Converter Hub and Pump Bushing Wear: The Loop That Kills Fresh Rebuilds

Nothing on a rebuild gets blamed as often, or as wrongly, as the front pump seal. Fluid shows up in the bellhousing a few weeks after the unit goes back in, the seal comes out, a new one goes in, and the same puddle is back. The seal was never the fault. The bushing behind it was.

The converter hub rides in a bushing pressed into the pump. That one bearing surface sets the centerline of the converter, carries the load it throws, and controls how much oil escapes out the front instead of doing work. When hub-to-bushing clearance opens up, four things happen at once: the converter wobbles and chews the seal lip, line pressure bleeds off where you cannot see it, the lube circuit downstream gets starved, and the added heat accelerates all three. Once that loop starts it takes the rebuild with it.

---

What the Pump Bushing Is Up Against

It is easy to see a bushing as a cheap piece of rolled bronze. The design brief is harsher. Sonnax lays out the envelope one of these has to survive: loads up to 1,400 psi, shaft speeds up to 9,500 rpm at the top planetary, stretches of running with low or no lube oil, and enough conformability to tolerate the misalignment that comes standard in a cast assembly. It also has to hold a specific clearance to the shaft, because that clearance sets the oil film thickness holding the two apart.

That clearance is not slop tolerance, it is a designed hydrodynamic gap. Too tight and the film cannot form, so the bushing runs metal-to-metal and in the worst case welds itself to the hub. Too loose and it is not a film any more, it is a leak path.

The Numbers Behind the Leak

Chrysler units make the clearest case, because Sonnax published the arithmetic. On the 40TE/41TE/42RLE/62TE family and the older A904 and 42RH, the OE pump bushing is machined in place after installation. That gives the factory a true centerline, but it also lets them run a wide tolerance on the pump housing bore — bore diameters vary by more than.003 in. casting to casting.

Drop a standard off-the-shelf bushing into a bore at the loose end of that range and you can end up with .008 in. or more of hub-to-bushing clearance. Sonnax measured the consequence: an average of ten times more oil loss than through a precision bushing sized to the bore. That is not a weep. That is a pressure circuit with a hole in it, hot, every second the engine runs.

The symptoms Sonnax attaches to that condition: pump seal leakage, pump noise, loss of prime, low pump volume, low cooler flow, TCC codes and switch codes. On Chrysler applications it names excessive hub-to-bushing clearance as a root cause of code 38, code 35 (loss of prime), TCC slip codes and front seal leakage. Every one of those gets chased somewhere else first — the pressure on the regulator valve, the cooler flow on a restricted cooler, the TCC code on the converter that just came back from the rebuilder.

The fix has a floor, too. Sonnax specifies its Chrysler pump bushing 12502-01 — steel-backed bimetal,.535 in. wide, 1.498 in. shaft diameter, 1.619 in. housing bore — only when the housing bore measures greater than 1.6175 in. Under that, you have a different problem and a different repair. Measure the bore.

Clearance Does Not Just Leak Out the Front

The front seal puddle is the obvious failure. The expensive one is quieter, because these bushings double as circuit separators. Where one straddles a crossover in a shaft or stator support, its clearance is the seal between two circuits. Open it up and apply oil finds release oil.

Ed Lee documented the textbook case in Transmission Digest. A 2001 Volvo with a 55-50SN came in leaking at the front. Soft-part overhaul, converter rebuild, and the only hard part replaced was the pump. It came back with no lockup and a P0740, and the converter got blamed immediately, as it always does. The real cause: that unit uses two input shaft diameters, 0.844 in. and 0.870 in., with two different pumps carrying matching stator bushings. The small shaft had gone into the pump built for the large one, and the resulting clearance could not hold the TCC apply circuit.

