On shim bucket style valve designs you will be adjusting the gap between the cam and the cam follower, also called a bucket. In some applications the gap will be between a rocker arm and the shim itself as in the CRF’s and KTM’s. The size of the gap is configured at the factory and has to do with heat expansion of the valve stem and oiling. For this reason set your valves at room temperature to factory specs. The most common factors that decrease clearance is valve face wear, seat wear and valve stem stretching. On a modern dirt bikes wear is accelerated by the higher RPM, dirt from and increased appetite for air and lightweight titanium race valves with small seat widths for more power. When there is an increase in gap it is usually due to carbon on the valve and seat face or cam and rocker arm component wear and in extreme cases a bent valve. As a rule of thumb, when you reduce the shim size more than two full sizes, .004" from the stock shim size it is time to machine the seats and replace valves. This is an indication that the seat and the valve is worn .004". The hard facing on most modern valves is less than .001" and when that=s diminished the valve wears extremely fast. You can=t face the valves on a grinder like in the old days because of the hard facing finish. Stainless valves can be lightly lapped in with a 500 or finer grit compound. Never lap Titanium valves or you will destroy the Nitrate or other special hard face finishes. This is another reason precision machining of the seat during valve replacement is critical. We see a lot of racers trying to troubleshoot poor running and re-jet when a worn valve or bad valve job is the culprit.
The proper feel on the feeler gauge is a slight amount of drag. Most factory specs for shim type valve trains give a tolerance of + or - .001". Factory shim sizes come in half sizes .001"/.025mm. Most available shim kit sizes are .002/.05mm sizing. A perfect adjustment of .005" would mean a .004 blade would slide in with no friction and a .006" blade would have to be forced. If your .006" blade slid in easy with no drag you should be able to go one shim size .05mm larger to get to the minimum gap of .004. Or a factory half shim size .025mm for the perfect adjustment of .005. I tend to lean towards the minimum for more valve lift and less spring harmonics but you have to check clearances more often. On many engines like the CRF450 the left intake wears faster than the right and just the opposite on the CRF250 so you might want to leave those at the maximum. Cam chain tension needs to be applied because the cam bearing has some play and this reduces the valve gap. Before you can set the gap you must have your engine on the compression stroke at top dead center. An easy way to tell when your on the compression stroke is when both the intake and exhaust valves have play. If you don=t trust the marks on the flywheel for TDC you can stick a probe through the spark plug hole and move the crank back and forth until you find top and then verify with the factory marks. If you have two marks on the flywheel the first one that comes into view while turning flywheel counter clockwise is the firing mark and the next one is the TDC. At TDC you will see marks on the cam sprocket that will line up with the head surface or a mark on the cam tower itself. On most twin cam models the top of the cam lobes will be point forward on the exhaust and to the rear on the intake and on single cam as in the CRF it will point away from the cam. From this point just line up the marks. If you can't get the marks to line up perfect you probably have a worn cam chain and guides. There is also a possibility that sprockets that are pressed on the cam may have slipped. In this case the marks can be on but the cam timing will be off. It is a good idea to put a locating mark on the sprocket and cam to verify their relation for future troubleshooting.
Common time consuming mistakes
Over tightening the cam holder top cover bolts. This can smash the aluminum in the journal causing cam seizure. It can even smash over into the cam follower bores and make it impossible to remove the buckets. You will also want to oil or use engine assembly grease in the cam journals if you cleaned the oil off the cam. Models that have these types of heads are KXF, RMZ, YZF type head. The head on these machines can run $500 or more so be careful and follow factory specs.
Don't drop the valve train parts into the bottom end through the cam chain opening. This includes dowels bolts, shims and half moons on the cam towers.
Make sure you have cams correctly timed before you kick it over. Slowly turn the engine over by hand at the crank and feel for sometime like a valve hitting the piston. Don't panic if there is a loud click just before TDC. That’s the compression release clicking off.
