long starting when cold on xc70: dealer says crankshaft thrust bearings

[jimmy57]

I cannot imagine what you could do to a starter to generate a large magnetic field that it doesn't do anyway unless the whole outer housing rusted off and it is just exposed wires.
OTOH I think a new crank sensor is a hell of a lot cheaper than a new engine. I would start with that and/or go to another shop. It has literally been decades since I even thought about thrust bearings, this seems a rather unusual defect. You might ask what they measured and how???

The flex plate carrying the perforated ring the crank sensor reads is steel. The starter is about 90 degrees away and 5-6 inches. The starter dragging due to bushings or other failure causes starter current to go up. Magnetic field is directly related to current flow. If the magnetic field is large enough to reach the snesor area then the the fluctuations in mag field will cause signal waves similar to what the sensor creates when perforations pass the sensor and its internal magnet is attracted to the dividers between perforations. Some starter faults do cause long crank or no start in white motors.

[jimmy57]

The starter pushes flex plate back and if it goes far enough the edge of the apertures in sensor surface affect the sensor and not just the dividers between apertures. Crank end play is pretty easy to check. If the crank is pried forward and then pried back and there is an audible clunk then you probably have an issue. The normal end play of .08-.27mm is enough that the oil keeps you from even having noticeable motion, much less be able to hear a clunk noise. The grounding is not what causes issues on these engines. The cleanliness or loose bolt between the trans and engine does cause it. The suppression of starter magnetic field is just adequate with the aluminum engine and trans. Loose bolts allow the suppression to fall just enough to upset that. The loose fasteners or grime do not cause Voltage drop issues. If the transmission has ever been out then the bolts or grime issue is certainly possible.

[abscate]

How many hours were charged for crankshaft endplay measurement? That involves removing the timing belt and putting a dial indicator down on the crank pulley at a minimum.

I am having difficulty imagining them do this.

[IslandV70]

Does anyone know the spec for starter current? That should be easy enough to check with a clamp-on meter. I have never measured my car, but my boat has current meters and draws 115-125 amps to start.

[537playing]

I read on the past that a Volvo model(perhaps a S80) has the same issue as yours and the solution was to cover the wire going to the starter or even re route it. The wire was throwing off the electrical measurements on the crank sensor.

[mrbrian200]

The area on the end of the crankshaft sensor that actually senses the holes on the flex plate is only a few milimeters wide, centered on the end of the sensor. End play would need to be 1/4 inch or more before before it started to degrade the sensor output. Is that even possible on an engine that still can run??? And why would it only be a problem on a cold engine when ambient temperatures drop below freezing? That whole business blaming it on endplay, on a well maintained engine that doesn't even have 100k miles on the clock, just doesn't add up if you ask me.

[mrbrian200]

If your chassis # (last six digits of VIN) is 254140 or under, for ~$25 you can try a different starter out of a pick and pull and give starter motor interference theory a whirl. If under that chassis# you can pull a starter out of virtually any Volvo L5 engine from 2007 going back to the mid 1990s. FCPeuro as well as many Volvo parts stores.. find the exact part for your car, then select 'what this fits' and viola! A magic list of other models/years you can pull a used part from.

I have doubts that the new crankshaft sensor is going to make a difference. But you never know. I recall several previous reports of this issue where replacing the sensor didn't do it.

[jimmy57]

Thrust wear happens but is rare. Trans fluid applied to torque converter gives a forward push on crank (depressing clutch on manual trans equipped car does the same) and if the induction hardening of the thrust radius of the main bearing (on white motors it is one bearing in from rearmost to make engine less long, if thrust is at rearmost then the back end of block has to pushed out to hold rear crank seal) then it can wear in time. The crank would go forward from this wear.

Crank end play is easy to check with no disassembly other than remove RF tire and bending that liner up and clamping it up. The dial gauge tip is placed on the crank pulley between the bolts on that flat area. A magnetic base can attach to the brcket fro the front end subframe engine mount. Pry the crank forward using subframe, zero dial gauge then push in and read dial gauge. .08-.27 mm is the spec. .002-.008 inch

[abscate]

You can probably just feel 0.002 while 0.008 will be a definite 'clunk'

[mrbrian200]

You can probably just feel 0.002 while 0.008 will be a definite 'clunk'

0.008, or even something 3 or 4x that amount couldn't possibly be enough to throw off this sensor. Even at 4x out of spec we're talking less than a millimeter of play here. I maintain that the active area of the sensor is only a couple mm in the center, and the holes on flexplate that the sensor 'reads' are at least 5x wide as that. If the crank sensor is that sensitive to such a 'minor' off-center misalignment (which I doubt) it needs to be redesigned.

I seem determined to shoot down some theory of crank end play being the culprit of long crank/hard starting (but only in cold weather??)

This sensor effectively would act as an antenna with a bi-directional conical shaped radiation pattern. Close to the sensor where the flex plate is, the radius may only be a few mm wide, but as it extends down the pattern would widen, possibly putting he starter motor within the 'beam width'. Hypothetically if conditions are 'just right' let's postulate one of the widings in the starter motor is arcing between the wires and could could make this hall effect sensor 'ring like a bell'. Again as you crank on the starter, the windings warm up, winding and arc current decreases, and the EM being pick up by the sensor drops below a threshold where the engine computer can make out the intended signal produced by the holes on the flexplate passing the end of the sensor.
If arcing is at issue you might be able to pick this up with a handheld AM/longwave radio, but it might be difficult to determine what level is 'abnormally high' compared to the normal amount of EM noise the DC motor brushes would produce on a 'good' starter.

I also have not discounted the possibility that there's nothing wrong with the starter, but that inadequate mating between the engine and trans, that normally forms a sort of faraday cage, might be responsible for issues with this sensor and a perfectly good starter.

Edit edit edit..
Now something else occurred to me... so the square wave output of this sensor probably tops out at a few hundred cycles per second. It goes without saying that the design engineers at Bosch would be well aware of this and likely design some sort of (likely passive) low pass filter right after the sensor output comes into the ECU, before A/D sampling. Lets say this low pass filter, probably not too dissimilar in design from a passive audio low pass filter has a failed HF shunt capacitor effectively disabling this circuit that is meant to filter out unwanted higher frequency EM interference. The guys at Xemodex who have probably reverse engineered these ECUs as well as anybody, might be able to answer whether this could be a possibility.
The crank waveform we can see in Vida would be after this analog signal has been digitally sampled, and could even be subject to some complex software 'interpretation' - in effect we're not looking at the actual (analog) sensor output or possibly even what the input side of the AD sampling converter sees.