Need advice on chasing down a rattle on my m20.

[gnmzl]

Digger, thank you for the detailed answer.
I understand now the part about the polishing and I will certainly ask Web Cams if the existing cam could be repaired and at what cost.
The thing about PtoV is that I measured it and was about 2.2 mm, which seemed plenty. I can't explain to myself for the life of me what lead to piston to valve contact, except I must have messed up the timing somehow.
Agreed on the differences that can only be made apparent with measurements and it is apparent to me that the next cam that goes into the engine should be carefully chosen.
What I intend to do now, in order to avoid disassembling the whole engine:
I have a lead on another m54b30 crank, zoomer's set of pistons is still around from his m20b30 attempt, currently there is a stripped m20b27 block on my engine stand and a cracked 885 head. Given the circumstances I can measure the P to V again.

If you want, I can measure the cam for you with a dial indicator and a degree wheel. Still haven't decided what to do with it, but since I have a stripped head laying around, it would be easy to measure if you would provide me the steps or a link how to do it.

[digger]

2.2 mm clearance is heaps, the schrick valve springs should be ok to stop float/bounce.

to measure i turn the head upside down, set clearances to zero with new rocker that runs on the fresh part of the lobe (remove the clip) and setup a dial indicator normal to the valve face (you know its normal when the peak lift registered is highest). i have a degree wheel attached to the stock cam gear that i made (printout out one from online the right scale and glued to a thin piece of mdf that i machined in a lathe that centres on the back of the gear ) and of a size so that you can still see the TDC mark on the head. you could measure at each tooth (peak and valley to get a fair idea) but may not give the best resolution of the ramps as its like every 7.5* of equiavalent crank rotation.

[gnmzl]

Got it, I'll start on making the degree wheel and measure the cam I have. I also have access to a catcams 304 and a catcams turbo application (forgot the durations), I could measure these as well and see if catcams have a pattern in making the form of the lobes. That way there could be a better comparison to the dbilas unit and ultimately make a better decision.
Is your cam database publicly available?

[digger]

some info is here. there are a few cams i havent added on yet

https://www.r3vlimited.com/board/for...es-and-rockers

[gnmzl]

So, finally found the time to do some stuff.
First thing was to make sure I didn't mess up the P to V clearance measurement the first time around.
Slapped a crank in a stripped block, put in a conrod and the prototype piston with some clay on top, torqued down a cracked head with valves, slid in my dbilas cam with two brand new IE Rockers, timed it and rotated the crap out of it. Then took the head off again to check P to V clearance.
I did this four times - twice with the cam wheel to zero, twice with 5 degree retardation. Sure enough - Intake 2,2mm clearance with the camwheel on zero, 2.6mm with the cam retarded 5 degrees. Exhaust has the opposite, 2.6mm on zero, 2.2 on 5 degree retarded. That's with 0.25 clearance on the rocker.

This is very confusing to me, what would cause the valve to float? Is it possible it bounces off the valve seat? Below is a small theory on that topic. Everything checks out and I really can't figure out what happened.
Would a loose timing belt cause that and if yes why? That's another thing I can't visualize - facing the engine, the crank rotates clockwise and turns the cam as well, regardless of the tensioner not being tight enough. Only when going on and then off throttle would it matter, cince it can skip a tooth, but that wasn't the case, my timing was spot on. Can anyone explain if P to V contact can be caused by a loose timing belt but not skipped tooth and why, I really can't figure it out.

Another piece of bad news: my second set of IE rockers is obviously not up to spec. I popped two rockers out the package to measure the cam as digger explained, but the part where the eccentrics go is so tight, I could barely make them fit and adjustment was a real pain with the nut and bolt fully loose. And one out of the two has a casting imperfection on it.

Couldn't measure the cam, because each time it reached the point after the peak lift it would just "fall down", several cam teeth off. That's with stock dual stock springs. Is this something that might have caused the valves to bounce off the seats, the "off" ramp on the cam profile being too steep

It seems I can't get a break these days.

Looking at the charts digger made, the Schrick 288 seems very appealing to me and based on my measurements for P to V I should be good if I buy that one. Contemplating on going with OEM Rockers, because I'm tired of this IE hiatus.

