Vaccum variable valve timing or camshaft phasing...

Hi folks,

Had some discussion on this before, where I suggested deflecting the cam belt or chain to alter camshaft timing on the fly. Well, couldn't think of a good way to do it then, but just came up with a way that it might be achieved using some off the shelf, or wreckers yard parts.

First off, the mechanism of deflection, best achieved I think with two idlers fixed with respect to each other, on an arm that can be moved back and forward. One would be on the tooth side of the belt, the other on the smooth side, ahead and behind the camshaft. They'd be set up to be deflecting the belt equally at rest, then when the arm is moved, tension is released on one side as it it taken up on the other, keeping constant stock belt tension. In this way, the system is somewhat failsafe in that if the actuator fails the belt should not jump, cam may be left advanced or retarded, but should run still until you fix it, unless you went really wild.

So with the idlers neutral, you'd set up your base cam timing. Probably want it advanced for lower down torque, with the adjustment going towards retarded for top end power.

Now, how do we move it? Well I was first thinking solenoids, but not expecting them to have the power, or be easy to control if powerful. I didn't think of hydraulics, until I heard it mentioned that limited hydraulics could be driven off the power steering pump. So I started thinking how to control that..... and I remembered that fuel pressure regulators are often vacuum controlled, AHA!!! So, if we get a suitable FPR, we can use vacuum to bias a hydraulic actuator to move the idlers back and forward. Set it up for low down torque and good economy at high vacuum and high end power at low vacuum! Would take some fiddling with to get it set nice I'd think, and would need other adjustment valves and probably an AFPR, but apart from the linkages and the idler arm, I think you can round up pretty much everything you need to do it.

So, might try this on the Escort eventually, which has known sweet spots for the cam timing, and should also remain non-interference in case I screw up, whereas the van's new motor is going to be interference, so want to have experience with it before I do it to that.

Might need to figure out how to bias ignition advance and retard though, since that will change WRT TDC if timed off the cam.

Road Warrior

like0

Comments

Vaccum variable valve timing or camshaft phasing...

daox's picture

Thats quite an interesting idea. I can see it working, although it would be hard to setup. You'll have to mount your idler arm very solidly to ensure that there is no variance in timing. Also finding extra slack in the belt/chain may prove hard to come by.

like0

Vaccum variable valve timing or camshaft phasing...

iburnh2o's picture

I've seen diagrams of such a setup though I can't remember just where.

like0

Vaccum variable valve timing or camshaft phasing...

RoadWarrior's picture

daox wrote:

Also finding extra slack in the belt/chain may prove hard to come by.

Well on the tensioning adjusters I've played with there's often about an inch of travel, and you really don't need a lot of movement of the belt to get quite a few degrees of timing change, typically one tooth on the belt can be 19* of cam timing, that's about 1/8 of an inch, .125. That's probably more than needed. If there's 14 inches say between cam sprocket and crank sprocket, and we stick the idler about in the middle... then we take a right angled triangle, where the hypotenuse will be the belt at max deflection, and is equal to 7.0625 (7 and a sixteenth 'coz the other sixteenth is on the other side) inches and the next longest side will be the baseline 7 inches, then we can use pythagoras theorum to tell us that deflection will be sqrt(7.0625^2 - 7^2) = ~ 0.94. I suspect that even restricted to half an inch, you'd get quite a good range. Off the top of my head I think -4 and +8 are the sweet spots on the Escort cam so would only need 12* there. But I guess if you have real trouble you could measure your belt, and try and find a slightly longer one. Chains I guess would be more trouble, and have less to play with from the get-go.

like0

Vaccum variable valve timing or camshaft phasing...

RoadWarrior's picture

Okay, can't for the life of me see why I said the idlers would be on opposite surfaces of the belt... late nights and allergy meds....

So, whoops, should be both on the inside of the belt which will be the toothed side..... guess you want two dead waterpumps or similar to pull the sprockets off.

like0

Vaccum variable valve timing or camshaft phasing...

RoadWarrior's picture

More thoughts on this.....

Recently been thinking about the Atkinson cycle and it's efficiency gain at light load. Wondering if one can get Atkinsonesque type light load FE by advancing the cam further than might seem "normal" under light load conditions. Yes one loses VE and torque, but most vehicles don't have CVT or a huge number of gear ratios to take advantage of that anyway, so not really any point using the fuel for that.

Wondering if there's a good way to match engine load to the advance required. Thinking it would have to be a combo of vacuum and throttle angle, but wondering about how mixture backflow would mess with vacuum readings. Thinking that fuelling might work out okay on the Escort due to the MAF system, it should suck in less air overall, so computer should just be giving it less fuel.

Thinking also that the motor might run like crap with extreme advance. So there would have to be a "normal" only slightly advanced position, to get it started and so that one doesn't get nasty stumble etc trying to move off. Maybe need similar conditions as DOD activation.

I can see a way to get it working for having default timing setting of "economy advance" and then retarding for peak performance as gas pedal is mashed, but I am just having a little trouble seeing a way to make it simple to go more advanced than "default".

Right about now, I'm thinking how nice it would be to have a twin cam head, manage the intake advance seperately from the exhaust.

For best efficiency, bearing in mind that one is doing the equivalent of dropping CR (sort of) or raising octane, then one would want ignition timing advanced, without firing up the intake of course.

So, think I might be seeing the need to get an electronic controller worked out for this. Need to ensure the right conditions, then advance to max, then bump ignition.

Now I'm seeing stellar MPG potential for the little old 'scort. Had been thinking I might get it to 45-50 highway with high comp, fuel heat, grooves, edging and porting, now thinking it could hit 60+ if I can implement this. Then there's aerodynamics... Which I think I could massage for another 25%... Would be absolutely stoked to get this into Insight/Prius type MPG figures...

like0