Discuss all things tuning in this section. News, products, problems and results.
I'll do my best to describe this. I'm stumped. working on fuel trims has been going great until attempted a light throttle pull to 6000rpm. to verify the cells up to 5000rpm. hit some sort of wall, stumbled and ever since the o2's show the fuel delivery is off in weird ways. Engine is a built 3.5l ecoboost, built 3.7l heads. running a p1sc procharger. at the time of the incident it threw p0373, home tooth count off. but it's done that before and it didn't affect driving like this. the code doesn't stay active. and I checked the wiring. there's some filtering that needs applied I think. so decided to work on that but when I went to run the car again to work on that it wouldn't rev in nuetral. just rolled all over and fueling was super lean.. On the road it would run rich in the 1500-2000rpm range. now the rpms roll and it runs super lean. the log included is a little weird. the first 90% is the road test and then the end of it I think is when I restarted and it wouldn't rev. tried re-recording it as a separate log but the 2500 logger started being weird. before restarting again, I'd like to have some ideas of things to try.
How is cam timing done on that engine? Any chance it has moved? That's the classic sign of a cam belt that is off-by-one tooth. It would happen occasionally to our Pinto race engines.
One theory I hoped against was it jumped timing. it uses 1 long chain to connect the two intake sprockets to the crank sprocket. I used the ford lock out tool on the cams, can't mis-time it. then marks on chain are lined up to one each on cam sprockets and the crank sprocket. then you install chains out from there to each exhaust sprocket. same process. then the tensioner should hold it, remove lock out tool. but it's not unheard of for it to jump timing at high rpm. it's a new engine with a new tensioner but who knows - maybe tensioner didn't seat right in the timing guide? getting at the timing chain is a huge hassle, whole front cover has to come off. before starting a tear down, would a compression test or leak down be worth while? which one? I have tools for both.
I would do a compression test first. If it was not really good, then I would do a leak-down to determine the cause.
ran the compression test. it was cold (intake blocks one bank so it's not easy to do a warm up and then run a compression. The battery was a bit weak so the numbers were lower than I expected but within 20%. the left bank (1-3) were lower than right bank (4-5). the engine is new but has a 100miles and hard passes on it. I'm not seeing something to worry about - is the difference indicative of one bank being off time?
Those are pretty low. Was the throttle open and all the plugs removed? If you think it's a speed issue, I would charge the battery and repeat the test.
A bit of a long shot, but is there a chance it was sitting for a bit before you took it for a road test - I was thinking it may have sucked some crap from the tank and fouled the filter.
Have you done a check of the fuel pressure - you don't seem to have mentioned it? If it's low, it could also have a regulator or pump issue?
I concurr the numbers are low. it's a harbor frieght compression tester so I'm skeptical of its accuracy. its a fresh, professional built bottom end with known good heads. the motor made 9lbs of boost at 6k which is in line with what the combo will do - should make 13lbs above 7500. looking at the difference in numbers was the goal. The intake is off the car so no throttle body. all the plugs are out. it should be about ideal for a test. I reran the #1 cylinder on a fresh battery and the number was identical.
the fuel pressure readings are in the log file. they track the map gauge flawlessly, even when its rolling because the idle is rolling. I agree checking the fuel filter is a good idea, I'll pull it tomorrow. the fuel in the tank did sit most of a year. tho its almost a fresh tank now. a lot got drained, installed the deatschwerks coyote return style kit, all new braided lines, new fuel hanger with two dw400 pumps one is new and one was in the car for a year tho the cars been down a lot doing this engine build. on 93 in the tank, the fuel need isn't stressing the fuel system yet. it's set up for e85 running more boost.
so I think it's a good call that tho the pressure is reasonable, there's something in the line blocking it. I'm also tempted to pull the injectors and flow test them. just bought a generic flow/cleaner. but that's a last resort. the car isn't running rough. just lean and doesn't want to rev. spark plugs reflect that as well. they're still fresh but showing the white like it's burning hot.
for verification, hooked up the leak down tester to all 6 cylinders. leak down was 20-30% across the board and generally matched the compression test values. the pcv valves appear to work as the oil catch can breather is where all the leak down air came from.
