Yes, the added fuel test is still super valid. If you have SDD or a scanner that shows live fuel trims, a small amount of added fuel will show up immediately, and make it pretty clear when you're in the right area. Vacuum leaks around the PCV, symposer, rear of intake etc are all found much easier doing this than smoke testing, in my experience.

 

@Robtrt8 What is the specific protocol for spraying MAF cleaner around the engine to find the leak?

Never done one but the smoke test seems easy to do from the YT videos I have seen and will detect the leak if it's there:

https://www.jaguarforums.com/forum/f...7/#post2835227

@RoverJoe Is the throttle body open or closed when the engine is off so that the smoke enters the SC?

@GerbilEngineer What about the blanking plate and vacuum plug for your removed symposer - could it be the source?

 

Definitely good advise. I haven't looked closely, but I did bring up the high pressure fuel line sensor values early on. They bounced around between the mid to high hundreds to the low thousands. From what little I have read, those values seemed to be within the bounds of normal. I would assume, in your case, that the sensor was reading habitually low and so the fuel flow calculations were making the computer think the car was getting way less fuel in the cylinder than it actually was. From what research I've done, 500-2000psi seems to be entirely reasonable.

I shall double check this evening.

 

My understanding, on current engines, is that you have to have real time logging to do the spray test. The engine computer adjusts the short term trims so fast that you won't get the change in RPM that you might in an older engine. Effectively you take a volatile spray -- I'd probably use a starter fluid spray, rather than a cleaning spray -- and give a little squirt to the area of a potential leak. With the intake in a vacuum state, the aerosolized fluid gets sucked in and provides extra fuel for combustion, so the computer sees the engine running richer than expected and trims the fuel intake down. As with all solvent sprays, don't go overboard and soak plastic and rubber. For the actual intake piping, it's pretty much impossible since its all plastic.

Throttle body is closed when off. It's opened by a motor and has a spring return. However, closed doesn't mean air tight. There is always a bit of an opening. You can see its normal, closed position in the picture below, and the gap at the bottom.


The symposer vacuum plate is on the back of the bank 1 intake plenum. It's obviously a potential place for a leak, though I'm pretty confident that it isn't the source. The body is gasketed to the intake plenum and the plug has thread sealant on it. Any leaks there will be minor and we are getting a large amount of extra air. I think a major leak there would still impact both banks, especially at idle, since this is a DI engine. If the MAF sensor is telling the engine computer a certain amount of air is going in, that air would be evenly split between the two cylinder banks. If bank 1 were getting extra air from a major symposer blanking leak, that would mean that more of the air, read by the MAF sensor, would be going into bank 2 so both banks would end up lean. I don't expect it to be there, but I won't rule it out either.

There is also a line that connects to the back of the bank 2 intake plenum. I think routes back to the intake, but I don't remember right off hand.

Interestingly enough, if it is small enough that the long trims start to drop at higher RPMs, then you can actually calculate how big a leak it really is and also whether it is before or after the supercharger. The computer is trying to keep the fuel air ratio at 14.1, measured at exhaust output (O2 sensors). The mass airflow sensors tell it how much air is going in and the high pressure fuel rail pressure lets it calculate how much fuel will go in for a given length of injector pulse. Nothing is perfect -- injectors age and get gunked and dwell open longer than when new, sensors age, leaks occur, etc -- so the theoretical math will almost never match reality and the primary O2 sensors will tell it which way to correct itself -- thats the very basics of fuel trim. At low RPMS, the total airmass is low, and intake vacuum is high and the supercharger isn't doing anything (probably in bypass -- I'll have to confirm). The extra air getting into the system will be a larger percentage of the total airmass and, as a ratio, the engine will be getting way more air than expected.

At higher RPMs, the airflow increases and, assuming the leak is not a gaping hole, the flow through the leak increases but not as fast since its highly constricted. As the RPMs raise, the leak contributes less and less to the total airmass. This is represented as a drop in the fuel trim since the actual air is closer to what the computer is expecting based on the MAF sensor values. Now, because this is a supercharged engine, as soon as the supercharger starts to contribute to the airflow, the intake vacuum starts to drop and will eventually go positive, becoming boost. If the leak is after the supercharger, then as the intake pressure drops, less air will enter via the leak vs normal intake and the fuel trims will start to drop more rapidly. In fact, at positive intake pressure, the leak would switch directions and the trim would switch from lean to rich. That line coming off the back of bank 2 and going to the left intake makes that a bit more complex. If the leak is before the supercharger, then the higher the boost, the more air sucked through both the normal intake path and the leak path. Because the leak is a constricted path, it won't necessarily increase at the same rate the regular intake will, so trims will decrease by some amount, as the ratio of air coming through the MAF sensor path and the leak changes. The larger the change, the smaller the actual leak is.

