Diagnosing "engine no start" problem

geogan

geogan

Petrolheads
Have an 03 STi Type UK. So the car was fine until Saturday when I turned the key and... nothing... just the sound of starter - no ignition. I drove over 100 miles a few days ago with no problem.

There is power, I can hear fuel pump priming up as normal when key is initially turned, and when started, the starter is turning the belt and main crank.

I have tested the crank sensor with multi-meter and it has resistance in normal range (i think) and I also tested starting without crank sensor in place and no difference.

I took out battery to make sure it wasn't that and left it charge overnight to full charge (its new battery only a few months old) but no difference next day.

I took out all four coil-packs to examine, but apparently from what I read from expert source (as opposed to some idiot youtube videos I found) it is not possible to check resistance on a 3-pin coil-pack, so no way to tell if they are good or bad.

I haven't removed or checked spark-plugs yet, but I can't imagine all four would go together, and one bad would not stop a car from starting. They are also quality Denso I put in as far as I remember.

I tried to buy a "spark tester" but doesn't seem to be any available to buy in Ireland (in stock), have to try and locate one online. Anyone know a good one that works with boxer engine layout? Is the light ones good enough or should I try the fancier "spark quality" screw type ones?

I also checked for any engine faults using OBDII reader connected to phone, and none, and no "check engine" light either.

Anyone know any other areas/parts related to ignition I might check?

I have heard that corrosion in power lines may drop voltage too low on coli pack to let it work. Can I test this somehow with multi-meter?
 
Last edited:
Fozzy said:
Check engine grounds as well.
Click to expand...

OK, will do, once I find out where they are. I am nearly certain this is some sort of electrical old-wiring problem (loose connection, corroded wires, voltage drop somewhere). But never seen such a severe effect before.

With the ignition on, should I be getting 12V to any of the coil-pack power plug pins I wonder?
 
When you say the starter is turning over is it turning over the engine or just the starter
 
When you say no ignition do you mean no lights in dash or that it's just not firing up?
 
scoobysub said:
When you say the starter is turning over is it turning over the engine or just the starter
Click to expand...

As far as I can see it is turning the main belt and the main crankshaft attached to it too as normal, just no spark and ignition. I recorded a video clip of it. Maybe I will upload the clip here somehow but don’t know if that is any use.
 
Last edited:
seanie said:
When you say no ignition do you mean no lights in dash or that it's just not firing up?
Click to expand...

All startup lights in dash, power to radio and everything looks and sounds as normal. No strange slowness, dimness, or noises like there is when battery low or flat.
 
geogan said:
All startup lights in dash, power to radio and everything looks and sounds as normal. No strange slowness, dimness, or noises like there is when battery low or flat.
Click to expand...

Can only be spark or fuel one would think. Is it getting fuel - pump priming, lines ok, no blockages?
 
Fozzy said:
Can only be spark or fuel one would think. Is it getting fuel - pump priming, lines ok, no blockages?
Click to expand...

Think so. Fuel pump priming sounds as normal. Was working fine a few days previous to this. I also have a fuel pressure regulator guage in engine bay so I can check that. I’d lay a bet it’s some electrical problem though. These old cars, old wiring and corrosion is starting to produce all sorts of weird effects.
 
Is there an “igniter” on the 2.0 STI engine with coil-on-spark system? Or is that just on the older engines which had the coils on top of the engine?
 
Checked the coil voltages and they were all correct at 12v, also verified fuel pressure to rails by reading Defi fuel pressure regulator dial. Can see this pressure increases from none to higher pressure when the key is turned and fuel pump primes up.

I think actually the problem might be the idle air control valve (again - it stuck a few years ago!), because the car now starts and runs for about a second or two before spluttering to a halt. I think it may be stuck closed. I did notice during idle recently, the revs not being stable which is another symptom of bad IACV.

Will be opening it up to examine and clean this today (it?s risky because opening it will destroy the gasket and sealent I used previously on this when I cleaned and lubricated it a few years ago). Will see how it looks from there. If it is not this, replacing it will require new sealent and drying/hardening time.

A bad IACV would also explain why car only attempts to start once per day when I tried it, as air slowly leaks into engine over night but one attempted start would produce vacuum in engine and no more starts.
 
if you pull vacuum hose off inlet manifold it will bypass IACV to see if it will run if you think it is stuck closed
 
it should still run with bad IACV if you give it the boot, IACV only comes into play at Idle, when up the revs air let in by IACV is negligible
 
seanie said:
if you pull vacuum hose off inlet manifold it will bypass IACV to see if it will run if you think it is stuck closed
Click to expand...

