Electronic valve control project

[Wiglaf]

This project is just what it sounds like: using electromagnets instead of a camshaft to control valves.

I'm in the beginning stages, and my current test subject is a used head for an M20 block. This is the BMW inline 6 that goes in the e30's. I picked this because I like the inline 6, and it's only 2 valves per cyl... but of course a head can be made for any motor.

the end product will be the head, all power+driver units, and a complete engine management computer to control the valves, fuel, and spark.

I would like the firmware and software interface to be free and open (as is the basis for the site). The computer could be DIY like we are used to but it will have to be faster than the megasquirt and probably what gets made on this site.

the power circuitry would not be DIY nor would I feel ok letting an average joe try to hack it, but the schematics would be no big secret either. If I can I'm building the power stuff into the head anyways so hopefully it just be a matter of sending out digital outputs.

so basically you would buy the head, buy or DIY the computer, then all the downloads/software/firmware is free and continuously being tweaked by the community. I feel that this is the best way since I am not a software engineer, and if someone wants to 2-cycle or 6-cycle the motor or whatever then they should have everything available to change the code. Although half the reason I am doing this is for 6-cycle operation so it would be an option from the start. Unless I get in a legal fight with Crower over the 6 cycle thing.

So now the part on why I'm posting here. I'm very experienced in analog power, and pretty good with mechanicals. I can get this thing built. Where I am weak is with digital, and when it comes to software/firmware I am kind of worthless. I imagine that this is the inverse of a lot of you, who are very good at the software/firmware+digital and then kinda shy away from the high-power and analog stuff. Not trying to put anyone down but we all have our strengths, and mine is not software.

This isn't going anywhere fast yet, I'm tied down financially and have to wrap up a turbo project. But hopefully in about 6~9 months I will have something ready to plug a computer into.
Let me know if you are interested, feel free to hit me with questions.

[Fred]

WOW, that is a seriously cool project :-)

I strongly suspect that our cpu will not be sufficient for your needs unless you use 3 of them (one for intake, one for exhaust, one for engine control).

How accurate do you need the timing of the valves to be? I imagine that it is a fairly fussy critical system timing wise.

I look forward to seeing your progress!

Admin.

[Mops]

Hm...

yeah, i've been toying with this idea for a while.

I came across this project on the internet, (cant remember the link), guy was experimented with small engines and electrically actuated valves. he actualy had several smal engines running with electric valves. I do not recall the details, but the channenge was to get the valves opening/closing fast enough to keep up with engine... If memory serves me right, the max sustainable rpm he as able to achieve was 3krpm... problem is... the siffer spring you use to close the valve, the more power you need to open the valve and the slower it goes... maybe actuate the valve electrically in both ways (by applying inverse current to close it) ???

BUT i'm thinking... with that valve design... the good thing is that you are not bound by common valvetrain limitations... you can place those valve actuators on the outside of the engine.... so you can make them big enough... but then usually big=slow...

i'd rather try with a single pot, 2 valve head, so in case it doesnt work, you havent wasted time/money/efforts into 6 sylinder design.

or maybe it would be feasible to DIY F1 style valvesprings operated by compressed air ? when done correclt, there's very little valvetrain loss because when lobe is pushing the valve down there' near no resistance and then air pressure is used to close the valve back up...

long story short... as much as i like tinkering.. to be successful it has to be cheap/easy to make/reliable....

maybe you can research rotary valvetrain... the only problem they have there is that it doesnt last too long.. if you can engineer that issue out then you are the man (worth millions).

aswell early in 20th century, they made those 'rotating sleeve valve engine', which IMHo is a great design, BUT they could never get lubrication to work properly and these didnt last long....

this 6 stroke BS is just a variation of opposed piston engine (design from early 20th century), used alot until these days in things like train big diesel engines. German Junkers (2ww bomber) was using *DIESEL* opposed piston engines that generated 1200hp from 1200lbp of weight. http://en.wikipedia.org/wiki/Opposed_piston_engine

[Jake Breyck]

What do you plan on useing for the solinoides to drive the valves? how reliable do you think they will be? and also how fast, i could see there being a functional rpm limit.

are you thinking about working in throttle control into this as well or will they have fixed valve timing and lift?

this would be extra neat if the timing and lift is veriable not only for the power tuning but the super simple intake system (velosity stacks welded to a flange, Done! No trottle body)

The extra power potental is great. not having to drive the cams alone would considerably increse power.


Other then mounting of the parts and disabling the oiling on the head what other modifacations would ned to be made?

Im very intreted and will follow this if you get a package that works i would be intrested in playing with it as well.

