T4 engine diagnosis
 

Excuse the vague title but this post covers quite a bit. Been having a running problem with my 2000 T4 for a while which I have been trying to pin down, and have come to a point where I could do with some advice. I have already used a lot of information on this site, so thanks to all posters.

The symptom I am seeing is persistently maxed out positive fuel trim (LT FTRM or LTFT) to which is added high positive short-term fuel trim in closed-loop. When in open-loop the lack of extra trim seems to make it almost cut-out and is undrivable until the oxygen sensor heats up. The codes the ECU throws are EFI-435 and EFI-512 (read with VOL-FCR) which only really indicate that the trims are out of range. So it seems the possibilities are: unmetered air in the combustion mix, too little fuel in the mix, or a faulty sensor reading giving the wrong ECU map.

Here's what I have tried so far:

• Suspected an air leak so got a vacuum gauge and tested values on the throttle body - perfect
• Repaced the fuel pump with a Walbro 255. The old one was whining loudly and had 120k miles so not a wasted exercise anyway. Thanks to 960kg for a great post on this one.
• Got a fuel pressure tester and performed all the tests detailed on Vadis - all values spot-on.
• Not satisfied with that result I routed a fuel line from the fuel rail into the driver's compartment and measured pressure while driving - this also transforms your V40 into a handy crematorium in the case of a crash. The fuel pressure was fine under all conditions, anyway.
• Wanted to get Jim on the forum here to do a professional cleaning job on my injectors, but unfortunately he didn't have a temporary replacement set and I don't have alternative transport. Set up a makeshift cleaning rig anyway and backflowed the injectors with solvent - probably not a patch on the proper service but should have removed large blockages. No change.
• Replaced the front oxygen sensor - no change.
• Checked all other sensor values with diagnostic tool (Uiftech U581 - highly recommended, btw). TPS, IAT, coolant temperature sensor, all good.
• Performed several tests on the MAF sensor as I presumed the fuel and air supply to be ok: Checked idle value against Vadis - right in middle of range
• Read somewhere that BHP * 0.8 can be used as a rule of thumb for maximum airflow. I got 156g/s flat out so looks good.
• Routed a cable from the MAF signal to a multimeter in the car to see values while driving - varied from 0.6 to 4.6 volts which is exactly as specified in Vadis
• Checked the exhaust upstream of the oxygen sensor for a hole causing a false lean reading - none found.
• Back to the air-leak idea, I set up a portable compressor on the air intake at the MAF to see if I could find a hole. Couldn't get any pressure to build up though and tried blowing through it myself - found all the air was coming out of the oil-filler cap and the dipstick tube. With either of these removed there was almost no resistance - the air just went straight through!
• Compression tested all cylinders, although admittedly this was a long shot give that the car runs ok in closed-loop.

So, can anyone confirm that very little air-resistance between the air-cleaner and the crankcase indicates a fault - the only route I can think of for this air is through an open PCV valve? Also, if the valve is blown, could this cause the symptoms I am seeing? Given that the crankcase pressure is fed to the intake manifold anyway, I don't see how the valve being open in the opposite direction would make a difference other than allowing boost pressure to enter the crankcase. Unless the PCV valve also provides some resistance to restrict the volume of air and this function is also blown out.

Any takers?

By the way, I now have quite a collection of gadgets amassed from this saga - if anyone needs any info on these or any of the procedures I have tried just shout:

• USB to OBDII cable
• OBDII diagnostic tool
• Vacuum/boost gauge
• Fuel pressure gauge
• Compression gauge
• Home-made injector cleaning rig

By way of bumping my long-winded post, just noticed on Vadis that there is no PCV valve in the parts catalog for this model, just what's called an oil trap which connects to the crankcase ventilation hose. Does this mean that the crankcase is open to manifold air at all times, or is there still supposed to be a check valve in the oil trap?

I know I could just take it off and see what's going on, but removing the intake manifold to get at it seems like a bit of an ordeal.

Unfortunately it's the only way to access it the PCV oil catch canister.

I wasn't quite sure I should reply to your thread as it seems you're well on top of your diagnosis and, to be perfectly honest, I don't understand half that fuel trim stuff you mention above. Although, there is one thing that springs to mind regarding a vacuum leak. I think it was Spangoolies had a similar problem with his 2.0T. It turned out to be the one-way valve located between the carbon canister purge valve and the inlet manifold. No harm to check that it's functioning I guess...

I would say your MAF is missreading or having low refresh rate. For the air leak check your manifold vacuum when idling. Can you post your datalog here?

