Alfa Romeo Forum banner

2003 156 JTS 2.0 selespeed cold starting problem

4809 Views 107 Replies 4 Participants Last post by  156and33
Hi
I’d welcome a few ideas on troubleshooting a morning starting problem with my old car that I need for work (done >250k klm)

Symptoms:
Used to always start first time, then took a while to fire up on first crank. It would stop cranking by itself even though the key was in the crank position (I thought this was weird). Now struggles to start first time, but cranks strongly until I release the key. Usually starts second go but sometimes not convincingly. Problem only occurs in the morning (20 Celsius), never any issues rest of the day.
Generally driving OK, no performance issues that I have detected, perhaps a the odd splutter
Battery tests ok with carbon pile load tester. Voltage drops from 12.6 to 12.3 over night. Having to keep on charger. Alternator is charging OK. Starting issues do not appear to relate to battery voltage
I know I have a main bearing seal leak and there might be oil in the bell housing… not sure if that can interfere with the crank sensor operation
No fault codes stored
Recent repairs:
Fuel rail pressure sensor replaced. Crank position sensor replaced after stalling/starting problems (last repair). New battery July 2020. Starter replaced Feb 2021. Alt replaced Sept 2017. Plugs not due to be replaced.
Cam belt due for replacement at the moment
Troubleshooting so far:
I thought maybe the fuel pump relay might be dicky and swapped the fan and fuel pump relay - no change
Ecuscan tells me the actuators and relays work, fuel pressure is regulated at ~50 bar, lambda signals from each bank are the same, battery voltage 13.7v at idle. Don’t really know what else to check.
I carefully sprayed some gas around the engine bay looking for keaks, no change in RPMs
Spark plugs were dry sooty black but that was after idling for 10 mins, electrodes OK
Sprayed the MAF with some dedicated cleaner
Thanks
Duncan
21 - 40 of 108 Posts

· Registered
Joined
·
4,934 Posts
Key status at MAR is correct.
Battery seems OK but I'm not so sure right now.
Load test on battery suggests it isn't so bad.
Residual current draw test suggests car is OK.

I'm not sure what starter motor the JTS has but after finding 2 listing, I expect the high compression engine should have the 1.4kW one and the 1.1kW one is for the TS.
That suggests the starter should draw 130amps and another few amps for other electrics. Say 150amps total.

You could measure engine cranking speed with MES. Ditto battery voltage. It may be best to graph them simultaneously. I'd expect 11+ volts and 300+ rpm.

I don't expect spark plugs to be bad at 50k km. New will still be better though. I'm guessing car has been in regular use recently and there have been no changes to your fuel buying habits.

I tried getting data but ran out of time due to a couple of issues. I still have a question about the battery though because 12.3 volts equates to around only 20% charged. That would mean a lot of power needs to be drawn overnight yet there is as yet no sign of a significant parasitic drain.
 

· Registered
Joined
·
4,934 Posts
Check battery terminal voltage.
Check battery voltage with earth onto body.
Check battery voltage with earth to engine.
Check voltage through diagnostic plug.

Any discrepancies point to a high resistance in the relevant earth points or through the positive supply.
I'd also check voltage at and through the ignition switch.
 

· Registered
Joined
·
163 Posts
Discussion Starter · #23 ·
Fruity
I just noticed that MES is showing Too High Latency warning and wonder if this may have something to do with no SPKADV and authorisation to start signals?

Vbatt = 12.78V after being connected to car overnight so no drop. This is unusual and the only thing I have done is replace all the ATO fuses near the steering wheel, and manually locked/unlocked the drivers door and bonnet latches when testing yesterday
Voltage between: (batt + and body -) and (batt + and engine -) = 12.78V
MES voltage at OBD plug about 0.2V down

I made a mistake and did not save the graph from the first failed start attempt but here are the two subsequent starts. I think your parameters are met and I recall that the first start was similar (but maybe the Vbatt dropped a little lower). Once the car started, the Vbatt recovered mostly
Rectangle Font Wood Personal computer Parallel


Rectangle Wood Font Parallel Pattern
 

· Registered
Joined
·
4,934 Posts
The latency too high warning can still occur even when latency setting in laptop is reduced to minimum. I don't put much weight behind that.