The same article shows how forgiving an air check can be. On a 4T65-E built with a 4T60-E drive sprocket support, air into the TCC apply port passed cleanly — enough flow got through the wrong-part clearance to fool the test. On the road it set a P0741 every week or two, only when lockup demand ran high. Even after the control valves and converter were replaced, TCC slip at 65% duty cycle sat at 400-500 rpm, down from 650-700, and still coded, because the actual leak was a clearance nobody had measured.

Material Is Not a Detail

Two bushings that measure identically can behave nothing alike. Material decides what happens in the low-oil moments every transmission has: startup, a hard corner with a low pan, the first thirty seconds after an install.

• Aluminum bushings carry a thin tin, silicon or lead layer on the load surface. It resists scoring well right up until it is damaged, which takes almost no time in a dry moment; once bare aluminum touches the spinning shaft it seizes fast. Sonnax is direct: aluminum is a poor choice anywhere lubrication is not immediate and continuous.
• PTFE bushings depend on how the PTFE got there. Sintered, or impregnated, means it is diffused into a bronze middle layer and locked in at the atomic level. Adhesive-bonded or thermal-sprayed PTFE is a coating, and coatings de-bond under load, which lets the bushing seize.
• Steel backing is close to non-negotiable. Without it, retention suffers and the bushing can distort under installation force — a fit problem you will not find until the unit is together.

The GM 6L family shows what that distinction costs. The 6L45, 6L50, 6L80 and 6L90 leave the factory with a PTFE-impregnated pump bushing GM does not sell separately; the OE path is buying a whole bell housing. Cheap substitutes, when they fail, delaminate before they seize, and the shed material goes everywhere the fluid goes — so whatever converter happened to be in the unit gets condemned for contamination it did not create. Sonnax's direct-replacement 104034A is PTFE-impregnated bimetal at 1.994 in. housing bore, 1.834 in. shaft diameter, 0.782 in. width. One caveat worth reading twice: not for 6L45 BMW models, which carry a bearing in that position rather than a bushing.

---

The Bench Inspection, In Order

1. Qualify the noise first

Converter bearing noise has a signature: absent in park and neutral, present in every drive range forward and reverse, louder as load goes up. Catching it at the noise stage instead of the metal-in-the-pan stage is the difference between a converter and a rebuild.

2. Measure the hub, do not just look at it

Mic the hub in at least three places along its length and around its circumference, looking for a step, a taper, or a polished band where the bushing rode. A hub with a visible wear step destroys a new bushing on the same schedule as the old one. Confirm you have the right hub for the pump, too. The Aisin AW TR-60SN / VW 09D shows two hub diameters inside one family: the bushing-style hub is 45.15 mm (1.778 in.) for its full length, the bearing-style hub is 44.95 mm (1.770 in.) and steps down to 44.30 mm (1.744 in.) where the pump gears ride. Mixing those, or mating a late pump to an early stator support and blocking the drain-back passage, blows the front seal out on a fresh build.

3. Check converter endplay and the pilot surfaces separately

ATRA's GEARS gives the field method: pick a shaft larger than the input spline and smaller than the stator spline, invert the converter so the impeller hub faces down, and push up on the input spline. Expect roughly .010 in. to.030 in. of movement on most units, with model-specific exceptions worth confirming with your converter supplier.

Endplay inside spec does not clear the converter, though. Sonnax's forensic teardowns make that point hard: three converters that all measured within OEM height and endplay spec had failed impeller-to-stator bearings anyway. On an early 4L80-E, the stator cap thrust surface looked fine and got reused — but the surface the bearing pilots on showed almost .070 in. of wear, letting the bearing run roughly .035 in. off center. Uneven loading, dead bearing. On an early Aisin-built Allison 1000, the separate bearing races wore into the stator and stator cap until the higher-loaded impeller-side bearing let go; Allison ended that one in model year 2006 with fully enclosed bearings.

Height and endplay are pass/fail on the stack, not on the surfaces. Check the pilot diameters and thrust faces on their own.

