AAZ: Rapping noise on deceleration

[Gromit]

Finally got this AAZ into my Kubvan. I was feeling good...

So I started the engine, let it idle a bit; oil gauge read good, faint tap noise that I suppose was an injector, generally sounded normal. I raised the engine speed a little, then let it drop:

As the engine decelerates it makes a loud rapping noise. Each time. Don't want to keep doing it. like a hammer on metal, (shudder).

What does such a noise on deceleration indicate?

[82vdub]

I'd start by checking the cam to crank timing position, and the tightness of the tensioner (and proper procedure according to the Bentley manual). There may be enough slack in the system and with just so slight being off that when the engine decelerates, the valves kiss the pistons.

[Gromit]

That's what I think it is.

I had to put an A1 type motor mount on, so of course I put a new t-belt and set the timing of cam and pump.

Before dis-assembly, fortunately, I checked the timing of both. At first I couldn't find the mark on the fly wheel, so I lined up the slot on the cam- still no mark. Eventually I found a thin white paint mark. With the FW on that mark, the cam was quite a bit off, but in the direction it would go if the t-belt had stretched. Pump timing at that mark was 74.5, not far from spec. So I went with that white line when I timed it.

Turning by hand I felt no interference, at idle it seemed fine. But the change in t-belt slack on deceleration could cause interference if the time isn't exact, eh?

I've found other things on this motor to make me doubt whoever has been working on it. .So I'm going to try to find TDC by removing an injector...

Anyone think that the valve interference yesterday could have damaged the valves?

[82vdub]

Here's a link to posts that have various TDC marks on the flywheel: viewtopic.php?t=7934

Also, I've identified a method to locate TDC pretty accurately and easily all external to the engine. There is the "dot" on the cogged lower crank pully that matches up with a dimple in the belt pully and then the 4 bolt holes that hold the pully to the cogged crank pully. I found that the 4 bolts that hold the accessory pully on the cogged pully appeared to be machined at 0, 90, 180, and 270 degrees to TDC with the "dot being about 120 degrees counterclockwise. You can line up two of the bolt holes with the dimple in about the 120 degree position and if you line up the two bolt holes parallel to the centerline of the cylinder, you'll be really close to TDC. I've used this several times as I work on my car by myself and it's a pain to try to find TDC while by yourself. By doing this, I can line up the bolt holes and then get out and look in the inspection hole for the TDC flywheel mark. It's typically only off true TDC by 1/4-1/2" or so simply by eyeing it up.

You can see the bolt holes and dimple I'm talking about in this pic. It's rotate a little too far (about 45 degrees too far) for what I'm describing above. I was changing the crank seal for this pic, not trying to line up TDC.

[Gromit]

Initial test suggests that the mark I timed this to is correct, and the cam looks about right.

For expediency, I clamped my dial indicator to a framework clamped to the vehicle's body, so of course the whole motor moves on it's mounts as I turn the crank, and that movement shows up on the dial. So I gotta set up something better now.

I'm trying to hurry. I'm working under old dead trees, storms are coming this afternoon, and a hurricane on Sunday. I'm trying to get this out from under the trees

Any thing else this could be?

[82vdub]

Did you check the cam position to the crank position? Not just to a dial indicator, which is checking the timing of the injector pump to the rest of the components. The timing of the injector pump wouldn't have anything to do with valves kissing pistons. Set the engine back to TDC and see where the slot in the end of the cam is positioned and if it's horizontal or tilted advanced or retarded.

[Gromit]

Problem is fixed. (Thanks 82 for your help!)

I set up a more accurate TDC detector, I made a wooden bracket fixed to the valve cover to hold the dial indicator. The TDC I got was not quite 1/16" ahead of the white line I was using as TDC. Soo... I reset the cam timing by that, what, .5° or so?

That did it. Wow, not any latitude there when timing the cam.

I was hurrying to get this done, darkness was coming, I'm working outside...
Hey! I discovered a really fast way of getting the main sprocket bolt off! Maybe some of you have discovered this one already. I didn't want my sprocket bolt off, though. . Now I gotta add that to the shopping list.

[Gromit]

But I would like your comments:

Could those 6-8 seconds of valve interference have done some damage that I should worry about?

Are the tops of my valves, miles down the road, going to pop off and fly through my pistons? What should I do at this point?

[82vdub]

I wouldn't worry about the valve to piston interferrence at this time.

[Gromit]

Well, I drove about 1000 feet, then heard it again, just for a moment.

The seller of this motor said this never happened, and I didn't hear it in the short time I heard it run (could not test drive). As I mentioned before, before I took the timing apart to change the mount, I checked the timing. The cam timing was off of the flywheel TDC mark (the reason I doubted that mark). I took a photo of the cam with fw @TDC, it lagged about 4°. That's 4° of cam, I guess that would be about 8° on the flywheel.

So if there was no problem with the cam at 4° behind, and there is when it is 0°...?

So I moved cam timing off, to maybe 1°. Now, when the cam is at 0°, I'm about 3/16" from the fw TDC mark. Timing work on a kubvan is no fun, the whole belt system is up against a structural wall.

I drove it maybe 1/2mile. I heard the rapping for a moment, but other than that moment it seemed fine.

My cam is retarded

So: do I retard it just a bit more, and test drive again? I need this running, I'm stranded here!

My gasket is a 2-notch. Probably I should remove the head, inspect, and put a thicker 3-notch gasket on. I'm at the frontier of my mechanical experience here.

