Valve Float

Valve Float

There has been some controversy over this point, so I have decided to post all the relevant comments made in the original thread!

Question By chris law
hi Ken , sorry to bother you again, i have done 1100 miles on my rebuilt engine and decided to try her out after running her in the result was two bent exhaust valves. I have just dropped the engine off at the bloke who rebuilt it who said he would rebuild it free of charge if the fault was down to him,and I asked him if he had skimmed the head .5mm and he said taht after looking at it he had realised he couldn’t take that much off the head so he left the head and machined the pistons flush to the top of the cylinder and that the compression was now 14 to 1. I can remember him saying the cam timing was at 108 and 112(I am not sure about the 108 100% but he definitely said 112 degrees I am also not sure which way round he had said they were). the bike is a 98 model with a 95 crank and high lift cams, what would be the best combination of timing and cam timing for fast road riding and would it be a good idea to fit heavy duty valve springs.

Post By Ken
As we discussed in a previous post, I am of the opinion you cannot mill or skim a ZX7 head .5mm or 20 thousand inch (.020). Actually you can butcher it all up, BUT its not feasible. It also brings the valves almost .020 closer to hitting. This is not my preferred method of increasing compression.

Piston to valve clearance MUST always be checked when building a engine. I have my own method of doing it,( I will explain this if anyone wants to know) and have specs that I call absolute minimum, and specs I call safe.(I will post these specs if there are any interest in them). Valves that are almost hitting, but are not, will hit at high RPM if the valves slightly float. I use Aftermarket “SINGLE” high RPM springs that must check out at 2 specified values on a RIMAC machine. A spring that is ever so slightly weak, will check out fine of free length measurement, but will fail you later. Ain’t expensive tools neat?

Cam timing numbers you posted do not sound right to me. Muzzy has always recommended 105/105. This is on all the 91-2001 models. I have found stock to be 107/104 on most models. I use 104 intake/107 exhaust for max power. These are not the best mid range numbers though. I call these Daytona numbers. I also use these numbers on street bikes. If it was a track bike, being raced on a short tight track, I would use 103 Intake/103-105 exhaust. The difference between the Daytona numbers and the tight track numbers is about 4-6 HP on top end, but the small numbers build much more mid range than the Daytona numbers about 10!! I use the method called LOBE CENTER, and are measured at 1mm of lift. I have no idea what method your mechanic used. So it may not be possible to compare numbers not knowing how he did it. The method I use is industry standard by shops like Muzzy, Vance&Hines, and most Nascar shops. It is my preferred method also, and allows me to compare other peoples numbers, otherwise comparisons are meaningless.

Ken H20s

Post By chris law
Hi Ken can you explain how you would check piston to valve clearance for me, could you also tell me what the best valve clearances would be. the top end was a bit noisy after the engine was rebuilt ,when I told the bloke he said they set them slightly on the slacker side to stop the valves hitting the pistons on full song.


Post By Bullitt
I am not Ken, but I can help.

Two ways to check clearance.

Clay method.
With the head off the engine, put clay onto the piston valve reliefs. Bolt on the head, attach the cams and chains and turn the engine over. Take everything off and measure the amount of squished clay. Voila! Measured clearance.

Time consuming though.

Check Springs.
With a fully assembled engine this is the way to go. Install light springs (very low tension) into the intake and exhaust valves of one cylinder. Turning the engine over and observing the opening and closing rates of the valves and also using a dial caliper along the way you can map out the exact clearances of the valves. The lighter tension springs allow you to easily compress the valves.

I hope I explained that well.


Post By chris law
Hi Ken just been to see my engine getting stripped, and guess what all the exhaust valves have been hitting the pistons. he said he can’t understand why they hit because there was 1.9mm of valve relief. the way he was talking as if it was not his fault as he had checked the cam timing several times, he also said that the cams wheels had not slipped and now says he set the cams at 106,104. when he machined the pistons he only machined the inlet pockets. i think he is trying to #@%$£#~@ me, what do you think.

Post By SigvartArctander
Chris, I thought the biggest risk of hitting any valves would be the exhaust valves? Why did he machine the inlet pockets then?


