When Unknown Repair History Strikes Your Bay
A customer just bought the car from an individual, with no pre-purchase inspection and very sketchy/limited information on service history, most of which was second hand word of mouth.
customers complaint: CEL, no power, engine won't rev past 2500 rpm, and heavy misfire
Upon arrival to service bay, faults were read out of entire car, lots of garbage in secondary systems about battery voltage, communications, etc...
Engine faults stored:
- p0011- intake valve timing, control performance bk1
- p0021- intake valve timing, control performance bk2
- p0345- camshaft sensor circuit, bank 2
During a visual inspection, it was noted that the timing cover had new silicone, and multiple things had been disassembled and shoddily put back together. Including bolts missing, different bolts used, things found loose, brackets missing, wiring that had been “repaired”, etc.. etc..
Looking at live data PID's for variable valve timing; BK1: INTV timing was -14.5 degree and BK2: INTV timing was -18.5 degree. Both pids were locked at idle with vehicle warmed up. Vehicle did run as customer described; 2500 rpm max , misfire, no power.
Based off just the live data I suspected an issue and initially thought maybe the flywheel had been clocked wrong as that is possible on this engine and will set similar faults. (visual inspection also found all bellhousing bolts had been removed at some point, "tooling marks")
My initial diag plan was to scope the timing signals and see what could be discerned as this is a quick and fruit full strategy. Allowing for the least intrusive testing, and providing a great deal of information.
I located a few known good CMP/CKP correlation waveforms for the engine (vg35de) from fellow techs online , and went and ahead and scoped both Camshaft position sensors (CMP) and the Crankshaft position sensor (CKP) to verify signals as well as possibly getting a standing on engine timing.
What I found was interesting, both banks were retarded as best I could tell based on the captured waveform, but the signal from BK1 CMP was garbage and sporadically hung at 5v for extended time frames. This is odd because the ECM had identified issue with BK2 CMP (p0345). The customer was informed that we had to replace the faulty CMP sensor in order to move further as we didn't know if the bad signal was causing the ecm to default to the retarded VVT settings ...Or whether it was just flat out creating logic errors. While scoping the CMP/CKP signals the vehicle overheated at idle on the bay, neither cooling fan came on but both were plugged in. INITIAL SCOPE
After BK1 CMP replacement, both CMP signals were now clean and direct copies of each other. But the correlation waveform then showed both banks were in fact retarded by about 16.7 degrees. This worked out mathematically based off the live data previously recorded. (BK1 14.5+ BK2 18.5= 33 degrees 33/2 = 16.5 degrees per crank rotation). Also interesting is the vehicle is now a CRANK NO START!?! SECOND SCOPE
Why will it not start, now that I have perfect CMP and CKP signal traces???..... Yet the engine was able to start and run with a faulty CMP sensor on the opposite bank than the ECM had set a fault for? BK2 set a CMP fault code, BK1 CMP sensor was actually found to be faulty by signal trace on scope capture.
This is when the investigation started to solidify. We thought maybe the original issue that caused timing chain cover to come off may have been water pump replacement, due to overheating (chain driven water pump), This could also explain the possible timing issue.
We call the customer and grill him ,just short of enhanced interrogation techniques, and he tells us he's the 3rd owner in a year and the car hasn't been able to be driven by him or the previous owner (2nd owner). The patchy story he lays out for us is: Owner 1 had gone in to engine to “fix something”, and it wouldn't run right afterwards. The 2nd owner bought the car off "craigslist" and was under the impression it needed a camshaft sensor (which he chose BK2 to replace) and a crankshaft sensor. Both of theses sensor did appear to be new when the vehicle made it to my service bay. The 3rd owner (current owner) was aware of the overheating issue and our water pump theory made sense to him based on all the details we brought to the phone conference.
Customer approved tear down. Once the timing cover was removed, we found a brand new water pump, and by a miracle of the gods, the timing chain had gone xxxxxx # of revolutions to then have the colored chain links match back up to the camshaft gear marks correctly (thankfully, and I can't stress how absolutely lucky this was ; 1 in a million plus!). What we found was the Crankshaft gear was one full link of chain off! BK1 CAM BK2 CAM CRANKSHAFT
It was finally proven that the owner #1 had been in the engine, and did in fact replace the water pump, obviously not following service info, as the whole cover had been removed instead of just the 2 access covers (for the main chain tensioner and water pump). Apparently Jumping the crankshaft timing 1 full link of chain during the process.
The main chain kit was replaced due to undesirable wear pattern on outside teeth, secondary chains were offered but declined, and were in good shape as well as the secondary tensioners anyway. All other bolts and issues related to a failed DIY “repair” were corrected.
It is believed that since the BK2 CMP was the only cam sensor originally putting out a plausible and repeating signal of any kind, the ecm only set that fault because the BK1 CMP had sporadic 5v climbs which sometimes had a leading edge matching up with the Crank TDC window on the waveform. This is just a theory though, based on recorded evidence and the fact that after we had 2 good CMP signals, the engine was then a Crank No Start. Does anyone care to comment or shed light on this theory?
Once the timing chain was replaced and timing set correctly, vehicle started and idled perfectly. With Live data for BK1: IN VVT 0 degree and BK2: IN VVT 0 degree ,at idle warmed up. LIVE DATA
Overheating was diagnosed after original complaint was resolved and found to be from a faulty fan assembly. One side fan barely had low speed, and the other side fan was dead on all 3 speed legs. Fan was replaced as well as the thermostat as a preventative measure. Engine runs and drives perfectly after repairs thru the entire RPM range.
This is just a case highlighting not only oscilloscope diagnosis, but making the most of the inspection, piecing together the problems, and attempting to approach a totally illogical situation with a semblance of logic and reasoning. I look forward to feed back and maybe some have input on the theory of the no start after both CMP sensors had plausible output
Thanks to Matt Wallace and Cody Gaddy for the known good waveforms!
Good stuff. What scan tool were you using when you saw "BK1: INTV timing was -14.5 degree and BK2: INTV timing was -18.5 degree" ? thanks
Actually was using the autel Maxisys. The slide in the article is just a list of live data, not a screenshot