There is a lot of nuance to reality, relative to what I wrote above -- and I think its right, but I reserve the right to say "ooops" later.

 

With the engine off the throttle body resting place is like 15%, so it's open enough for air to pass freely. We smoke them from the intake tubes all the time with no issues.

 

This evenings troubleshooting session was.... interesting. Data was gathered.

To the best of my knowledge, there is not an air leak. I went through about half a can of starter fluid (50% ether blend), with not a wiggle in any of the fuel trims. I was surprised, to say the least. I realized later that I should have shot some in the actual air intakes in the front bumper, to make sure I got some sort of response. Shall try and do that tomorrow.

The high pressure fuel rail is sitting about about 440psi at idle. A quick rev will make it shoot as high as 2000psi. At 2000 RPM, with no load, it's 750-800psi. Have not been able to confirm expected lower value.

At idle, with nearly 40% on the long term trims, and near zero on the short term, O2 sensor 1, on both banks shows almost 0 current -- something like 2uA, which to my understanding means the fuel/air ratio is spot on. This is a warm start graph.


Long term trim creeps up as soon as the O2 sensors start providing data.

I was going to note that the sort term trims were behaving oddly as there didn't appear to be much adjustment in them. See below.


Then I realized it was just SDD being stupidly unhelpful. The scale on the short term charts is +/- 100%. 5% variations are probably a single pixel or so in the chart. I'll make sure and adjust the scale next time.

The following graph has the first 40seconds at idle. Then I bring the RPMs up to about 2000, smoothly and hold for a few seconds. Then blip the throttle a couple of times and hold at 2000-2250 RPM, blip and let it return to idle. After that, its some more mid rpm and four quick blips, mid rpm, one more blip and back to idle.


The throttle blips show up in the O2 sensor 1 channels. The short term trims aren't in screenshot (and are probably useless due to vertical scale anyway), but I wouldn't expect them to show much different.

Of interest in the above -- the long term trims drop to the 5-7% range if engine RPMS are kept at above about 1500RPM. It is only at idle and fast idle RPMs that the long term trims jump so high. This made me wonder if it takes a hot start to make it throw codes. The long term trims seem to start around 10% and it it immediately idled down, they may not rise fast enough to keep the engine from going lean. The short term trims can only do so much. However, after three warm/hot starts, I still couldn't trigger a code. It's possible it also needs actual load on the engine... I was just idling in the garage in park.

More data collection tomorrow.

 

I had a forehead slapping moment this morning. If you read back, I noted that SDD picked a useless vertical scale for the short term fuel trim graph. If you read even further back, I started out with "sprayed a bunch of starter fluid and got no change in trims". It dawned on me this morning that with the bad vertical scale, any response in the short term trims, due to the starter fluid, wouldn't show up. So this morning I re-did some tests before work.

This is what it looks like when you spray starter fluid in each of the intake openings behind the front bumper.

You can see the two huge dips in the short term fuel trim as the fluid vapors hit the cylinders.

And now, this is what it looks like when you find the (or an) intake air leak.

The dip for 90-110 seconds is when I soaked the right lower intake. The lesser dips from 50-55 and 10-20 seconds are more targeted bursts at individual junctions of the lower intake tube that connected to the MAF sensor section. Looking at the compression clamp on the connection between the lower intake tube and the MAF housing definitely looks like it didn't tighten it fully.

I'll get get that corrected this evening and maybe I get lucky...

 

Well there's your problem....



Lower intake pipe is not seated on the MAF sensor housing. That would do it. Loosened the clamp and the airbox mount and got it seated. Started to tighten down the clamp and right after I started to get some tension...




I'm guessing that one side of the solid band, where it meets the drive fitting, had already sheared and my tightening it was enough to do in the other side. Put a damper on my evening.

It is surprisingly hard to find 9mm wide, 70-90mm hose clamps. For the record, it looks like the original is a Norma TORRO Wave Spring 70-90mm. Norma part #01377708080. I have a regular stainless steel 9mm band, 70-90mm clamp arriving Sunday (theoretically). I'll head to the dealership in the morning and see if they stock them. Unlikely as I expect it comes with lower intake pipe, but I'll take a shot since they are only about 10 minutes away. Assuming they don't have it, I'll transfer the wave spring piece over to the new clamp.

 

Glad you found it, nice diag work!

 

Bit more on this saga.

First, the original wave spring clamps are impossible to find. Well, not impossible, but I'd have to buy a box of 250 and the lead time is 16-18 weeks. However, the wave spring can be transferred to a regular 9mm wide band clamp.

New on the left, old and busted on the right.



The band needs a slight deformation, where the spring sits, so that the spring generally continues the round profile of the band, rather than shortening it. Bit hard to describe but really its just this.



Slide the wave spring on and you are good to go. No telling how many patents I just violated...



Put it back in and still had a bit of a leak, though not nearly as bad. What is wrong with this picture?