You can’t always trust your local shop to do valve seat machining
Read the following so you can ask educated questions before having your valve seats refinished. Most of tuners including myself in the off road race scene have work mostly on 2-stroke engines. Most of my experience on 4 strokes came from the XR's and KTM race 4-strokes. These machines don't even compare with the new high RPM, Formula 1 style engines. For a short while after the first YZ400 came out in 1998 I couldn't understand why my old Neway manual valve seat cutting tools, the motorcycle shop standard, where not cutting concentric seats. I soon realized it was flex in the pilots because the smaller diameter guides. The heads used on the newer multi valve sport and off road race bikes run very small diameter valve stems and guides that is placed a lot farther from the valve seat. This is due to the huge ports needed for flow volume and also for an unobstructed flow around the valve stem. If a flexible, loose or tapered pilot tilts or bends at the guide it creates an exaggerated larger degree of error at the seat surface creating an oval or non-concentric seat. A round valve face won’t seal in an oval seat. The result is a loss in compression and gasses will be push into the ports in the opposite direction creating more airflow, mixture and carb tuning problems. Valves dissipate their heat through contact with the seat and if the seat is oval it will only come in partial contact resulting in burnt valves. We are talking an out of round seat by tenths of thousandths. A human hair is about .002” and acceptable race tolerance is about .0005 about one quarter the width of a hair or less. I bought high quality carbide pilots at $145 each to replace the more flexible tool steel. This helped but was still unsatisfactory and usually by the time I cut all 3 angles and attempted to get reasonable concentrically after re-cutting after each test the seat was deep enough in the head to decrease the shims several sizes. This also reduces compression when the valve sinks deeper into the head and destroys airflow. Also the seats are softer on many of the newer race bikes so it’s real easy to cut to deep and cleaning up mistakes are costly. I decided to try a quality automotive machine shop that had built lots of race engines and I explained the special needs. I even supplied the special smaller 3 angle bits and pilots. I think they would have had to practice on a lot of motorcycle heads to figure it out. Most automotive machinist found that it’s not cost effective to take the time to cut a precision seat. One of the machinists even mentioned that you can make a few mistakes on a V8 and you don’t really notice it, unlike a single cylinder motorcycle. I then bought a Serdi Micro and all the special tooling and it proved to be acceptable after learning its shortcomings and even at that it took forever and found it would have been more cost effective to buy new heads unless it had a high dollar port job. Now after several years of searching for the perfect solution and spending thousands on wasted tooling, talking for hours with race engineers, exploring high tech internet sites and dealing with machine equipment representatives I have come up with a pretty good idea of what it takes to cut near perfect valve seats. It=s not cheap, it take a lot of patience and a meticulous understanding of precision machining and a hands on feel for the machine your using. I decided on the Rottler SG7 with a lighter weight air float spindle head and dual straight taper tungsten carbide precision pilots. I also bought a lot of precision options for cutting deaths and duplicating jobs.
Seat machining 101
A pilot shaft is fit into the valve guide in order to achieve a center axis for the cutting tool. The actual cutting tool can be a grinding stone, a multiple blade carbide cutter system or a multi angle single blade cutter and the newest design being a single cutting point that is CNC controlled to cut the angles. Just remember that all of these cutters work off the centerline axis through the pilot. If a straight or live pilot has play in the guide it will tilt. If a taper pilot is used it can tilt just by design. The best is a dead pilot that starts in guide using the straight design and has a light taper to take up any clearance. A live pilot turns in the guide and a dead one is tight an does not turn. Multi blade cutters or stones are normally turned with a hand held motor or crank above the pilot and therefore can’t have a rigid top support on the pilot. Supporting the pilot from above and below in the guide gives more support, less flex, and keeps a more accurate centerline.
Machining factors on concentrically tolerances
The newer heads run a valve with a smaller diameter stem. Consequently the pilots used are also smaller creating the possibility for more flex when cutting. To reduce this flex pilots have gone from tool steel to carbide reducing the flex by about 10 times. The cutter and pilot will try to follow the old seat diameter even if it has become elliptical from wear or less precise previous seat cutting. Grinding wheels and multiple blade angle cutters try to create a naturally more concentric diameter because they even out the pressure at several cutting points or 360 degrees as in a stone. Unfortunately these type cutters are only supported with the pilot in the lower guide and not the top top of the pilot. this creates a huge possibility for lateral pressure on the pilot while turning it by hand or by a hand held motor which created an off center or oval seat. The other set back by these now ancient methods is that they can only cut one angle at a time. Whenever one angle is cut the width of the adjoining angle is decrease creating a painstakingly long drawn out time to create a good valve job that aligns on the face of the valve properly and is exactly the same dimensions and depth as the seat next to it. The current cutter is a single blade that has all the angles and radius built in so you can duplicate in the next seat. Sense the cutting edge puts all its pressure against the seat in one point it try to flex the pilot in the opposite direction and will follow the old seat. To reduce flex and cut on center the pilot on this type of cutting is also anchored at the top and both axis should be in perfect alignment. A carbide pilot has to be used especially on the smaller diameter guides. The blade is adjusted in or out depending on the location needed to center on the valve face. Once this is set you really only need to measure the depth of the cut toward the guide so that it is the same a the seat next to it. This keep the valve shim size the same. To deep of a cut will diminish spring tension, reduce shim size and also lower the compression because the valve will sit deeper in the head. Every time a mistake is made you have to cut deeper to clean it up. Its really easy to make mistakes especially when using non-top supported pilots and having to measure cut widths when using single angle cutters.