[zaq123]

Yes, I checked during assembly, measured the clay while on the piston, it was about 2.2 mm. Only picture I have though



I saw marks for inlet, but the camera wouldn't wiggle enough for me to check the exhausts :(
Only thing I can think of - me being stupid and actually advancing the engine instead of retarding it. That's quite possible and only logical explanation.

you mentioned P to V was measured and "about 2mm". IN? EX? Is there a possibility that step was rushed thru hence the problem? Unless you really ADV it much...there should be ADV/RET room with the save clearance.
In any case this sucks....hopefully you will get it sorted

[digger]

yes cam can bounce off seat many reasons, loose lash can cause this as the valve isnt set down on the seat gently, bad cam profile, incorrect spring loads for the profile and rpm. when you spin the head the resistance is higly variable at different points (firing pulses are regular but not constant as well), this sets of torsional oscillations so the belt flexes back and forth, so a loose belt can cause issues with timing but normlly id expect skipped teeth to get p to v contact.

yes the stored energy in the spring will try and rotate the cam and close the valve automatically (this is why stiffer springs dont consume much hp ave for sliding friction or loss of lift due to deflection is its stored energy) you just need to provide resistance to stop this but its easier to use a light checking spring or just the inner spring.

did you check the cylinder that had the issue or just number 1?

[zaq123]

So, finally found the time to do some stuff.
First thing was to make sure I didn't mess up the P to V clearance measurement the first time around.
Slapped a crank in a stripped block, put in a conrod and the prototype piston with some clay on top, torqued down a cracked head with valves, slid in my dbilas cam with two brand new IE Rockers, timed it and rotated the crap out of it. Then took the head off again to check P to V clearance.
I did this four times - twice with the cam wheel to zero, twice with 5 degree retardation. Sure enough - Intake 2,2mm clearance with the camwheel on zero, 2.6mm with the cam retarded 5 degrees. Exhaust has the opposite, 2.6mm on zero, 2.2 on 5 degree retarded. That's with 0.25 clearance on the rocker.

This is very confusing to me, what would cause the valve to float? Is it possible it bounces off the valve seat? Below is a small theory on that topic. Everything checks out and I really can't figure out what happened.
Would a loose timing belt cause that and if yes why? That's another thing I can't visualize - facing the engine, the crank rotates clockwise and turns the cam as well, regardless of the tensioner not being tight enough. Only when going on and then off throttle would it matter, cince it can skip a tooth, but that wasn't the case, my timing was spot on. Can anyone explain if P to V contact can be caused by a loose timing belt but not skipped tooth and why, I really can't figure it out.

Another piece of bad news: my second set of IE rockers is obviously not up to spec. I popped two rockers out the package to measure the cam as digger explained, but the part where the eccentrics go is so tight, I could barely make them fit and adjustment was a real pain with the nut and bolt fully loose. And one out of the two has a casting imperfection on it.

Couldn't measure the cam, because each time it reached the point after the peak lift it would just "fall down", several cam teeth off. That's with stock dual stock springs. Is this something that might have caused the valves to bounce off the seats, the "off" ramp on the cam profile being too steep

It seems I can't get a break these days.

Looking at the charts digger made, the Schrick 288 seems very appealing to me and based on my measurements for P to V I should be good if I buy that one. Contemplating on going with OEM Rockers, because I'm tired of this IE hiatus.

There is a lot that can be an issue. Custom M54 piston variances (are you measuring with the same damaged piston), Accuracy of the assembly (this time vs the first time). My previous post above was based on the first page of this thread....didn't realized there are 3 pages already. But it does appear that P to V process was skimmed through during the assembly the first time. I would really pay attention to ALL details the second time vs getting crazy investigating what has happened.

I don't think the valve float was your issues.. Was the engine primed before the start up? Did you see oil sprayer squirting oil during the priming process? If wasn't primed, that cam/rocker wear could be the result. IE rockers...post a picture of imperfections...tight eccentric fit is not a big deal ass long as it fits...splitting hairs here ....but tight fit is potentially better that very loose as there are less stress on the rocker when your eccentric is fully tightened...

