics, has a cam angle sensor for each cam. These are 3 wire, whilst crank is two wire, cannot swap.
The car has been in the workshop for a couple of weeks, the connectors have been unplugged, checked, tested and re-plugged many times. The testing I did was with a pin in to the cable insulation, this tests the connector as well, as it was done on the body side of the plugs. Good pictures on the CRO, just got to go and do some exact voltage level readings.
It appears the pulses may be outside what the ECU will recognise.
Why?????

RSR 555 · Post by **RSR 555** » Mon Oct 28, 2013 11:49 am

Very strange

Ok, so it has AVCS on the inlet cams. Can you remove the rocker cover and check that both inlet cams (old and new) look the same.

Is there a possibility that the engine is not exactly the same as the one removed?

tambox · Post by **tambox** » Mon Oct 28, 2013 1:18 pm

Yep done that.
A visual inspection was done, the only question I still have about them, is are the pulse slots in the same location, relative to the timing gears.
Clutching at straws now, but looking for anything.
What are the differences between engines of that type, that are supposed to be the same?
Trying to find online some genuine details on the 04 manuf MY05 turbo Forester, nothing yet, still looking.

The OEM UCU analyser shows pulses coming through on the cam sensor, in its scope mode. It shows when you turn the ignition on, the crank sensor goes high, then stays there as it cranks, on the OEM ECU scope readouts.

The guys that tested it and said the ECU was ok, could not give any test system trigger voltages. They have a magic black box they plug it into and it tells them if is ok or not.

Going back tomorrow to get a bit more serious with it, as the workshop is giving up.

tambox · Post by **tambox** » Mon Oct 28, 2013 1:59 pm

Got what is supposed to be an OEM manual for it, 3616 pages, bit of light reading.
Having never worked on an AVCS vehicle, to this level before, the book says it has 1x crank sensor -7 - +7v, 1x cam sensor -7 - +7v, LH & RH AVCS sensors -7 - +7v, these are figures @ idle.
Do they sound right?

steptoe · Post by **steptoe** » Mon Oct 28, 2013 4:37 pm

Greg, maybe suggest an upgrade to an EA81 -only electrical problem with them is a $100 module or a fuse link

RSR 555 · Post by **RSR 555** » Tue Oct 29, 2013 11:04 am

tambox wrote:Got what is supposed to be an OEM manual for it, 3616 pages, bit of light reading.
Having never worked on an AVCS vehicle, to this level before, the book says it has 1x crank sensor -7 - +7v, 1x cam sensor -7 - +7v, LH & RH AVCS sensors -7 - +7v, these are figures @ idle.
Do they sound right?

So did you read the whole manual last night? lol

I've only rebuild the AVCS engines, never had to fault find one.

The wave signal is quite large when running but at idle I wouldn't expect those little sensors producing +7v. I could be wrong though.

I'm not sure what other differences the engines have but there is definitely something that is telling the ECU the wrong things.

Also, the ECU and everything to do with the wiring side of it, is it all the original stuff from the running car that drove in?

tambox · Post by **tambox** » Tue Oct 29, 2013 6:36 pm

Skipped the boring bages, went straight to engine diagnostics.

All the original stuff has been put back in.
Did a few measurements with the small CRO and multimeter, at the ECU pins.
Everything looked good and met specs, except the crank angle sensor. It pulsed at about 2vpp when cranking, then jumped to 4vpp, when the engine fired. The crank angle pulses were very messy and varied in voltage.
Looking at the ingitor and injector pulses, they are not constant, couple here, nothing, then another, nothing, then a few more. Oh for a dual trace scope so I could see injector relative to spark, might have to take one next time, if we can get some fault finding info on one of these.

The crank angle sensor was about 1mm away from the pulley tabs, gave it a trim so it sat about 10 thou from the tabs. The pulse voltage is now about 4vpp at crank and goes to 8vpp when it fires, much tidier and even pulses now.

Still the same.

The fault finding in the manual is great, replace part"x" as indicated by the fault code, then while its idling, check this and this. We have no fault codes.
The engine had the same pre code, just different engine number, so it should be the same.
According to the Subaru ECU analyser, the exhaust temp is 370 deg, one cam advances and retards, the other stays at zero. Appart from those, all the other levels look good, while cranking. Without better quality test gear, I dont believe only one readout, I need it confirmed, the scantool was happy with everything.
Trying to get some info from people who actually know about fault finding these, there aren't many.

Jonno, maybe if we put two turbo EA81's in it, to give the same power, it might be simpler than this.

steptoe · Post by **steptoe** » Tue Oct 29, 2013 8:57 pm

Once-upon-a-time, I would have suggested the dealer until I met a mechanic working in a big dealership. Dealer diagnostic computer ? You kidding? was his reply. Engine rebuilds done on upturned garbage bins. Another brand of car, dealer was suggesting customers problem was with either the cat or afm, customer left to decide what avenue to take FFS !, so she went aftermarket exhaust - no fix! I suggested she press them to borrow afm off another car in the yard to determine if AFM was suspect - she came back with a box of beer after her Foz was made to run right with Jonno diagnostics.

Then learnt another dealer (well, let's pretend it was) sends hard work outside to smaller , more experienced? shops !!

RSR 555 · Post by **RSR 555** » Tue Oct 29, 2013 10:11 pm

Exhaust temp is 370deg? Sounds like the engine is 1/2 running on idle??

Cams shouldn't move until higher RPM, so if they are changing. this means they could be getting stuck. You may need to remove the filter inside the AVCS control valve. But this is stange as this happens to cars that are running, not ones they can't start.

I'd make sure you have the original Timing Belt Crank Pulley and the original Crank Sensor and make 100% sure that the original wiring loom is attached.

There really is no other reason why this wouldn't start.. just look at the basics of making an engine run

tambox · Post by **tambox** » Wed Oct 30, 2013 7:12 am

We are at the stage now where the basic signals are correctly being sent to the ECU, but the ECU is confused with something as it tries to start.
The simplest way to test any further will be with a newer OEM ECU analyser that will look at exactly what the computer is trying to do.
It can be investigated further with a dual trace CRO, but having never been this far into finding a fault on one of these and we cannot get detailed info on what we should see.
The other option is to get another car the same, then compare/swap bits.

As the electronic signals seem good, the mechanics of the motor are being carefully checked today. Which makes sense, as the only thing that has been changed is the motor.

Glad I'm not paying for this one.

RSR 555 · Post by **RSR 555** » Wed Oct 30, 2013 10:47 am

Would be very handy to have one of the same model and swap bits over to try out. Sorry I can't help you any more

Would be nice to know what it turns out to be and I'm glad I'm not paying the bill too

NachaLuva · Post by **NachaLuva** » Wed Oct 30, 2013 1:19 pm

steptoe wrote:Greg, maybe suggest an upgrade to an EA81 -only electrical problem with them is a $100 module or a fuse link

Lol

tambox · Post by **tambox** » Thu Oct 31, 2013 10:28 am

80, 90, 110, 120 PSI.
The "good" engine is being sent back to the engine supplier, with the rest of the car, for them to sort it out.

Why was was it only sparking on two cylinders??
It seems the ECU cycles through a start up proceedure over and over, no fault codes as the rpm is to low when cranking.

Will see if it starts when the engine is fixed.

steptoe · Post by **steptoe** » Thu Oct 31, 2013 6:08 pm

I admire the member that suggested go back to the basics fuel , spark and compression.
Gotta love the wreckers that sell "good engines" Even my low compression 7.7:1 EA82T's run 130 psi, and the Vortex 150 ! How's an EJ go so low ? Cooked and shrunk the pistons?