Fuel pressures and AFRs all over the place

[Ryan.G]

Just to add also i have only mapped one race car which had a oil temp correction map so that when the oil temps went above XXX then it added extra fuel to cool the cylinders and set a lower rev limit to protect the engine.

[Tricky-Ricky]

Did it not occur to them to check and replace the water temp sender? if its not that then I'm with Ryan, a badly scaled/faulty correction map.

[Dash Rendar]

Okay, so now I need to look at the map, ecu, or temperature sender? I always had the ECU on my list of potentially faulty parts, but the apparent fuel pressure problem threw me off the track for a long time! (Everything else in the car has been changed, so I may as well carry on!)

 

I'm happy for you to all laugh at me, but I can only go on the advice I've been given. That's why I turn to the likes of Mark and you guys! What else can I do...

[Ryan.G]

Did it not occur to them to check and replace the water temp sender? if its not that then I'm with Ryan, a badly scaled/faulty correction map.

 

Its not going to be the water temp sender Dick. As that correction will only ever add fuel. And when it was orginally mapped with homer i have logs showing a good 89c all the time. As well as when i checked the problem over first.

 

Mark is going to send me the stats he has done showing the car going leaner with increasing oil temperature (starting at 70 and going off the scale by 82), and that the AFRs start to come back as the oil temps cool in the same fashion

 

Ryan

[bignum]

It won`t be the actual oil temperature, more like as the motor gets warm so does the oil, you know coinsides.

[absz]

Might be worth checking your aeromotive fpr again for vac leaks under boost conditions. i had a problem with my fpr were the fuel pressure kept changing. turn out to be the adjuster nut causing a vaccum leak on boost above 0.4bar. cured by using thread sealant.

 

you can see the air bubbles from the fpr under pressure.

[Dash Rendar]

It's not the FPR. I've tried three different units. And all the vacuum lines have been replaced. And besides... the fuel pressure remains constant with constant throttle, and is rising with boost as it should. Once the car gets hot and the AFRs go lean off the scale, the fuel pressure on idle remains at a steady 38psi. (Just as it was when cold.) At least... that's what Mark tells me. I'm picking the car up on Saturday, so I'll see for myself.

[b_have]

Ok, my 2p's worth.. When your wb shows off the scale lean, does the car run like a dog/PFC det warning flash? I am linking the coolant from the exhaust compromising your wideband reading from wherever it coming from, maybe a turbo cartridge leak? As the engine warms up, coolant pressure increases, evacuating coolant into the exhaust.

[Muffleman]

Very much my thoughts Charlie.

[Homer]

Hopefully Darren can confirm this, but from what I understand the car appears to run fine except for the occasional Det warning at VERY high load (it did this occasionally when I had it too). Edit - not that I think Darren has used the car in anger since the problems began.

 

The AFR gauge is the one thats showing the lean condition, whether it's the one in the car or the Ryan's one.

 

Mark also mentioned that he thought one of the "turbo" gaskets is leaking (from a conversation with Darren this afternoon, it appears this is the one between the manifold and turbo).

 

So, we have two possible sources of O2 getting into the exhaust and causing a lean reading on the AFR gauge:

- coolant from a cartridge leak

- air from a gasket leak

 

One other posibility is a head gasket failure, however this seem unlikely as there is no sign of carbon in the coolant and the leakdown tests showed a healthy engine.

 

As the lean condition appears only once the engine is at full operating temperature this would tie with the two possibilities above (thermal expansion causing either to fail).

 

Thats the situation as I see it now.

[Dash Rendar]

This is really good stuff! Makes a lot of sense.

 

It's exactly like Darryl says... The car doesn't seem to run like a dog when the AFR goes lean off the scale. In fact, it drives fine. I get the occasional det on full boost, but we never finished the mapping under boost because of all these issues. There didn't seem much point... While I was under the impression the fuel pressure was dropping, I wasn't going to boost it anyway.

 

I like this theory of coolant contaminating the AFR reading. It makes a lot of sense! Again, I think this might have been one of the early theories that got lost/forgotten in light of the fuel pressure red herring.