Now that you have an idea on how these different cutting methods are used we will focus on all the factors that come into play in order to reduce the out of round and non-centered seat. Lets start with the pilots. A tapered dead pilot can't move in the guide but it can tip or be tipped due to it's nature and the wear in the guide. A live straight pilot can only tip as far as the angle created from the upper and lower edge of the guide and the clearance of the pilot to the guide. A dead pilot will have more of a tendency to cut concentric seat if it doesn't tip while cutting and if the spindle in the machine which is live doesn’t have any play. With a dead pilot system the valve stem would have to angle itself or flex when seated if the pilot was off centered even though the seat diameter was perfectly concentric. Normally with a dead tapered pilot system you will only need one or two pilot guide diameter. Say your valve guide is 5mm and not two worn you might just have to stock a 4.98 and a 5. In a life pilot system you want a tight fit but it has to turn so you might need a 4.96 to a 5.1. Each carbide pilot can cost $125 or more. Now you are asking yourself witch is the best system, dead or live. Most of the high volume automotive machine shops tend to like the taper pilot system. They are faster and easier to set up and can stay within factory specs. Race shops, on the other hand usually prefer the live pilots because they obtain a higher degree of accuracy. Especially shops that deal in smaller high performance multi valve heads. You have to remember that any tolerances in the pilot to guide and the spindle will show up equally in the concentricity of the seat. A standard automotive valve job on either machine may have one or two thousands concentricity and still only create 10 to 15% leak down. A leak down test is performed by filling the cylinder at TDC on the compression stroke with a steady supply of pressure, say 100 lbs and what leaks off is the amount escaping through the valve seats, piston rings or gaskets. 0-5% perfect for race engines, 10-15% usually won=t change tuning to much but will be down on performance. 20% or more means the engine needs to be rebuilt depending on the source of the leak. Valve clearance less than 0 will also cause leak down so check that first. Leak down or head problems on a single cylinder race machine that doesn=t have a few extra good cylinders to cover up a problem is even more critical on a perfect valve job.
The perfect head
In order for a head to be considered a race head you would have to have to have all the seats concentric to within .0005" or less. That's one half of a thousandth. The seats would also have to be smooth without chatter marks because you can't lap in Titanium valves with a nitrate or some other micro hard surface coating. For racing and not necessarily endurance you would probably see a .8mm intake and a 1.0mm exhaust seat width. The valve cools mostly through contact with the seat and so you need a bit more on the exhaust and smaller for better flow on the intake. This is also were concentricity comes in. If your valve is only touching part of the seat it won't transfer heat near as fast. You will be using new valves because you can't cut the face of modern valves or you will loose the hard facing and they all should be exactly the same. Sense all the valves are identical the shim sizes would be the same because the race machinist to each seat to the same depth. The guides would reamed to spec or at least within wear limits before machining the seats. All of this would be done most likely on live pilot machine and most likely a Serdi. Set up is just as critical as the actual machine at from this point on. In order to keep super tight tolerances the machinist would not only have to know the feel and familiarity of the machine he=s working on but also use the tightest pilots, have zero clearance in the spindle bearings and be able to align the pilot perfectly between the guide and the spindle. On a dead pilot machine the spindle ball mechanism would have to be free of play. Most high dollar machines have a air float system for the motor/spindle and some even on the bed where the head is clamped. Certain machines have heavier floating parts that can overcome the strength of the smaller pilots used in small multi valve car, karting and motorcycle race engines. The spindle on most modern seat and guide machine can cut on an angle to adapt to the angle of the guide. Gravity from the weight of the spindle can also off center the pilot to the seat. Therefore its better to tip the head until the guide is in line with the pilot and spindle. A lighter spindle/motor/floating head is better for alignment on small motors. Certain companies are building equipment with this in mind but most will still do heavy automotive heads. The most common 3 angle seat is a 60/45/30. The angle the valve comes in contact with is almost always 45 degrees. There are hundreds of differently cut bits including 3 and 5 angle and radius for racing. Most machines use a similar tool holder layout and will accept the original Serdi style tooling.
The end result
You can gain 10 to 15% more horsepower between an acceptable and a real race prepared valve job. The better seal you have on the valves the higher the compression will be. Especially at low RPM when the compression has more time to leak out. For those racers that are on and off the throttle as in motocross and tight course racing the largest gain from a perfect seat is in the lower RPM range. After that the flow characteristics kick in from the angles of the seat and the porting. The valves will cool better and have less hot spots in the seat because they are touching the seat all the way around. If you pick the right multi angle cutting bit for a better flow you will pack a lot more mixture into the chamber. Carburetor tuning is more consistent because the valves are not constantly bouncing around the seats. Valve spring harmonics can also be reduced due to the way the hits the whole seat at the same time. Imagine a valve hitting the high part of a seat that was not cut perpendicular to the centerline of the guide. The stem of valve actually has to flex when the head of the valve is force to conform with the seat surface under pressure. This eventually weakens the valve and the head can break off causing catastrophic damage.
Top race tuners know that given a certain engine size and compression, the 5mm area that includes the seat and the valve is the most critical in achieving the greatest amount of power and yet it is usually overlooked. If you can't put it on a shelf in every performance shop in the country you can't really market it. Most mechanics and race shops don't have the equipment to do a standard valve job so they opt to sell bolt on performance or sell new heads if needed.