Post By Koh
Valves float at very high RPM, causing valves to stay in fully open position for a spilt moment before closing, but piston will keep moving constantly at all RPM, Maybe that would be the cause of valves hitting piston. By Bullitt <> on Sunday, October 07, 2001 – 10:07 pm: valves float because they are vibrating from the impact of closing.
They aren’t getting ‘stuck’ open.
It used to be called valve ‘flutter’


Post By SigvartArctander
No Josh,
Valve float is what Koh says, they stay open to long and get hit by the piston.
What you are describing, I don’t know if it is called valve flutter, but it is a fault too. I have heard something about “valve bounce”…
This can be caused by a to steep cam slope, on the cam closing ramp, this won’t give the valve a controlled “landing” in the valve pocket and it will bounce back. This will also quickly destroy the valve pockets.


Post By Ken
Valve float IMO is when they open and then never fully close before they start to open again. This is usually but not always caused by weak valve springs and or to much rpms for the springs design.

The cam lobe opens and lifts the valve and the valve should be following the cam closed. But the weak spring cannot keep up or because of to many rpms for the springs design, and hence the valve begans to open again before its ever fully closed. Then of course the valve hits the piston. BOOOM!

This is why so many people BLOW UP motors with aftermarket ignitions. They turn the rev limiter up, and BOOOOOM. If it could safely handle more the factory would have gave it to you. If you feel they left a little for a warranty margin, then you use that up, and a tad more, BOOOOOOM!

I have found that springs that measure OKAY per the free length specs given in the shop manual may be on the verge of failure. It takes a hydraulic spring tester to actually measure the spring at 2 lengths. The length of the spring at “Installed height”, the amount the spring is compressed when the valve is installed or closed, then open height, the amount the valve spring is compressed at maximum cam lift. One spring can pass one test but not the other.

The springs should spec out at 40lbs at 1mm of compression. +/- 10%, and 120 lbs at full lift +/- 10%. It takes a special tester for this and special tools to measure installed heights. The tester runs about 600 dollars and is called a Rimac machine. Other than my basement garage, I have never seen one of these in a motorcycle shop. All Nascar shops and most REAL Hi Pro car shops have them. All the motorcycle shops do is measure free length. Its no wonder so many motorcycles blow up from valve springs.

This is one of the real tricks in building engines. As a car mechanic once told me, anyone can do A BRAKE JOB, BUT doing one that “LASTS” is the key. Anyone can build a Big HP motor, but build one to last, now that impress’s me.

Ken H20s

Post By james
Ken not to disagree but every auto and engine class and book I have read says valve float is when the spring does not have enough force to keep the valve closed at the seat. the valve bounces on the seat. I have never heard of the valve not starting to close before the cam starts to open it again. I would think it would try to close but not have enough to keep it closed. Now I have never dealt with a engine that revs as much as my 7 but all my race engines run around 7-7500 rpm. if I am wrong sorry I have read your post on here many times and you sound vary informed on these bikes and I have read about your nascar back ground but if I read your post right I think you may be wrong.


Post By Ken
Perhaps I am wrong, or simply didn’t explain it well enough, so I will look it up in some engine building book and get back with you and a better worded explanation of what I am talking about. Where I was wrong was in stating that it blows up motors. Sometimes YES, Sometimes NO! I may have made it sound like it always does. I do say that valve float is when the valve never fully closes before it starts to open again because the spring can’t keep up. IMO, WHAT YOU ARE DESCRIBING IS CALLED VALVE “BOUNCE”. Lets look at it like this. You have a spring on your front door. You slowly push the door wide open, and slowly let it close shut without ever taking your hand off of it. This is normal. But push it open real fast, and retract you hand back faster than the door closes, then repush the door back open before it fully shuts, that is valve float! And that is the best I can explain it. I will go look it up now!

Ken H2Os

Post By Ken
VALVE FLOAT: as described by Rick Voegelin out of his book, the step by step guide to “Engine Blueprinting”, and POWER SECRETS by Smokey Yunick.

