Electronic valve control project

[GartnerProspect]

Have you considered winding your own solenoids?

I remember once upon a time having a way to calculate the force based on certain dimensions, voltage and gauge of wire. If you keep the valve springs I can only imagine that you're going to need a pretty hefty solenoid.

[Wiglaf]

Have you considered winding your own solenoids?

I remember once upon a time having a way to calculate the force based on certain dimensions, voltage and gauge of wire. If you keep the valve springs I can only imagine that you're going to need a pretty hefty solenoid.

that's sort of the plan.
the higher voltage should keep the size of the coils down but they will still be pretty serious, just not as absurd as the 12v coils. Sorting this out is my first priority, as soon as the other turbo car is back on the road. :p

[GartnerProspect]

I was sketching out some designs and have been leaning towards a push-pull design. Also tossed around the idea of a rocker arm as well. Without a heavy duty valve spring you could get away with a very light rocker which would negate the typical downfalls of using one.

With a rocker and a push-pull arrangement you can keep the gap small and throw short which should improve cyclic response. Of course I'd still include a lightweight valve spring just to default the valve to the closed position.

With a high voltage system, essentially springless design, and taking advantages of the strengths of a solenoid by using a rocker. I think the overall concept has merit. However, I'm also one to realize that if it was feasable with shade tree technology that the OEMs would have jumped onboard by now... Really curoius to see how it goes.

[Wiglaf]

i'd like to avoid rockers if possible.. you're right that most of the time we won't need that much force if there isn't an insane spring to fight, but we will have to lay down that sort of force to avoid float at higher RPM. but at idle/low rpm we should be able to reduce those losses quite a bit.

and don't underestimate the lengths that OEM's will go to avoid changing the 12V voltage system, it's a pretty big deal when you consider that it would snowball into millions of dollars just in safety testing, let alone parts distribution and everything. Just look at how long LED's have been around before they started using them, and most that do still slap regular bulbs in for the turn signals. They HAVE had working electronic valve prototypes.

[GartnerProspect]

As I'm sure most people know know valve float is caused by the inertia of the valve train overcoming the force of the spring. What lots of people don't realize is how much inertia the actual valve spring itself has. Typically it's one of the heavier single components and while it doesn't ALL move, lots of it does. It's what led to the development of the beehive springs. They can eliminate valve float without increasing springrate simply by reducing intertial mass of the valvetrain.

By eliminating the spring and using effectively two solenoids per valve you should never encounter valve float until you exceed the cyclic rate of the solenoid. This limit is going to exist whether you use a single acting system with a traditional valve spring or the two-way setup I'm describing. (though I suspect that the size and strength of the solenoid needed for one-way may actually have a slower reaction time)

Just like a fuel injector, using a peak and hold driver and low impedence coil will help with cyclic time. As will closing the airgap and keeping the plunger throw to a functional minimum. And keep in mind air gap shape largely influences the reaction of the solenoid with a flat faced plunger offering the most peak force at close range, but conical shape offering the most linear force response over longer travels (at a significantly lower peak force) Also, because the plunger HAS to be a ferrous (read: massive) materia, and the total availible force is a function of the mass of the plunger, it makes sense to reduce the travel and utilize a (albiet more complicated) rocker design.

I usually prefer to keep things simple though, which pulls me away from the rocker based design. With a direct actuated push-pull valvetrain, and a low impedence coil on a 12v high current peak and hold driver I think a functional model could be done. I just think it's going to be RPM limited more than a rocker design. In my head some of this comes down simply to a larger more powerful solenoid requiring more time for the magnetic flux to build, but upping the voltage should fix that too... I think. I'm no electrical engineer so take all of this crap as you will!

If I wasn't working 70 hours a week and already doing more homework on this stuff than I ever did in school I might have time to read more about solenoids!

[Wiglaf]

well, that's the biggest problem right there.. the more megnetic force we need, the more ferrous material there has to be, which is more weight, so we need more force... but don't jump to the conclusion that it will be a lump of iron. At the rate of open/close for 10krpm, we're dealing with times in the microsecond range.. pushing the effective control frequency to the 50~100kHz range. For transformers we use ferrite for switching in that range, but for a solenoid app it might be bunk. I really just need to wind up some experiments.

[GartnerProspect]

10krpm is crazy talk!

Thankfully, 95% of automobiles out there don't rev past 6k. So for proof of concept you're dealing with a little more millisecond for total valve cyclic time... BUT, you do have another 3ms to charge the coil. So I think if your valvetrain mass was limited to bare bare minimum it's doable with worldly materials. That's the only glimmer of hope I've found so far

But I'm a hands on kinda guy too so wind up a 'noid, grab the scope, and get testing!!

[Fred]

10krpm is crazy talk! :)

Thankfully, 95% of automobiles out there don't rev past 6k.

You won't find it in my garage unless it spins at least 7k.

8k is reasonable. 9k pushing things, 10k still achievable.

longevity starts to decline though.

[GartnerProspect]

I'm all about 10k. The sound can not be matched by anything.

I just don't see solenoids reacting fast enough to make it happen

[rsafier]

The Sturman Industries valve system I think went up to like 12000rpm, tho they are marketing for diesels right now as that is were the most money is. They use an oil pressure system to move the valves, but still solinods eventually help control that flow. Its been a while since I read all their patents, but they did have a electromechanical only system only but have moved on. I would assume the mass of a direct solinod does make higher RPM operation too hard. using the fluid pressure to help things out allows them to have a small control solinode that probably can respond much quicker and lessen the mass of the valvetrain.
Still from an amature level that adds a hell of a lot more complexity.