I'll pull the fuel filter element next.
With that leakdown, taken with the compression test - 20-30% 'may' be acceptable with large ring gaps, but it's certainly on the high side, and a leakage test on the fresh engine, and after a couple of runs, would have given you some idea on what would be 'normal' - I think you may looking at much bigger problems than 'just' the fuelling - that leakage may still be acceptable but did you verify where the leakage was coming from?
You can buy cheap boroscopes that work with smart phones, so before going much further I'd suggest having a good look at the bores, as I think you may be unpleasantly surprised.
There are a few other things that will give a 'lean' lambda, such as a faulty brake server/boost valve (might have mentioned that), or intake or exhaust leaks that may have developed during the drive (ditto), with a common problem being a light misfire that isn't picked up.
Gord, I concurr that there's something "global" that is at play. working through the problem, agree The fueling is a symptom of something. looking for that before fiddling with the base fuel table as the table is almost spot on. I'm mechanically inclined but new to engine diagnostics. its weird for a vacuum leak. it rolls instead of hunts but I will cap off the brake booster and see if it tames down. block and tackle in steps to see what changes the outcome.
Appreciate the ideas of what to try next as I was stumped.
The ring gap is on the larger size as it's designed for a lot of boost and the engine is cold from sitting. I can see the point that the engine may have some bore wear - though that seems to be a separate issue to why the engine isn't running 100% doing neutral revs
other things from the current list: took out the fuel filter. it was a little dirty, a few pleats had caked but far from enough to block flow. cleaned it up. going back to a global issue - unless all 6 injectors are clogged, it doesn't act like an injector issue. i.e. one bank of 02 is way off and the engine is rough.
to test where things are after all of the work done to this point, put the intake back on and fired it up. ran great again, free rev in neutral no problems, fueling match commanded.
comparing the traces, the MAP showed vacuum was not decreasing as throttle opened during the trouble run. now it is again. very smooth changes that followed the throttle traces. where as before the MAP values were jagged and stayed flat.
I have no idea how an engine could do that mechanically unless something 3" in size was physically in the inter cooler tube blocking it. which there doesn't appear to be - it went back on the car after the engine went in recently and it was stored in crates.. the other culprit I can think of is there was something small in the reference line to the MAP sensor and unhooking everything cleared it? it's brand new vacuum rated 1/4" so it shouldn't be choking off but I could theorize that a small item got stuck in there and was pulling the line closed at a certain vacuum value on the map sensor side and increasing throttle would just suck it harder shut preventing the change in vacuum from reaching the map sensor.. has anyone seen that before?
unhooking the brake booster raised the map value, the fueling went up (lambda down) and it revved up til it reached equilibrium at commanded lambda accordingly. so ruling out vacuum leak out as it doesn't fit the facts from the log of the lean condition. I'll post back if it pops back up again on next drive, hopefully tonight.
Attached a screen grab from mega log viewer. the top two trace boxes are self-explanatory. the lower one is a calc - wideband lambda minus target lambda. the pattern is repeatable that when the cam overshoots and then over reacts, on the way back up to target the map sensor shows an increase in vacuum which leans out the mixture.
So to get to this data a discovery was made. the haltech downloads everything in memory when a dump is pulled. not just the latest. so the data was becoming increasingly skewed as it included all the previous sessions as the fuel was dialed in. so I kept pulling fuel.
last night did four laps, clearing memory after each dump and after these revisions it drives/revs much better. but it still has these lean spikes that appear to be related to the cam position error. watching the VVT courses it looks like the VVT progressive values are too high as it way over shoots and comes back. is that a fair assessment?
update: going to close this as the main issue of fueling is addressed by new data logging method. going to open a new one relating to the lean spikes when intake VVT doesn't match commanded.