Thanks guys for the replies,

@Baffler - funny you should mention that check valve as I have just replaced that part a couple of weeks ago. The valve was blown through, I'm assuming from sustained boost pressure. The replacement part was a pipework assembly which the guy in Volvo parts couldn't even identify and my wallet probably wouldn't sustain, so made a new assembly up with some fuel line and a couple of generic check valves from eBay. Will post up the "fix" when I get a chance. Didn't solve the problem though, and also tried temporarily disconnecting the EVAP system altogether to no avail.

Took me a while to get my head around the fuel trims as well so here's what I understand from it so far. When the engine is in open-loop it is using a map in the ECU to "guess" the fuelling (injector period) for the current conditions (mainly RPM and load measured by MAF). When the engine goes into closed loop (lambda sensor up to temperature, not idling and not wide-open-throttle) the lambda sensor will give the ECU an indication of rich or lean running and then the ECU will trim the map to bring the combustion ratio to the stochiometric point. The amount of this trim is the amount the guess of the map in open-loop is off the mark. To further complicate things, one trim setting is stored over longer periods (LT FTRM or LTFT) and is also used in later in open-loop and one is instantaneous (ST FTRM or STFT), but they both contribute to the overall trim in closed-loop. So in my situation the LTFT is at 20% and this is the maximum value, so the ECU has learned that the map is underfuelling by this amount; when in closed loop the STFT adds another 5-20% and then the car runs ok. So all this means is without closed loop operation the car would be running very lean.

@LevL - I had myself fairly convinced that it was the MAF as well, but tested it to death and looks spot on. The only diagnostic specified in Vadis was to check the Idle air flow for running temperature = 8 - 13 Kg/hour. My OBD reader gave around 3 g/s == 10.8 Kg/hour so fine by this measure. I didn't think this was comprehensive enough so I tapped the signal wire and read the voltage while running - 0.6V to 4.6V, also perfect. Then tried the BHP measure I found above which may be folklore but also gave a near-perfect reading. Really wanted it to be the MAF but doesn't look like it. Also for the air-leak, I measured the vacuum at the throttle body and gave a stable 20", with proper response to a throttle blip. I presume this would show up a leak in the crankcase ventilation system as well, so might be barking up the wrong tree there?

......first of all i am not rude but slightly blunt and straightforward....you are really trying to hard, ok so all those measurements you have done in various ways shows that you know what you are doing but they get in the way of what is happening in a practical way....out of all the words that you have typed there is only this sentence that says you have a problem.....

......."When in open-loop the lack of extra trim seems to make it almost cut-out and is undrivable until the oxygen sensor heats up"............

....in open loop surely you are almost flat out anyway as that is the reason it is open loop to get more fuel in for the fast speed or performance as when in closed loop the O2 sensor will be in charge to sort out the mixture.....the O2 sensor is practically hot as soon as you start the engine as it has it`s own heating circuit ...that`s why they are the four wired type.....and they don`t rely on outside heating appliances like in the 960 model to encourage quick heating, as this model has six cylinders and the quicker the sensor is hot the sooner it will use less fuel.....however at full speed or throttle the O2 sensor is of course hot and this is the problem you are talking about, almost cutting out....

.....i am sticking my neck out here but your problem is not what you think the O2 sensor not at operating temperature but i do believe it is your Walbro 255 petrol pump!!!!......if you read and read on the S40 concepts site you will see that there are a lot of people have used this model and have suffered "cut outs" with low fuel in the tank heights even some at half full tank......i did my research on that site to find out that the 342 model pump is the one to use because of it`s offset spigot as the std. pump has....ok. so this is the only way in for the fuel but at least it has the provision of using the fuel in the plastic pot the pump is in when under full performance, the fuel will automatically get in the pot with this pump past the one way valve at the pot entrance because the 342 model has an offset spigot to allow it in, and once in cannot get out, so then the fuel is sucked in through the offset spigot position towards the rear of the tank or car so that under severe acceleration or even very low tank heights of fuel the engine will still run normally because it has a sustained pressured fuel supply..........What your 255 is doing is almost blocking the pot entrance one way valve because of it`s central larger suction spigot, so therefore fuel is being used from the petrol which is sloshing around in the tank and not from the pot which it is designed to do, so when you are cornering fast or slow, unless you have a fullish tank of fuel the pump is grabbing at air in between the fuel sloshing around so therefore it carries through that when the air gets to the injectors the engine wants to cut out!!! .....
.....which is more or less in so many words is what you have also said....

......"The codes the ECU throws are EFI-435 and EFI-512 (read with VOL-FCR) which only really indicate that the trims are out of range. So it seems the possibilities are: unmetered air in the combustion mix, too little fuel in the mix".....

