Uh oh, it's not the timing chain guides

[curtmcd]

Some side to side play is normal except in the thrust bearing. You want to push and pull them (up and down for example). That'll be harder to do in situ.

There doesn't seem to be any play whatsoever on the brackets if I try to push upward on them with a screwdriver or pound on the screwdriver. Would it not require a fair amount of play to create such a loud rattle?

I ordered a $40 wireless borescope from Amazon that should be here tomorrow. It seems like a very useful thing to own regardless.

[Santaclaus4]

Up and down play of even 1MM is enough to destroy the engine. The bearing clearance for engine bearings will be in the neighborhood of 0.002". There is an actual range of course and it may be a llittle different from that but that'll give you an idea how little movement you can have and still have a bad bearing.
I haven't built an engine in years so I'm a little rusty on the bearing tolerances, etc. Maybe some of the guys that have more recent experience will chime in. The point is that by the time you can see the bearing is damaged it just may be too late. However, that being said...I've polished crank journals with an emery cloth soaked in oil and installed new bearings and had the engine run and sound fine for years after that without even taking the engine out of the car. I just had the oil pan off and removed the bearing caps.

[geargrinder]

There doesn't seem to be any play whatsoever on the brackets if I try to push upward on them with a screwdriver or pound on the screwdriver. Would it not require a fair amount of play to create such a loud rattle?

I ordered a $40 wireless borescope from Amazon that should be here tomorrow. It seems like a very useful thing to own regardless.

So as Santa says it's really the opposite. A tiny slop is plenty for the force of the crank and combustion to make noises...

[JimLev]

As Santa has said the side to side clearance is required on the rod bearings, should be at least a couple of thousands of an inch. Even if it's a lot more that isn't the source of the noise.
You really need to remove the rods end caps to clean the oil of both bearing halfs to see if there is any up/down play.
The oil will take up some of the excessive gap (if there is one) unless the gap is really big, but then the knock would sound like a Tommy gun.
Although it's pretty rare it could be the wrist pin.
If you can't find anything conclusive and end up putting it back together and still have the noise I'd disconnect one coil at a time. That will tell you if it's really coming from the rotating assembly and from which piston.

[itsbrokeagain]

Since you have the caps exposed, unbolt one and pull the end off and inspect the bearing. I've never heard a rod sound that loud at idle, they usually have been making some noise for quite some time at various rpms, until the tolerance gets bigger and bigger and starts knocking at a lower and lower rpm.

[curtmcd]

Unfortunately, it would be too difficult to remove the visible bolts from underneath. I could predict parts falling out and having to disassemble the whole shebang!

I got a borescope, but there was little to see. I could see the top of the piston through the spark plug hole, and the bottom of the piston through the oil pan. The cylinder wall seemed OK.

At this point, I've decided to go ahead and reassemble. I timed it this evening. I'll try running it with the belts off, and pulling coils one by one as suggested.

One question regarding VANOS timing: why the continuity meter? It only conducts at the far left of the rotation anyway (full CCW).

[JimLev]

The meter continuity doesn't always work due to dirty metal, don't worry about it if you can't get a reading, just make sure you turn the inner section of the vanos hard fully CCW.

[R Shaffner]

Bummer. I still hope you can find an easy fix. I like Jim's idea about isolating cylinders, once it's all back together.

I still wonder about the noise from bad rod bearing and whether it would be 1x or 2x the engine speed. (Seems to me that a loose piston would knock in both directions.) If it were my car, I'd consider getting it back together enough to crank with the starter to listen and watch (carefully) from underneath with the pan off. I'd particularly want to look and listen to the oil pump in motion.

I don't want to hijack your thread, but perhaps others can add some insight here. I always heard that worn bearings could cause a drop of oil pressure, especially at idle, because they let the oil escape from the pressurized passages easier. Since then I've read that a slipped bearing could block the passage and cause low pressure (which I don't understand).

In any event, if your bearings look pretty good from underneath it seems that they are unlikely to be reason for the sudden drop in pressure. (Still wondering if the two symptoms are related.)

[geargrinder]

One question regarding VANOS timing: why the continuity meter? It only conducts at the far left of the rotation anyway (full CCW).