Page 94, Engine Blue Printing,

Stock Valve springs work well on stock motors. But a racing or high performance motor requires something stronger. The valve springs are responsible for keeping the valvetrain under control. A performance cam profile opens the valve faster, holds them open longer, and closes them quicker. You don’t have to be a camshaft engineer to understand that this dramatically increases the spring requirements. If the spring does not have enough tension to ensure the valvetrain precisely follows (remember my screen door example, KW)the cam lobes, then the valves will “FLOAT”, usually with dire results without the valve ever fully closing. This results also in the exhaust valve never being able to cool by making contact with it’s seat.

Power Secrets: page 86-87

It is more difficult for the valve spring to counterbalance the valve train and start to push everything back in the opposite direction on the valve closing phase. When things reach the point that the spring fails to exert enough countering force during the transition, the cam will momentarily loose contact with the valve. (again remember my screen door example). Racers usually describe this condition as valve float, but no matter what you call it, if it happens in a high speed racing engine, you are going to wind up with a bunch of broken pieces. The destruction occurs when the spring generates enough force and starts pushing like hell in the opposite direction, and when the valve train comes back together its like hitting the pieces with a big hammer. The forces generated in current high lift valve trains are tremendous, and when you over rev the motor for the springs design, and push the valvetrain to the point of “valve float”, you are going to break something in a hurry.

OKAY, I think this describes exactly what I have been describing. VALVE BOUNCE??? I don’t think this happens in a motor that has no abnormal problems, and does NOT occur from over revving or weak springs. I do not believe that valves can bounce on and off the seat like a basketball. I have never seen this.

Ken H2Os

Post By SigvartArctander
I just need to get this straight, Ken.
What you say, is Chris`s exhaust valves got bent by hitting the pistons, and this was caused by the cam pushing the valve open again before it closed fully (your hand pushing the door open before it closes) ????


By hineycrack
The mass of a valve, moving up and down at extremely high velocity due to high engine rpm, can be more than a too weak or worn valve spring can handle. the valve spring’s job is to keep the valve closed and/or make the valve follow the eccentric of the cam lobe. if the spring is too weak for the mass of the valve at a particular velocity, then the valve will not follow the cam profile, i.e. the valve will be open (to varying degrees) when it isn’t supposed to be. this can lead to the piston and valve colliding.

Post By SigvartArctander
Why I think your wrong:

Let`s take Chris`s engine, his bent exhaust valves and how it could have happened.

When a valve looses contact with the valve (you retract your hand back faster than the door), (due to a weak spring, over rev or a race cam with an aggressive closing ramp), it will float in thin air. After a micro second the spring will try to push the valve back on the cam, but the cam ramp height is reducing at a faster rate than the spring propelled valve can handle.
This is causing a delay in closing the exhaust valve, relatively to the raising piston.
So what should have been 1.9 mm of piston to valve clearance on Chris`s engine (1.9 is usually enough, so there is probably something more wrong…), is now reduced to -0.5mm maybe, and the raising piston will hit the closing (but not closed enough) valves

As a general rule, the closest point of piston to valve clearance during the rotation of an engine, is between 15 and 5 crankshaft degrees before the piston reaches top dead center.
This is the area where the piston will hit the valves if they are floating, or if the cam timing is wrong, Chris…

If the floating valves should clear the raising pistons, due to deep valve pockets or some “mild” cam timing numbers, they would still be in a “free fall” against the valve pockets. Usually they get a soft landing guided by the cam closing ram, but not this time. Hitting the valve pockets like it’s been hit “with a big hammer” “as described by Rick Voegelin out of his book”, “you are going to break something in a hurry”.
Hitting the valve pocket so hard will make the valve BOUNCE back a fraction, “resulting in that the exhaust valve ALMOST never being able to cool by making contact with it’s seat.”

If the valve where to float so much that it wouldn’t make contact with the cam before the cam started it’s opening ramp, then the valve where to float for approx 60 degrees crank rotation, and that’s a long time floating.
And if the valve where to float for so long and began to open before its ever fully closed, it wouldn’t reach the piston because the piston is over halfway to the bottom by now (ca 80 degrees left until it reaches bottom). I base this on the exhaust opening and closing numbers for a ZX 7 or 9 engine.


This site is kindly funded by the politically correct folks at

Visit the Red Monkey ZX-7R Owners Club

They are very helpful as long as you fill in your profile – your favourite dish had better be roasted nuts if you fail to adhere to this one and only site rule!

Print Friendly, PDF & Email