There is a pocket molded into the groove around the pipe. The worm screw housing needs to be in that pocket. It does two things -- first, it keeps the band from rotating as you tighten it and second, it allows the housing to sit flush against the pipe when you tighten it. If the housing is not in that pocket, it's really hard to get the junction to seal. It won't be a huge leak, but it will leak. I flipped the lower right band so it fit in the pocket, tightened it down, then went and check the rest and adjusted those that weren't sitting right. Even took out the left headlight so I could check the left side. The same lower band wasn't fully tight there either, though it wasn't causing much of a leak. It was sitting correct and mostly tight.

And yet, I still have a leak. Not as bad mind you -- it trims out to 30-32% rather than almost 40%, but still a big leak.



Unfortunately that leaves less easy things to fix as potential causes. And the first thing I checked turned out to be the cause.
Here is was the fuel trim is supposed to look like.



You can see the short term trims jump when I rev the engine, but the long term trims are pretty much flat, and right around 4%. So what is the problem?

One, or maybe both, of the mid-load PCV diaphragms are leaking air -- badly. The above graph is with the PCV hose disconnected from the throttle body and the inlet capped. Just to confirm it wasn't the hose, I also disconnected the hose from the valve cover and capped the end of the hose, while leaving the throttle end connected. Same thing.

And we that, the Jag is staying home for next week's race. Just out of time. Fixing this will require pulling the charge air cooler, and chances of me getting the PCV diaphragm caps off without snapping a clip are effectively zero.

I replaced the diaphragms, when I had the engine apart, since they are about impossible to do with everything together, the parts aren't terribly expensive, and they are a wear item. I am curious to see the cause as there isn't much to screw up unless the diaphragms are just too thin so they aren't sealing well.

 

You can do the PCV caps without removing the charge air cooler. The tabs will break when they come out, but I was always installing new kits anyway which come with new caps.

 

I figured I'd give it a go regardless. I assume the process it to release four clips you can access then lift the cap and just let the two hidden clips break.

Yes, I'll be putting new caps on.

@RoverJoe Do you put anything on the diaphragm lips to promote a good seal? Motor oil being the prime choice. I note from looking at pictures that the originals I pulled out looked pretty "wet" around the entire inner part of the lip, as does the housing.





When I installed the new ones, I wiped down the housing to make sure there wasn't any crud in it and then installed the diaphragms dry.

 

I don't put anything on them, a bit of silicone spray maybe to help them clip into place as i'm doing them with the engine fully assembled and in the car, but no in general I would use a pick or small screwdriver to snap off a few tabs until I can rotate it and lift it out, then set the new rubber in, carefully align the cap/spring, and press it down until all the tabs click in. Done a bunch of them in place and never had an issue but I could see how rolling an edge or pinching the diaphragm could happen for sure.

 

Thanks. That's pretty much what I did on the install. Will be curious to see what I find. Will be a couple weeks before I can report back.

 

Yes I have replaced both of the diaphragms with the SC on.
Here is my thread. It's not an easy job but I got it done. Maybe with practice like RoverJoe you could get faster.

PCV Diaphragm Replacement

One big tip I got is to use a steel ruler to hold down the spring while trying to get the plastic cap in place.
One thing that came up and was never answered is are there any differences between the 2 springs? I saw a post saying there was but my replacement ones had 2 identical springs. No problems so far and it's been several thousand miles.

If you have not seen this I also replaced that terrible back heater hose with a shorter cheaper version. Fit's fine and I am in debt to whoever posted that originally with the replacement part number. The new hose assembly eliminates that terrible bolt going into the back of the engine that has almost zero access!

It's documented in the above thread with pictures and part numbers. Take a look and see if you like it.
.
.
.

 

When I took the original diaphragms out, I had one short and one long spring. My kit came with two of each, so I just matched the lengths when I put the new ones in.

The ruler tip is a good one.

 

I believe you need both springs, as there's a break over point where it opens under certain pressures and does not under others. I always got the same thing out of the new parts bag as what came off the engine so I never had to mess with, just put the whole new kit on as it came.

 

GerbilEngineer thanks for looking and now what goes where? Long spring or short in the front PCV cap? What you found is what I read after I did my PCV stuff. I just swapped them as both of my springs appeared to be the same. But I installed RKX aftermarket ones because I wanted the better diaphragm material.
Nothing wrong I guess as it's been 3K miles with no problems or codes.
.
.
.

 

I am new to this forum. Ive been doing a lot of reading to familiarize myself with my newly purchased F-Type sports car. Now I see this trash. Really .... ?!?!?!??!

 

There's a triangle icon with an exclamation point at the right of the header for the post. Use that to report the post. The moderators will take it down. They're usually pretty quick about it.

Also, please do not quote the actual spam, as that can persist after the original is taken down.