Its an odd "overkill design feature" by BMW IMO. Ignore it. Its only CCW tho' because that's all you care about. That's the direction the hubs need to be pinned fully when cranking down the bolts. We don't care about anything else really.

The meter continuity doesn't always work due to dirty metal, don't worry about it if you can't get a reading, just make sure you turn the inner section of the vanos hard fully CCW.

In addition, when the VANOS are rebuilt using the 'homebrew' methods that we all use now (the Beisan method), pressing the hub seems to often distort the plastic inside enough to make the pin no longer contact. At least that seems to be what anecdotal evidence suggests.

Bummer. I still hope you can find an easy fix. I like Jim's idea about isolating cylinders, once it's all back together.

Agreed! Might be frustrating to test that then need to tear it all down again but it is what it is...

I still wonder about the noise from bad rod bearing and whether it would be 1x or 2x the engine speed. (Seems to me that a loose piston would knock in both directions.)

I think you'd find its 'on average' 1x. The rod is in compression most of the time...
Intake - tension
Compression - compression
Power - compression
Exhaust - compression

So... on 1st revolution there's one 'transition', and at end of 2nd revolution there's one transition, meaning - assuming it knocks during the transition, an average of once per rev. Granted the periods are asymmetrical, so if thats the case, it'd be a loping rhythm of 180 degrees / 540 degrees (knock knock, knock knock, knock knock, etc.), but on average... 2 knocks per 2 revolutions... But your ear wouldn't be able to hear that difference that's for sure and a simple frequency measuring application or tool would probably not indicate the difference but would average pulses per second. You'd see it on an oscilloscope though. You could probably 'scope the knock sensor signals and see it pretty well.

I've read that a slipped bearing could block the passage and cause low pressure (which I don't understand).

I am guessing they mean a blocked bearing passage would cause low pressure to that bearing.

[R Shaffner]

All great points, GG, as usual.

I have no personal experience with bad rod bearings, so I'm still learning here...

Would the knock be heard while cranking the engine w/o starting it?

If one pulled the plug on a cylinder with a bad rod bearing, then what? Would the knock be louder? Would it then be 2x? Would that also be a way to tell that it was a rod bearing, and which cylinder it was in?

[BimmerBreaker]

Rod noise will be present when the engine cranks, but may be harder to hear until the engine is idling
If you pull a coil on the affected cylinder the noise should get quieter as there is no longer combustion occuring in that cyl to force the piston down which in turn causes it to slap the crank (if the rod bearing is whats bad)

[curtmcd]

vinyldude had used audio analysis to determine the noise is 12 Hz (720 RPM).

When slowed down to 15% speed, it sounds more like a "buzz clank clank" pattern repeating at 4 Hz.
Here is an a recording of that.
Is there something in the engine that sort of goes in threes?

[geargrinder]

Did you read my previous point at all..? 180/540? Knock knock, knock knock, etc? Is that not pretty much "buzz clank clank"? ("Buzz" obviously being the background noise between the knocks...)

[TheAngryBear]

It’s like Hunt for Red October in reverse...

If it is a knock (still hopeful for otherwise) would it be possible to detect during timing when turning by hand? Or is it only under pressure?


2001 750iL DD74441
Stock

RIP: 2003 540iA Sport GS56111
H&R front springs, Ultimate Cup Holder, Euro Dash & Armrest, Grom, BavSound Stage1

[curtmcd]

Did you read my previous point at all..? 180/540? Knock knock, knock knock, etc? Is that not pretty much "buzz clank clank"? ("Buzz" obviously being the background noise between the knocks...)

Yes, thanks I appreciated that very much, and intended the additional info specifically for you. My point was, if the rod was shifting, wouldn't the pattern be a 4-cycle instead of a 3-cycle?

Since idle with A/C off is 750 RPM (12 Hz) and the noise base frequency is 12 Hz, there is a single buzz or clank per revolution. Bizarrely, we observe a uniform "buzz clank clank" pattern, 3 noises every 3 revolutions (12 Hz), so this 3-cycle pattern repeats at 4 Hz.