Are there other causes of valve interference? Carbon build up caking the piston top, or on the valve preventing seating? Valve not seating due to damage? Weird belt behavior? This motor has the TDI-style spring-loaded tensioner. Could crank/rod/piston-pin wear do this? Wrong cam? -It has the wrong flywheel I think -027 105 273 D goes on a 1.8 gasser I think.

(Anyone know a good mechanic for this engine in Maine?)

[Gromit]

As I mentioned, I checked the timing of this engine before removing timing components for a motor mount install. Here is a photo of the cam timing slot: the engine is at TDC here.

I printed this and checked it with a protractor: it seems to be about 4° retarded, which would be 8° BTDC (correct?)

Is it even possible that an engine running with the cam timing like this would not have damaged exhaust valves?

Having now read a lot from this website I'm now of the opinion that my intermittent valve clash is from one or more valves that are sticking open because they are damaged from having been previously so out of time.

I will take the head off next, after giving the seller of this "good running" engine an opportunity to take it back now. Or can a cam be this far off time without damage?

The question of how far off timing can be before damage is done is probably of general interest so I'll put this on a new topic...

[82vdub]

The slot in the cam should be parallel with the head when timed properly. It does look retarded slightly. What about shimming the rear of the cam lock tool to make the cam 2-3 degrees advanced and see what that does. I mean force the cam timing to the advanced side a bit. Not sure if there'll be any issues with this, but just a thought. What it sounds like is that there's enough slack in the timing system where the valves are kissing the pistons, when on deceleration, but not on acceleration. That's my theory on advancing the cam slightly.

I think if there was damage enough to have valves that are sticking open, you would have the noise constantly. There is no room for a valve that's open in the chamber. And, if turning the engine by hand, you should feel the pistons hitting the valves.

What else do you know about the motor? You may have a situation where your rod bearings are a little worn, and moving enough when the piston changes directions where the valve kisses the piston. Could this noise be something else then a valve hitting a piston? Worn wrist pin? It is also quite possible that a previous owner installed a head gasket that's too thin and the valves are hitting the pistons. Something that playing with timing may never solve. Or, it could be solvable but take forever finding that exact spot where everything fits just fine.

[Quantum-man]

As I mentioned, I checked the timing of this engine before removing timing components for a motor mount install. Here is a photo of the cam timing slot: the engine is at TDC here.

I printed this and checked it with a protractor: it seems to be about 4° retarded, which would be 8° BTDC (correct?)

Is it even possible that an engine running with the cam timing like this would not have damaged exhaust valves?

Having now read a lot from this website I'm now of the opinion that my intermittent valve clash is from one or more valves that are sticking open because they are damaged from having been previously so out of time.

I will take the head off next, after giving the seller of this "good running" engine an opportunity to take it back now. Or can a cam be this far off time without damage?

The question of how far off timing can be before damage is done is probably of general interest so I'll put this on a new topic...

Hi,
Interesting topic.
I think engines can survive/operate near that point, dependant upon valve head position inside it's countersunk resting place.

Can I first say, wrt timing, always base your setting up by having a known reference point on the flywheel for TDC that you can work from. In the event of corrosion or strange flywheels, one method is to use a dropped valve, and bisect the two marks created when moving crank clockwise, and then anticlockwise until valve felt.
To drop a valve I use a flat screwdriver, slightly bent at the start of the flat part, and push it inbetween the cam and the follower on #1 valve, when engine is operating at #4 TDC.
In otherwords you will be simulating a new cam lobe on the back of the zero lift area of #1 cam lobe. I pick the exhaust one for convenience. Bisect the result to give a true TDC allowing for any slack in the bearings.

On to your projectL
The cams are listed as being 1mm lift @ +/- 5 degrees cam for the exhaust and inlet respectively.
The amount of play at TDC isn't the issue, because when the camlobes come into operation, they open or close at a faster rate than the piston does. This means that the clash is invariably off TDC.
An experiment you can do if you have not yet dissembled, is to set flywheel to #1 TDC, then lever down with the screwdriver tool on the cam follower measure valve gap .with feeler gauges. Knowing that on the flywheel, each flywheel tooth is 2.73 degrees of rotation, you can make a series of measurements.

Do it for about 4 teeth either side of TDC and you will see a variation of available gap....

[Gromit]

Hi 82vdub

Yea, that "forever finding that exact spot" was what I was doing. As for other causes of the noise, yes, I don't know the sound of valves vs wrist pin. All options sound bad, the noise sounds bad.

The noise was on deceleration when I first started it. But running the motor in the process of trying to find the timing position that eliminates this noise, it really seems to be set off randomly. Thats why I thought maybe a valve jams in the guide, then gets rammed shut. But I'm just a glow-plug .

I'm coming to this thinking because It seemes like the as-found cam timing is retarded enough that valves probably were clashing. Maybe as the belt gradually stretched the valves gradually pounded recesses into the pistons. Only opening it up will tell.

Hello Quantum-man, your info about the cam lift is interesting, thanks. I suppose it suggests that if my 4° measurement is right, maybe no damage has been done by this. Re known flywheel reference mark, I did doubt the mark. This seems to have the wrong flywheel (it is part #027 105 273 D, seems to be a gas 1.8?). I checked it with a dial indicator through an injector hole, and the paint mark is very close to TDC if not dead-on. Never heard of the dropped valve method, clever one.

And measuring valve movement is a good idea. I'm used to adjusting valve lash on my previous 1.6 motors. I suppose it is similar for the hydraulic. So there should be some tiny amount of play at TDC if the valve isn't actually touching the piston.

I'm grateful for the help, thanks.

[Fatmobile]

It's easier to see how much it's off by if you lock the cam and take a picture of the timing mark on the flywheel.