Need advice on chasing down a rattle on my m20.

[gnmzl]

Managed to reach our workshop yesterday and snapped a few pics as well as a short video.

Here's the imperfection I am concerned about. That's one of my brand new IE rockers, never used set.




Here's a short video of how tight they are compared to the other set I used up until now in the engine.
Note, that the eccentrics were bought brand new from a BMW dealer, I tried this with a used set too and the result was the same.
In the video, the rocker I take out the eccentric of is one of the ones that got damaged with the cam and I unpack one from the brand new set, just to prove a point it hasn't been messed with.

https://vimeo.com/389922493

As you can see, that's VERY tight, I can barely fit the eccentric in it.


Went ahead and took a pic of the oil pulse groove at the back of the dbilas cam, same as what the german guy did. Sure enough, it was off. As commented, hard to see in the picture, but I measured it to be about 1.5mm off. When installed and looking through the hole for the tube, it's visibly misaligned, but I doubt this could lead to oil starvation, about 4/5 of the hole is not blocked.

About the rockers - even if I shoot out an email to IE, I doubt that they would refund or do anything about it. These were purchased back in 2015 by a friend's brother in the States, then shipped over here to another friend, who held on to them until he decided he wasn't building his m20 and sold them off to me in the original unopened package. It was different from the one I received my other ones in, but both sets were bought directly from IE. My original set came in a big plastic bag, which was then hot pressed over each individual rocker, you can see it in the only picture I have from back then



These rockers were all packaged separately in plastic bags which in turn were all thrown together in a big plastic bag.

Any thought, has anyone had a similar experience? What kind of package did your good IE rockers come in?

[digger]

the location of the casting flaw is not in a critical area. try deburring the edges of the slot with a thin fine file or deburring tool

[gnmzl]

I thought so too, but I'll still locate this rocker on one of the easiest to replace locations.
Anyway, I was busy trying to figure out stuff and I have some conclusions. Either one or both of the below can be true

1. The pistons were not machined 0.25 more as I thought they are
2. My head and block were skimmed a bit more than I asked

What leads me to believe that - I did the P to V test with the solder wire and here is what I found.

First off, I made several identical pieces of solder wire, using one of the cracked heads laying around.




Then I cut up a ruined old rocker shaft and used the good part of it to do this



Stuck the solder wires in the spark plug hole and here are the results

Cylinder 3, cam wheel retarded 4 degrees



Cylinder 1, again 4 degrees retarded



Now Cylinder 1 with zero degrees on the cam wheel



This is very far off from the measurements I remember and what I previously got from the "dummy" engine. The "dummy" engine block is skimmed 0.05 and the head is almost like new, the four marks are all perfectly visible. But that still should keep me in the safe, 1.5 is the minimum clearance required AFAIK.

So, it is what it is. I assume the pistons kissed the valves during the period of time when the cam belt was loose and there was some minimal play on the crank sprocket due to the jesus bolt collar not being shaved down enough. Both these thing are not corrected.

Now I need to purchase a new cam, if I can run the IE rockers that would be great, otherwise I need to get a set of OEM ones.
For the sake of it, I took pictures of the oil grove on the last cam journal.

This is my dbilas 292



This is a catcams 304 I have access to.



As you can see, the dbilas is set off towards the front of the car and the catcams is set off in the other direction, both not exactly on spot.

Now I need to talk to Schrick to see if they corrected that issue on their camshafts.

[gnmzl]

Forgot to mention, when doing the solder wire check, rocker eccentric to cam lobe clearance was set to zero.

[gnmzl]

Thought I should update this, for future reference to anybody facing similar problems.

So in summary: I sent and exchanged a bunch of emails with dBilas, Schrick and Gladen Motorsport (Schrick official retailer). Schrick answered within a week, stating and assuring me the oil passage placement was corrected and no longer an issue. So far so great - sent an email go Gladen Motorsport to ask them if there was any chance for me to get an old stock cam with the issue. No answer up to this day. Emailed dBilas to notify them about the issue. What they did was reply very quickly and remained engaged throughout but they blamed the damage on the cam and rockers on the rockers themselves and offered me to fix the placement of the oil passage and all the lobes for the total price of (shipping included) ..... 240 euro. New, the cam costed me 350 plus 10 euro shipping. Didn't make any sense to me, so risked it and ordered a Schrick 288 from Gladen MS, which arrived last week.
My car had a birthday on March 8th and turned 35 years old. So it was the most appropriate day of all to make it run again.

Started by unboxing the Schrick 288 (very nicely packaed)



Immediately stuffed it in my cracked 885 head



All is well with the oil passage



From here on, I didn't take pictures, since I wanted to get it done. But basically the order of events:
Put the Schrick 288 in the engine, slap on the mockup rocker with the cut shaft and set clearance to zero - do the solder iron test again. Results were 1.4 mm with the cam wheel set to zero and a smidge over 1.9 with 4 degrees retardation. Left it with 4 degrees retardation and started working on making the eccentrics fit in the rockers. Took a deburring tool as per digger's advice and a 400 grit sandpaper - all went smooth and not much was actually needed to make em fit. Installed them and verified timing 3 times. Next on the list was priming the oil pump really good and I made sure oil was coming out of all orifices AND I rotated the engine to see if it would come out the tube as well. It did, so it was time to quickly put everything together, prime the oil pump once more and slap the cover on it, followed by the air box.