[gnmzl]

It is possible that I rushed this step, I only said about 2.2 mm because that was my recollection. Talking about IN side, because that's where the contact happened.
However, in my previous post the statement is made after rechecking this just yesterday eve, granted in a spare block and with a spare head and the prototype piston created for this project (actual pistons are 0.25mm lower), but the head and block are shaved more than the ones in my car, a total of 0.5mm more. So, after this latest check:

Intake 2,2mm clearance with the camwheel on zero, 2.6mm with the cam retarded 5 degrees. Exhaust has the opposite, 2.6mm on zero, 2.2 on 5 degree retarded.

This seems plenty to me, that's why I'm asking for a reality check by you guys.

Granted, I'm a noob in cam measuring and couldn't figure it out for the life of me, but since I rotated the mockup engine with this very cam in it and the P to V checks out, I am clueless.

There's something else I just found out and didn't know about oiling on the m20 head. The drip/spray bar depends on getting oil from the grooves on the first and last cam journals. Here's a similar issue like mine that a fellow german e30 enthusiast had:

https://e30-talk.com/topic/108333-klackern-nach-motor%C3%BCberholung-m20b27-video/

Post number 6 contains the important pictures for the ones that don't speak German.

So, another thing to check for me is if these line up with the tube mounting points. That would add some insight on the destroyed rockers and cam lobes, if these grooves wouldn't line up.

[gnmzl]

yes cam can bounce off seat many reasons, loose lash can cause this as the valve isnt set down on the seat gently, bad cam profile, incorrect spring loads for the profile and rpm. when you spin the head the resistance is higly variable at different points (firing pulses are regular but not constant as well), this sets of torsional oscillations so the belt flexes back and forth, so a loose belt can cause issues with timing but normlly id expect skipped teeth to get p to v contact.

yes the stored energy in the spring will try and rotate the cam and close the valve automatically (this is why stiffer springs dont consume much hp ave for sliding friction or loss of lift due to deflection is its stored energy) you just need to provide resistance to stop this but its easier to use a light checking spring or just the inner spring.

did you check the cylinder that had the issue or just number 1?

Ok, that's a valid point, I checked number 1 and 2, but not number 3 (the problematic one).
I'll do that next time.
Bad cam profile was the suspect, hence me trying to measure it with your method, to see how it compares to the other ones in your graph.

There is a lot that can be an issue. Custom M54 piston variances (are you measuring with the same damaged piston), Accuracy of the assembly (this time vs the first time). My previous post above was based on the first page of this thread....didn't realized there are 3 pages already. But it does appear that P to V process was skimmed through during the assembly the first time. I would really pay attention to ALL details the second time vs getting crazy investigating what has happened.

I don't think the valve float was your issues.. Was the engine primed before the start up? Did you see oil sprayer squirting oil during the priming process? If wasn't primed, that cam/rocker wear could be the result. IE rockers...post a picture of imperfections...tight eccentric fit is not a big deal ass long as it fits...splitting hairs here ....but tight fit is potentially better that very loose as there are less stress on the rocker when your eccentric is fully tightened...

The damaged pistons are still in the engine, I was using the initial piston that was created to see if it would work. They are basically identical, only the ones in the engine are 0.25 shorter (so safer in regards of P to V)
Engine was primed before initial startup BUT the 20 minute cam breakin process was not done, due to various reasons.
I will take a picture of the imperfection, but the fitment of the eccentric is so tight, that it's not possible to adjust it when the rocker is installed.

[zaq123]

It is possible that I rushed this step, I only said about 2.2 mm because that was my recollection. Talking about IN side, because that's where the contact happened.
However, in my previous post the statement is made after rechecking this just yesterday eve, granted in a spare block and with a spare head and the prototype piston created for this project (actual pistons are 0.25mm lower), but the head and block are shaved more than the ones in my car, a total of 0.5mm more. So, after this latest check:

Intake 2,2mm clearance with the camwheel on zero, 2.6mm with the cam retarded 5 degrees. Exhaust has the opposite, 2.6mm on zero, 2.2 on 5 degree retarded.

This seems plenty to me, that's why I'm asking for a reality check by you guys.

Granted, I'm a noob in cam measuring and couldn't figure it out for the life of me, but since I rotated the mockup engine with this very cam in it and the P to V checks out, I am clueless.

There's something else I just found out and didn't know about oiling on the m20 head. The drip/spray bar depends on getting oil from the grooves on the first and last cam journals. Here's a similar issue like mine that a fellow german e30 enthusiast had:

https://e30-talk.com/topic/108333-kl...-m20b27-video/

Post number 6 contains the important pictures for the ones that don't speak German.