Voltage and cranking speed seems OK.
The spark advance data is a way to ensure the spark is occurring when it should. I do not know why your system does not display it. Linked to that I kind of question the cam sensor. That would ensure the spark happens on compression/ignition rather than exhaust/induction. You don't have a cam sensor fault code though because you'd have mentioned that.

That goes back to everything looks fine but it doesn't start. That's why I'd like the spark testers to verify if plugs are sparking.

So far we are left with what we can see is seemingly OK but still the car sometimes won't start. I think I'd start looking more at the ignition switch. It may be that certain module functions can operate without ignition being switched to MAR but that statement seems wrong if the starting authorisation is OK. No doubt this is an elusive problem but it may be worth simply swapping the main ECM relay to rule that out.
 

· Registered
Joined
·
163 Posts
Discussion Starter · #25 · (Edited)
The Owners Manual does not show an ECU relay and the only similar one would be the electronic injection relay in the engine bay. There are two additional relays hidden near the fuse panel but not listed, and I know one is for the low beams.

I have not come across a cam angle sensor yet, where is that located, behind the can sprockets? Might be time to test it. How is that best done (simple resistance check or do I need to dig out the oscilloscope and back probe the connector?).

I have one more scan tool I could use tomorrow to see if SPKADV and authorisation data are picked up. This is the one I keep in the boot to help with replacing coil packs when they fail away from home,

It will be a while for the spark plug testers to get here from the UK. Luckily I will now not be needing the car for work until Feb because of COVID impacts

I forgot to mention that I have not changed petrol stations or fuel type but the car spent most of 2021 sat in the garage on a battery charger with a full tank of fuel.

I just fitted new tyres and windscreen and have the cam belt service parts sitting in a box ready to go, so am committed to getting this fixed. Might get a new camshaft sensor while I am at it. Thanks again for the support but let me know when you think it is time to take it to an auto electrician?
 

· Registered
Joined
·
4,934 Posts
I don't think it is time of an auto sparky.

The cam sensor is mounted at the exhaust cam pulley around the 8 o'clock position. The wiring connector is mounted on top of the inlet manifold, I think. I'm sure the housing is blue. Hope I'm not getting confused with a TS though. I don't think that is the issue as a fault code should readily appear.

Big clue is petrol. I think your fuel may have gone off and lost volatility. I'd drain a bit from the low pressure connector at the inlet manifold. Put a thimble or half an egg cup of the stuff in a container (not combustible) or even concrete and light it. It should instantaneously cause the whole lot to burn. If you can see the flame travel, it is just not volatile enough. That's why I say the ground. Ideally concrete with low porosity.
 

· Registered
Joined
·
163 Posts
Discussion Starter · #28 · (Edited)
More testing
Vbatt = 12.78V after connected to car overnight. Seems like current drain has ceased
Cranking: SPKADV BTDC starts at 4, goes negative for a second and then stabilizes around 10 at 900rpm
Revving: 1000 rpm, 12 deg, 2000 39 deg, 3000 42 deg

I back probed the camshaft position sensor: earth OK, Vsupp = 5Vdc
MAR: Vsig = 0.11Vdc
Idling: Vsig = 6Vac at 30 hz
2000 rpm 80hz
3000 rpm 130hz

I cannot find any timing data or specs on the cam sensor to check these values against
 

· Registered
Joined
·
4,934 Posts
There is a similar thread but with a JTD. The last advice on that was to try a crank sensor. I'm not going to argue with experience. I thought that if it shows a plausible engine speed and spark advance whilst cranking, that should mean both good inputs and outputs. It may not be the case though and no matter how unlikely I think it may be a crank sensor, I think the cost of one is worth it for the time and a trip to a sparky.

If you have an oscilloscope, by all means, have a close look at the wave form whilst cranking on starter. Typically if an indictive pulse sensor is starting to break down, there is a pause and a lull in the rising sine wave form before it suddenly overcomes a high resistance and then largely follows the form it should produce. Because the amplitude is proportional to engine speed, is is possible a glitch in the signal during cranking can cause issues.

I don't like suggesting 'try a part' but it's going that way. There are bits where I am not sure whether the supplied results were whilst cranking or engine running.
 

· Registered
Joined
·
4,934 Posts
Sorry, I must have forgotten that the crank sensor was changed.
I struggle when things done are not simply listed.

