Valve float on a datalog
 

Hi

Does anyone here know what valve float might look like on a datalog?
How would it affect the AFR? I am chasing a RPM wall (5800) and can see nothing on my logs, but am not really sure what to look for. I want to eliminate any mechanical possibilites, as the tune-up seems fine.

Thanks

Re: Valve float on a datalog
 

Not positive but would expect to see quite a bit of shift on the AFR, depending on what valves are floating, INT. or EXH. I've floated a valve or 2, and on my datalogger it looks much like I'm on the 2step, wasn't logging AFR at the time. What application, and what is your AFR doing...as well as what is your fuel and ignition doing (adding or removing fuel same for ignition) running in closed loop? 5800 sounds a bit low for valve float on most any engine, most completely stock engines will easily rev beyond that point.Again more info please.

Re: Valve float on a datalog
 

OK...you asked for it. Its a 5.7 Hemi with a 76mm turbo, running 12 psi boost. I know this combo doesn't belong here, but I also know that some of the smartest racers are on here so I am hopeing to tap into your brain trust. The AFR is pretty much flat at 12.5 when this occurs. The boost and MAP are also steady. I am in open loop.

Re: Valve float on a datalog
 

O, well I have no experience with turbo cars, however 12.5 sounds a bit on the lean side for 12 PSI of boost IMO. Have you tried richening the map in the area of your issue? I'll bet Ed Wright has tuned a hair dryer car or 2 hopefully he'll chime in here. Do you have any idea of what kind of valve spring pressure you have? Again on boosted stuff I'm merely guessing, however I'd be hard pressed to think you're in valve float at that engine speed. Do you have a map to run unboosted, you could revert to that map and pull off the charged air tube and try it.

Re: Valve float on a datalog
 

11.5-1 is safe. 12.0-1 is pushing it. I would never let one out the door at 12.0-1, certainly not 12.5-1. You can keep leaning it until it eats it's self, and power keeps going up.

Re: Valve float on a datalog
 

I've never had a blown engine float valves and seen AF ratio change. I have seen one with too little spring pressure that had blower pressure keeping the intake valve open too long. It was blowing the valve off the seat. It acted like the cam was too big. The intake did finally close, but way late and it made no power and did some stupid surges. With a large valve and 12 lbs boost you had better have way more spring than you think is normal. Do the math 12 PSI x area of the backside of the valve... that just gets you started back to holding the valve on the seat... I just did a little quick math and with a 2.2 valve you need a min of 46 additional lbs just to hold the valve on the seat. That is over your traditional seat pressure.

Re: Valve float on a datalog
 

I believe your O2 would show a big difference because of a incomplete combustion. Its been a long time since i have used a gas analyzer, but there would be alot of oxygen going out the headers.

Kirk

Re: Valve float on a datalog
 

I'll agree with Adger on this one.

My gut says AFR won't change during valve float.

The valve train would be noisier when valve is bouncing off its seat.

Re: Valve float on a datalog
 

It's not valve float, but the intake valve blown off or held off the seat by intake pressure before the cylinder pressure pushes the valve shut on the compression cycle. Then normal combustion takes place. I should have been more clear on my comment about 02 readings. I've never had any 02 sensors on an engine that had this problem, so I don't know how 02's would respond. I do think they would be normal because the problem with the valve off the seat occurs during the compression event & before combustion, not during overlap. It is Ex Valve Float during overlap that will mess so much with 02 readings.
I think Ed was on somethng with the A/F ratios. I've seen the same thing with a BB blown gas engine I did quite a lot of development work on. I called it Fat when it made good/best power.
My thoughts are you might be looking at two problems. Intake spring pressure and being lean. Don't be shy to address both or others. It is human nature to look for one problem, not a couple. Good luck
One other thought: What is the timing curve look like when & after the boost comes in?

Re: Valve float on a datalog
 

Thanks for your help. The timing is at 20 degrees and steady when it happens. I will try richening up next time. The valve spring info makes alot of sense, and agrees with what my go-to guy was suggesting. Guess I'll be ordering up some springs soon.

