Rocker shaft float

[steve_wicks]

My 81 R65 has developed a top end noise on the right pot so I took a valve cover off. Valve clearances seem okay (I don't own a feeler guage!) but I did notice a bit of float on the rocker shaft.

My Haynes manual doesn't say much about rockers apart from mentioning a special tool is needed.

But, is it likely to make a sort of loose tappet noise and why would it come loose or develope play in the rocker shaft?

[vestandpants]

Steve, buy a set of feeler gauges When we are talking in the 'thou then that fraction is a lot. Set them up right and see if it cures it.

When I rebuilt the top end on mine I don't recall the rockers having any float although I will check them next friday when I reset the tappets.

[Andy-Gadget]

Rocker end float is the source of most of the top end noise on a boxer, unless adjusted properly.
If there is up and down play on the shaft, the rocker acts like this.
At idle the rocker has time fall down under gravity, only to have the push rod, which comes up at an angle, push the rocker up to the top of its play, before it then starts to open the valve.
IMHO this is the source of the rattling of tappets heard on most boxers
BMW eventually fixed it on the later air head boxers by using shims to control the end float.

But on the early boxers some "careful" adjustment is required for a nice quiet top end, warning, this fix involved hitting things with hammers  

I use a socket that will fit over the rocker shaft and sit on the pillow block that the head stud goes through.
I push the rocker into the push rod, to take any play out of the rocker in rotation, and then try and move it up and down, if it moves at all I use the socket and a hammer to "tap" the pillow block towards the rocker.
I alternate between the top pillow block and the bottom one.
The goal is no perceptible free play up and down, but free to move the rocker around the shaft.
I stop when I can't feel any play, but can see the oil film moving in and out as the rocker is moved up and down.

I then set the tappet clearance, as moving the rocker up and down greatly affects the actual settings.

BTW Guzzi's don't suffer this problem as the push rod is pushing the rocker down, not up, and Guzzi fit light springs to the rockers to control the movement along the shaft.
And the early boxers didn't suffer as well, for roughly the same reason, a down side to NOT seeing the push rod tubes  

[nhmaf]

You'll want to slightly loosen the nuts on those studs which hold the blocks in place to facilitate nudging them, and putting so much sideloading on those studs (which hold your cylinders & head together).
Afterwards, tighten up to correct torque in an X-pattern and recheck clearance.   I just loosen the nuts about 1/2 a turn or so, to relieve some of  the tension on the stud.   It is generally good practice to just loosen and retorque the heads at valve adjustment time anyhow.   I also recommend using the lower end of the torque spec (25 to 26 lb-feet) - don't go above this or you have some potential for pulling studs out

[steve_wicks]

It all started when I bought a new helmet and thought something had suddenly worked loose. I suppose that's why I thought it had happened over night - I'd just never heard it before because of the old helmet.

It is a relic from my car racing days and is over 10 years old, a good ole tight fitting Arai GP2 complete with nomex fireproof lining etc, but a very small slot to look out of.

As for a feeler guage, well I can't remember when I last used one, but I'll pop out a buy one although I did find a slide rule in a box of junk a year or so ago!!

[Justin B.]

I wish I could eyeball within a thou, I wouldn't have had to waste all that money on a drawer full of mics and calipers!

[steve_wicks]

What I mean't when I said valve clearances seem okay and I don't own a feeler guage was that there was no obvious big clearance or difference between inlet or exhaust, but there was a lot of play on the inlet rocker shaft which I assumed (rightly or wrongly) shouldn't be so.

As I've said before, the Haynes manual is pretty much next to useless on this.

[Landlubber]

Steve, You have the problem in hand, it is end play that makes most of the noise, definately get a set of feeler gauges, I doubt the slide rule would fit in the gap, but if it does, then you really do have serious end play!

Just do as they say above, loosen off the studs slightly and wack the blocks, the oil squashing bit is very important and absolutely essentilal to allow for thermal expansion. It shows there is clearence, but not much.

[steve_wicks]

I'll be doing the job tomorrow, and I have been warned to go very easy on loosening the head studs and then re-tightening them.

But, I would like to know why I've got float on the rocker shaft. Surely something has come loose, although nothing seems loose, but I haven't checked the torque on the studs yet. Should I do that before I loosen?

[Andy-Gadget]

I don't loosen the studs when doing the clearance adjustment, it is all "impact" driven, But I leave it up to you if to loosen or not.

As to where the increased clearance came from I have no idea, Much as I like to treat diseases rather than symptoms I will continue to treat the symptom of top end rattle with the proven fix of removing the excess clearance.

Now that I think about it, it has been my observation over the years as an engineer, that as the clearance on some oscillation system increases, the point is reached when the increasing amount of travel available will make the Joint clearance increase rapidly at some point.
This might be what has happened?

BTW, if left untreated, the bottom needle rollers will "punch" a hole through its end and all the needles will come out through the gap in the pillow block, and collect in the rocker cover, and some will make it to the sump, which means past the followers, but they seem to do this with no passing damage.

I'll be doing the job tomorrow, and I have been warned to go very easy on loosening the head studs and then re-tightening them.

But, I would like to know why I've got float on the rocker shaft. Surely something has come loose, although nothing seems loose, but I haven't checked the torque on the studs yet. Should I do that before I loosen?

[Bob_Roller]

BMW also has shims in varying thickness available, I shimmed mine about 70,000 miles ago.

It has been mentioned that the shims shouldn't be used on the 'old style' rocker arms, but I have had no problems as of yet.

[nhmaf]

I think that the blocks and the holes through which the studs pass have a certain amount of"tolerance" or slop designed in, to accommodate manufacturing variances as well as thermal expansion/contraction.  I hypothesize that under many heating/cooling cycles it is possible for the blocks to "slip" slightly if the stackup of machining tolerances is right.

I like to slightly loosen the nuts on the studs so that if the surface of the block isn't perfectly parallel to the inside surface of the nut, my whacking the block down doesn't effectively "wedge" it in even tighter against the nut, which may have the effect of increasing the tension in the rod and possibly pulling a thread in the case.  Since I am loosening and retorquing the nuts in the head anyways when doing a valve adjustment I don't see this as a wasted effort.  To each his own - I am sure that there are better ways that I don't know about.

[msbuck]

So do you re-torque the heads for EVERY valve adjustment?

[Bob_Roller]

The BMW manual states that the head should be torqued every time you do a valve clearance adjustment.

I don't know if this referring to backing the nuts off to zero torque then complete the 3 step torque sequence, but that is the way I've been doing it since getting the bike.

Some people claim, to just check the final torque, either the lower 25 ft/lb, or the original 29 ft/lb setting.

[nhmaf]

Yes, I do check them each valve adjustment, per the manual.  And, to get an accurate reading, you do have to back off the nuts to at least a significantly lower torque, if not "0" torque, and bring it back up.  That's how I was taught, anyhow.

YMMV