Any way to get a first stab at VVT in the code?

[Paul_VR6]

Got the new motor in and it uses VVT on both the intake and exhaust cams. Apparently even when "full on" it uses PWM ~80% to keep the current down. I was going to use the boost controller for one cam control in open loop. Is it possible to copy this feature so that there's two PWM rpm/load tables? Thanks!

[muythaibxr]

I could probably do an open-loop VVT controller just to help you get a start with it. However, if I understand how most of these systems work, they almost have to be done in a closed-loop fashion to get any kind of repeatable results.

Ken

[Paul_VR6]

Open loop is exactly what will get me rolling. At the very least I can do on/off with just the duty table. I know on an earlier version of this motor the exhaust cam was just on/off with 0/80% duty but later models made it fully variable. I really want to see if the VVT logging feature will work out, and try and pin down details on this cam sensor setup.

[muythaibxr]

Anyone who wants VVT supported should probably send me specs for how their particular engines work.

It appears that while many factories did this similarly, none did it the same.

Ken

[Matt Cramer]

Was the VW VVT controller a single-acting solenoid with just one oil output, or a spool valve that sends oil through two separate channels? The spool valves would be just about impossible to control in open loop, as they control the rate of cam advance rather that the amount.

[muythaibxr]

That's exactly what I was referring to in saying that open-loop control wouldn't be possible Matt.

Ken

[Paul_VR6]

As far as I can tell there are two oil feeds to the camshaft through the vvt solenoid, one passage for advance the other for retard, and a bleed port off the end. There is definitely no mechanical means of returning the cam to rest, springs, etc.

Trying to figure out how it will all work. At fixed duty the valve should be at a set physical position, and the main oil feed to the solenoid should be split into some fixed ratio between them (perhaps untrue?). As long as that ratio isn't 1:1 the cam will continue to move, what I'm unsure of is there a specific duty cycle that hits that ratio (50% for example if the valve is actually balanced). It would be nice to test all this without the motor running and the cam cover off so I can actually see what's happening. Maintaining the proper oil pressure might be hard, however.

Here are some plots from factory control of this motor:

RPM (X) vs MAF (Y) vs Cam Advance Request Angle (Z)



RPM (X) vs MAF (Y) vs Intake Camshaft Duty Cycle (avg) (Z)



RPM (X) vs Load (Y) vs Cam Advance Actual (Z)



I had a good plot of cam angle vs duty cycle with that data but I must have not hosted the image. I'll try and dig that up tomorrow, I vaguely remember it containing something worth looking at.

I know open loop isn't ideal but I can at least work out cam sensor trigger lead with vvt active/inactive, as well as the funky trigger wheel setup as well. There has to be power to be made even with crude control.

[muythaibxr]

I don't know... it's going to be basically impossible to control that setup in any kind of consistent manner open-loop. It might be worse than not controlling it at all.

Ken

[Paul_VR6]

Well it only needs to be consistent for about 12 seconds and at WOT I'll start with the off/80% scheme first and go from there. I just need an output other then the boost controller, and have it able to come out on the MS3 header pin for VVT.

[Matt Cramer]

Well it only needs to be consistent for about 12 seconds and at WOT

Unless you need it to be consistently at full advance or full retard, I can pretty much guaranty you that open loop won't stay at any other spot for that long.

[Paul_VR6]

Well it's better then nothing, right?

[Paul_VR6]

https://share.acrobat.com/adc/adc.do?do ... 0451d37408

Yeah closed loop for real control.