Have we seen oil level sensors leak?

[turbodan]

The turbo 528 is insisting on leaking oil. I recently pulled the original engine and installed a fresher M20, which had been treated to a new rear main, oil pan gasket and a few others. All existing leaks were resolved however a new leak has appeared at the oil level sensor. I had R&R'd this sensor and installed a fresh o-ring, so this led me to believe it may have been leaking through the sensor body. I've never seen this but who knows.

I drained the new oil back into it's original 5 quart jug, removed the sensor from the old engine which never leaked, and installed it with a new o-ring onto the pan. This sensor now leaks as well. I suspect there may be some issue with the oil pan, which appears to have been replaced at some point due to remnants of part number labels. It seemed to be in great condition but this is a very simple mating surface that just isn't sealing. Unless I have two bad sensors, one of which I know was good, I must have a defect on the pan.

Have we ever seen one leak through the sensor? I think both of these are almost certainly fine.

[Mike W.]

With only M30 experience I have not had any leaks in half a dozen or more. They appear to be the same design, albeit with a different part number, different length I would guess. I have had problems with them always showing low on oil, but not leaking. Not that they couldn't leak, I mean oil pressure switches certainly leak, but they're under a lot more pressure, not that 200F oil by itself isn't enough to weep out of not much of a space.

Also given that I don't recall ever seeing it here before, I'd guessing it's a pan problem. I believe there's kits for detecting cracks in castings, might that be an option?

[Federico]

Maybe it's the oring. Have you checked that it sits proud of the groove in the sensor, to provide some squish? Can't really trust new parts these days. Also the proper oring should be dark brown, not black.

[turbodan]

This is one of the stamped steel M20 pans, very ductile. A crack is possible but I would be surprised. The mating surface could be deformed, that is a possibility. Aluminum would crack if it's bent but steel will show no obvious signs of a slight deformation.

The o-ring is the correct part specific for this application.

[1st 5er]

This is one of the stamped steel M20 pans, very ductile. A crack is possible but I would be surprised. The mating surface could be deformed, that is a possibility. Aluminum would crack if it's bent but steel will show no obvious signs of a slight deformation.

The o-ring is the correct part specific for this application.

I may have a pan if you need another.

On topic, I can't recall a leaking level sensor,
plenty of other places though.

[Mike W.]

OK, with much thought I've got a longshot. I had forgotten that etas had a steel pan, but I still have an idea. It's steel, but thin. They have a boss, flange, whatever, welded on as they're not going to bolt to sheet metal. But they made a lot of these, it's not like they had a guy running a bead all the way around. Or at least I don't think they did. Probably some sort of spot weld, which is ok, you seem to be the first with a problem with leakage. But a spot weld is not a great weld. So, with etas being a weakness for me, I'm guessing the pan to flange connection is less than perfect. OK, I'm reaching, but a thought.

[turbodan]

Here's a pic from the engine that just came out:



The pan is rolled over into a sort of flange with the three threaded studs spot welded in place. A very simple design that I've never had problems with until this one. I do understand why the B25 pan moved the sensor to the top though. Better to avoid a large hole in the bottom of the sump.

[Mike W.]

... I do understand why the B25 pan moved the sensor to the top though. Better to avoid a large hole in the bottom of the sump.

That's funny, on earlier M30s such as an '80 E24 I once had, had the oil level sensor mounted on the side of the pan. For E28 era M30s they moved it to the bottom. I do agree with you though, a big hole in the bottom doesn't seem like a great design, although as much oil is running down and being thrown around while running it might not make much difference.

[austin8753]

i have seen them leak from both the oil level sensor housing, and where it would affix to the oil pan. had to thoroughly RTV a persistent leaker once.

[turbodan]

Drained and disassembled for another attempt:





I see no obvious issues. There doesn't seem to much of a mark from the oring between 12 o clock and 4 o clock in the second pic, perhaps the pan is distorted here and it is relieving the compression on the oring.

This time with hondabond:



Wouldn't be surprised if it still leaks.

[Shawn D.]

On M30s, the studs are screwed into threaded blind holes, so there can't be any leaks via the studs, but these appear to pass through the pan. My bet is that the seal is good, but the leak(s) is(are) via the studs.

[turbodan]

On M30s, the studs are screwed into threaded blind holes, so there can't be any leaks via the studs, but these appear to pass through the pan. My bet is that the seal is good, but the leak(s) is(are) via the studs.

Certainly a possibility. I have driven it a bit since the repair and though it is far better than it was, I can see the first signs of seepage beginning to appear on the nuts. Better to have sweating nuts than dripping nuts I guess. I should have gooped up the studs too, I bet that would have sealed it up completely.

At this rate I doubt it forms a drop per week. If it doesn't get worse I'm happy with it.