...same thing, but different happened to me once many years ago in the mid `70`s......i had this Lotus Cortina which was really great and decided to do a service....changed all the oil, new contact points, 4 new Champion Plugs,set the ign.timing, great i thought she will really go now.....well half way up the road it started to misfire at higher revs. i went to work all week with it misfireing, it was really getting on my nerves, i had replaced all the leads, new coil and it was the same......one of my mates who worked at Ford Dagenham said use Ford plugs not them Champions, but i said they are brand new.....anyway i decided to change the new plugs with newer plugs and the car went like a rocket no problems at all.....i have not bought Champion plugs since and never will.......i am saying all this just because when one fits new parts to the car they can still be inferior or be the wrong application........

.....Hope this helps?....

......just found these pics. of pump ends so you can see the difference......it isn`t the 255 but has the same central entrance, like the one on the left and the 342 is similar to the other two, of course the plastic spigot has to be attached but it shows the offset entrance for the fuel....

I'm no great authority on engine function but that sounds completely wrong to me. The air intake is a closed system to all intents and purposes feeding the turbo, to the intercooler, to the intake manifold, through the valves and into the cylinders. At no time can any air escape to the cam cover unless you have a leaking valve seal or to the crankcase unless you have leaking piston rings.

If you were using very high pressure it might get past but that wouldnt indicate a fault, just more air pressure than the parts can contain.

it might be worth reading up on the 'leak down compression test' to see if you have accidentally done the test and found a fault, seems doubtful.

The rear lambda sensor is the one which measures the content of the burned exhaust gas, if that was not working properly you would have an engine warning light and the ecu would default to a rich stock fuelling map iirc. Afaik the stock fuel map is slightly rich and the various sensors allow the ecu to lean down the mixture safely and to bring it inline with emissions regs ?

You said that the long term fuel trim is used to adjust the short term fuel trim which would mean the open-loop running should be fuelled correctly. If it was doing what you say and adding excess petrol you would have sooty plugs on idle at a guess ?

Have you checked the temp sender as i have read several times these tell the ecu the engine is cold and this adds more fuel to the mix. I expect the temp sender could be removed and resistance tested in a pan of warming water. High resistance when cold and low resistance when hot iirc.

It would seem that the exhaust mounted lambda sensor is the likely fault due to this being the sensor which regulates the fuel correction, which is your fault code.

Trying to decphier your problem from your description, is it poor cold running performance ?

What were your compression readings on the cylinders when the engine was warm ?

EDIT

Wow, took me ages to type that. lol

960KG if the fuel pump was at fault then the fuel pressure reading would be wrong when the car was playing up, has the OP found the pressure reading was stable when the car was poorly performing or not ?

Thanks 960kg - and agreed, probably too much detail, but I will counter the fuel pump theory with some more ;-) The pump I put in is exactly the one you recommended in your post - I think the 255 designation is a family while the GSS342 is the subtype or some such:

http://cgi.ebay.ie/ws/eBayISAPI.dll?...=STRK:MEWNX:IT

And you can be as rude as you like, given that you saved me €€€ on the Volvo fuel pump with the information you provided.

Anyway, I also did all of the fuel pressure tests in Vadis and no issues, along with trying the fuel pressure tester in the driver's cabin (!) to see the readings under full flow - also fine.

When I said open loop I wasn't being clear enough, I meant only open loop prior to running conditions - I have checked this on my OBD scanner and it does take a couple of minutes for the system to go CL. Then it drives ok under almost all conditions, idle, WOT, etc.

I'm glad you agree with my suspicion, Old Painless, that being able to blow clear from the air cleaner intake through to the oil filler cap with little resistance is odd to say the least. And, as a compression test showed no variation between the cylinders, I don't think it's a major mechanical fault. That is what got me to the crankcase ventilation system. Could there be any other route this air is taking?

I haven't really looked into the rear lambda sensor as a cause as I thought this was mainly to test the condition of the cat - will check that out.

The exact symptoms I am seeing are: when I start in the morning it will generally idle fine (I cleaned the IAC valve a while ago and works very well now). If I touch the throttle at all the engine will stutter, drop revs to almost cut out and then recover with the IAC again. This reduces as it heats up, and is gone when the OBD shows CL. The remainder of the drive is fine regardless of conditions, but the fuel trims are off the charts.

I performed the compression test with the throttle closed, which I gather now is the wrong way to do it. Anyway, the values I got were 12.5 bar on 3 cylinders and 13 bar on the other.

A few more posts like this and we can publish this as a novel.

I think the short-term trim is used to adjust the long term trim, and the closed loop running is fuelled correctly with the trims, which means the open-loop is too lean! Anyway the plug condition seems ok - very dry with a whitish deposit, no electrode damage. The car was running cool for a while before I replaced the thermostat so could have caused some fouling. I am putting in new Volvo plugs this evening anyway so will see if that makes any difference - the current ones are also OEM.

The coolant temperature reads 90C when warmed up with the new stat so looks ok also.