Well I removed and "unseized" the IACV and replaced on car and it started up first time two days ago. Everything was back to normal and felt even more responsive than normal. Drove for two days then drove 30 miles away today and parked in carpark.

When I came back car was DOA again. No idea what the **** problem is now. I don't understand how fixing the IACV fixed it for two days then there is a sudden return of problem. I even removed IACV again to check and it was still free and loose.

I had no tools with me to remove or replace sparkplugs or test anything else. This kind of thing needs replacement parts to just rule out failed parts and that costs money. To test basic I would need new IACV, sparkplugs, coils, maybe cam sensors and crank sensor.

I did tests with multi-meter and the coils were still getting 12V.

We also tried a push rolling start but nothing.

Also tried ECU reset by unplugging battery and hitting brake.

And the fact that whatever the problem is can just suddenly appear/disappear instantly is worrying.

That and the fact there are absolutely NO check engine codes coming back at all according to my OBDII phone software connected to ECU. That is what I don't understand. How can a car be totally DEAD and yet the ECU sees NO errors?!?

So car is right now stuck on its own 30 miles from my house, and I am back home :-(
 
Last edited:
Water temp sensor? They go bad from time to time meaning wrong signal gets sent, car thinks it?s warm when it?s cold etc
 
Fozzy said:
Water temp sensor? They go bad from time to time meaning wrong signal gets sent, car thinks it?s warm when it?s cold etc
Click to expand...

No idea at this stage. Once I get car home I will not be driving it anywhere again until the cause is confirmed. I will test spark when I get spark strength tester if I can get it to work in this engine. A lot of them expect the simple "spark plugs sticking out of top of engine" type. Will try changing spark plugs if it shows weak spark or coil somehow.

Of course it could be any of the other sensors involved in ignition too.

Might get one of those handheld oscilloscopes that show the signal pulses to sensor (or from it) to confirm a signal
 
Last edited:
Tested coil pack wiring and verified “0v, signal, 12v” to three pins of each pack. When engine is cranked the centre signal pin gets about 3v average on multimeter (showing pulse voltage average there, haven?t got actual time domain signal reader)

I presume this means the ignition fuses are not blown since coils recieve voltage when power is on?

Tested with spark strength tester and it seems that there is *no* spark at all on all coils, so couldn’t be all coils gone bad at the same time. The tester also allows adjustment of distance between terminals so I tried various from a few millimetres to 2cm. These coils are supposed to spark between 2-3cm in air.

Removed and tested the cam and crank sensors and both have about 2Kohm resistance which is supposedly correct. Also tested continuity of sensors to battery ground.

Tried starting with crank sensor out and no difference. Forgot to try with cam sensor out. That will be next test. Also have to find ground connections and check for corrosion in case this is something to do with it.
 
Last edited:
UPDATE:

Had to replace entire starter motor with a new one, because old one stopped cranking correctly (must have burnt out the solenoid).

Also replaced spark plugs with new NGK PRF7B, and changed fuel filter.

Checked wiring loom integrity (voltages/resistance) from coils, cam, crank sensors all the way back to ECM module pins following service manual.

Also tried new MAP sensor (?120+) but that was not compatible without some settings remap so had to put old one back in for now.

Also took apart front of engine to access the timing belt and verified it looks OK - still lined up cams when crank mark is centered. Also verified the crank teeth are still all there and OK.

Finally gave up and ordered a new crank and cam sensor, even though I can see valid signal coming from them using oscilloscope on ECM pins, and OBDII monitors sees "RPM" signal during crank of about 220prm, so that tells me the crank sensor is working, but just running out of ideas here.

Also bought laptop so I could use ROMRaider to log ECU, but can't get it to work with my own cheap OBDII cable (getting error on ECU Init command) - and no way I am paying 200+ Euro for the Tactrix cable. Anyone got an old Tactrix cable they want to sell?

By the way, once or twice every few days I get a slutter start and a few sparks, then it goes dead again, and only crank.

But fuel pressure regulator still says 4bar pressure on fuel when Walbro primes up so I don't think its fuel pressure.

Only thing I didn't test/check is injectors, but I can't see how four could just suddenly go at same time on one day, when they just drove 30 miles only 10 minutes beforehand...

Still absolutely zero pending codes, and just old historic stored codes in OBD scanner.
 
You must log in or register to reply here.
Back
Top