Dave

Re: Valve float on a datalog
 

I'm no expert on boosted stuff,or anything else as far as that goes,but in your description you said 5800 was like hitting a wall, I would check springs first. n/a engines will act like that if intake valves are leaking,blows compression back up runner,basically blows fuel back up runner,messes up carb signal,etc
usaully they will get a certain rpm and act like has a governor,had one once on dyno would get to 6500 and shut off like you killed ign.
If you are blowing valves open could act the same as leaking valves.
I'm gonna say that if valves are blowing open you could add fuel and A/F ratio would probaly not change like it should,it probaly would'nt richen as much as it should.
Mike Taylor 3601

Re: Valve float on a datalog
 

Ok, 20 degrees of timing with 12 PSI of boost? Again I'm certainly no expert on boosted engines but seems high to me, certainly at 12.5 AFR, have you ever looked at the plugs on this thing? While I'm all in on the weak valve spring theory, maybe as Adger had eluded to you have a few issues instead on merely one. BTW is this a new engine, I mean fresh engine and have you spoken to the engine builder? Who engine controller are you using? Joe

Re: Valve float on a datalog
 

The plugs look good. The engine builder.....that would be me! Yes its a fresh build

Re: Valve float on a datalog
 

Ok, then can we assume you assembled the heads and if so what did you use for springs, XXX's lbs. @ XXX installed height. This alone should indicate whether you have enough spring on it thus solving 1 mystery.Has this engine been tuned on either an engine or chassis dyno or did you start with a base map? Joe

Re: Valve float on a datalog
 

Which brand is your ECM ? What rpm limits are in the calibration ? Silly me, but I ran my engine on the dyno with a low rpm chip in the MSD, took a few pulls to figure that out. Opps !

Re: Valve float on a datalog
 

I'm using a MS3x for ecm. The rev limiter is set for 7000 rpm. No dyno time (yet)
The valve springs are stock 100lbs@ 1.7" and 325lbs @ 1.5"
Valve lift is .550"

Re: Valve float on a datalog
 

I am not expert but I guessing a 100pds on the seat is way too light. Not knowing alot about your setup you need to spend some money on valve train parts. Is this a street strip combo?

Re: Valve float on a datalog
 

100 lbs on the seat is not even remotely enough!!!!!!!
And your 12.5 lbs of boost is taking off 39 lbs, so you are left with 61 lbs of seat pressure only!!! Float city!!!!!!!

Re: Valve float on a datalog
 

We typically run 140 seat 400/425 open on street N/A hyd roller and you need to raise that,coming up with spring probably won't be big deal,the problem is when you start going over that whether the lifters will handle it. When start going over 140/400-425 a lot of lifters will start compressing,there are lifters that will take more pressure,but don't think stock ones will take it.

You'll probably need 180 ish seat 450-475 open.
Don't know all info that need to know,these are just ballpark recommendations

Mike Taylor 3601

Re: Valve float on a datalog
 

With a low lift like you have and the low spring installed ht. you need a spring that has a lot more seat, like 200+ and a rate that is very tough ending up with at least 550 to 600 open. That may be hard to find. Then when you get enough spring you need a real stiff pushrod. Those engines have a nice short pushrod with good angles. I would use at least a .125 wall and if it will fit a 3/8 to 1/2 double taper .150 wall pushrod. You need to remember that a boosted engine has trouble opening the valves. The pushrods in an A/fuel engine that I've helped with look like drillpipe.
BTW those are just a good starting point for spring pressure. You could need more because of cam profile and valve/valve gear weight.

Re: Valve float on a datalog
 

Could you share more info on your car and engine combo?

I have an 05 Magnum with 5.7 hemi and have thought about forced induction of some kind. But the computer controls intimidate me some.

Are you using a factory or aftermarket computer?

Also, how reliable is the transmission and rear end when more power is made?

Re: Valve float on a datalog
 

Chris:
It's a '73 Dart Swinger 5.7 with a 76 mm turbo. Powerglide with 8 3/4 Mopar diff. I use a MegaSquirt ver. 3x. The set-up and tuning has been difficult, but hopefully will pay off. The Megasquirt worked great with my old motor, and for thousands of other folks too (www.msefi.com) I assume you mean the factory trans and diff behind your Magnum. If so, the trans can be beefed up considerably. The diff., however has its limits. the heavy hitters swap in a 8.8 ford center section ($$$). Check out www.lxforums.com for more info.

Dave