Electronic valve control project

[Wiglaf]

i'm using 10k as a sort of guideline/goal in terms of component selection for the operation frequency, computer, etc. You guys are probably right that the solenoids themselves won't be able to keep up, but if i keep 10k as the goal then what comes out should be just plain fine at 5k.

the hydraulic approach seems like it would be great for making lots of force in a small area, but consider the weight of all that fluid compared to a valve spring, particularly in light of what you said earlier. seems like we'd be shooting ourselves in the foot going any faster than 6k.
plus the added complexity.. i would rather just try to get the computer and solenoid approach to work first then step up to that if we really have to.

[GartnerProspect]

Oil I don't know about. Seems like the valve would have to have a pretty big orifice to allow enough fluid through. And that to me spells latency.

Plus ,If you forgo the valve spring all together then it's no longer a matter of force but just a matter of latency. Though this seems like an idea only I like

Air might be another alternative but then you have to generate high pressure compressed air and carry reserve for startup...

I agree, keep it simple. Build a working example at 2krpm and work up from there. Once the basic parts are in place changing and trying new things is way easier. I think most of will come down to lightweight valvetrain and highly optimized solenoids.

And this is why I'd also suggest starting on a single or twin cylinder bench motor. But that's just me.

[Wiglaf]

the oil would be for the solenoid itself, most would stay up top just like the normal cam arrangement uses.

there's a few reasons that I want to keep a spring, but it doesn't have to be nearly as hard as they are now.
First off it's a fail-safe for losing power or something, the valve just stays closed where it won't get mashed.
having a reasonable spring in there is part of how i intend to get a simple variable lift control, just control the current to the opening coil. if it's not there, then you have to do some push-pull stuff.. i think that would be less predictable, especially if the valve moves very easily, a little more or less air turbulence might leave us with a different result than anticipated.. unless you had a sensor and closed-loop the control. but that would be adding some complication..

I do intend to have a way to measure and log how far the valve opens for development purpose but it won't be closed loop. I had in mind just a sliding potentiometer, it's cheap and would work well for dev but not have much of a lifespan for car use.

[Fred]

I wonder if you could somehow use a hall effect device in raw form and measure the strength of the perpendicular element in it? perhaps that could tell you distance fairly accurately once calibrated? Just a random thought :-)

[jbelanger]

Do you really need variable lift if you have complete control of the opening and closing time? You might need to start closing the valve a tiny bit before if you have more lift but it seems simpler than controlling lift. Unless you control turbulence with the lift or you have an interference setup, I don't see the need to have variable lift.

Jean

[Fred]

I think you may find you get better torque at lower lifts at lower rpms. Just guessing though really :-)

[GartnerProspect]

If you use a light spring you're going to have problems with float. I definitely agree in any case a spring needs to be there, but I'm talking like 10lbs/inch only used for keeping the valve closed by default. You really can't escape having to deal with inertia. Adding the mass of the solenoid plunger (and the pushrod for your solenoid), is going to mean you'd need an even stiffer spring than you'd normally need in a cam-over-valve setup.

I like the idea of variable lift though. I'm not sure how consistent any current limiting based concept is going to be though. Too many varaibles for me. Temp changes, slight manufacturing or stress differences is solenoids and valvesprings... Which now that I think about it are all things that greatly affect the system regardless.

With the cam you have an absolute profile. The only thing that needs to be perfect is the cam. The rest of the system doesn't matter as long as that lobe profile is legit. (with peak RPM use the exception) With the electronic system that changes quite a bit. No more mechanical, absolute, invariable guidance. Just magnetic flux in a coil. Valve reaction time and opening speed will varry quite a bit with all the other variables.

Just something else to overcome

[jbelanger]

I think you may find you get better torque at lower lifts at lower rpms. Just guessing though really

Actually, that's what I'm not sure about. Is it the lift or the shorter duration? With a cam lobe, you don't have much of a choice to have low lift for a short duration but if you don't have this limitation what's the best choice?

Jean

[Fred]

Honda seem to think that both lower duration AND lift are the way to go, but that *could* be to produce less torque and use less fuel I guess? As much as honda only make three good cars (as opposed to good engines of which they make plenty) NSX, S2K, Beat, as an engineering company, I have a lot of respect for them.

[jbelanger]

But that's what I'm saying. You have to go with a lower lift if you go with a lower duration because of the cam profile. If you remove the constraint of the cam profile, would you get better results with a lower lift or not? I'm not sure that the lift does have a significant impact compared to duration and timing.

Jean