Does this knock sound familiar to anyone?

#1

Hey everyone. My spec e30 project is happily back off the back burner, trying to get it running right now. Unfortunately since buying the car, it has had a pronounced single knock or tick, half engine speed, every rotation, regardless of rpm or cold vs warm.
First thing I did after bringing it home was adjust the valves, with no improvement. Recently rechecked them to find still in spec
.
In the last month I did a new timing belt and tensioner(needed belt and suspected bearing was bad), water pump, rotor, cam seal. Knock still there.

Leak down test yesturday (engine cold), suspecting a bent valve, showed 0.5%-1.5% on all cylinders.

Ran the car today with no valve cover to hopefully catch something I missed or at least localize the sound but I’m still at a loss. Pushed on rockers, pulled plug wires one by one, was unable to change the knocking sound. I know it’s towards the front two cylinders. Is it still a bent valve closing slowly but sealing? Some kind of a worn rocker or cam bearing?

#2

I struggled with the exact same issue last year. After blowing a head gasket, I sent the head into the machine shop to get decked and rebuilt. Reassembled, then had almost identical knock. Tinkered and replaced a lot of valve train components, but ultimately discovered that the head was shaved just past tolerance and pistons 2 and 3 where kissing the head. Swapped in a thicker head gasket and knock disappeared. I have video of the knocking, PM me and I’ll send it to ya

#3

Interesting, thanks for the new place to look. Should be able to measure the head height while it’s still in place.

#4

Well, so I was able to get a decent measurement on the head height and it is factory original, unmachined.

Is it possible to have a slightly bent valve that may be slow to close but still seals well in a leakdown test?

#5

I would say probably not. The idea of “slow to close” sounds kinda unlikely. But a compression test would be a better check of your theory than a leakdown test. The former is nicely dynamic so should catch the “slow close”. A leakdown is too static.

Rich Bratton did a bunch of testing re. shaved heads. He found that shaving the head is a bad idea. The mechanical timing retardation that occurs costs you more hp then the compression you gain.

#6

Thanks Ranger, I was thinking I’d try a compression test next, see if it shows anything different. The head looks to measure unshaved. Haven’t found an easy way to measure the deck height of the block yet, but besides the noise, the engine seems to be proving more and more healthy with everything I check.

Is it normal for the rockers to move laterally on the shaft as the cam engages them? The #1 exhaust rocker moves maybe 0.050" front to rear every exhaust cycle.

#7

What’s the history on the head? Has it been off the car and rebuilt?

It sounds like the piston is kissing the head.

#8

Would the tapping be half of engine speed if it’s a piston? Maybe on exhaust stroke but not compression?(Dont know if that’s possible) It is for sure tapping once per cam rotation.
I don’t have a history on the head other than it doesn’t look to be decked (measures full height).

#9

Can guides be worn so far that it’s slapping side to side?

#10

From playing back the video at half speed, I think you are correct that the noise only happens once per cam rotation. That information doesn’t really inform much though. The noise could be a lot of thinks to include.

  1. Extremely damaged valve guide; sometimes chunks will just fall apart.
  2. Valve to piston contact from a bent valve; this doesn’t jive with your leak down results.
  3. Week or failed valve spring causing valve float and piston to valve contact.
  4. Deformed piston or collapsed skirt. Piston loading varies greatly between exhaust and intake strokes with the most pressure during the intake stroke. If a piston were deformed, it could rotate and tap the cylinder head or it could just be bouncing around the cylinder under high loads. I’ve had an IE HD rocker arm fail and destroy two piston skirts before. Side note, don’t use IE HD rocker arms.
  5. Piston to head contact because of reduced deck height or head thickness. This could also only happen on the higher pressure intake stroke because of piston rotation even if the piston isn’t deformed. It’s not uncommon for only one piston to have interference after a cylinder head service.
  6. Excessive rod bearing clearances. Perhaps the piston is traveling higher in the cylinder on the exhaust stroke because of excessive bearing clearance.

A few questions for you.

  1. How confident are you in your leak down results and what pressure did you use? At 100 PSI, most new engines don’t test within your range, .5% to 1.5%.
  2. Have you checked compression? You clearly have fuel and spark.
  3. How are you measuring your cylinder head thickness? I normally measure in several places with the head on a workbench; I’m not sure how you could measure with confidence on an assembled and installed engine. You should measure at room temperature, not a hot engine.

Regardless of your problem, removing the head will be enlightening. If you haven’t done it before, now is a good time to learn.

Rich Bratton

#11

Hey Rich, super appreciate the detailed response, giving me some additional motivation to just go ahead and start opening things up. This is my 1st E30 but have been fabbing, wrenching, and tracking cars for almost 20 years. Don’t have a huge amount of experience diagnosing mechanical engine noises, almost everything I do is external to the engine so you guys and this forum have been a big help introducing me to the m20.

To answer your questions:

  1. Fairly confident. First time using a new longacre tester, so could be the tool, but the procedure went easy and as usual.
  2. Have not checked compression yet. Yes all cylinders firing from what I can tell pulling a plug wire one at a time.
  3. I found a spot at the front right of the head where I could clearly see the head gasket and block deck, measured with the depth gauge on a caliper. Cold engine. But yeah, only able to measure in one spot.

I will pull the head when I get a few hours of play time, and likely the pan at the same time (good excuse to install a scraper anyway, right?) and report findings back. Hopefully it will be an obvious culprit.
Thanks again for the insight
Chris

#12

Before pulling the head, do a visual inspection on your valve springs. Look for a broken or missing coil, then press on each one with a BIG screwdriver. If one of the springs is weak, you’ll feel it. They should be really hard to push down.

#13

Springs looked and felt good from what I could see. Finished pulling the head this morning. Maybe found contact between the #1 piston and exhaust valve. Maybe the #3 as well. Might be hard to see in the pictures but the valve relief in the piston looks to be cleaned/ indented silver into the black buildup. Edges of the corresponding valves look cleaned as well. No damage from what I can see, just like the black has been knocked off. All other cylinders have even buildup/blackness. Is this what light valve to piston contact looks like?

#14

20190831_115543
20190831_115543.jpg2138x2851 2.76 MB

#15

20190831_115644
20190831_115644.jpg2138x2851 2.45 MB

#16

20190831_121431
20190831_121431.jpg2138x2851 2.1 MB
Closer look at relief

#17

Good update.

Can you see the dimples on the head casting?
Found a picture.
image

#18

Yep, all dimples are intact and nice and deep

#19

Broken spring would be my next guess. Slow closing. It does looks like a good amount of surface work has been done on the valve and piston. Otherwise, look how far each piston extends past the surface.

#20

Thanks. I will double check piston height. Going to remove the spring and valve and check them and the guide next. Also have the pan off so I’ll probably check the rod bearing(s). It’s all open, might as well take my time and go through everything.