[Mops]

WOW, that is a seriously cool project

I strongly suspect that our cpu will not be sufficient for your needs unless you use 3 of them (one for intake, one for exhaust, one for engine control).

How accurate do you need the timing of the valves to be? I imagine that it is a fairly fussy critical system timing wise.

I look forward to seeing your progress!

Admin.

not any more the spark timing.... 1 degree crank rotation accuracy would be acceptable, 0.1 degree would be more than enough. after all all you need is to send signal to open the valve and then send signal to close the valve....
problem is that it will be hard to physically build valves that are light and strong (titanium maybe) and open as fast as required.... somebody correct me... for how long the valve is open at 6krpm ? 3-4ms ? that valve would have top fully open AND close by then....

[AbeFM]

I hate to be pessimistic, so let's put it this way: If you fail, at least you won't be alone.

Didn't ford work on this for a while? Anything that's fast enough seems not to last very long. Anyway, I'm a HUGE fan, that and the linear actuator, fully EM suspension BOSE is working on are about the two coolest things since sliced bread when it comes to cars. The only true innovations in a long long time.

You didn't put your location, tell me you live in socal and I'll come by and help - I'd love to get shocked into a crispy critter by something that awesome.

I've thought electrically actuated rotary or spinning disc valves would be a lot easier to do, but... I have no clue.

The air pressure F1 thing is super awesome, never heard of it, but it could DEFINITELY be made electronic I would think.
-Abe.

[Fred]

Pneumatic actuation is pretty much the only way to spin to 20k rpm. That's been around for a while :-)

Billzilla has some thoughts on that on his webpage (excellent small toyota reference) :

http://www.billzilla.org

not sure of the exact url.

[ababkin]

Very cool discussion guys

I myself also did not know about the pneumatic F1 thing

Just a random idea: why not use solenoids and pneumatics together to extract benefits from both?:
- have a small electronic solenoid (for fast speed/response) controlling the pilot pressure, which in turn controls the big pneumatic valve (to generate strong mechanical force), which moves valves.

one will need an air compressor for this though, which will be needed anyways (for brake booster and other stuff that relies on vacuum) if no throttle-body is used (like on newer bimmers, that control throttle through the intake valve lift)

To the OP: you are absolutely right on strength-weakness thing - i think i am the complete opposite of your skill set

Alex

[Wiglaf]

to answer your questions.. I cannot find an off-shelf solenoid that is suitable. I will wind my own coils. the valve itself shouldn't have to be made of crazy exotic material or anything, but will be heavier. It will have a spring return similar to regular valves but i would like to make that lighter and use a return coil to avoid float at higher rpm.

each coil will have a variable current driver so that the current is proportional to the lift against the spring. This is also to control the open/close speeds. I imagine calibrating this to a table or equation, should be as consistent as the springs are. (so yeah throttle body is gone)
A lot of throttle we could probably do through atkinson cycle.
http://en.wikipedia.org/wiki/Atkinson_Cycle
http://en.wikipedia.org/wiki/Miller_cycle

This reason (aside from noise) is why i don't like pneumatic much. I've dealt with air driven industrial/automation stuff and it's very on/off in nature. fine for a 20krpm f1 car i guess. I am going to try and design for 10krpm even though this particular motor might not like it.

I am not blocking off oil and coolant to the head, I anticipate needing to cool and lubricate the coils.

i've considered the rotary valvetrain but it doesn't offer the adjustability and options that controlling each individual valve does. also there is a whirlwind of patents surrounding that.

I agree that the bose suspension is awesome. I've considered attempting such control but the power needs of that setup are extreme. I think in that case an air over oil hydraulic control would be good.

BMW and bosch (i think) successfully made an electronic valve prototype in the 90's. it was a 12V system that was very loud (no soft letdown of the valve) and the coils were bulky and had heat problems.
I don't know if ford tried, but they would have had the same trouble. The power supplies to control this sort of power back then were very bulky.

I believe that even with the newer wave of power supply tech, there is no way that this would ever work on a 12V system, the current is just too high, and the field strength comes so much easier with higher voltage. There will be a higher voltage subsystem involved. I think I can keep it safely encased in the head with the drivers, if not then it will be limited to 48V to keep it classified as touch-safe low voltage.

.. what else.. I'm in Cleveland, Ohio.

mops, I meant the crower version of the 6-stroke:
http://en.wikipedia.org/wiki/Crower_six_stroke
this would require direct-injection of the water though, and this head I picked out only has holes for the spark plugs. so even more modification is necessary to get me there.

[Fred]

I was going to suggest that blocking coolant flow was a bad plan, but your 6 stroke link answered that :-)

Interesting stuff. I'm very much looking forward to seeing your progress.

Admin.