This is the type of stuff that we all pay good $$ to read about on KR-net! Thanks! This was well worth my hefty annual subscription fee! -- Ross More of this, and less of the other stuff please! ----- Original Message ----- From: veedu...@aol.com List-Post: krnet@list.krnet.org Date: Wed, 22 Oct 2003 20:53:35 EDT To: kr...@mylist.net Subject: Re: KR>VW Engine Rockers
> In a message dated 10/22/03 3:12:57 PM Pacific Daylight Time, > wa...@hispeedwireless.com writes: > > > Hey all this is why I used to rebuild my VW in my sandrail every year and > > now since being converted to the R.S Hoover way of Do it right and maintain > > it. I have not needed an overhaul since 97. > > > > ------------------------------------------------------------ > > Now my hat don't fit :-) > > ------------------------------------ > > Actually, it isn't too difficult to get the VW valve train geometry dead-on, > it just takes a bit of time. I'm told Santa is bringing me a digital camera > this Christmas so I'll be able to post some pictures of the tools I've made > for > setting up valve train geometry. 'VW Trends' published a rather lengthy > article titled 'Dialing in Your Cam' which describes the first stage of the > process. Once you've got the cam dialed in and have measured the actual lift > at the > cam the remaining pieces of the puzzle fall into place. > > (If you're into engines there's a lot more poop about PORSCHE valve train > geometry than you'll ever find for the VW. You CAN set-up a VW properly... > but > the factory never did. Close enough was good enough.) > > The whole idea behind setting up the valve train geometry is to find the > point of minimum loss in the process of turning an upward push, delivered at > an > angle of about three degrees, into a downward shove, delivered at an angle of > about nine and a half. The two basic principles involved are the fact that > the > push-rod and valve can only move in a linear fashion (or nearly so) while the > rocker arm can only move through an arc, with the input arc typically having > a > shorter radius than the output (that is, the rocker ratio is 1.1 to 1, 1.25 > to 1 and so forth. Stock is 1.1, after-market rockers go up to 1.45 and just > to make things interesting, sometimes the rockers actually come pretty close > to > those numbers :-) (Nowadays, you gotta check EVERYTHING.) > > The tricky bit is HOW you go about matching the mid-point of the linear push > (or shove) to the tangental mid-point of the arc. > > Think about it for a minute. Push the DOOR open with a broom stick... Broom > stick travels pretty much in a straight line but the door has to swing thru > an arc. No matter WHAT YOU DO the path traveled by the broomstick will > always > be LESS than the arc traveled by the door. That is, the broom stick travels > on the chord of the arc whereas the door follows the radius of the arc. > > Bottom line, there's always some loss. Your job is to arrange things so as > to give the LEAST loss, which also translates as the MOST motion at the > valve. > So you find out how much the cam pushes the follower and the follower pushes > the push-rod and the rocker swings through an arc... and you raise the rocker > up or lower it down so that the mid-point of its arc perfectly matches the > mid-point of the linear travel of the valve and you adjust the length of your > valve stems (if needed) and the length of your push-rod (ALWAYS needed) and > then > you run the engine on the test stand for five or six hours and discover > everything has changed anyway... :-) > > Actually, you crank in a bit of Texas Windage when you do the set-up, so that > as the engine wears in and you come back and re-torque the heads and tweak > the valves, the anticipated amount of wear will have caused the tangental > point > to be within a gnat's ass of dead on, or so close that it's well within the > range of your adjusting screw. (Gnat's ass is a technical term meaning > five-tenthousandths of an inch if the wind is from a south and its a Tuesday. > Work > with me here; you'll pick it up as we go along.) > > I suppose that's really the tricky part, knowing the amount you have to stay > on the high side of spec so that as the engine wears in it runs sweet instead > of sour. A lot of that part of valve train geometry won't make a lot of > sense > because nowadays, with "Volkswagen" parts coming from all over the world, > from factories Volkswagen never heard of and without any quality control to > speak > of, the initial wear-in coefficient is based on guesses and good intentions. > > But ANY attention you pay to your valve train geometry will usually be > rewarded with SOME improvement in your volumetric efficiency. On the other > hand, > I've seen lots of engines that gave away as much as 25% of their lift at the > valve from being thrown together right out of the box. > > The engine will tell you. Trouble is, most folks don't understand what it > sez. > > --------------------------------------------------------------- > > Didja see how Mike did the set-up on his Corvair? That photo of the > roller-rocker just kissing the valve stem? For VW's I made a jig so I can > set a valve > at its computed half-height. Then I paint the stem with Dykem and have a > scriber on the rocker arm so that when it's set at the proper height it draws > a > line across across the face of the valve stem. An optical comparitor shows > me > how close I've come to the center of the valve stem for that particular > setting. > > That probably sounds sorta gim-crack but I haven't found any better way of > doing it. Actually, it's not too bad of a method for a guys like me with > greasy > fingernails instead of engineering degrees because I'm not actually MEASURING > anything in quantified terms, I'm simply COMPARING things, peering through a > glass to see if the line is too high, too low or just right. > > Raising the pivot point of the rocker will cause the line to fall a little > lower, lowering the tower will raise it. Once you get it dead on you go back > and re-run the numbers to see if the input is close enough to tangental. If > not, you can use a different valve or a different tower. The height of the > tower > and the length of the valve stem don't really matter (within reason) so long > as the mid-points of the linear travel exactly match the tangental points on > the arcs. And keep in mind, you've got four rockers on the same shaft; > you're > seeking HARMONY rather than an identical numerical value. > > Then it hums. 'Free' horsepower. More power for the same fuel. Wears at a > slower rate. Runs a lot smoother becuase the valve train stresses are more > evenly distributed... or something. It also causes you to grow long, blond > Surfer God hair allll over your bod-dee. > > Actually, I was just kidding about that last part. (As he tosses his long, > blond, Surfer God hair back from his noble sunburnt brow. Being a native > Californian isn't as easy as it looks.) > > Engineering types get off on what happens when you DON'T pay attention to the > valve train geometry, as in calculating how fast the engine will trash > itself, or how low the efficiency can fall before EPA will sue your ass back > into > the Stone Age. The Stone Age is where you'll find all those VW 'experts' who > never mention valves at all, telling the Yuppies, "This engine will RUST OUT > before it WEARS OUT." > > Uh, right, John. > > -R.S.Hoover > > _______________________________________________ > see KRnet list details at http://www.krnet.org/instructions.html -- ___________________________________________________ OperaMail free e-mail - http://www.operamail.com OperaMail Premium - 28MB, POP3, more! US$29.99/year Powered by Outblaze