If every two revolutions the rod should clank once into tension and once back into compression, I would expect a 4-cycle pattern such as:
- clank clank silent silent (180 + 540)
- buzz clank silent silent
- buzz clank clank silent
Also, the noise base frequency ought to be 4 noises every 2 revolutions (24 Hz), and this 4-cycle pattern should repeat at 6 Hz.

That's why I wonder what engine function might repeat every 3 revolutions (as opposed to 2).

[geargrinder]

If every two revolutions the rod should clank once into tension and once back into compression, I would expect a 4-cycle pattern such as:
- clank clank silent silent (180 + 540)
- buzz clank silent silent
- buzz clank clank silent

HerpderpWTactualF are you talking about. Ramblings of a madman, brother.

Where are you getting "4 noises every 2"? Where are you getting those other 2 cycles?
If it is indeed at the transitions, its 2 noises every 2 revolutions, that's what I've told you and assplained very clearly.

In internal combustion terms, each "cycle" = 180° of crank movement. Therefore a 4-'stroke' or 'cycle' engine has only 720° of rotation before everything repeats.

180° + 540° = 720° = full cycle of the engine, now it all starts repeating. Nothing else happens after that, it repeats.

Do you think you have some kind of 12 cycle engine!?!? and if so how come you think it clanks differently in the second 2 series of 4-cycles!?!?!

And how do you 'expect' this?!?

- clank clank silent silent (180 + 540) << this is 720° of revolution, now it just repeats...
- buzz clank silent silent << WTF are these? what cycles are happening here?
- buzz clank clank silent << (same WTF here?)

You seem to be that kind of dude who's just set upon making something simple way way more complicated by over thinking it.

Crank turns 180° - Intake-Compression transition - KNOCK (or CLANK as you seem to be pedantic about saying)
Crank turns another 539° 59' -no sound - but which general background noise I'm pretty sure will turn up on a slowed down audio recording as 'buzz'..
Crank turns 1' - Exhaust-Intake transition - KNOCK (or CLANK as you seem to be pedantic about saying)
^^^ we have now seen 720° - 2 revs of the crank. NOW IT JUST REPEATS...

I started with 'Intake' for my own pedantic purposes but if you want to start with the 'downtime'... "buzz clank clank, buzz clank clank" (and again if we have to stick with your terminology).

Now I very well could be wrong about whats makin the noise and how its manifesting rod knock, but your crazy weirdo 12-cycle 2160° concept makes no sense... BTW your wav file link no worky so nobody has any idea what that sounds like.

[JimLev]

Wave file worked for me, sounds like an intro to a heavy metal tune.
Get that thing back together and start disconnecting coils.

[R Shaffner]

If you're like me, you'll want to try everything you can think of trying that might help, while you still have things apart.

I found this video as a way to check for rod knock while the engine isn't running. You can try it before you have everything back together. (Though it involves turning the engine by hand, so you'll need the valve train back together, of course.)

https://www.youtube.com/watch?v=pYleKxjl0uY

Just another thought.

[Santaclaus4]

Yes, thanks I appreciated that very much, and intended the additional info specifically for you. My point was, if the rod was shifting, wouldn't the pattern be a 4-cycle instead of a 3-cycle?

Since idle with A/C off is 750 RPM (12 Hz) and the noise base frequency is 12 Hz, there is a single buzz or clank per revolution. Bizarrely, we observe a uniform "buzz clank clank" pattern, 3 noises every 3 revolutions (12 Hz), so this 3-cycle pattern repeats at 4 Hz.

If every two revolutions the rod should clank once into tension and once back into compression, I would expect a 4-cycle pattern such as:
- clank clank silent silent (180 + 540)
- buzz clank silent silent
- buzz clank clank silent
Also, the noise base frequency ought to be 4 noises every 2 revolutions (24 Hz), and this 4-cycle pattern should repeat at 6 Hz.

That's why I wonder what engine function might repeat every 3 revolutions (as opposed to 2).