To get exactly what is going on, the car starts 2nd try?
It cranks fine.

Have you simply tried turning the ignition on (MAR) for 2-3 seconds to prime the low pressure fuel system? Possibly do that twice before trying to start. Result?

Have you diagnostically measured fuel pressure in the high pressure fuel rail before trying to start it? An initial graph of fuel pressure may be useful taken at the point of first cranking.

You stated it starts 2nd try. How long do you let it crank 1st try? How long does it crank 2nd try before starting?
 

· Registered
Joined
·
163 Posts
Discussion Starter · #33 · (Edited)
Generally starts second try but not always straight away, or easily. Sometimes it attempts to fire but does not quite get there, sort of splutters and stops as if the fuel tank is empty (but it isn’t). Yesterday the car cranked strongly for 10s but did not even look like starting. Started second go but that was not very convincing as if the engine was flooded (3-4s). For the rest of the day it usually starts easily (1-2s)

Fuel pressures (low and high) seems OK as tested in post #7. Two primings in MAR makes no difference

I might video it this morning and see if that can be uploaded here.

I will switch out the fuel injector relay and see what happens.

The crank sensor did not fail outright. One day the car failed to start when my wife was by herself, and a fault code logged (the car started eventually). I reset the fault log and it then drove OK for several weeks until I replaced the sensor. Luckily I had a new unit in my spare parts box as I knew this was a critical sensor.
 

· Registered
Joined
·
4,934 Posts
I have another couple of ideas.
Firstly, look at the cylinder compression pressures. (Hopefully you can get a tool to test). I'd like to know if there is a significant difference between first thing in the morning and with a fully warm engine.

Another thing to look at is the fuel trims. Look at this thread for reasonable details;
 

· Registered
Joined
·
163 Posts
Discussion Starter · #35 ·
Not checked compression on this engine yet, fear of getting the adaptor stuck down the hole. Can I pull the injector relay when doing this ? I was starting to wonder if it was a mechanical issues with valves stuck open.

I guess you mean the long term fuel trim values for each bank?
 

· Registered
Joined
·
4,934 Posts
Yes. I don't know what data there may be for short term fuel trims given that the O2 sensors will be offline. It's at this point I wish there was more information about the map and adaptive learning.

I did wonder about the valves. What distance has your JTS covered? That can only be a guide though as conditions and quality of fuel make more difference. Couple to that I read on BITOG that modern oil which is made more viscous with VII tends to be responsible for stuck piston rings and sticky deposits to the point that people seem to get good results going back to a 15w/40 or 20w/50 to free the rings. It makes @johnlear 's choice of Pennrite 15w/60 look very intelligent.
 

· Registered
Joined
·
163 Posts
Discussion Starter · #37 · (Edited)
About 250k klm
Always used 10w60, either Selenia or Castrol Edge
Fuel has been 98 RON

Cannot see any way to upload a video file to show behaviour

Just came back from a 1 hr drive and collected this data when idling in garage A/C on:
Eng temp 93
Air temp 53
STFT B1 0.00
STFT B2 0.00
LTFT B1 -2.34
LTFT B2 -3.12
SPKADV 15
TPS 9%
MAF 10.25
Vbat ecu 13V
 

· Registered
Joined
·
4,934 Posts
OK. Nothing badly wrong there but a couple of things are a little unusual but I must remember you are in summer heat so probably mid-high twenty odd degrees.

The air temp is fairly high but somewhat expected due to heat soak.
MAF is rather low. It is lower than I've seen a 1.6TS at idle.
LTFT are leaning the mixture a little.

Your car has covered a fair distance and I'm guessing the engine hasn't been apart. I'm guessing you have the JTS hum of restricted airflow by 5000rpm. Maybe earlier?

At this point I'd definitely like to know what compression pressures are- especially when engine is cold.

If you like, you could try to get a maximum MAF sensor reading but it may need an assistant and I don't think your engine would sound nice at 6500rpm with a wide open throttle. It will probably sound like it is in its death throes.

With all the information so far, I think a temporary loss of compression is what you are experiencing. It is not unusual for direct injection engines to experience this.
Not a definite though. Just an idea.
 
21 - 40 of 108 Posts
Top