So, another thing to check for me is if these line up with the tube mounting points. That would add some insight on the destroyed rockers and cam lobes, if these grooves wouldn't line up.

see my previous post...assemble and HAND PRIME the engine until you see oil coming out where it supposed to ..... before you put the key into the ignition....Electric drill does very nice job simulating the oil pump.

I think the variances in the cam grove is a potential issue worth looking into but most likely not. Hard to tell from the pic in that German forum...but it looks like the last the oiling grove is at about the same distance from the front of the cam.

[digger]

Ok, that's a valid point, I checked number 1 and 2, but not number 3 (the problematic one).
I'll do that next time.
Bad cam profile was the suspect, hence me trying to measure it with your method, to see how it compares to the other ones in your graph.



The damaged pistons are still in the engine, I was using the initial piston that was created to see if it would work. They are basically identical, only the ones in the engine are 0.25 shorter (so safer in regards of P to V)
Engine was primed before initial startup BUT the 20 minute cam breakin process was not done, due to various reasons.
I will take a picture of the imperfection, but the fitment of the eccentric is so tight, that it's not possible to adjust it when the rocker is installed.

you wont get the accuracy plotting it out to pickup issues with the grind that might cause dynamic issues without proper cam dr equipment. the measurement is about knowing whats installed and then working out if its ideal for your engine or if there is a better candidate

the IE ones i have were fairly tight to, doesn't make the adjustment process very fun.

[gnmzl]

you wont get the accuracy plotting it out to pickup issues with the grind that might cause dynamic issues without proper cam dr equipment. the measurement is about knowing whats installed and then working out if its ideal for your engine or if there is a better candidate

the IE ones i have were fairly tight to, doesn't make the adjustment process very fun.

Thanks, this will save me some time trying to plot out the cam with my medieval tools. I'll try and make a video of how tight they actually are. And also take a pic of the casting imperfection.

see my previous post...assemble and HAND PRIME the engine until you see oil coming out where it supposed to ..... before you put the key into the ignition....Electric drill does very nice job simulating the oil pump.

I think the variances in the cam grove is a potential issue worth looking into but most likely not. Hard to tell from the pic in that German forum...but it looks like the last the oiling grove is at about the same distance from the front of the cam.

For sure, it will be primed before startup, as it previously was too. There weren't any oiling issues at the time oil squirted away from the expected spots. You are right, it is hard to tell, the guy says it's only slightly off, by 2.5 mm which is hard to see in pics, especially of a round object. Even so, that's about half the diameter of the hole for the oil tube, so it's not like there wouldn't be any oil.


I found a post by digger about Z58 ENEM rockers and looked them up. Unfortunately was not able to find any decent feedback on those, and as they cost more than the Schrick I'm reluctant on ordering. Anyone got some decent info on these?

[digger]

Thanks, this will save me some time trying to plot out the cam with my medieval tools. I'll try and make a video of how tight they actually are. And also take a pic of the casting imperfection.



For sure, it will be primed before startup, as it previously was too. There weren't any oiling issues at the time oil squirted away from the expected spots. You are right, it is hard to tell, the guy says it's only slightly off, by 2.5 mm which is hard to see in pics, especially of a round object. Even so, that's about half the diameter of the hole for the oil tube, so it's not like there wouldn't be any oil.


I found a post by digger about Z58 ENEM rockers and looked them up. Unfortunately was not able to find any decent feedback on those, and as they cost more than the Schrick I'm reluctant on ordering. Anyone got some decent info on these?

Z58 is an ENEM camshaft, the rockers are KM brand i'm not a fan the pad is a totally different profile and the valve lift suffers





catcams



ive had good luck the MM peened and inspected rockers, OE+ quality

[gnmzl]

I see, thanks for the info.
Well, currently I'm leaning towards good OEM rockers and a Schrick 288. Max lift may not be what it could with the IE rockers, but it's a good and safe enough option, what would you say?

Also, can anyone tell me if I understand the solder wire method for P to V contact checks: take a thick solder wire (above 3mm) and stick it through the spark plug hole, find a way to keep it steady. Rotate crank at least two revolutions, take out solder wire and check lowest point.
That's what I want to do next, check the P to V on the existing assembled engine, without taking it apart (yet).