Holy friggin cow...that made my head hurt.
Wachoo 'talkin 'bout???
GG's explanation is spot on. I was gonna add a little more but I think this horse is dead. Now either pull the bearing caps off and look at them or put the engine together and troubleshoot a running engine. We're waiting...Lol

[geargrinder]

Holy friggin cow...that made my head hurt.
Wachoo 'talkin 'bout???
GG's explanation is spot on. I was gonna add a little more but I think this horse is dead. Now either pull the bearing caps off and look at them or put the engine together and troubleshoot a running engine. We're waiting...Lol

Yeah... I prob have overposted already.

Sub-delete time. Leave it to you guys to carry on.

[JimLev]

You can only get to the front 2 rod caps unless you remove the upper oil pan.

[qcdstick]

This is what rod knock on a M62 sounds like at idle.



Sounds to me you are idling a fair bit higher, or the rod is so loose it's getting a "two fer" bouncing twice per stroke. My rod bearings looked about like what bimmerbreaker's did on the last page.

Hope it's something else.

[curtmcd]

Sigh, no gg, I have not overthought this. My question arose immediately upon hearing the noise. Then I took way too long to explain it because I need help.

You seem to agree with this:
- clank clank silent silent (180 + 540) << this is 720° of revolution, now it just repeats...

Knowing that the audio contains one loud noise per 360 degrees, how can you match that to the engine cycle? My car is performing a waltz.

If you can't hear the .wav file then here it is on video, but the volume is quite a bit lower.


- - - Updated - - -

Now either pull the bearing caps off and look at them or put the engine together and troubleshoot a running engine. We're waiting...Lol

I decided not to open the heads so I won't be able to pull off the caps. At this point I would rather get past it and reclaim my garage space. Either the noise was something in the chains/guides/accessories, or the car is done. Yeah, lolz me. Currently the timing covers are back on.

[curtmcd]

This is what rod knock on a M62 sounds like at idle.

Sounds to me you are idling a fair bit higher, or the rod is so loose it's getting a "two fer" bouncing twice per stroke. My rod bearings looked about like what bimmerbreaker's did on the last page.

Hope it's something else.

Thanks, I had actually found your video already. Analysis of your video shows 450 RPM (7.5 Hz) which is closer to normal but low. To me the sounds are pretty different.

Played at 15% speed, your noise is a 2-cycle of "buzz clank". The buzz is one rev and clank is the next. Unlike mine, that makes sense!

If you're like me, you'll want to try everything you can think of trying that might help, while you still have things apart.

I found this video as a way to check for rod knock while the engine isn't running. You can try it before you have everything back together. (Though it involves turning the engine by hand, so you'll need the valve train back together, of course.)

https://www.youtube.com/watch?v=pYleKxjl0uY

Just another thought.

Thank you, I will try this method!

[R Shaffner]

OK, now I'm probably over-posting, but I ponder these things while driving around in my 540. Plus, many of us are trying to learn here and others might read this down the road.

Curtmcd: From the video link I shared above, I gather the idea is to have the crank pulling the rod and piston away from you, to put the crank-to-rod gap (if there is any) on your side of the bearing. That way, if there is a gap you might notice it when you push on the piston. I imagine the same thing would happen when pushing on the bearing fitting from below. (If the crank were pulling the rod down, you wouldn't be able to push on the rod bearing fitting and notice any play (any gap would be on the opposite side). If the crank were pushing the rod up, then you might notice the play when you push on the bearing fitting.

Again, Jim's suggestion about removing the spark sounds like a winner. There are several videos of this online. Here's one:

https://www.youtube.com/watch?v=JIOdS7XbHys

However, this begs another question about frequency. The loud knocking that we hear when the spark is present suggests the combustion stroke is doing it. (I can't think of another explanation.) That suggests the loud rod knock from just one cylinder would be every-other revolution -- half the engine rpm (as in the M62 video above). So if the frequency of the loud knocking is closer to engine rpm, it might be something else (or two rods knocking, of course, but then you'd get a reducing knocking when removing the spark from 2 cylinders -- seems unlikely).

Personally, I keep coming back to the oil pump. When I look at the diagram on Real OEM, it looks like the sprockets driving the pump are about the same size, which means the pump has about the same rpm as the engine.