 

I don't think it's the 'turbo' gasket, as this is a new issue. I guess it could be coolant from the cartridge, but - on the other hand - I dont lose much coolant. I was under the impression that the exhaust was actually venting a lot of condensation. (Mark tested the efflux and said it had no coolant in it.)

[Scooter]

next question must be can you run the oil and water cooled turbo with just oil cooling and the water pipes linked up or blanked off?

[Sean1933]

Really sorry if i offended mate. I seriously though it was a windup as i couldnt see the relation.

 

However thinking about it now, if it is ECU and oil temp related then it has to be the oil temp correction setting within the apexi map. Personally if it were me i would bring the car to running temp, then reset all oil temp correction figures to zero (obviously having written them down first!) to see if it still pulls fuel.

 

The other side of the coin is what were the oil temps when originally mapped, i assume ryan can confirm these were at normal operating temperature (from datalogs?) therefore no oil temp correction was effecting the AFR's being set?

 

just my 2p

[Homer]

I don't think it's the 'turbo' gasket, as this is a new issue. I guess it could be coolant from the cartridge, but - on the other hand - I dont lose much coolant. I was under the impression that the exhaust was actually venting a lot of condensation. (Mark tested the efflux and said it had no coolant in it.)

 

 

This is a very good bit of info.

 

If the exhaust test was done while the car was experiencing the lean condition and turned up no coolant it can't be a cartridge or head gasket (since either would result in a positive exhaust gas test). However, the fact that you *are* losing coolant suggests a very small leak somewhere. Are there any signs it's losing fluid from any of the lines (particularly the turbo water lines).

 

IMO this might indicate O2 getting into the exhaust causes the AFR gauge to run lean.

 

To be honest, the amount time the engine has been apparently running lean, I'd have expected it to have shat itself by now. However the recent leakdown test showed pretty much the same as those taken while the car was BPU (They were actually a pecentage better)

[Tricky-Ricky]

Its not going to be the water temp sender Dick. As that correction will only ever add fuel. And when it was orginally mapped with homer i have logs showing a good 89c all the time. As well as when i checked the problem over first.

 

Mark is going to send me the stats he has done showing the car going leaner with increasing oil temperature (starting at 70 and going off the scale by 82), and that the AFRs start to come back as the oil temps cool in the same fashion

 

Ryan

 

Doh! good point thinking about it even when they pack up it just sends the ECU very rich, so what about the correction map?

also never thought the PFC could be oil temp corrected?

[Tricky-Ricky]

Ok' date=' my 2p's worth.. When your wb shows off the scale lean, does the car run like a dog/PFC det warning flash? I am linking the coolant from the exhaust compromising your wideband reading from wherever it coming from, maybe a turbo cartridge leak? As the engine warms up, coolant pressure increases, evacuating coolant into the exhaust.[/quote']

 

I see where your coming from, but i would have thought it would need an awful lot of water to affect the wideband reading, and then it would tend to make it read rich rather than weak, or screw up the readings completely, and very likely kill the sensor.

[Ryan.G]

Really sorry if i offended mate. I seriously though it was a windup as i couldnt see the relation.

 

However thinking about it now, if it is ECU and oil temp related then it has to be the oil temp correction setting within the apexi map. Personally if it were me i would bring the car to running temp, then reset all oil temp correction figures to zero (obviously having written them down first!) to see if it still pulls fuel.

 

The other side of the coin is what were the oil temps when originally mapped, i assume ryan can confirm these were at normal operating temperature (from datalogs?) therefore no oil temp correction was effecting the AFR's being set?

 

just my 2p

 

There is no Oil temp correction map on the Powerfc.

[Ryan.G]

I see where your coming from, but i would have thought it would need an awful lot of water to affect the wideband reading, and then it would tend to make it read rich rather than weak

 

Charlie is right it will make it read leaner as the oxygen content of the H20 is released

[Dash Rendar]

Really sorry if i offended mate. I seriously though it was a windup as i couldnt see the relation.