It cranked a bit longer than expected but in the end it started and roared into life, did the "break-in procedure" for new camshafts. Was a bit skeptical about it and peeked under the cover afterwards to make sure all went well - it did.
Apart from the casing of the MAP sensor melting due to prolonged heat exposure all went well and by the sound of it, I think the engine will be happier with this camshaft.

Once I figure out a way to route a vacuum line without drilling the firewall so I can use the Ecumaster's built in MAP sensor, it will be off to the dyno for retuning - we'll see the actual difference between the dbilas 292 and the Schrick 288.

Thank you all for your input, it was highly appreciated.

[digger]

wow, that must be the first time someone has ever packaged a m20 cam half decently.....

a vacuum line can be fed through the firewall right at the side of the car near the battery tray (early models)

did you do power run on the dyno with the 292 previously ? on paper they should be very similar subject to how you dialed them in degree wise. you might find the cam will like it better straight up. based on your measurements it looks like there is enough clearance

[gnmzl]

Yeah, I was impressed with the packaging as well, that's why thought I'd share. Apparently Gladen Motorsport are not ones to reply to emails, but they package their stuff correctly.
About the vacuum line - I was thinking about going that route too, either making a hole in the power cable grommet or punching out the AC line block off plate. Just got the central locking vacuum line off a Volkswagen and will try it in the evening, see how it goes.

Yes, the dyno sheet posted previously is with the 292 dbilas cam, set to zero degrees. No power run with it on 4 degrees retarded, since I only drove it like that for about 100 km and parked it.
Clearances seem to be good, yes and this time I made damn sure the cam belt is tight on there. Basically checked everything at least twice before turning the key.

Tomorrow I need to make the insurance, get through tech inspection and then make an appointment with the tuner. Very interested in the differences on the power curve.

[dvallis]

Was just reading this thread and almost blew coffee out my nose when I got to post 7

"Plot twist: Turns out I didn't retard the engine, but myself."

Too freaking funny.

[GabiJohnson]

I really don't want to reassemble the engine without finding the cause of that.

Another thing that comes to mind - the woodruf key on the crank sproket potentially could have failed, causing the timing mark to align, but the crank would be on the incorrect position.

[gnmzl]

Was just reading this thread and almost blew coffee out my nose when I got to post 7

"Plot twist: Turns out I didn't retard the engine, but myself."

Too freaking funny.

Happy to make one smile with my stories of being careless :D

I really don't want to reassemble the engine without finding the cause of that.

Another thing that comes to mind - the woodruf key on the crank sproket potentially could have failed, causing the timing mark to align, but the crank would be on the incorrect position.

That happened too yes, it has now been corrected with a new woodruff key, new sprocket and the jeesus bolt shim has been slimmed down more to the point the "collar" is visible only if you knew it was there.. I think I mentioned it earlier in the thread. Maybe not, too lazy to reread.
Essentially while chasing down this rattle I found many issues with the engine and I'm kind of glad I was on this hunt for so long and corrected them before something worse happened.


Anyway, today was dyno day. We used a different dyno, but it was callibrated exactly as the one we used before, verified by strapping on different cars that were initially tuned on the original dyno, then showed same results on this one.
Here's what we were interested in:

Schrick 288 with air filter.



Dbilas 292 WITHOUT the air filter. (before failure)




When I removed the air filter this time, with the Schrick 288, whp bumped up to 213, torque remained almost the same. No pic though, si it didn't happen, right?

Schrick 288 has more torque at revlimiter, torque comes just a tad later than with the dbilas 292, but we need to keep in mind the cam wheel is now 4 degrees retarded.

All in all, both cams appear to be very similar, however on the road the Schrick 288 makes all the difference. No jolting while cruising with light trottle around 1500 rpm, power is delivered smoother and I can't explain why, but that dip at 2900 rpm is not that noticable while driving/accelerating. It's still there but not that noticable and not annoying. Car feels somewhat quicker than with the 292 dbilas. Idle is also better.
One thing to note is that now the vacuum is significantly less at idle, maybe the lobe separation is not exactly as advertised on the dbilas 292 (it's bigger in reality) and hence the difference.

So, there you have it - don't buy dbilas, buy Schrick. You'll have a better driving feel and you won't have oil starvation.

Pic on dyno

[digger]

those two curves 288 vs 292 look uncannily the same with the peaks and troughs etc. on paper the cams are almost the same and it seems like they are in reality to...
Since you're in Bulgaria those would be flywheel numbers converted from the wheels right as that seems to be universal in Europe?

[gnmzl]

Hello digger, those are actual numbers on the wheels. Only a few dynos over here provide this service. It was imperative to know what the car actually puts down on the pavement.

[ForcedFirebird]

Yeah, those are definitely WHP numbers, fairly average for the type of build - TQ looks amazing, tho!

If you looks a little closer, the dip actually corresponds to the rich dip you have in the AFR graph. Typically the O2 reading is about 300rpm lagging behind the power RPM since the wide band is generally plugged in the tailpipe, not the headers. I wouldn't be surprised if the power would smooth a little if you flattened that AFR. The m20 (885) head was designed to run on the lean side - they are most happy in mid-13's. All of those pulls show in the 11's, about 10% too much fuel (unless you were adding fuel for knock).

[digger]

the torque is in Nm

[ForcedFirebird]

the torque is in Nm

Ohhh the joys of converting only HALF of the SAE measurements. grrr. Curious why?

Anyone in the 21st century should be grabbing their .winpep or .wp8 files from the dyno operator and they can be manipulated for a graph in any correction, smoothing factor etc...