4. Read the shaft journal, in RMS

A shaft that looks polished can still eat bushings. Sonnax gives targets you can hit on a bench: bushing load surface under 25 micro-inches RMS, shaft journal 8 to 12 RMS, housing bore 60 to 125 RMS for retention. Grind chatter — waviness left by the manufacturing grind — is the sneaky one. A 4L80-E main shaft journal with chatter ridges wore out bushings repeatedly until the ridges were knocked down.

The fix is emery cloth and a light touch, and grit maps to finish predictably:

• 36 grit — 160 RMS
• 60 grit — 98 RMS (use on the housing bore)
• 120 grit — 58 RMS
• 180 grit — 34 RMS
• 240 grit — 17 RMS
• 320 grit — 14 RMS
• 400 grit — 10 RMS (use on the shaft journal)

You are not removing material, you are kissing the surface. On a journal, rotate the cloth around the axis of rotation, never lengthwise, and clean every lube hole afterward.

---

Installing It So It Survives

• Use an arbor press, not a driver and a hammer. Shock load deforms precision bushings, and the distortion becomes a clearance problem you cannot measure once the pump is together.
• Dress the leading edge of the chamfer before pressing into an aluminum bore, so the bushing does not gall the bore going in.
• Lightly lube the bushing before pressing.
• Retain high-load bushings with sleeve retainer, or stake them carefully, so they cannot spin in the bore, and run the fluid the bushing was designed for — the film thickness that clearance assumes is a property of that fluid at temperature.
• Fix your grounds. Stray current that cannot reach ground through its designed path finds another route, and bushings are a common casualty. Check chassis grounds for voltage drop on any unit that has eaten bushings without an obvious mechanical cause.

Putting the Converter Back In

GEARS lays out the sequence that protects the bushing you just pressed:

• Inspect the pilot pocket in the crank or flexplate adapter for wear, clean it, then a light coat of grease. The 4L80 and A4LD/4L55 are known for trouble here. Tap the flexplate with a wrench while you are there: it should ring like a bell, and a buzz means look for a crack.
• Prefill with one pint to one quart of the correct fluid. Most remanufacturers do not presoak the clutch lining, so the prefill is what gives it soak time and lubricates the bearings on first rotation.
• Lubricate the pump seal and bushing with fluid, not grease.
• Some units, the 6T70 and 6F50 among them, need the transmission held vertical for converter install. On a 45RFE, rough installation cracks the plastic turbine hub-to-cover thrust washer.
• Pushed fully home, the flexplate-to-pad gap should land between 1/8 in. and 1/4 in. on most units.
• Start all converter bolts before final tightening, and never run them with an impact. Overtightening, or a bolt that is too long, damages the lockup clutch reaction surface.

While You Are In There: Endplay

Bushing life and endplay are the same conversation, because endplay decides how thrust loads land on the surfaces a bushing is trying to keep aligned. On the 4L60/4L60E family, Sonnax's Randall Schroeder puts total endplay at .005 in. to.036 in., with .020 in. to.025 in. preferred, set by the selective shim under the bearing on top of the input drum. Check gear set endplay separately first, from the rear hub up to the front planetary snap ring: no published spec, but more than .030 in. between output shaft and rear hub means look harder, especially if a drive shell or planetary was swapped. With a new pump cover gasket, subtract .005 in.; with no gasket, add.005 in.

---

The Short Version

A front seal leak on a recent rebuild is a measurement problem, not a sealing problem. Before the new seal goes in: mic the hub, mic the bore, confirm the bushing is right for that housing, check the pilot surfaces separately from endplay, put the shaft journal at 8 to 12 RMS, press the bushing square with an arbor press, and prefill the converter. Every one of those has a number attached, and every one costs less than the comeback.

Working through the rest of the unit? Our thrust washer and endplay guide, line pressure testing walkthrough and TCC shudder diagnosis cover the circuits a worn bushing quietly bleeds.