 

That's okay. None taken. This issue has gone on so long that it's causing me to occasionally misplace my sense of humour!

 

If the exhaust test was done while the car was experiencing the lean condition and turned up no coolant it can't be a cartridge or head gasket (since either would result in a positive exhaust gas test). However' date=' the fact that you *are* losing coolant suggests a very small leak somewhere. Are there any signs it's losing fluid from any of the lines (particularly the turbo water lines).[/quote']

 

The exhaust test was done on cold idle, not during the lean condition.

 

There is a very small coolant leak. It only seems to cause the level in the expansion tank to drop, at a rate of maybe an inch every couple of months. Once the expansion tank empties, the coolant level in the radiator itself doesn't seem to drop any more. (Bizarre.)

 

However, it used to have a worse leak from the hard rad pipe. (I'm told they're renowned for this.) At the time, it had a 1.3 bar rad cap on it. I'm not fond of these, as I think it was a 1.3 bar cap that resulted in me needing to buy a new radiator on my old car. I swapped it for a stock cap, and the coolant leak has dropped dramatically. I don't think the stock radiators were meant to hold coolant at 1.3 bar.

[Ryan.G]

So what about the correction map?

 

After 50c there is no correction. Set it this way when first saw the problem.

 

Trust me, this is not a map or ecu related problem.

 

Im going to think about this abit more over the weekend on the way to newcastle as something is getting effected by heat but there are alot of very good points being made.

 

Ryan

[Mike B]

another stupid q, is the wideband sensor accurate.. I guess you must have already checked afr with a secondary sensor.

 

is the egt higher than normal? (should be if it's really leaning out.. right?)

[Homer]

The exhaust test was done on cold idle, not during the lean condition.

 

I can see there are a lot of people replying to this thread so will just add that the exhaust sniffer test did not help (in this case) confirm anything as it wasn't done while it was apparently running lean. It needs to be done under the 'lean' condition to show if there is a leak when its fully up to temp.

 

So in summary, the potential for a coolant leak from the cartridge is still a possibility.

 

Great responses so far, it's certainly getting closer to finding the real issue here.

[Dash Rendar]

Charlie is right it will make it read leaner as the oxygen content of the H20 is released

 

Sorry if this is a silly question, but how? I'm presuming exhaust gases dont make an electrolytic environment, and I'm pretty sure exhaust gases dont get hot enough to cause fission! (It's been nearly 10 years since my biochemistry days, and I have a memory like a sieve!)

[Dash Rendar]

another stupid q, is the wideband sensor accurate.. I guess you must have already checked afr with a secondary sensor.

 

is the egt higher than normal? (should be if it's really leaning out.. right?)

 

The wideband sensor was replaced with a new one, as when we first saw this problem, it was assumed to be misreading. The new one behaved the same and Ryan subsequently confirmed it was reading the same as his Innovate.

 

In summary, I'm confident it's reading accurately. I'm not so confident about the environment it's reading (i.e. the possibility of water contamination).

 

I dont know about the EGTs. It's the one gauge I haven't bought!! I'm not sure if Ryan was able to measure this when mapping? It's a good question though. If the AFRs are apparently lean, but the EGTs are fine, would that further support the theory that the AFRs are in fact fine and are just misreading?

[Homer]

Just thinking about this from another angle.... I'm sure you've not done this recently, but lets say you get it fully warmed up (i.e. seeing lean condition on the AFR gauge) and took it out for a drive, set the BC to off (1 bar) and go for a WOT run in 4th. Do you get a det warning flash up? If it doesn't detect det, what is the afr reading at the max load point (about 5.3krpm)

 

It's slightly risky to try this as it means you're relying on the knock sensors to tell you to back off...

 

Try again at 1.2 and then 1.4 bar provided there are no det warnings.

 

The reason for suggesting this is to ascertain if you're seeing an afr reading that would *certainly* cause det, but no knock is being found by the sensors.

 

Risky approach though!